Synthesis and Characterization of [n]Cumulenes

Synthesis and Characterization of [n]Cumulenes

––––––––––––––––––––––––––––––––

Synthese und Charakterisierung von [n]Cumulenen

Der Naturwissenschaftlichen Fakultät der Friedrich-Alexander-Universität

Erlangen-Nürnberg

zur Erlangung des Doktorgrades Dr. rer. nat.

vorgelegt von

Johanna Agnes Januszewski

aus Bytów

Als Dissertation genehmigt

von der Naturwissenschaftlichen Fakultät

der Friedrich-Alexander-Universität Erlangen-Nürnberg

Tag der mündlichen Prüfung: 2.12.2014

Vorsitzender des Promotionsorgans: Prof. Dr. Jörn Wilms

Gutachter: Prof. Rik R. Tykwinski PhD

Prof. Dr. Jürgen Schatz

Prof. Dr. Peter R. Schreiner

Die vorliegende Arbeit entstand in der Zeit von Juni 2010 bis August 2014 am Institut

für Organische Chemie (Lehrstuhl I) der Friedrich-Alexander-Universität (FAU)

Erlangen-Nürnberg. Meinem Doktorvater Prof. Rik R. Tykwinski PhD gilt besonderer

Dank für das interessante Dissertationsthema und seine Unterstützung.

Für meine Familie

(v.a. Tata Jan, Mama Krystyna, Kasia,

Wiesia, Darek, Babcia Aniela & Ciocia Ula)

Kocham was

Dla człowieka,

podobnie jak dla ptaka,

świat ma wiele miejsc,

gdzie można odpocząć,

ale gniazdo tylko jedno.

Oliver Wendell Holmes, Sr.

Kurzzusammenfassung

Die vorliegende Arbeit beschreibt die Synthese von zwei ungeradzahligen Serien von

[n]Cumulenen (n = 3, 5, 7, 9, 11 und 13) mit 3,5-Di-t-butylphenyl- und Mesitylendgruppen.

Zahlreiche Charakterisierungsmethoden werden präsentiert, darunter spektroskopische

(NMR- und UV/vis-Spektroskopie), elektronische (Cyclovoltammetrie) sowie strukturelle

Analysen (Röntgenstrukturanalyse). Die erhaltenen Ergebnisse werden zusätzlich durch

theoretische Berechnungen bestärkt. Die hergestellten Cumulene werden untereinander als

auch mit einigen bereits literaturbekannten Cumulenen verglichen und ihre optischen,

elektronischen und strukturellen Eigenschaften untersucht.

Die Synthese von kürzeren [n]Cumulenen (n < 9) war erfolgreich; ab n ≥ 9 zeigten die

Cumulene jedoch eine Stabilitätsgrenze auf. Dennoch wurden Synthesewege für [9]-, [11]-

und [13]Cumulene durchgeführt, die jedoch keinen eindeutigen Nachweis für die Bildung der

längeren [n]Cumulene, mit n = 11 und 13, ergaben. Um die vorhandene Stabilitätsgrenze zu

umgehen, wurden verschiedene Optimierungsansätze vorgenommen. Des Weiteren wurden

Makrozyklen zur Bildung von Cumulenrotaxanen eingesetzt. Die daraus folgende bessere

Abschirmung der instabilen Cumulenkette zeigte bereits beim [9]Cumulenrotaxan, im

Vergleich zu seinem analogen „nackten“ Vertreter, eine erhöhte Stabilität auf. Somit konnten

weitere mögliche Oligomerisierungs- bzw. Zerfallsreaktionen vermieden werden.

Die UV/vis-Spektroskopie von Cumulenen zeigt mit Verlängerung der Cumulenkette,

d.h. mit größerem n, eine Rotverschiebung der kleinsten Energieabsorption λmax und somit

eine Verkleinerung der optischen Bandlücke. Diese Beobachtungen stimmen mit den

Cyclovoltammetrieergebnissen überein, in denen die elektronische Bandlücke mit der

Verlängerung der Cumulenkette ebenfalls verkleinert wird und ähnliche Energiewerte besitzt

wie die optische Bandlücke. Die Röntgenstrukturanalyse bestätigt die bereits für

Polyinsysteme bekannte Bindungslängenalternanz (BLA) und zeigt eine Senkung der BLA

mit Verlängerung der Kumulenkette als auch eine langsame Annäherung an einen Plateauwert

auf.

Abgesehen von der Synthese und Charakterisierung von [n]Cumulenen wurde auch

ihre Reaktivität mittels verschiedener Reaktionen untersucht. Eine davon basierte auf

Additionsreaktionen des Elektronenakzeptors Tetracyanoethylens (TCNE) an die

Kumulenkette verschiedener [5]Cumulene als auch eines [7]Cumulens. Die Reaktion

zwischen dem [5]Cumulen mit 3,5-Di-t-butylphenylgruppen und TCNE ergab sehr

interessante Produkte, u.a. Cyclobutanderivate, Radialene und Dendralene. Die Produkte

konnten durch Additionsreaktionen mit Br2 und ROH funktionalisiert werden. In einem

weiteren Projekt wurde ein funktionalisiertes [5]Cumulen als HCl-Addukt durch die Reaktion

des [9]Cumulens mit 3,5-Di-t-butylphenylgruppen und SnCl2 sowie HCl in CH2Cl2

hergestellt. Zuletzt ergab eine thermische Dimerisierungsreaktion eines [5]Cumulens ein auf

einem Cyclobutanring basierendes Dimer mit vier exocyclischen Alleneinheiten, das durch

Röntgenstrukturanalyse nachgewiesen werden konnte.

Abstract

This thesis deals with the synthesis of two odd-numbered series of [n]cumulenes with

n = 3, 5, 7, 9, 11, and 13 containing 3,5-di-t-butylphenyl and mesityl endgroups. Several

characterization methods have been performed including spectroscopic (NMR- and UV/vis

spectroscopy), electronic (cyclic voltammetry), and structural analysis (X-ray

crystallography). The findings are additionally supported and discussed by the use of

theoretical calculations. The synthesized cumulenes are compared to each other as well as to

several literature known cumulenes, and their optical, electronic, and structural properties

have been investigated.

The synthesis of lower [n]cumulenes (n < 9) was successful; higher cumulenes,

however, showed a stability limit at n ≥ 9. Nevertheless, synthetic approaches to [9]-, [11]-

and [13]cumulene were performed giving no definite confirmation for the formation of

[n]cumulenes with n = 11 and 13. To overcome the stability limitations, several reaction

optimizations, as well as incorporation of macrocycles in order to form cumulene rotaxanes,

were applied. As a result, [9]cumulene rotaxanes already showed a higher stability than their

„naked“ representatives leading to a better shielding of the unstable cumulene chain through

the macrocycle and thus preventing further oligomerization reactions and decomposition.

UV/vis spectroscopy of cumulenes reveals a red-shift of the lowest energy absorption

λmax and thus a reduction of the optical band gap energy by increasing chain length n. These

results are consistent with the findings in cyclic voltammetry measurements, in which the

electronic band gap also decreases with increasing chain length having similar energy values

as the optical band gap. X-ray crystallographic analysis confirms bond length alternation

(BLA), which is already well-known for polyynes, and reveals a decreasing BLA with

increasing cumulene chain length n, tentatively approaching a plateau value.

Aside from the synthesis and characterization of [n]cumulenes, reactivity was

investigated via several approaches. The first one included addition reactions of the electron

accepting tetracyanoethylene (TCNE) molecule to the cumulenic chain of several

[5]cumulenes and one [7]cumulene. The reaction of a [5]cumulene with 3,5-di-t-butylphenyl

endgroups and TCNE resulted in very interesting products including cyclobutane derivatives,

as well as radialenes and dendralenes. These products could be further functionalized by Br2

and ROH addition reactions. In another project, a functionalized [5]cumulene was obtained by

treatment of a [9]cumulene containing 3,5-di-t-butylphenyl groups with SnCl2 and HCl in

CH2Cl2 leading to the formation of a HCl adduct. Finally, thermal dimerization reactions of a

[5]cumulene resulted in the formation of a cyclobutane based dimer with four exocyclic allene

units, which was confirmed via X-ray crystallographic analysis.

Table of Contents

1. Chapter I. Introduction to cumulenes ............................................................................. 1

1.1 Definition of the one-dimensional carbon allotrope carbyne ....................................... 1

1.2 Cumulenes as one possible isomer of the carbon allotrope carbyne ............................ 2

1.2.1 Cumulenes in history and nature .......................................................................... 3

1.2.2 Cumulenes in organometallic chemistry .............................................................. 4

1.2.3 Structural differences of cumulenes ..................................................................... 6

1.3 Synthesis of [n]cumulenes ........................................................................................... 7

1.3.1 General cumulene synthesis ................................................................................. 7

1.3.2 [3]Cumulenes ....................................................................................................... 8

1.3.3 [4]Cumulenes ..................................................................................................... 10

1.3.4 [5]Cumulenes ..................................................................................................... 12

1.3.5 [6]Cumulenes ..................................................................................................... 14

1.3.6 [7]Cumulenes ..................................................................................................... 15

1.3.7 [9]Cumulenes ..................................................................................................... 16

1.4 Reactions of [n]cumulenes with n ≥ 5........................................................................ 17

1.4.1 Miscellaneous reactions ..................................................................................... 18

1.4.2 Cycloaddition and oligomerization reactions ..................................................... 21

1.5 Motivation and goals of the doctoral thesis ............................................................... 26

1.6 References .................................................................................................................. 27

2. Chapter II. Cumulenes – Synthesis of tetraarylcumulenes [n]tBuPh and [n]Mes .... 36

2.1 Synthesis and structure of tetraarylcumulenes [n]tBuPh and [n]Mes....................... 36

2.1.1 General aspects and motivation .......................................................................... 36

2.1.2 Synthesis of the [n]tBuPh cumulene series ....................................................... 38

2.1.2.1 Synthesis of the bis-(3,5-di-t-butylphenyl)methanone endgroup ................... 38

2.1.2.2 Synthesis of [3]cumulene [3]tBuPh ............................................................... 39




2.1.2.6 Synthetic approaches to [11]cumulene [11]tBuPh and [13]cumulene

[13]tBuPh ...................................................................................................................... 53

2.1.3 Synthesis of the [n]Mes cumulene series ........................................................... 56

2.1.3.1 Limitations of “common” synthetic pathways: Toward the synthesis of

precursors to [n]Mes ..................................................................................................... 56

2.1.3.2 Synthesis of [9]cumulene [9]Mes ................................................................... 57




2.2 Summary and conclusion regarding the stability of [n]cumulenes ............................ 64

2.3 Experimental part ....................................................................................................... 66

2.3.1 General procedures and methods ....................................................................... 66

2.3.2 Experimental data and compound characterization............................................ 67

2.4 References .................................................................................................................. 88

3. Chapter III. Cumulene rotaxanes – Synthesis and stability of [n]tBuPh rotaxanes . 91

3.1 General introduction to rotaxanes .............................................................................. 91

3.2 Polyyne rotaxanes as motivation for cumulene rotaxane formation .......................... 93

3.3 Introduction to cumulene rotaxanes: Motivation and target ...................................... 95

3.4 Synthesis of rotaxane precursors and the appropriate cumulene rotaxanes ([9]tBuPh

rotaxanes) using three different macrocycles ....................................................................... 96

3.5 Stability of [9]cumulene rotaxanes and comparison to [9]tBuPh ........................... 100

3.6 Synthetic approach to higher [n]cumulene rotaxanes (n > 9) .................................. 103

3.6.1 Synthetic approach to [11]cumulene rotaxane ................................................. 103

3.6.2 Synthetic approach to [13]cumulene rotaxanes including UV/vis spectroscopy

studies .......................................................................................................................... 106

3.7 Summary and conclusion ......................................................................................... 111

3.8 Experimental part ..................................................................................................... 112

3.8.1 General procedures and methods ..................................................................... 112

3.8.2 Experimental data and compound characterization.......................................... 113

3.9 References ................................................................................................................ 116

4. Chapter IV. Characterization of [3]-, [5]-, [7]-, and [9]tBuPh including [9]tBuPh

rotaxanes and comparison to different series of [n]cumulenes ........................................ 118

4.1 UV/vis spectroscopy ................................................................................................ 118

4.1.1 Introduction ...................................................................................................... 118

4.1.2 UV/vis spectroscopy of [3]-, [5]-, [7]-, and [9]tBuPh ..................................... 120

4.1.2.1 General observations .................................................................................... 120

4.1.2.2 Influence of cumulene chain length .............................................................. 121

4.1.2.3 Influence of endgroups ................................................................................. 122

4.1.2.4 Conclusion including comparison of the band gap of cumulenes ................ 124

4.1.3 UV/vis spectroscopy of [9]cumulene rotaxanes and comparison to [9]tBuPh 124

4.2 X-ray crystallography of [n]cumulenes and discussion of bond length alternation

(BLA) .................................................................................................................................. 127

4.2.1 Introduction ...................................................................................................... 127

4.2.2 General observations ........................................................................................ 128

4.2.3 Bond angles ...................................................................................................... 130

4.2.4 Bond lengths ..................................................................................................... 131

4.2.5 Torsional angles ............................................................................................... 133

4.2.6 Bond length alternation .................................................................................... 134

4.3 Theoretical studies including comparison to UV/vis spectroscopy and BLA analysis .

.................................................................................................................................. 138

4.3.1 Influence of twist angles on BLA and electronic absorption energy ............... 138

4.3.2 UV/vis spectroscopy – Theory and experiment ............................................... 142

4.4 Electrochemistry (cyclic voltammetry) including comparison of the electronic band

gap (Eele) to the optical band gap (Eopt) .............................................................................. 144

4.4.1 Introduction ...................................................................................................... 144

4.4.2 Cyclic voltammetry of [3]tBuPh, [5]tBuPh, and [7]tBuPh ............................ 144

4.4.3 Comparison to known cumulene systems ........................................................ 146

4.4.4 Cyclic voltammetry of a [9]cumulene rotaxane ............................................... 149

4.4.5 Electronic and optical band gap of [n]tBuPh .................................................. 150

4.4.6 Comparison of electrochemical properties of cumulenes and polyynes .......... 152

4.5 NMR spectroscopy of [n]cumulenes ........................................................................ 153

4.5.1 Introduction ...................................................................................................... 153

4.5.2 13C NMR spectroscopy of [9]cumulene rotaxanes and their precursors .......... 154

4.5.3 13C- and correlation NMR spectroscopy of [n]tBuPh (n = 3, 5, and 7) and

[9]cumulene rotaxanes .................................................................................................... 156

4.5.4 Discussion and comparison .............................................................................. 162

4.6 Summary and conclusion ......................................................................................... 164

4.7 References ................................................................................................................ 165

5. Chapter V. Reactions of [n]cumulenes ........................................................................ 169

5.1 Addition reaction of [5]tBuPh with tetracyanoethylene (TCNE) ............................ 169

5.1.1 Motivation and objective .................................................................................. 169

5.1.2 Target, synthetic pathway, and test reactions ................................................... 174

5.1.3 Mechanistic studies and characterization by UV/vis spectroscopy and X-ray

crystallography ................................................................................................................ 182

5.2 Addition reactions of other cumulenes with TCNE ................................................. 189

5.2.1 [7]tBuPh cumulene .......................................................................................... 189

5.2.2 [5]MeOPh cumulene ....................................................................................... 196

5.2.3 [5]oTol cumulene ............................................................................................. 201

5.3 Addition reaction of a [9]cumulene with HCl.......................................................... 202

5.3.1 Synthesis of [5]cumulene 5.31 ......................................................................... 202

5.3.2 Characterization of [5]cumulene 5.31 via UV/vis spectroscopy and X-ray

crystallography ................................................................................................................ 203

5.4 Dimerization of [5]tBuPh ........................................................................................ 206

5.4.1 Synthesis of the dimer of [5]tBuPh ................................................................. 208

5.4.2 Characterization of the dimer of [5]tBuPh ...................................................... 211

5.4.3 X-ray crystallographic data: Discussion and comparison ................................ 212

5.5 Conclusion and summary ......................................................................................... 213

5.6 Experimental part ..................................................................................................... 215

5.6.1 General procedures and methods ..................................................................... 215

5.6.2 Experimental data and compound characterization.......................................... 216

5.7 References ................................................................................................................ 225

List of Figures

Figure 1.1 Schematic depiction of carbyne and homologous series of polyynes and cumulenes

as model compounds for carbyne. .............................................................................................. 2

Figure 1.2 Naturally occurring [3]cumulenes. .......................................................................... 4

Figure 1.3 Three common forms of organometallic cumulenes................................................ 5

Figure 1.4 Triferrocenyl[n]cumulenium salts (n = 2, 4, 6, 8) including mesomeric polyynic

structures. ................................................................................................................................... 6

Figure 1.5 Axial chirality and cis-trans isomerism of cumulenes. ........................................... 6

Figure 1.6 Schematic depiction of major structural classes of [n]cumulenes discussed in this

thesis, where n is the number of cumulated double bonds in a chain constructed of n + 1

carbon atoms. ............................................................................................................................. 8

Figure 1.7 Selected cumulene complexes with diverse coordination patterns. ....................... 20

Figure 1.8 Three possible structures for a [3]H iron carbonyl complex (cumulenic bonds that

are relevant for coordination are marked red). ......................................................................... 21

Figure 1.9 Diradical mesomeric structures as suggested by Hartzler for reactions of a

[5]cumulene. ............................................................................................................................. 22

Figure 1.10 Dimerization products of [5]cumulenes. .............................................................. 24

Figure 2.1 Structures of selected [n]cumulenes. ..................................................................... 37

Figure 2.2 Comparison of hybridization of the outermost carbon atom in a polyyne (left) and

cumulene (right) chain. ............................................................................................................ 38

Figure 2.3 UV/vis spectra taken during attempted conversion of precursor 2.31 to [13]tBuPh

(in Et2O). .................................................................................................................................. 56

Figure 3.1 Definition of a [2]rotaxane. .................................................................................... 91

Figure 3.2 Three common methods for rotaxane formation: a) capping, b) clipping, and c)

slipping. .................................................................................................................................... 92

Figure 3.3 Active template method for rotaxane formation. ................................................... 93

Figure 3.4 Polyyne rotaxanes reported by a) Gladysz and b) Anderson/Tykwinski. .............. 94

Figure 3.5 Three macrocycles, compounds 3.1, 3.2, and 3.3, used in the synthesis of

cumulene rotaxanes. ................................................................................................................. 96

Figure 3.6 DSC scan of [9]cumulene rotaxane 3.9. .............................................................. 102

Figure 3.7 DSC scan of [7]tBuPh. ........................................................................................ 103

Figure 3.8 UV/vis spectra taken during attempted conversion of precursor 3.14 to

[13]cumulene rotaxane 3.13 (in Et2O). .................................................................................. 108


[13]cumulene rotaxane 3.17 (in Et2O). .................................................................................. 111

Figure 4.1 Electronic effects based on odd- and even-numbered, as well as alkyl- and aryl

endcapped [n]cumulenes, demonstrated schematically with canonical structures for [4]- and

[5]cumulenes. ......................................................................................................................... 119

Figure 4.2 UV/vis spectra of [n]cumulenes: UV/vis spectra of a) [n]tBuPh and b) [n]Mes.

Both sets of spectra were measured in Et2O and normalized to the most intense low energy

absorption. Spectra of c) [n]Ph (in benzene) and d) [n]Cy (in Et2O). Spectra of [n]Ph and

[n]Cy were adapted with permission from reference 1. Copyright 1964 John Wiley & Sons.

................................................................................................................................................ 121

Figure 4.3 Qualitative UV/vis spectra (in Et2O) of the [9]cumulene rotaxanes 3.8, 3.9, and

3.10 as well as the “naked” [9]cumulene [9]tBuPh (a quantitative spectrum was recorded for

3.9, see right axis). ................................................................................................................. 125

Figure 4.4 Quantitative UV/vis spectra of [n]tBuPh (n = 3, 5, 9, in CHCl3) and [9]cumulene

rotaxane 3.9 (in Et2O). ............................................................................................................ 127

Figure 4.5 Description of bond lengths and twist angles using [9]Mes (ORTEP drawings with

20% probability level): a) structure of [9]Mes including carbon labeling, b) planes defining

the twist angle (front view), and c) planes defining the twist angle (side view). Carbons C11–

C16 define the grey-colored plane, while carbons C11, C21, and C1–C5, as well as the

appropriate symmetric atoms define the blue-colored plane.................................................. 129

Figure 4.6 Bond angles of the cumulene chain in [n]tBuPh and [n]Mes with n = 3, 5, 7 and

n = 5, 7, 9, respectively. ......................................................................................................... 130

Figure 4.7 Illustration of possible intramolecular C–H/π-interactions of [3]tBuPh. ............ 131

Figure 4.8 X-ray crystallographic structures of [3]tBuPh, [5]tBuPh, [7]tBuPh, [5]Mes,

[7]Mes, and [9]Mes (ORTEP drawings with 20% probability level). ................................... 132

Figure 4.9 Bond lengths (in Å) of the cumulene chain of the series [n]tBuPh, [n]Mes, [n]Ph,

and [n]Cy. .............................................................................................................................. 133

Figure 4.10 BLA values (inset) versus chain length n (lines are only a guide for the eye). . 137

Figure 4.11 Optimized geometries of [9]tBuPh and [9]MePh cumulenes. .......................... 138

Figure 4.12 BLA calculation of [9]Ph versus the aryl twist angle. ....................................... 139

Figure 4.13 Left: [9]tBuPh cumulene: B3LYP-MOs for equilibrium geometry (aryl twist

angle of 32°). Right: [9]Mes cumulene: B3LYP-MOs for SCS-MP2/def2-TZVPP equilibrium

geometry (aryl twist angle of 49°). ......................................................................................... 140

Figure 4.14 Hartree-Fock-MOs for [9]Ph with an aryl twist angle of 90°. .......................... 141

Figure 4.15 UV/vis spectra of one possible conformer of [9]Ph (D2 symmetry) in dependence

of the phenyl twist angle calculated at CC2/def2-TZVPP//SCS-MP2/def2-TZVPP level. ... 142

Figure 4.16 Calculated and experimental UV/vis spectra of [9]MePh/[9]tBuPh (top) and

[9]Mes (bottom) cumulenes. The twist angles are 31° for [9]MePh, as well as 49° for [9]Mes.

All theoretical UV/vis spectra have been computed at the CC2/def2-TZVPP//SCS-MP2/def2-

TZVPP level of theory. .......................................................................................................... 143

Figure 4.17 Cyclic voltammogram of [3]tBuPh. 0.1 M electrolyte (Bu4NPF6) in CH2Cl2.

Ferrocene (Fc) has been used as the internal standard. Scan rate: 150 mV/s......................... 145





Figure 4.20 Cyclic voltammogram of 0.001 M [5]Ph. 0.2 M electrolyte (Bu4NPF6) in

CH2Cl2, referenced to SCE. Scan rate 200 mV/s. The graphic is adapted from ref[20]. ....... 147

Figure 4.21 Cyclic voltammogram of rotaxane 3.9. 0.1 M electrolyte (Bu4NPF6) in CH2Cl2.


Figure 4.22 Plots of electronic (left) and optical (right) band gaps (Eele and Eopt, respectively)

versus 1/n for the cumulenes [3]tBuPh, [5]tBuPh, [7]tBuPh, and 3.9 (band gap data taken

from Table 4.7). ...................................................................................................................... 152

Figure 4.23 Carbon atom labeling of precursors and [9]cumulene rotaxanes for NMR

spectroscopic discussion. ....................................................................................................... 154

Figure 4.24 Comparison of 13C NMR spectra of [9]cumulene rotaxane 3.9 and precursor 3.6

(dotted lines highlight the assignment of signals in the spectrum of [9]cumulene rotaxane 3.9

that result from precursor 3.6, which is present as an impurity). ........................................... 155

Figure 4.25 13C NMR spectra (165–100 ppm region) of [9]cumulene rotaxanes 3.9 (top) and

3.8 (bottom). ........................................................................................................................... 156

Figure 4.26 Decoupled and coupled 13C NMR spectra of [3]tBuPh including the

corresponding HMBC NMR spectrum (aryl region). ............................................................ 158





Figure 4.29 Expansion of the HMBC spectrum of [9]cumulene rotaxane 3.9 (inset: relevant

correlation signals between H2 and C1). ............................................................................... 161


correlation signals between H2 and C1). ............................................................................... 162

Figure 4.31 Plot of 13C NMR carbon chemical shifts versus the number of double bonds n for

[n]tBuPh (n = 3, 5, 7) and [9]cumulene rotaxanes 3.8 and 3.9. ............................................ 163

Figure 5.1 HOMO and LUMO of TCNE and [5]tBuPh. ...................................................... 173

Figure 5.2 Performed reaction of [5]tBuPh with TCNE, monitored by TLC analysis. ........ 175

Figure 5.3 1H NMR spectrum of product A (inset: expansion of the aryl region). ............... 176

Figure 5.4 13C NMR spectrum of product A, allene signal highlighted; * = Et2O. ............... 176

Figure 5.5 Top: Overlaying of product A (in CH2Cl2) with EtOH and MeOH giving adducts

5.12 and 5.13, respectively. Bottom: ORTEP drawings (20% probability level) for compounds

5.12 and 5.13. ......................................................................................................................... 177

Figure 5.6 Identification of product A as cyclic [3]dendralene 5.14..................................... 178

Figure 5.7 Left: Identification of product B as [4]radialene 5.15. Right: ORTEP drawing

(20% probability level) for compound 5.15. .......................................................................... 178

Figure 5.8 Conversion of product C to A (compound 5.14) and B (compound 5.15), and the

associated TLC analysis. ........................................................................................................ 179

Figure 5.9 Conversion of product C with Br2 affording [4]dendralene 5.17; ORTEP drawing

(20% probability level) for compound 5.17. .......................................................................... 180

Figure 5.10 Identification of product C as cyclobutane 5.11. ............................................... 182

Figure 5.11 Computed relative energies (kcal/mol) of the products 5.11 (product C), 5.14

(product A), and 5.15 (product B), as well as the hypothesized product 5.10a, in comparison

to the reactants TCNE and [5]tBuPh. Calculations based on DFT including (red), and without

(blue) dispersion interaction corrections. ............................................................................... 186

Figure 5.12 Quantitative UV/vis spectrum of radialene 5.15 (in CHCl3); Inset: λmax values for

5.15 as a function of solvent. .................................................................................................. 187

Figure 5.13 Reaction of [7]tBuPh with TCNE, and the associated TLC analysis. ............... 190

Figure 5.14 Qualitative UV/vis spectrum of product A measured in hexanes/CH2Cl2 (eluent

from column chromatography). .............................................................................................. 191

Figure 5.15 IR spectrum of product A. ................................................................................. 192

Figure 5.16 1H NMR spectrum (recorded in CD2Cl2) of product C after addition of HCl. .. 193

Figure 5.17 Qualitative UV/vis spectrum of product C before (black curve) and after (red

curve) addition of HCl measured in hexanes/CH2Cl2 (eluent from column chromatography).

................................................................................................................................................ 194

Figure 5.18 Qualitative UV/vis spectrum of product D measured in hexanes/CH2Cl2 (eluent

from column chromatography). .............................................................................................. 195

Figure 5.19 Reaction of [5]MeOPh (containing pink and baseline spot) with TCNE, and the

associated TLC analysis. ........................................................................................................ 197

Figure 5.20 Products B–D with appropriate ORTEP drawings (20% probability level). ..... 198

Figure 5.21 Second and third reactions of [5]MeOPh with TCNE, and the associated TLC

analysis, in comparison to the first reaction. The precursor contained a) only the pink spot and

b) the pink and baseline spot. ................................................................................................. 199

Figure 5.22 Reaction of [5]oTol with TCNE, and the associated TLC analysis. Proposed

product structures are shown as A, B, and D. ........................................................................ 201

Figure 5.23 Qualitative UV/vis spectrum of [5]cumulene 5.31 (in Et2O). ........................... 204

Figure 5.24 Two canonical structures of [5]cumulene 5.31. ................................................. 205

Figure 5.25 Scale-up of [5]tBuPh. ........................................................................................ 207

Figure 5.26 Thermal reaction of [5]tBuPh in toluene, and associated TLC analysis. .......... 209

Figure 5.27 [4]Radialene 5.32 as the desired product from the thermal reaction of [5]tBuPh

in toluene. Additionally, the possible head-to-tail and head-to-head dimers are shown. ...... 209

Figure 5.28 Qualitative UV/vis spectrum of the pink reaction mixture (A) as presented in

Figure 5.26 (in toluene). ......................................................................................................... 211

List of Schemes

Scheme 1.1 Synthesis of [3]cumulenes based on the carbene/carbenoid route. ........................ 9

Scheme 1.2 Synthesis of [3]cumulenes based on reductive elimination of trihaloalkanes. ...... 9

Scheme 1.3 Synthesis of [3]cumulenes based on acetylenic diol derivatives. ........................ 10

Scheme 1.4 Synthesis of [4]Fc based on a diol derivative. ..................................................... 10

Scheme 1.5 Synthesis of [4]cumulenes from diesters. ............................................................ 11

Scheme 1.6 Synthesis of [4]cumulenes from [3]cumulenes. ................................................... 11

Scheme 1.7 Synthesis of a [4]cumulene via carbene trapping with an olefin. ........................ 12

Scheme 1.8 Synthesis of [5]cumulenes based on carbenes/carbenoids. .................................. 13

Scheme 1.9 Synthesis of unsymmetrical substituted [5]cumulenes based on carbenoid

precursors. ................................................................................................................................ 13

Scheme 1.10 Synthesis of [5]Me from [3]Me. ........................................................................ 14

Scheme 1.11 Synthesis of [5]cumulenes based on acetylenic diol derivatives. ...................... 14

Scheme 1.12 Synthesis of [6]Fc. ............................................................................................. 15


Scheme 1.14 Synthesis of the bis[7]cumulene [7]pPh. ........................................................... 16


Scheme 1.16 Oxidation products of [5]tBu. ............................................................................ 18

Scheme 1.17 Examples of cycloaddition and dimerization reactions reported for

[5]cumulenes. ........................................................................................................................... 22

Scheme 1.18 Trimerization reaction of a [5]cumulene. ........................................................... 25

Scheme 1.19 Trimerization reaction of [5]Ph. ........................................................................ 26

Scheme 2.1 Synthesis of ketone 2.2. ....................................................................................... 39

Scheme 2.2 Synthesis of terminal acetylene 2.4. ..................................................................... 39

Scheme 2.3 Synthesis of precursor 2.5 and reductive elimination to [3]tBuPh. ..................... 40

Scheme 2.4 Synthesis of [5]tBuPh. ......................................................................................... 41

Scheme 2.5 Synthetic approaches to precursors to [7]tBuPh, compounds 2.10 and 2.11

(lithiation route). ....................................................................................................................... 42

Scheme 2.6 Synthetic approach to the precursor (2.13) to [7]tBuPh (“mixed” homocoupling

route). ....................................................................................................................................... 43

Scheme 2.7 Synthetic approach to the precursor (2.18) to [7]tBuPh (FBW route). ............... 44

Scheme 2.8 Rearrangement reaction of ketone 2.16 to triyne 2.18, followed by a reductive

elimination to [7]tBuPh. .......................................................................................................... 45

Scheme 2.9 Synthetic approaches to triynes 2.18 and 2.22 (precursors to [7]tBuPh) using

bistrimethylsilyltriyne 2.19. ..................................................................................................... 47

Scheme 2.10 Synthesis of [9]tBuPh. ....................................................................................... 48

Scheme 2.11 Synthesis of tetrayne 2.14. ................................................................................. 49

Scheme 2.12 Synthetic approach to [11]tBuPh. ...................................................................... 53

Scheme 2.13 Synthetic approach to [13]tBuPh. ...................................................................... 54

Scheme 2.14 Unsuccessful approach to mesityl acetylene 2.33. ............................................. 57

Scheme 2.15 Synthesis of [9]Mes. .......................................................................................... 58


Scheme 2.17 Reaction of terminal acetylene 2.42 with EtMgBr and ethyl formate. .............. 62


Scheme 2.19 Pd-catalyzed homocoupling reactions giving precursor 2.49. ........................... 64

Scheme 3.1 Synthesis of polyyne rotaxanes. ........................................................................... 95

Scheme 3.2 Formation of rotaxanes 3.4, 3.5, and 3.6 via an active metal templated

homocoupling reaction. ............................................................................................................ 97

Scheme 3.3 Formation of rotaxanes 3.4 and 3.6 via an active metal templated Cadiot-

Chodkiewicz heterocoupling reaction. ..................................................................................... 98

Scheme 3.4 Synthesis of cumulene rotaxanes 3.8, 3.9, and 3.10............................................. 99

Scheme 3.5 Synthetic approach to [11]cumulene rotaxane 3.11. .......................................... 105



Scheme 5.1 Reaction of [3]cumulenes with TCNE in CH2Cl2 at rt....................................... 170

Scheme 5.2 Conversion of cyclobutane 5.3 to 5.7 and 5.8. ................................................... 170

Scheme 5.3 TCNE addition to [5]Fc giving cycloadduct 5.9. ............................................... 171

Scheme 5.4 General cyclization/cycloreversion method giving donor-substituted TCBD

derivatives. ............................................................................................................................. 171

Scheme 5.5 Expected cyclization/cycloreversion reaction of a [5]cumulene with TCNE to the

unsymmetrically substituted [3]cumulene 5.10. .................................................................... 172

Scheme 5.6 Reaction of a push-pull [3]cumulene with EtOH and Br2, respectively. ........... 179

Scheme 5.7 Conversion of product B (compound 5.15) to [4]dendralenes 5.19 and 5.20;

ORTEP drawing (20% probability level) for compound 5.19. .............................................. 181

Scheme 5.8 Overview of compounds identified from the reaction of [5]tBuPh with TCNE

including further addition reactions. ...................................................................................... 183

Scheme 5.9 Summary of a mechanism of the reaction of a [3]cumulene with TCNE as

suggested by Kawamura and coworkers. ............................................................................... 183

Scheme 5.10 Proposed mechanism for the conversion of 5.11 to 5.14 and 5.15. ................. 184

Scheme 5.11 Proposed concerted mechanism for the addition reaction of bromine to 5.11 and

the stepwise addition of ROH to 5.14. ................................................................................... 185

Scheme 5.12 Synthetic approach to [5]cumulene [5]MeOPh. .............................................. 200

Scheme 5.13 Synthesis of [5]cumulene 5.31 via reductive elimination of tetrayne 2.14 in

CH2Cl2, and the associated TLC analysis. ............................................................................. 203

Scheme 5.14 Thermal dimerization of [5]cumulenes reported by Hartzler and Iyoda. ......... 207

Scheme 5.15 Thermal trimerization of [5]Ph reported by Kawamura. ................................. 208

List of Tables

Table 2.1 Reductive elimination of 2.25 under various conditions. ........................................ 52

Table 2.2 Methylation reaction of 2.42 under various conditions. .......................................... 60

Table 3.1 Comparison of the qualitative stability of [9]tBuPh and [9]cumulene rotaxane 3.9

when kept under an argon atmosphere at rt. ........................................................................... 101

Table 4.1 Lowest energy absorption λmax (in nm) and energy values Eg (in eV) of [n]tBuPh,

[n]Mes, [n]Ph, and [n]Cy (in Et2O). ..................................................................................... 122

Table 4.2 UV/vis spectroscopic data (λmax in nm) of selected [5]cumulenes with different

endgroups. .............................................................................................................................. 123

Table 4.3 UV/vis spectroscopic data (absorption wavelengths in nm) of [9]tBuPh and

[9]cumulene rotaxanes 3.9, 3.10, and 3.8 (in Et2O). .............................................................. 126

Table 4.4. Aryl twist angles of aromatic ring relative to cumulenic framework. .................. 134

Table 4.5 Selected bond lengths (Å) for cumulene series [n]Ph, [n]Cy, [n]tBuPh and

[n]Mes, as well as theoretically calculated values for [n]H including BLA data. ................ 136

Table 4.6 CV data of [5]cumulenes [5]tBu, [5]tBu/Ph, [5]EtPh, and [5]tBuPh. ................ 149

Table 4.7 Selected UV/vis spectroscopic and electrochemical details including optical (Eopt)

and electronic (Eele) band gap values for [3]tBuPh, [5]tBuPh, [7]tBuPh, and 3.9. .............. 151

Table 5.1 Natural population analysis (NPA) charges of the cumulenic C-atoms in

[5]cumulene [5]tBuPh. .......................................................................................................... 173

Table 5.2 Bond lengths (Å) of radialene 5.15 and selected radialenes known from literature.

................................................................................................................................................ 189

Table 5.3 Bond lengths (Å) and bond angles (°) of [5]cumulene 5.31. ................................. 206

Table 5.4 Bond lengths (Å) and bond angles (°) of [4]radialene 5.32 (left) and Iyoda’s

[4]radialene 5.33 (right). ........................................................................................................ 213

List of Abbreviations and Symbols

Å Angström

Ac acetyl

ACN acetonitrile

Ad adamantyl

An 10,10-dimethyl-9,10-dihydroanthracenyl

anal analytical

anhydr anhydrous

APPI atmospheric pressure photoionization

aq aqueous

Ar aryl

ATR attenuated total reflection

BLA bond length alternation

br broad

Bu butyl

calcd calculated

CCDC Cambridge Crystallographic Data Centre

cm centimeter(s)

cmpd compound

CSD Cambridge Structural Database

CuAAC copper(I)-catalyzed azide-alkyne cycloaddition

CV cyclic voltammetry

Cy 2,2,6,6-tetramethylcyclohexyl

d doublet

d day(s)

deg degree(s)

decomp decomposition

DWCNT double-wall carbon nanotube

δ chemical shift

∆ heat

DCTB trans-2-(3-(4-t-butylphenyl)-2-methyl-2-propenylidene)malononitrile

DFT density functional theory

DSC differential scanning calorimetry

E potential

ε molar extinction coefficient

EI MS electron impact mass spectroscopy

ESD estimated standard deviation

ESI electrospray ionization

Et ethyl

EtOAc ethyl acetate

equiv equivalent(s)

eV electron volt(s)

FBW Fritsch-Buttenberg-Wiechell

Fc ferrocenyl

Fl fluorenyl

FT-ICR Fourier transform ion cyclotron resonance mass spectrometry

g gram(s)

h hour(s)

HMBC heteronuclear multiple bond correlation

HOMO highest occupied molecular orbital

HRMS high resolution mass spectrometry

Hz Hertz

i iso

IR infrared

irrev irreversible

J coupling constant

kcal kilocalorie

λ absorption wavelength

λmax lowest energy absorption wavelength

L liter(s)

LUMO lowest unoccupied molecular orbital

m multiplet (NMR)

m medium (IR)

µ micro

m meta

m-CPBA meta-chloroperoxybenzoic acid

M formula weight

M molar

MALDI matrix-assisted laser desorption ionization

Me methyl

Mes mesityl

mg milligram(s)

MHz megaHertz

mL milliliter(s)

mmol millimole(s)

mol mole(s)

Mp melting point

MS mass spectroscopy

mV millivolt(s)

m/z mass-to-charge ratio

n-BuLi n-butyllithium

NBS N-bromosuccinimide

NIR near-infrared

nm nanometer(s)

NMR nuclear magnetic resonance

o ortho

ORTEP Oak Ridge thermal ellipsoid plot

p para

PCC pyridinium chlorochromate

Ph phenyl

ppm parts per million

Pr propyl

q quartet

quant quantitative

qurev quasireversible

Rf retention factor

ref reference

rt room temperature

s singlet (NMR)

s strong (IR)

s second(s)

sev several

Sub suberyl

supertrityl tris(3,5-di-t-butylphenyl)methyl

t triplet (NMR)

t tertiary

TBAF tetrabutyl ammonium fluoride

TCBD tetracyanobutadiene

TCNE tetracyanoethylene

temp temperature

TFE trifluoroethylene

THF tetrahydrofuran

THP tetrahydropuran

TIPS triisopropylsilyl

TLC thin layer chromatography

TMEDA N,N,Nʹ′,Nʹ′-tetramethylethylenediamine

TMS trimethylsilyl

TOF time-of-flight

Tol tolyl

UV ultraviolet

UV/vis ultraviolet-visible

vw very weak (IR)

w weak (IR)

1

1. Chapter I. Introduction to cumulenes†

1.1 Definition of the one-dimensional carbon allotrope carbyne

The three- and two-dimensional carbon allotropes represent the insulating sp3-

hybridized carbon allotrope diamond and the conducting sp2-hybridized carbon allotropes

graphite and graphene (as well as the recently discovered fullerenes and carbon nanotubes),

respectively. Both allotrope classes are well-investigated and show interesting electronic,

physical, and optical properties. Resulting applications are used in all fields of chemistry

ranging from molecular wires over thin film transistors to solar cells, etc.1–6 Aside from these

carbon allotropes, one example, the one-dimensional sp-hybridized carbon allotrope carbyne,

is missing. In contrast to the three- and two-dimensional allotropes, the one-dimensional

allotrope carbyne has rarely been investigated and characterized. The existence of carbyne has

been a topic of much and sometimes controversial7,8 discussion over the years.6,9–15 The

natural existence of this one-dimensional allotrope has been proposed in, for example,

meteorites,16,17 interstellar dust,18 and shock-compressed graphite,19 as well as terrestrial plant,

fungal, and marine sources (in the case of polyynes).20–22 Carbyne is synthetically not readily

available or isolable like other carbon allotropes. Evidence of carbyne formation has been

provided by a variety of processes, e.g., by solution-phase synthesis,13,23,24 laser irradiation of

graphite,25 or gas-phase deposition methods.26 Whether naturally occurring or produced in a

laboratory, carbyne remains rather poorly characterized as a material, and specific properties

of carbyne are thus often difficult to define. There are, nevertheless, quite a number of

fascinating properties and resulting applications predicted for carbyne based on unusual

electrical and optical nature of sp-hybridized carbon,27,28 such as nanoelectronic or spintronic

devices,29 nonlinear optical materials,30–33 and molecular wires.34–36 Recent theoretical

calculations have also suggested that under tension, carbyne could be twice as stiff as the

stiffest known materials, such as carbon nanotubes, graphene, and diamond.29,37 Furthermore,

† A version of this chapter has been published: J. A. Januszewski, R. R. Tykwinski, Chem. Soc. Rev. 2014, 43,

3184–3203, see http://dx.doi.org/10.1039/C4CS00022F - Reproduced by permission of The Royal Society of

Chemistry; Portions of this chapter have been published: J. A. Januszewski, F. Hampel, C. Neiss, A. Görling, R.

R. Tykwinski, Angew. Chem. Int. Ed. 2014, 53, 3743–3747.

2

even though carbyne might be extremely stiff against strain, it might also allow for

deformation through bending without influence on properties.38

1.2 Cumulenes as one possible isomer of the carbon allotrope carbyne

In principle, two possible forms of carbyne exist. On one hand, carbyne can be built up

from sp-hybridized carbon atoms with alternating triple and single bonds forming a series of

semiconducting molecules called polyynes. On the other hand, cumulenes, composed of

consecutive double bonds, define the alternative form of carbyne and are assumed to possess

metallic behavior (Figure 1.1).39,40 Which form is preferred remains a so far unanswered

question that still needs to be investigated.

Figure 1.1 Schematic depiction of carbyne and homologous series of polyynes and cumulenes

as model compounds for carbyne.

Considering the synthetical point-of-view, polyynes have been studied intensively

over the past decades.41–48 The longest polyyne that has been synthesized contains 44 carbon

atoms in a chain constructed of 22 acetylene units.41 Due to the fact that instability increases

with increasing chain length, this polyyne is endcapped with bulky endgroups shielding the

most reactive outer acetylene units of the chain. The experimental results illustrate that this

compound has not yet achieved a carbyne-like “status” when comparing the UV/vis

spectroscopic data, which show that no saturation of the optical band gap has been reached.

Extrapolation from the spectroscopic data predicts saturation of properties at the point of 48

acetylene units in order to form a carbyne-like compound, i.e., where elongation through

additional acetylene units has no effect. In contrast, the investigation of [n]cumulenes (where

n is the number of cumulated double bonds in a chain constructed of n + 1 carbon atoms) as

model compounds for carbyne has barely been discussed in the literature. To date, the longest

3

cumulenes to be synthesized and studied are [9]cumulenes, i.e., molecules with nine

consecutive double bonds in a chain of 10 carbons.49–52 On the basis of structure, a

[9]cumulene is only approximately equal to the length of a rather short polyyne (a tetrayne).

Unlike polyynes of this length, however, [9]cumulenes show dramatic instability under

ambient conditions.

In conclusion, regarding the practical results achieved in laboratory, polyynes seem to

be more accessible compared to cumulenes. Recent theoretical predictions, however, show

that polyynes and cumulenes are linked to each other, more than initially expected. Yakobson

and coworkers,29 for example, predict a transition from cumulenic to acetylenic geometry as

carbyne is stretched. This effect goes along with an increase in bond length alternation (BLA),

energy of Peierls distortion, and band gap. Thus, a transition from a metallic to an insulating

state could be reached, suggesting an interesting property for carbyne-like compounds as

conducting polymers. The reverse process has also been predicted, and studies suggest a

transition from a polyyne structure to a cumulene form, for example, under UV

photoexcitation53 leading to a new photogenerated species that is cumulenic. Another study of

the transition from a polyynic to a cumulenic form is suggested to proceed via a charge

transfer between metal nanoparticles such as silver and gold connected by phenyl-substituted

oligoynes.39 Consequently, the observed rearrangement of the polyynic structure to a

cumulenic-like structure could result in possible applications as tunable electronics. Finally,

theoretical calculations typically predict higher stability for the polyyne form of

carbyne.29,54,55 This prediction is also consistent with practical results. An ultimate assignment

of carbyne to either the polyyne or the cumulene form, however, remains undetermined.

1.2.1 Cumulenes in history and nature

The history and thus the interest in research of polyynes have started in the end of the

19th century and the beginning of the 20th century represented by pioneers like von Bayer56

and Dupont.57 Later, from 1950 on, Jones,45,58 Walton,43,44,59 and Bohlmann41,60 have played a

huge role in this field. The chemistry of cumulenes, on the other hand, has started with its

most famous pioneer, Richard Kuhn, in the 1930s.61 It is Kuhn, who has lent the cumulene the

name („Kumulen“) that is commonly used to date. Before 1938, however, some evidence of

cumulenes have been observed. Allenes ([2]cumulenes), the smallest cumulenes, containing

two cumulated double bonds, have been investigated starting with the unknowing formation

4

of an allene by Burton and Pechmann in 188762 (the allenic structure was revealed 67 years

later by Jones63). The first [3]cumulene, the tetraphenylbutatriene has been accidentally made

by Brand64 in 1921. The first [5]cumulene, also a tetraphenyl-substituted cumulene has been

formed by Kuhn61 in 1938, and since then, the study of cumulenes has reached their zenith,

mostly investigated by the groups of Kuhn, Cadiot, and Bohlmann, especially in the 1950s

and 1960s.50,51,65–69

Cumulenes, mostly as allenes, have also been observed in and extracted from natural

sources70 from the 1920s to the 1960s by, e.g., Staudinger,71 Celmer,72 and Jones.73–75

[3]Cumulenes are also present in nature, but rarely. The first „natural“ [3]cumulene has been

isolated by Bohlmann et al. from Conyza bonariensis, and its instability allows handling only

in solution or in crystalline form at −70 °C, otherwise it tends to polymerization.76 Three

additional [3]cumulenes, also extracted by Bohlmann, have followed between 1966 and 1971,

and are the last examples discovered to date (Figure 1.2). All four [3]cumulene natural

products show high instability, thus, cumulenes are not only less common in nature compared

to the polyynes, but also less stable.20,77,78

Figure 1.2 Naturally occurring [3]cumulenes.

1.2.2 Cumulenes in organometallic chemistry

Aside from the application of cumulenes in organic chemistry, this class of

compounds has also been extensively used in organometallic chemistry, as metallacumulenes

(metallated cumulenes), i.e., coordinated directly to metal centers, or as cumulenes that

possess organometallic endgroups, e.g., ferrocenyl groups. Three common forms of

organometallic cumulenes are presented in Figure 1.3. Finally, several “unusual” cumulene

complexes exist that show a more exotic coordination of the cumulene chain to the metal

centers, such as multiple coordinations or bimetallic systems.79–83 All organometallic

cumulenes are presented in more detail, regarding the reactivity of these compounds, in

Section 1.3.1.

5

Figure 1.3 Three common forms of organometallic cumulenes.

While the smaller metallacumulenes appear quite numerous in the field of

organometallic chemistry as vinylidene or allenylidene metal complexes84–86 containing a

M=C=C or a M=C=C=C unit, respectively, complexes with higher cumulenes remain rare to

date.87,88 In the case of higher metallacumulenes, pentatetraenylidene complexes82,89–91 are

reported most frequently, while there are only few analogues of the lower butatrienylidene

complexes92 and just one heptahexaenylidene93 complex known although it has not been

isolated. Considering application possibilities, metallacumulenes have gained attention as

potential precursors for optical or electronic materials, such as e.g., molecular wires

(as bimetallic metallacumulenes),83,91 or in synthesis for important catalytic reactions.82,86,88,91

In the case of ferrocenyl-substituted cumulenes, Bildstein94 is the pioneer in this field

after forming a series of ferrocenyl[n]cumulene, ranging from n = 2–6 as well as

ferrocenyl[n]cumulenium salts with n = 2, 4, 6, 8 (Figure 1.4). This class of compounds is

positioned between “metallacumulenes” and typical “organic cumulenes” and shows redox

activity, donor properties, and reasonable stability that can be attributed to the ferrocenyl

endgroups.

6

Figure 1.4 Triferrocenyl[n]cumulenium salts (n = 2, 4, 6, 8) including mesomeric polyynic

structures.

1.2.3 Structural differences of cumulenes

Depending on the number of double bonds n in a cumulenic chain, [n]cumulenes show

different stereochemical properties. Even-numbered [n]cumulenes (n = even) show axial

chirality and thus optical activity in the case of adequate choice of endgroups. In contrast,

odd-numbered [n]cumulenes (n = odd) can possess cis-trans isomerism (Figure 1.5).95 This

difference not only highly influence the properties of these two classes of cumulenes, it also

has a big impact from the synthetic point-of-view. While odd-numbered cumulenes can be

synthesized rather easily using several standard synthetic methods, even-numbered cumulenes

are more difficult to obtain due to synthetic accessibility of the precursors, and no general

synthetic route can be applied.

Figure 1.5 Axial chirality and cis-trans isomerism of cumulenes.

7

1.3 Synthesis of [n]cumulenes

1.3.1 General cumulene synthesis

Since Kuhn’s early work on cumulenes in the 1930s,61 quite a number of synthetic

approaches have been developed to provide [n]cumulenes of various lengths. For the shorter

analogues with n = 3–5, various common methods exist, mainly because of greater stability of

these cumulenes when compared to longer [n]cumulenes (n > 5). Better stability of the

cumulene product also fosters greater functional group tolerance, and many more shorter

cumulenes have been synthesized with n < 5. The synthesis of longer cumulenes has benefited

greatly from the pioneering advances for assembling acetylenic compounds made by the

groups of Bohlmann,41,60 Walton,43,44,59 and Jones,45,58 as well as others.96–98 From this era,

acetylene compounds, and more precisely oligoyne diols,60,99 have emerged as most

convenient synthetic precursors for [n]cumulenes.

The following overview gives a summary of synthetic methods for cumulenes and is

presented based on ascending molecular length, starting from [3]cumulenes and concluding

with the longest [n]cumulenes known to date, the [9]cumulenes.100 Standard synthetic

pathways as well as alternative synthetic approaches are presented giving mainly symmetrical

substituted cumulenes (four identical endgroups), although some exceptions are found. It is

also noteworthy that the synthesis of tetraalkyl-substituted [n]cumulenes is typically more

difficult when compared to tetraaryl[n]cumulene, primarily due to stability of the products.

Finally, the assembly of even-numbered [n]cumulenes (n = 4 and 6) is more complicated than

odd-numbered [n]cumulenes (n = 5, 7, 9) often due to synthetic accessibility of the

precursors. Metallacumulenes are not included in this synthetic summary.88,101–104 The

synthetic routes to [2]cumulenes, i.e., allenes, are not covered in this thesis. In addition, only a

few synthetic methods for [3]cumulenes are depicted, since too many individual approaches

have been reported to date, and this work has been reviewed.105–107

To facilitate discussions in synthesis and reactivity sections in the introduction as well

as the main part of the thesis, Figure 1.6 offers a legend that introduces the endgroups of

cumulenes and the associated nomenclature.

8

R

R R

R

m

t-Bu

t-Bu

[n]Fl [n]An

R

R

=

Ph

Ph

[n]Ph

t-Bu

t-Bu

[n]Cy[n]Ad

[n]Mes[n]tBuPh

[n]Sub

Fe

Fe

[n]Fc

[n]Me[n]tBu

[n]MeOPh

OMe

OMe

[n]cumulenes(n = m + 2)

H

H

[n]H

[n]EtPh [n]pPh

[n]oTol

Fe

[n]Fc/Ph

[n]tBu/Ph

[n]MePh

Figure 1.6 Schematic depiction of major structural classes of [n]cumulenes discussed in this

thesis, where n is the number of cumulated double bonds in a chain constructed of n + 1

carbon atoms.

1.3.2 [3]Cumulenes

One of the common routes to [3]cumulenes is the metal catalyzed dimerization of

carbenes/carbenoids (Scheme 1.1), as summarized in a review by Stang.108 These can be

accomplished with Cu(I) catalyst for aryl or alkynyl substituents as described by

Diederich,109,110 as well as Kunieda and Takizawa.111 On the other hand, Iyoda112,113 has used

a Ni(II) catalyst to form [3]cumulenes. Tetraalkyl[3]cumulenes can be made directly from a

lithium carbenoid as reported independently by, for example, Köbrich,114 Komatsu,115 and

Oda.116 Stang, on the other hand, successfully has formed a variety of alkyl-substituted

[3]cumulene via the carbenoid route using ethynylvinyl triflates as precursors.117,118

9

Scheme 1.1 Synthesis of [3]cumulenes based on the carbene/carbenoid route.

An alternative method to generate carbenoids for the synthesis of [3]cumulenes is

based on elimination of trihaloalkanes (Scheme 1.2). Direct elimination using a Cu/Cu(I)

catalyst system gives [3]cumulene products in moderate yields (ca. 40–60%).111 Brand and

coworkers,119,120 on the other hand, describe a route that proceeds through a dichloroalkene

intermediate formed through reduction of the trihaloalkane leading to higher overall yields

(>85%) than the direct elimination although it requires three distinct synthetic steps.

Scheme 1.2 Synthesis of [3]cumulenes based on reductive elimination of trihaloalkanes.

The most common synthetic route to [3]cumulenes is based on the reduction of an

acetylenic diol or diether derivative (Scheme 1.3). The precursor is usually a

Li- or Mg-acetylide, 49,50,57,61,94,121,122 which is used in an addition reaction with a ketone that

defines the endgroups. In cases where the free alcohol may not be compatible with subsequent

synthetic steps, it can be blocked by formation of, for example, a methyl ether through

trapping with MeI.49 From either the diol or diether, aryl-substituted [3]cumulenes can be

formed directly via reduction with SnCl2,49,94 although P2I4

61 or HI and I2122 have also been

used. Tetraalkyl[3]cumulenes are usually formed via halogenation, followed by reductive

elimination (Scheme 1.3).50,123,124

10

Scheme 1.3 Synthesis of [3]cumulenes based on acetylenic diol derivatives.

1.3.3 [4]Cumulenes

Bildstein and coworkers have reported the formation of [4]Fc based on an adaptation

of the diol approach described above for [3]cumulenes (Scheme 1.4).94,125 A homopropargylic

ether is formed, lithiated, and then added to diferrocenylketone to complete the carbon

skeleton. Treatment with acid (HBF4) results in the stabilized (and isolable) [3]cumulenium

intermediate, which can then be converted to [4]Fc through base induced elimination.

Bildstein’s approach is conceptually similar to that reported earlier by Nakagawa and

coworkers,126 in which two different endgroups have been introduced in order to explore

optical activity and racemization of [4]cumulenes.

Scheme 1.4 Synthesis of [4]Fc based on a diol derivative.

11

Another possibility to form tetraaryl[4]cumulenes has first been described by Kuhn127

and later by Karich and Jochims128 and relies on the intermediate formation of

dibromo-1,4-pentadienes from the appropriate unsubstituted dienes (Scheme 1.5). The diene

is then converted to the [4]cumulene in good yield through base-induced elimination.

Scheme 1.5 Synthesis of [4]cumulenes from diesters.

[3]Cumulenes can be converted to [4]cumulenes via addition of dichlorocarbene to a

[3]cumulene, followed by rearrangement or reductive elimination, depending on the structure

of the precursor (Scheme 1.6). Jochims and coworkers highlight the potential effectiveness of

this route through formation of [4]Ad, where MeLi gives quantitative yield in the reduction

step. Their route also provides rare examples of mixed dialkyl/diaryl-endcapped

[4]cumulenes.128 Irngartinger and Götzmann have developed a slightly modified version of

this general protocol to synthesize [4]Cy, using Zn in the final reductive elimination step

(Scheme 1.6).129

Scheme 1.6 Synthesis of [4]cumulenes from [3]cumulenes.

[4]Cumulenes can be obtained via carbene trapping as demonstrated by le Noble and

coworkers (Scheme 1.7).130,131 The dialkylpentatetraenylidene intermediate is formed and

reacts in situ with tetramethylethylene to give the [4]cumulene, which quickly converts to a

12

radialene through dimerization (in the case of R = Me). The analogous reaction with

2-adamantylpentatetraenylidene, on the other hand, forms the stable [4]cumulene.

Scheme 1.7 Synthesis of a [4]cumulene via carbene trapping with an olefin.

1.3.4 [5]Cumulenes

Of the higher [n]cumulenes (n ≥ 5), [5]cumulenes are by far the most studied, and

there are thus a number of efficient routes that have been developed for their synthesis.100 In

analogy to the syntheses described for [3]cumulenes, dimerization reactions of carbenes have

also been used to give [5]cumulenes (Scheme 1.8).67,121,132–135 Starting with a terminal

acetylene and a leaving group in the propargylic position, reaction with base produces a

carbene intermediate, which leads to the [5]cumulene via dimerization. An alternative

carbenoid route to [5]cumulenes has been described by Kollmar and Fischer,136 in which the

vinylidene carbenoid is generated directly from a haloallene (Scheme 1.8). It is interesting to

note that the influence of endgroups in this reaction is likely enhanced versus the analogous

reaction to give [3]cumulenes, considering the mesomeric stabilization of this intermediate

carbene (Scheme 1.8).

13

Scheme 1.8 Synthesis of [5]cumulenes based on carbenes/carbenoids.

Unsymmetrical [5]cumulenes can be synthesized through trapping of vinylidene

carbenes, as reported by Stang and coworkers (Scheme 1.9).118,137,138 More specifically,

elimination of a diyne vinyl triflate forms the intermediate carbene, which can be trapped by

either addition to electron rich alkenes (i.e., tetramethylethylene and cyclohexene) or M–H

bond insertion using R3MH (M = Si, Ge). In some cases, the resulting [5]cumulene is not

stable and isomerizes or polymerizes during the reaction.

Scheme 1.9 Synthesis of unsymmetrical substituted [5]cumulenes based on carbenoid

precursors.

[5]Cumulenes can be formed based on the reaction of a [3]cumulene with

dibromocarbene as described by Skattebol (Scheme 1.10).139 The addition of dibromocarbene

14

to tetramethyl[3]cumulene gives the bicyclopropylidene product, and the subsequent

rearrangement reaction, induced with MeLi, gives the unstable [5]Me product.

Scheme 1.10 Synthesis of [5]Me from [3]Me.

Assembly of [5]cumulenes based on acetylenic diol precursors is quite common

(Scheme 1.11), in part because the necessary precursor, a diyne diol, can be efficiently formed

via a number of routes. Often mimicking strategies described for [3]cumulenes, acetylenic

diols are thus readily assembled through, for example, the addition of a Li- or Mg-acetylide to

a ketone. Alternatively, oxidative homocoupling of propargyl alcohol derivatives is usually

quite efficient. With the diyne diol in hand, conversion directly to an aryl-substituted

[5]cumulene is accomplished via reduction with P2I4,61 CrCl2,

65 or SnCl2.49,68,140 In the

majority of recent studies, SnCl2 is the reductant of choice and usually gives good yields. In

the case of alkyl substitution, conversion of the diyne diol to the corresponding dihalide with

PI3,50 PBr3,

50,140 HBr,140 or HCl140 is required, which is then followed by reductive elimination

using Zn or n-BuLi.

R

R

OH

R

R

HO2

H

R

R

HO

M M2

R

R R

R

R

R

X

X

R

R

R = aryl

R = alkyl

reductive

elimination

Zn or n-BuLi

X = Br or Cl

reductiveelimination

halogenation

O

R R

M = Li, MgBr

R = alkyl, aryl

homo-

coupling


1.3.5 [6]Cumulenes

To date, only one synthesis has been reported for a [6]cumulene, namely [6]Fc

(Scheme 1.12). Bildstein and coworkers141 have shown that a diyne diether can be readily

15

assembled via an acetylenic cross-coupling reaction, and the [6]cumulene is then realized

through an elimination process similar to the synthesis of the [4]Fc. It is worth mentioning

that after the first elimination step, an unusual, air-stable cumulenium salt is obtained, and the

stability is explained by the presence of four ferrocene donor groups and the unsaturated

cumulene chain. Unfortunately, the resulting air-sensitive [6]Fc product cannot be isolated,

but UV/vis spectroscopy confirms the formation of the cumulene framework.

Scheme 1.12 Synthesis of [6]Fc.

1.3.6 [7]Cumulenes

To our knowledge, only seven tetraaryl[7]cumulenes and one tetraalkyl[7]cumulene

have been reported to date49,50,66,69,142 including the [7]cumulenes formed during this thesis

research. All [7]cumulenes have been assembled from the corresponding triyne diol precursor

(Scheme 1.13), which is typically formed as described above for [3]- and [5]cumulene

syntheses. An alternative approach to the requisite diol has recently been developed in this

thesis, using a carbenoid Fritsch-Buttenberg-Wiechell (FBW) rearrangement143–145 in

combination with Colvin’s reagent146–148 to form the triyne framework.49 This synthetic

pathway will be covered in more detail in the main part of the thesis (see Chapter II).

Following the trends established by [3]- and [5]cumulene syntheses, formation of the

tetraalkyl[7]cumulenes requires conversion of the diol to the dibromide, followed by

reductive elimination with Zn. In contrast, the tetraaryl[7]cumulenes can be formed directly

by reduction with either P2I4 or SnCl2. In most cases, the [7]cumulenes tend to decompose

quickly, both in solution and the solid state.

16


Perhaps the most structurally interesting [7]cumulene reported to date, [7]pPh, has

been synthesized by Cadiot and coworkers through a variation of the diol approach as

outlined in Scheme 1.14.69 In this case, the acetylenic diol precursor has been assembled using

a Cu-catalyzed heterocoupling between a terminal diacetylene and a bromoacetylene

derivative. While the final product cannot be isolated, UV/vis spectroscopy confirms

formation, with a lowest energy absorption at 700 nm that is red-shifted versus that of all

other [7]cumulenes (see Chapter IV).

Scheme 1.14 Synthesis of the bis[7]cumulene [7]pPh.

1.3.7 [9]Cumulenes

As a result of instability of the final product, very few [9]cumulenes have been

successfully assembled, and there are only five examples to be found in the literature,

including four tetraaryl- ([9]Ph, [9]Sub, [9]tBuPh, [9]Mes)49,51,52 and one

17

tetraalkyl[9]cumulene ([9]Cy).50 From that, two of the [9]cumulenes, [9]tBuPh and [9]Mes

have been formed in the group of Tykwinski.149 The synthetic pathway to [9]cumulenes is,

predictably, analogous to that of [5]cumulenes, except diynes provide the precursors for the

dimerization, rather than terminal alkynes (Scheme 1.15). As per usual, tetraaryl derivatives

are formed directly from the diol via reductive elimination using P2I4 or SnCl2, while the

tetraalkyl[9]cumulene [9]Cy must be synthesized from the dibromide, via reductive

elimination with Zn. Isolated yields have not been reported for any of the [9]cumulenes due to

the inability to isolate the unstable products.


1.4 Reactions of [n]cumulenes with n ≥ 5

The reactivity of [n]cumulenes with n = 2 and 3 has been reviewed by Diederich,150

Chauvin,105 and Ma.151 The increasing instability in longer [n]cumulenes (n ≥ 5), however,

has limited the reactions of these molecules. Thus, reactions of longer cumulenes have been

rare and are limited almost exclusively to reactions of [5]cumulenes. Due to its multitude of

double bonds, members of this class of compounds show many attractive positions for

addition and/or cyclization reactions, and products thus vary in symmetry and conjugation

depending on the regiochemistry of the addition leading to unique structures as outlined

below.

18

1.4.1 Miscellaneous reactions

Simple reactions of [5]cumulenes have been reported such as hydrogenations134,152 and

partial hydrogenations153,154 of the cumulene core using H2/Rh/alumina and Al/Hg or the

Lindlar catalyst. The oxidation of a [5]cumulene via epoxidation has been described by

Crandall and coworkers (Scheme 1.16).155 For example, reaction of [5]tBu with m-CPBA

gives a cyclopropanone intermediate, which goes on to give an allenic ester as the product.

Epoxidation with dimethyldioxirane gives the cyclopropanone as a stable product, which can

then be used to form [4]tBu through either thermal or photochemical loss of carbon

monoxide. The [4]cumulene [4]tBu has also been subjected to oxidation with m-CPBA, and

this reaction also gives the cyclopropanone product.

Scheme 1.16 Oxidation products of [5]tBu.

Theoretical predictions regarding the reactivity of cumulenes have been recently

reported, particularly concerning oxygen sensitivity with respect to the carbon allotrope

carbyne.156,157 Using density functional theory calculations, Moseler and coworkers report that

reaction of O2 with the cumulene chain can cause cleavage, followed by repeated shortening

of the chain through additional oxidation and loss of CO2.158

A variety of metal complexes can be formed through the reaction of an electrophilic

metal with the π-rich skeleton of a [5]cumulene (see Section 1.1.3. and Figure 1.7).100 A

number of unusual cumulene complexes with unique structural properties are outlined below

in Figure 1.7. Stang, for example, has reported about rhodium and platinum complexes of [3]-

and [5]cumulenes.159–161 Similar to Stang, Werner and coworkers also have reported about

rhodium complexes including [5]cumulenes.162,163 Complexes of [5]Ph with rhodium prefer

bonding to the β-bond when triphenylphoshine is used as a ligand.159 The analogous system

with (i-Pr)3P ligands shows rhodium bonded to the γ-bond at low temperature (the kinetic

19

product), while complete conversion to thermodynamic product with Rh-complexation at the

β-bond is achieved upon warming162,163 (for description of β- and γ-bonds of cumulenes see

Scheme 1.17). Complexation of (Ph3P)2Pt to [5]Ph reveals similar behaviour, namely an

equilibrium between the kinetic complex at the γ-bond and the thermodynamic complex at the

β-bond.159 Another metal center for cumulene complexes, namely iron, has been used by

Nakamura79,80 and King81 as iron carbonyl compounds. The reaction of [5]tBu with Fe2(CO)9

or Fe3(CO)12 gives a mixture of the mono- and dinuclear iron complexes,81 while Iyoda and

coworkers have shown that under the appropriate conditions using non-sterically demanding

endgroups, the reaction of either [5]H or [5]tBu with Fe3(CO)12 can be forced all the way to

the tetranuclear iron complex.79,80 In contrast to the typically coordinated cumulenes, Suzuki

has presented a new class of cumulene complexes, where the cumulene coordinates to the

metal, i.e., zirconium to form a metallacycle including an endocyclic acetylene unit.164–167

Finally, Suzuki and coworkers have shown that [5]cumulenes can be trapped with the low-

valent zirconocene–bisphosphine complex [Cp2Zr(PMe3)2] to give the very strained

zirconacyclopent-3-yne products.165–167

20

Figure 1.7 Selected cumulene complexes with diverse coordination patterns.

To conclude, cumulene complexes can be obtained in a variety of unique compounds

with interesting coordination behavior. Furthermore, this metal complexation can provide a

higher stability of the cumulene framework compared to the bare cumulene structure.

Nakamura has shown, for example, that the usually unstable unsubstituted [3]cumulene [3]H

can be stabilized through formation of a stable iron carbonyl π-complex.168 Three possible

structures for this [3]cumulene complex have been proposed (Figure 1.8). Due to absence of

an X-ray structure, the correct structure cannot be assigned to the unsubstituted [3]cumulene

iron carbonyl complex.

21

Figure 1.8 Three possible structures for a [3]H iron carbonyl complex (cumulenic bonds that

are relevant for coordination are marked red).

1.4.2 Cycloaddition and oligomerization reactions

The most commonly investigated reactions for [5]cumulenes are cycloadditions,

including cumulene oligomerization and either the addition of alkenes or alkynes. To date, no

general synthetic pathway for reactions of cumulenes has been reported, however, Scheme

1.17 summarizes the basic reactivity of [5]cumulenes concerning cycloaddition reactions.

Regarding the addition of alkenes to cumulenes, on one hand, it has been reported that

tetrafluoroethene (TFE) attacks the central γ-bond of [5]tBu to give the symmetric

cyclobutane derivative 1.1.121,133 On the other hand, Bildstein and coworkers have shown that

cycloaddition reaction of [5]Fc with either tetracyanoethylene (TCNE) or C60 (at a 6,6-ring

junction) occurs at the β-bond, which affords the unsymmetrical cyclobutane derivatives

1.2.132 Noteworthy, all three known examples in Scheme 1.17 utilize electron deficient

alkenes, but nevertheless two different reactivity patterns are clearly operative, i.e., at the

β- or γ-bond. Bildstein suggests that addition to the β-bond is the thermodynamic reaction

pathway, while the alternative, addition to the γ-bond to give the symmetrical adduct, is

described as the kinetic reaction pathway,169 similar to the situation described for metal

complexes in Section 1.3.1.

22

Scheme 1.17 Examples of cycloaddition and dimerization reactions reported for

[5]cumulenes.

Hartzler, on the other hand suggests that the [2 + 2] addition probably occurs by way

of a thermally accessible diradical of the cumulene, and thus cycloaddition reactions at the

central double bond might be expected, especially if the terminal carbon atoms are sterically

hindered by substituents such as t-butyl (Figure 1.9). This is consistent with the experimental

results, which show that addition of the highly reactive reagent tetrafluoroethene occurs at the

γ-bond.

Figure 1.9 Diradical mesomeric structures as suggested by Hartzler for reactions of a

[5]cumulene.121

23

Two examples of alkyne addition to [5]cumulenes have been reported, and both

reactions use highly activated acetylenes to give the symmetrical [2 + 2] cyclobutene adduct

1.3 via reaction at the central γ-bond (Scheme 1.17).121,132 The nature of the cumulene varies

significantly in these two reactions, from R = t-Bu to R = Fc, but both authors suggest that the

addition to the γ-bond is observed because of steric hindrance resulting from endgroups.

The reaction of [5]cumulenes often results in a formal cycloaddition between two

cumulene molecules. Most commonly, such reactions proceed either thermally or via a Ni-

catalyzed reaction (Scheme 1.17). Thermal dimerization is often observed for cumulenes with

bulky alkyl substituents, giving a symmetrical vinylidene-substituted [4]radialene 1.4, i.e., a

cyclobutane ring that possesses four equally substituted exocyclic allene units (Scheme

1.17).170,171 This has been first demonstrated by Hartzler and coworkers for [5]tBu, i.e., when

[5]tBu melts, the resulting liquid resolidifies with the formation of the radialene product.121 In

a later report by Iyoda, it has been suggested that the dimerization reaction occurs at the

central, γ-double bond, due to the crisscross nature of the structure (transition state) that

would be required for a thermal [2 + 2] reaction.140 It is interesting to ponder why the thermal

reactions of tetraaryl[5]cumulenes do not seem to follow a similar pathway in the solid state,

i.e., dimerization to give a radialene. It might be due to molecular structure and stronger BLA

(see Chapter IV), or perhaps steric factors based on the endgroups. Alternatively, it is also

conceivable that favorable intermolecular stacking interactions of the aryl endgroups might

prevent a crisscross orientation that would be necessary for a thermal [2 + 2] reaction.

Aside from solid state reactions, radialenes 1.4 are also formed from alkyl-substituted

cumulenes in solution under Ni-catalysis.140 With slightly less bulky alkyl endgroups,

Ni-catalyzed head-to-tail dimerization at the β-bond results in the formation of [4]radialenes

1.5, while tetraaryl[5]cumulenes afford the deep blue head-to-head dimers 1.6 (Scheme 1.17).

The authors suggest that the bulkiness of the terminal substituents in the [5]cumulenes

controls the course of metal coordination, and leads to the selectivity observed in the

oligomerization reactions.140

The final mode of dimerization reaction for [5]cumulenes has been documented

independently by the works of Stang172 and Scott,134 and later by Hopf and coworkers with

the unsubstituted [5]H.173 In these three cases, the lack of sterically encumbered endgroups

permits reaction at the α-bonds, with concomitant rearrangement of the cumulene framework

to give a butadiyne moiety (Figure 1.10). Scott has reported that the Cu(I)-catalyzed

cyclodimerization of [5]Me leads to the symmetrical 12-membered ring 1.7.134 Stang and

coworkers have shown that no metal catalysis is necessary, and macrocycle 1.8 is isolated in

24

31% yield as the only viable product.172 In the case of 1.8, a radical is suggested, which

avoids the necessity of a symmetry forbidden [6 + 6] thermal cycloaddition. Finally, Hopf and

coworkers report the formation of macrocycle 1.9,173 conceivably through dimerization of the

parent system, [5]H, although the intermediate presence of [5]H has not been established.

Figure 1.10 Dimerization products of [5]cumulenes.

Similar to dimerization, [5]cumulenes also show to formally undergo trimerization

reactions, although in each of the two reported cases, dimerization precedes formation of the

trimer. The first example from Kawase et al. is outlined in Scheme 1.18 and forms a novel

tricyclobutabenzene derivative from a precursor in which a [5]cumulene links two quinone

rings.174 The authors suggest a mechanism in which the [5]cumulene is converted to a radical

intermediate via oxidation. This intermediate then dimerizes to a cyclobutene intermediate,

and subsequent addition of the third [5]cumulene unit gives a dicyclobutene intermediate.

Finally, further oxidation and a final cyclization step give the tricyclobutabenzene product.

25

Scheme 1.18 Trimerization reaction of a [5]cumulene.

The second trimerization example describes the cyclotrimerization of [5]Ph to a

tricyclodecadiene derivative as reported by Kawamura and coworkers (Scheme 1.19).175 The

authors suggest that the reaction initiates with a solution-state dimerization of [5]Ph to give

the symmetrical [4]radialene. A third equivalent of [5]Ph adds to this intermediate and gives

the Diels-Alder adduct, which is ultimately converted to the product via an electrocyclization

reaction. The reaction yield is quite reasonable (up to ca. 70%), as long as the concentration

of [5]Ph is >13 mmol·L–1. The product is rather stable, and subsequent Diels-Alder or

electrocyclic reactions are not observed. This overall reaction is particularly unusual since it

relies on the thermal dimerization of a tetraaryl[5]cumulene at the γ-bond, which is a

reactivity pattern typically reserved for [5]cumulenes terminated with sterically bulky alkyl

groups.

26

Scheme 1.19 Trimerization reaction of [5]Ph.

1.5 Motivation and goals of the doctoral thesis

As described in Section 1.1, the structure of carbyne is on the one hand fundamental to

its properties, and on the other hand not well understood. Thus, investigation of carbyne might

be best approached through the rational synthesis and study of molecules with defined

structure. There is a growing fundamental interest in long [n]cumulenes as model compounds

for the sp-carbon allotrope carbyne, and insight from the study of [n]cumulenes of defined

length should allow researchers to compare and contrast properties of the cumulenic and

polyyne versions of carbyne. The unique π-electron structure of cumulenes provides

distinctive electronic and optical properties that suggest fascinating opportunities in molecular

electronics and materials science. Key to exploiting this potential is the development of more

stable cumulene structures, and the synthetic methods to realize these targets. Regarding the

reactivity of cumulenes, a variety of possible synthetic transformations might exploit the

reactive cumulenic π-system, particularly in the field of cycloaddition reactions giving unique

compounds as candidates for precursors to unprecedented interesting and potentially useful

conjugated structures.

This thesis deals with the design, synthesis, characterization, and reactivity of

cumulenes as model compounds for carbyne. The instability of [n]cumulenes has no doubt

slowed progress on their synthesis and study. Therefore, to enhance the stability of longer

cumulenes, two bulkier endgroups, 1) the 3,5-di-t-butylphenyl group, a modified version of

27

the so called “supertrityl” group, and 2) the mesityl group have been incorporated into the

cumulene system to achieve better shielding of the cumulene chain. These two series of

cumulenes, and consequently their synthesis, characterization, and the resulting properties

will be discussed in Chapter II and IV, respectively. Another alternative to enhance the

stability of longer cumulenes is given in Chapter III and describes the formation of cumulene

rotaxanes that include a phenanthrene-based macrocycle designed to shield the cumulene

framework. Finally, the reactivity of several cumulenes is presented in Chapter V, including

addition reactions and dimerization reactions as well as the appropriate characterization and

discussion of the obtained results.

1.6 References

1 F. Diederich, Y. Rubin, Angew. Chem. Int. Ed. 1992, 31, 1101–1123; Angew. Chem.

1992, 104, 1123–1146.

2 A. R. Murphy, J. M. J. Frechet, Chem. Rev. 2007, 107, 1066–1096.

3 J. E. Anthony, Chem. Rev. 2006, 106, 5028–5048.

4 F. Diederich, P. J. Stang, R. R. Tykwinski, Acetylene Chemistry: Chemistry, Biology,

and Material Science, Wiley-VCH, Weinheim, 2005.

5 M. M. Haley, R. R. Tykwinski, Carbon-Rich Compounds: From Molecules to

Materials, Wiley-VCH, Weinheim, 2006.

6 E. H. L. Falcao, F. Wudl, J. Chem. Technol. Biotechnol. 2007, 82, 524–531.

7 P. P. K. Smith, P. R. Buseck, Science 1982, 216, 984–986.

8 For an entertaining opinion on the subject of carbyne, see the comments by Sir Harold

Kroto:

http://www.rsc.org/chemistryworld/Issues/2010/November/CarbyneOtherMythsAboutCa

rbon.asp

9 H. O. Pierson, Handbook of Carbon, Graphite, Diamond and Fullerenes – Properties,

Processing and Applications, William Andrew Publishing/Noyes, New Jersey, 1993.

10 S. Kim, Angew. Chem. Int. Ed. 2009, 48, 7740–7743; Angew. Chem. 2009, 121, 7876–

7879.

11 F. Cataldo, Polym. Int. 1997, 44, 191–200.

28

12 Y. P. Kudryavtsev, S. Evsyukoc, M. Gusevca, C. Babaev, C. Khvistov, in Chemistry

and Physics of Carbon, ed. P. A. Thrower, Marcel Dekker, New York, 1997, vol. 25,

1–99.

13 R. J. Lagow, J. J. Kampa, H.-C. Wei, S. L. Battle, J. W. Genge, D. A. Laude, C. J.

Harper, R. Bau, R. C. Stevens, J. F. Haw, E. Munson, Science 1995, 267, 362–367.

14 Y. Tobe, T. Wakabayashi, in Acetylene Chemistry: Chemistry, Biology, and Material

Science (Eds.: F. Diederich, P. J. Stang, R. R. Tykwinski), Wiley-VCH, Weinheim,

2005, chapter 9.

15 F. Cataldo, Polyynes: Synthesis, Properties, and Applications, Taylor & Francis, Boca

Raton, FL, 2006.

16 A. G. Whittaker, E. J. Watts, R. S. Lewis, E. Anders, Science 1980, 209, 1512–1514.

17 R. Hayatsu, R. G. Scott, M. H. Studier, R. S. Lewis, E. Anders, Science 1980, 209,

1515–1518.

18 A. Webster, Mon. Not. R. Astron. Soc. 1980, 192, 7–9.

19 A. El Goresy, G. Donnay, Science 1968, 161, 363–364.

20 A. L. K. Shi Shun, R. R. Tykwinski, Angew. Chem. Int. Ed. 2006, 45, 1034–1057;

Angew. Chem. 2006, 118, 1050–1073.

21 Y. Pan, T. L. Lowary, R. R. Tykwinski, Can. J. Chem. 2009, 87, 1565–1582.

22 B. W. Gung, C. R. Chim. 2009, 12, 489–505.

23 M. J. Wesolowski, S. Kuzmin, B. Wales, R. Karimi, A. A. Zaidi, Z. Leonenko, J. H.

Sanderson, W. W. Duley, Carbon 2011, 49, 625–630.

24 Y. Sato, T. Kodama, H. Shiromaru, J. H. Sanderson, T. Fujino, Y. Wada, T.

Wakabayashi, Y. Achiba, Carbon 2010, 48, 1673–1676.

25 A. Hu, M. Rybachuk, Q.-B. Lu, W. W. Duley, Appl. Phys. Lett. 2007, 91, 131906.

26 L. Ravagnan, P. Piseri, M. Bruzzi, S. Miglio, G. Bongiorno, A. Baserga, C. S. Casari,

A. Li Bassi, C. Lenardi, Y. Yamaguchi, T. Wakabayashi, C. E. Bottani, P. Milani,

Phys. Rev. Lett. 2007, 98, 216103.

27 R. B. Heimann, S. E. Evsyukov, L. Kavan, Carbyne and Carbynoid Structures,

Kluwar Academic Press, London, 1999.

28 C. H. Hendon, D. Tiana, A. T. Murray, D. R. Carbery, A. Walsh, Chem. Sci. 2013, 4,

4278–4284.

29 M. Liu, V. I. Artyukhov, H. Lee, F. Xu, B. I. Yakobson, ACS Nano 2013, 7, 10075–

10082.

29

30 S. Eisler, A. D. Slepkov, E. Elliott, T. Luu, R. McDonald, F. A. Hegmann, R. R.

Tykwinski, J. Am. Chem. Soc. 2005, 127, 2666–2676.

31 A. Lucotti, M. Tommasini, D. Fazzi, M. Del Zoppo, W. A. Chalifoux, R. R.

Tykwinski, G. Zerbi, J. Raman Spectrosc. 2012, 43, 1293–1298.

32 I. Kminek, J. Klimovic, P. N. Prasad, Chem. Mater. 1993, 5, 357–360.

33 S. Ermer, S. Lovejoy, D. Leung, J. Altman, K. Aron, R. Spitzer, G. Hansen, Proc.

SPIE Int. Soc. Opt. Eng. 1990, 1337, 89–98.

34 S. Ballmann, W. Hieringer, D. Secker, Q. L. Zheng, J. A. Gladysz, A. Görling, H. B.

Weber, ChemPhysChem 2010, 11, 2256–2260.

35 P. Moreno-Garcia, M. Gulcur, D. Zsolt Manrique, T. Pope, W. Hong, V. Kaliginedi,

C. Huang, A. S. Batsanov, M. R. Bryce, C. Lambert, T. Wandlowski, J. Am. Chem.

Soc. 2013, 135, 12228–12240.

36 N. D. Lang, Ph. Avouris, Phys. Rev. Lett. 1998, 81, 3515–3518.

37 Y. Zhang, Y. Su, L. Wang, E. S.-W. Kong, X. Chen, Y. Zhang, Nanoscale Res. Lett.

2011, 6, 577.

38 I. E. Castelli, P. Salvestrini, N. Manini, Phys. Rev. B 2012, 85, 214110.

39 A. Milani, A. Lucotti, V. Russo, M. Tommasini, F. Cataldo, A. Li Bassi, C. S. Casari,

J. Phys. Chem. C 2011, 115, 12836–12843.

40 M. Weimer, W. Hieringer, F. Della Sala, A. Görling, Chem. Phys. 2005, 309, 77–87.

41 W. A. Chalifoux, R. R. Tykwinski, Nature Chem. 2010, 2, 967–971.

42 F. Bohlmann, Angew. Chem. 1953, 65, 385–389.

43 T. R. Johnson, D. R. M. Walton, Tetrahedron 1972, 28, 5221–5236.

44 R. Eastmond, T. R. Johnson, D. R. M. Walton, Tetrahedron 1972, 28, 4601–4616.

45 E. R. H. Jones, Proc. Chem. Soc. 1960, 199–210.

46 Q. Zheng, J. C. Bohling, T. B. Peters, A. C. Frisch, F. Hampel, J. A. Gladysz, Chem.

Eur. J. 2006, 12, 6486–6505.

47 M. I. Bruce, B. K. Nicholson, N. N. Zaitseva, C. R. Chim. 2009, 12, 1280–1286.

48 T. Gibtner, F. Hampel, J.-P. Gisselbrecht, A. Hirsch, Chem. Eur. J. 2002, 8, 408–432.

49 J. A. Januszewski, D. Wendinger, C. D. Methfessel, F. Hampel, R. R. Tykwinski,

Angew. Chem. Int. Ed. 2013, 52, 1817–1821; Angew. Chem. 2013, 125, 1862–1867.

50 F. Bohlmann, K. Kieslich, Chem. Ber. 1954, 87, 1363–1372.

51 F. Bohlmann, K. Kieslich, Abh. Braunschweig. Wiss. Ges. 1957, 9, 147–166.

52 W. Ried, W. Schlegelmilch, S. Piesch, Chem. Ber. 1963, 96, 1221–1228.

30

53 M. M. Yildizhan, D. Fazzi, A. Milani, L. Brambilla, M. Del Zoppo, W. A. Chalifoux,

R. R. Tykwinski, G. Zerbi, J. Chem. Phys. 2011, 134, 124512.

54 S. Yang, M. Kertesz, J. Phys. Chem. A 2008, 112, 146–151.

55 M. Kertesz, J. Koller, A. Ažman, J. Chem. Phys. 1978, 68, 2779–2782.

56 A. von Bayer, Ber. Dtsch. Chem. Gesell. 1885, 18, 2269–2281.

57 G. Dupont, Ann. Chim. Phys. 1913, 30, 485–587.

58 E. R. H. Jones, M. C. Whiting, J. B. Armitage, C. L. Cook, N. Entwistle, Nature

(London) 1951, 168, 900–903.

59 R. Eastmond, T. R. Johnson, D. R. M. Walton, J. Organomet. Chem. 1973, 50, 87–92.

60 F. Bohlmann, Chem. Ber. 1951, 84, 785–794.

61 R. Kuhn, K. Wallenfels, Chem. Ber. 1938, 71, 783–790.

62 B. S. Burton, H. V. Pechmann, Chem. Ber. 1887, 20, 145–149.

63 E. R. H. Jones, G. H. Mansfield, M. C. Whiting, J. Chem. Soc. 1954, 3208–3212.

64 K. Brand, Chem. Ber. 1921, 54, 1987–2006.


66 R. Kuhn, H. Zahn, Chem. Ber. 1951, 84, 566–570.

67 P. Cadiot, Ann. Chim. [Paris] 1956, 13, 214–272.

68 R. Kuhn, H. Krauch, Chem. Ber. 1955, 88, 309–315.

69 J. Rauss-Godineau, W. Chodkiewicz, P. Cadiot, Bull. Soc. Chim. France 1966, 9,

2877–2884.

70 A. Hoffmann-Röder, N. Krause, Angew. Chem. Int. Ed. 2004, 43, 1196–1216; Angew.

Chem. 2004, 116, 1216–1236.

71 H. Staudinger, L. Ruzicka, Helv. Chim. Acta 1924, 7, 177–201.

72 W. D. Celmer, I. A. Solomons, J. Am. Chem. Soc. 1952, 74, 1870–1871.

73 J. D. Bu’Lock, E. R. H. Jones, P. R. Leeming, J. Chem. Soc. 1955, 4270–4276; 1957,

1097–1101.

74 J. D. Bu’Lock, E. R. H. Jones, P. R. Leeming, J. M. Thompson, J. Chem. Soc. 1956,

3767–3771.

75 E. R. H. Jones, P. R. Leeming, W. A. Remers, J. Chem. Soc. 1960, 2257–2263.

76 F. Bohlmann, H. Bornowski, C. Arndt, Chem. Ber. 1965, 98, 2236–2242.

77 F. Bohlmann, H. Burkhardt, C. Zdero, Naturally Occurring Acetylenes, Academic

Press, New York, 1973.

78 F. Bohlmann, H. Bornowski, C. Arndt, Fortschr. Chem. Forsch. 1962, 4, 138–272.

31

79 A. Nakamura, Bull. Chem. Soc. Jpn. 1965, 58, 1868–1873.

80 M. Iyoda, Y. Kuwatani, M. Oda, K. Tatsumi, A. Nakamura, Angew. Chem. Int. Ed.

Engl. 1991, 30, 1670–1672; Angew. Chem. 1991, 103, 1697–1699.

81 R. B. King, C. A. Harmon, J. Organomet. Chem. 1975, 88, 93–100.

82 S. Szafert, P. Haquette, S. B. Falloon, J. A. Gladysz, J. Organomet. Chem. 2000, 604,

52–58.

83 H. Lang, Angew. Chem. Int. Ed. Engl. 1994, 33, 547–550; Angew. Chem. 1994, 106,

569–572.

84 J. P. Selegue, Coord. Chem. Rev. 2004, 248, 1543–1563.

85 M. I. Bruce, Chem. Rev. 1991, 91, 197–257.

86 A. B. Antonova, A. A. Ioganson, Russ. Chem. Rev. (Engl. Transl.) 1989, 58, 693–710.

87 M. I. Bruce, Chem. Rev. 1998, 98, 2797–2858.

88 V. Cadierno, J. Gimeno, Chem. Rev. 2009, 109, 3512–3560.

89 D. Peron, A. Romero, P. H. Dixneuf, Organometallics 1995, 14, 3319–3326.

90 R. W. Lass, P. Steinert, J. Wolf, H. Werner, Chem Eur. J. 1996, 2, 19–23.

91 G. Roth, H. Fischer, Organometallics 1996, 15, 1139–1145.

92 K. Ilg, H. Werner, Chem. Eur. J. 2002, 8, 2812–2820.

93 G. Roth, H. Fischer, Organometallics 1996, 15, 5766–5768.

94 B. Bildstein, Coord. Chem. Rev. 2000, 206–207, 369−394.

95 H. Fischer, in The chemistry of alkenes (Ed.: S. Patai), John Wiley & Sons, New York,

1964, pp. 1025–1159.

96 W. Hunsmann, Chem. Ber. 1950, 83, 213–217.

97 H. H. Schlubach, V. Franzen, Justus Liebigs Ann. Chem. 1951, 573, 105–109.

98 H. H. Schlubach, V. Wolf, Justus Liebigs Ann. Chem. 1950, 568, 141–159.

99 C. L. Cook, E. R. H. Jones, M. C. Whiting, J. Chem. Soc. 1952, 2883–2891.

100 M. Ogasawara, in Science of synthesis. Cumulenes and allenes (Ed.: D. Bellus), Georg

Thieme Verlag, 2008, vol. 44.1, pp. 9–70.

101 P. Aguirre-Etcheverry, D. O’Hare, Chem. Rev. 2010, 110, 4839–4864.

102 C. M. Che, C. M. Ho, J. S. Huang, Coord. Chem. Rev. 2007, 251, 2145–2166.

103 C. Coletti, A. Marrone, N. Re, Acc. Chem. Res. 2012, 45, 139–149.

104 D. Touchard, P. H. Dixneuf, Coord. Chem. Rev. 1998, 178, 409–429.

105 L. Leroyer, V. Maraval, R. Chauvin, Chem. Rev. 2012, 112, 1310–1343.

32

106 P. Cadiot, W. Chodkiewicz, J. Rauss-Godineau, Bull. Soc. Chim. Fr. 1961, 2176–

2193.

107 C. Bruneau and J.-L. Renaud, in Comprehensive Organic Functional Group

Transformations II (Eds.: A. R. Katritzky, R. J. K. Taylor), Elsevier, Oxford, 2005,

vol. 1.20, pp. 1019–1081.

108 P. J. Stang, Chem. Rev. 1978, 78, 383–405.

109 J.-D. van Loon, P. Seiler, F. Diederich, Angew. Chem. Int. Ed. Engl. 1993, 32, 1187–

1189; Angew. Chem. 1993, 105, 1235–1238.

110 A. Auffrant, F. Diederich, C. Boudon, J.-P. Gisselbrecht, M. Gross, Helv. Chim. Acta

2004, 87, 3085–3105.

111 T. Kunieda, T. Takizawa, Chem. Pharm. Bull. 1977, 25, 1809–1810.

112 M. Iyoda, H. Otani, M. Oda, J. Am. Chem. Soc. 1986, 108, 5371–5372.

113 M. Iyoda, N. Nakamura, M. Todaka, S. Ohtsu, K. Hara, Y. Kuwatani, M. Yoshida, H.

Matsuyama, M. Sugita, H. Tachibana, H. Inoue, Tetrahedron Lett. 2000, 41, 7059–

7064.

114 G. Köbrich, H. Heinemann, W. Zündorf, Tetrahedron 1967, 23, 565–584.

115 K. Komatsu, H. Kamo, R. Tsuji, K. Takeuchi, J. Org. Chem. 1993, 58, 3219–3221.

116 H. Kurata, S. Muro, T. Enomoto, T. Kawase M. Oda, Bull. Chem. Soc. Jpn. 2007, 80,

349–357.

117 P. J. Stang, T. E. Fisk, J. Am. Chem. Soc. 1980, 102, 6813–6816.

118 P. J. Stang, Acc. Chem. Res. 1982, 15, 348–354.

119 K. Brand, D. Krücke-Amelung, Chem. Ber. 1939, 72, 1036–1047.

120 K. Brand, A. Busse-Sundermann, Chem. Ber. 1950, 83, 119–128.

121 H. D. Hartzler, J. Am. Chem. Soc. 1971, 93, 4527–4531.

122 E. Bergmann, H. Hoffmann, D. Winter, Chem. Ber. 1933, 66, 46–54.

123 M. Iyoda, K. Nishioka, M. Nose, S. Tanaka, M. Oda, Chem. Lett. 1984, 131–134.

124 J. Salkind, A. Kruglow, Chem. Ber. 1928, 61, 2306–2312.

125 B. Bildstein, M. Schweiger, H. Kopacka, K.-H. Ongania, K. Wurst, Organometallics

1998, 17, 2414–2424.

126 M. Nakagawa, K. Shingu, K. Naemura, Tetrahedron Lett. 1961, 22, 802–806.

127 R. Kuhn, H. Fischer, H. Fischer, Chem. Ber. 1964, 97, 1760–1766.

128 G. Karich, J. C. Jochims, Chem. Ber. 1977, 110, 2680–2694.

33

129 H. Irngartinger, W. Götzmann, Angew. Chem. Int. Ed. Engl. 1986, 25, 340–342;

Angew. Chem. 1986, 98, 359–361.

130 W. J. le Noble, S. Basak, S. Srivastava, J. Am. Chem. Soc. 1981, 103, 4638–4639.

131 S. Basak, S. Srivastava, W. J. le Noble, J. Org. Chem. 1987, 52, 5095–5099.

132 B. Bildstein, M. Schweiger, H. Angleitner, H. Kopacka, K. Wurst, K.-H. Ongania, M.

Fontani, P. Zanello, Organometallics 1999, 18, 4286–4295.


134 L. T. Scott, G. J. DeCicco, J. Org. Chem. 1980, 45, 4055–4056.

135 P. Cadiot, A. Willemart, Bull. Soc. Chim. France 1951, 100–104.

136 H. Kollmar, H. Fischer, Tetrahedron Lett. 1968, 40, 4291–4294.

137 P. J. Stang, M. Ladika, J. Am. Chem. Soc. 1980, 102, 5407–5409.

138 P. J. Stang, M. Ladika, J. Am. Chem. Soc. 1981, 103, 6437–6443.

139 L. Skattebol, Tetrahedron Lett. 1965, 26, 2175–2179.

140 Y. Kuwatani, G. Yamamoto, M. Oda, M. Iyoda, Bull. Chem. Soc. Jpn. 2005, 78,

2188–2208.

141 B. Bildstein, W. Skibar, M. Schweiger, H. Kopacka, K. Wurst, J. Organomet. Chem.

2001, 622, 135–142.

142 H. Fischer, W. D. Hell, Angew. Chem. Int. Ed. Engl. 1967, 6, 954–955; Angew. Chem.

1967, 79, 931–932.

143 S. Eisler, R. R. Tykwinski, in Acetylene Chemistry: Chemistry, Biology, and Material


2005, chapter 7.

144 W. A. Chalifoux, R. R. Tykwinski, C. R. Chimie 2009, 12, 341–358.

145 W. A. Chalifoux, R. R. Tykwinski, Chem. Rec. 2006, 6, 169–182.

146 E. W. Colvin, B. J. Hamill, J. Chem. Soc., Perkins Trans. 1 1977, 865–869.

147 J. Kendall, R. McDonald, M. J. Ferguson, R. R. Tykwinski, Org. Lett. 2008, 10, 2163–

2166.

148 P. Bichler, W. A. Chalifoux, S. Eisler, A. L. K. Shi Shun, E. T. Chernick, R. R.

Tykwinski, Org. Lett. 2009, 11, 519–522.

149 [9]Mes was synthesized by Dominik Wendinger, while [9]tBuPh was synthesized as

part of this thesis.

150 P. Rivera-Fuentes, F. Diederich, Angew. Chem. Int. Ed. 2012, 51, 2818–2828; Angew.

Chem. 2012, 124, 2872–2882.

34

151 S. Yu, S. Ma, Angew. Chem. Int. Ed. 2012, 51, 3074–3112; Angew. Chem. 2012, 124,

3128–3167.

152 L. T. Scott, G. J. DeCicco, Tetrahedron Lett. 1976, 31, 2663–2666.

153 R. Kuhn, H. Fischer, Chem. Ber. 1961, 94, 3060–3071.

154 R. Kuhn, H. Fischer, Chem. Ber. 1960, 93, 2285–2289.

155 J. K. Crandall, D. M. Coppert, T. Schuster, F. Lin, J. Am. Chem. Soc. 1992, 114,

5998–6002.

156 F. Cataldo, Polym. Degrad. Stabil. 2006, 91, 317–323.

157 C. S. Casari, A. Li Bassi, L. Ravagnan, F. Siviero, C. Lenardi, P. Piseri, G. Bongiorno,

C. E. Bottani, P. Milani, Phys. Rev. B 2004, 69, 075422.

158 G. Moras, L. Pastewka, M. Walter, J. Schnagl, P. Gumbsch, M. Moseler, J. Phys.

Chem. C 2011, 115, 24653–24661.

159 L. Song, A. M. Arif, P. J. Stang, J. Organomet. Chem. 1990, 395, 219–226.

160 P. J. Stang, M. R. White, Gerhard Maas, Organometallics 1983, 2, 720–725.

161 M. R. White, P. J. Stang Organometallics 1983, 2, 1654–1658.

162 H. Werner, R. Wiedemann, N. Mahr, P. Steinert, J. Wolf, Chem. Eur. J. 1996, 2, 561–

569.

163 I. Kovacik, M. Laubender, H. Werner, Organometallics 1997, 16, 5607–5609.

164 This topic has recently been reviewed, see: N. Suzuki, D. Hashizume, Coord. Chem.

Rev. 2010, 254, 1307–1326.

165 N. Suzuki, D. Hashizume, H. Yoshida, M. Tezuka, K. Ida, S. Nagashima, T. Chihara,

J. Am. Chem. Soc. 2009, 131, 2050–2051.

166 N. Suzuki, N. Ohara, K. Nishimura, Y. Sakaguchi, S. Nanbu, S. Fukui, H. Nagao, Y.

Masuyama, Organometallics 2011, 30, 3544–3548.

167 N. Suzuki, D. Hashizume, H. Koshino, T. Chihara, Angew. Chem. Int. Ed. 2008, 47,

5198–5202; Angew. Chem. 2008, 120, 5276–5280.

168 A. Nakamura, P.-J. Kim, N. Hagihara, J. Organomet. Chem. 1965, 3, 7–15.

169 B. Bildstein, O. Loza, Y. Chizhov, Organometallics 2004, 23, 1825–1835.

170 H. Hopf, G. Maas, Angew. Chem. Int. Ed. Engl. 1992, 31, 931–954; Angew. Chem.

1992, 104, 953–977.

171 M. Gholami, R. R. Tykwinski, Chem. Rev. 2006, 106, 4997–5027.

172 M. Kaftory, I. Agmon, M. Ladika, P. J. Stang, J. Am. Chem. Soc. 1987, 109, 782–787.

173 C. Werner, H. Hopf, I. Dix, P. Bubenitschek, P. G. Jones, Chem. Eur. J. 2007, 13,

35

9462–9477.

174 T. Kawase, Y. Minami, N. Nishigaki, S. Okano, H. Kurata, M. Oda, Angew. Chem.

Int. Ed. 2005, 44, 316–319; Angew. Chem. 2005, 117, 320–323.

175 N. Islam, T. Ooi, T. Iwasawa, M. Nishiuchi, Y. Kawamura, Chem. Commun. 2009,

574–576.

36

2. Chapter II. Cumulenes – Synthesis of tetraarylcumulenes [n]tBuPh and

[n]Mes‡

2.1 Synthesis and structure of tetraarylcumulenes [n]tBuPh and [n]Mes

This chapter describes the synthesis of two series of tetraarylcumulenes containing

either 3,5-di-t-butylphenyl or mesityl endgroups, [n]tBuPh and [n]Mes, respectively (Figure

2.1). The main emphasis of this chapter is put on the series of [n]tBuPh cumulenes, whereas

only a short summary and several synthetic features are presented for the [n]Mes cumulene

series.§ The synthetic steps outline the formation of [n]tBuPh and [n]Mes with n = 3, 5, 7,

and 9. Furthermore, synthetic improvements and approaches to higher cumulenes are

presented. Regarding the [n]Mes cumulene series, discussions of synthetic challenges and

thus, deviations from standard synthetic methods, are given. Finally, stability issues of

cumulenes in solution and the solid state are discussed.

2.1.1 General aspects and motivation

As already mentioned in Chapter I, compared to polyynes, the study of cumulenes has

lain essentially dormant since early work1,2 reported by Kuhn3,4 and Bohlmann.5,6 Thus, many

unanswered questions remain about the physical properties of this intriguing class of linear

molecules. To date, UV/vis spectroscopy has been the most useful method for the

characterization of cumulenes,7 and reported analyses of cumulenes document a lowering of

the lowest energy absorption (λmax) as a function of cumulene length, such as that for known

cumulenes [n]Ph and [n]Cy (n = 3, 5, 7, 9, Figure 2.1).4–6 Changes in λmax versus chain length

n are, obviously, intricately dependent on the structure and on the degree of bond length

alternation (BLA, defined as the bond length difference between the two central-most double

‡ Portions of this chapter have been published: J. A. Januszewski, D. Wendinger, C. D. Methfessel, F. Hampel,

R. R. Tykwinski, Angew. Chem. Int. Ed. 2013, 52, 1817–1821. § The chemistry of [n]Mes was introduced during my doctoral thesis and continued by doctoral student Dominik

Wendinger who synthesized the series of [n]Mes. In addition, Christian D. Methfessel performed the synthesis

of [5]Mes (bachelor thesis under my supervision).

37

bonds of the cumulene chain). Recent theoretical studies predict that the BLA for cumulenes

will rapidly approach zero (BLA ≤ 0.01) with increasing length of the cumulene chain,8–10 i.e.,

Peierls distortion is essentially absent.11 Experimentally, X-ray crystallographic analysis

would provide an opportunity to confirm or refute theoretical trends in BLA as a function of

cumulene length. Unfortunately, very few solid-state structures have been reported for

cumulenes, and no crystallographic data for [n]cumulenes with n > 5 are available.12

Figure 2.1 Structures of selected [n]cumulenes.

To better understand the properties of cumulenes, a series of stabilized derivatives has

to be formed that enables the study of physical properties even for longer [n]cumulenes

(n > 5). Therefore, a bulky endgroup, i.e., an analog of the “supertrityl” endgroup has been

applied for the synthesis of cumulenes, since it has been predicted to shield the reactive

cumulene chain as in the case of longer polyyne chains.13 For cumulenes, however, the

supertrityl endgroup is “modified” possessing only two aryl groups instead of usually three

because of the hybridization of the outer carbon atoms in a cumulene chain (Figure 2.2). More

specifically, the outer carbon atoms are sp2-hybridized in the cumulene chain, while the

analogous carbons are sp3-hybridized in the polyyne chain.

38

Figure 2.2 Comparison of hybridization of the outermost carbon atom in a polyyne (left) and

cumulene (right) chain.

The formation of a cumulene typically derives from an oligoyne that possesses a

leaving group in the terminal propargylic position, such as OH or OMe. The final synthetic

step then involves reductive elimination, usually performed with SnCl2 under acidic

conditions. This approach has been followed for the current syntheses.

2.1.2 Synthesis of the [n]tBuPh cumulene series

2.1.2.1 Synthesis of the bis-(3,5-di-t-butylphenyl)methanone endgroup

The synthesis of [3]-,[5]-,[7]-, and [9]tBuPh cumulenes has been adapted from the

known general synthetic approaches that have already been discussed in Chapter I and are

here presented in more detail. The endgroup for this series is based on acetylide addition to a

ketone unit, and thus, the ketone has been the first target.

3,5-Di-t-butylphenylbromide was converted to a Grignard reagent and further treated

with ethyl formate to afford secondary alcohol 2.1 in 89% yield (route a, Scheme 2.1).

Alcohol 2.1 was then oxidized to ketone 2.2 with pyridinium chlorochromate (PCC) in 99%

yield. Route a, however, was not reproducible due to problems concerning the formation of

the Grignard reagent. Alternatively, route b was attempted involving lithiation of the 3,5-di-t-

butylphenylbromide with n-BuLi followed by addition to ethyl formate (Scheme 2.1). This

reaction provided a lower yield of 2.1 (57% versus 89%), but it was much more reproducible.

In addition, ketone 2.2 was formed simultaneously in 14% yield by route b, but the products

could be easily separated via column chromatography.

39

Scheme 2.1 Synthesis of ketone 2.2.

2.1.2.2 Synthesis of [3]cumulene [3]tBuPh

The synthesis of [3]cumulene [3]tBuPh started with the formation of trimethylsilyl-

protected alkyne 2.3 through the reaction of a Li-acetylide with ketone 2.2 followed by

treatment with MeI (Scheme 2.2). It was necessary to form a methyl ether group instead of the

free hydroxy group to allow for further synthetic steps. Compound 2.3 was desilylated with

K2CO3 in MeOH/THF (20:1) to afford terminal acetylene 2.4 (Scheme 2.2).

Scheme 2.2 Synthesis of terminal acetylene 2.4.

With the terminal acetylene 2.4 in hand, the precursor 2.5 was synthesized by

lithiation of 2.4 and addition of the resulting acetylide to ketone 2.2 (Scheme 2.3). Reductive

elimination of 2.5 using anhydrous SnCl2 and HCl in Et2O under an inert atmosphere gave

pure [3]tBuPh as bright yellow solid in 84% yield. In the case of the stable lower

40

[n]cumulenes (n = 3, 5), it was not necessary to use anhydrous and oxygen-free conditions,

but the yield and purity of the products were usually increased by this practice.

Scheme 2.3 Synthesis of precursor 2.5 and reductive elimination to [3]tBuPh.


The synthetic route to [5]cumulene [5]tBuPh was similar to that used for [3]tBuPh

and started with the formation of acetylene 2.6, which was obtained by addition of lithiated

trimethylsilylacetylene to ketone 2.2 (Scheme 2.4). Desilylation of 2.6 gave terminal

acetylene 2.7, and subsequent oxidative homocoupling under Hay conditions

(CuCl and TMEDA) afforded the precursor 2.8 in 98% yield. Diol 2.8 was reduced to the red

[5]cumulene [5]tBuPh in 74% yield using the same conditions (SnCl2 and HCl) as given

above for [3]tBuPh.

41

Scheme 2.4 Synthesis of [5]tBuPh.


The preliminary synthetic steps to achieve the precursor to [7]tBuPh differed

compared to that of cumulenes [3]tBuPh and [5]tBuPh. Several approaches were attempted

and have been summarized below. The first three pathways (Schemes 2.5–2.7) were adapted

from earlier studies,14 but were also examined in greater detail during my doctoral research

period.

The lithiation route in Scheme 2.5 was based on the

triisopropylsilyltrimethylsilyltriyne 2.9 which was selectively lithiated via exchange at the

trimethylsilyl group. The resulting Li-acetylide was added to ketone 2.2 affording triyne 2.10

or 2.11, depending on the absence or presence of a final MeI addition, respectively. Triyne

2.10 could only be obtained from the reaction as an impure mixture, and efforts to purify 2.10

failed. In contrast, triyne 2.11 was isolated pure, but the following desilylation reactions with

CsF failed.

42

Scheme 2.5 Synthetic approaches to precursors to [7]tBuPh, compounds 2.10 and 2.11

(lithiation route).14

The “mixed” homocoupling route in Scheme 2.6 was attempted via the reaction of

monoyne 2.7 and diyne 2.12 under Hay conditions (CuCl and TMEDA). Herein, all three

possible products were formed, i.e., compounds 2.8, 2.13, as well as 2.14, the precursors to

[5]tBuPh, [7]tBuPh, and [9]tBuPh, respectively. Preliminary analysis of the reaction showed

that the desired product 2.13 was produced in much lower yield compared to the

homocoupling products 2.8 and 2.14. Unfortunately, separation of the products via column

chromatography or recrystallization failed. Even the use of different functional groups, i.e.,

OH group in alkyne 2.7 and OMe group in diyne 2.12 did not afford a better separation for the

products.

43

Scheme 2.6 Synthetic approach to the precursor (2.13) to [7]tBuPh (“mixed” homocoupling

route).14

The synthesis described in Scheme 2.7 employed the conventional Fritsch-Buttenberg-

Wiechell (FBW) rearrangement reaction of a dibromoolefin. Initially, secondary alcohol 2.15

was obtained via a Grignard exchange reaction between 2.4 and EtMgBr, following by

addition of ethynyl Grignard to ethyl formate (Scheme 2.7). Oxidation of 2.15 with PCC gave

ketone 2.16 in 67% yield. Ketone 2.16 was treated with CBr4 and PPh3 under Ramirez

conditions in order to give dibromoolefin 2.17. Dibromoolefin 2.17 should then be converted

in a rearrangement reaction to triyne 2.18 using n-BuLi in hexanes (FBW rearrangement

reaction). Unfortunately, compound 2.17 could not be synthesized, not even with additional

amounts of Zn that has normally accelerated reactions of this type.15

44

Scheme 2.7 Synthetic approach to the precursor (2.18) to [7]tBuPh (FBW route).14

It was not possible to synthesize the precursor to [7]tBuPh via lithiation- or coupling

reactions. An alternative route was developed based on ketone 2.16. Triyne 2.18 was formed

from 2.16 using lithiated trimethylsilyldiazomethane (Colvin’s reagent) as a reagent to

generate a carbene/carbenoid intermediate (Scheme 2.8), a version of the FBW

rearrangement.16–18 The use of Colvin’s reagent avoided the synthesis of the dibromoolefin

intermediate as described in Scheme 2.7. Unfortunately, the application of Colvin’s reagent

also had several disadvantages, such as low yields, bad reproducibility, as well as high

toxicity of trimethylsilyldiazomethane. In addition, the purification process was laborious

leading to loss of product and lowered reaction yield. Nevertheless, after reductive

elimination of precursor 2.18 under inert conditions, [7]tBuPh was obtained as stable solid

via careful crystallization by overlaying of a CH2Cl2 solution of [7]tBuPh with MeOH. This

45

procedure reproducibly afforded deep purple needles of [7]tBuPh in a yield as high as 44%.

The [7]cumulene was, however, not stable in solution or as amorphous powder and started to

decompose within days/weeks. In contrast, the crystalline solid of [7]tBuPh was indefinitely

stable.

Scheme 2.8 Rearrangement reaction of ketone 2.16 to triyne 2.18, followed by a reductive

elimination to [7]tBuPh.

During my diploma thesis14 and the beginnings of my doctoral research, the

rearrangement reaction using trimethylsilyldiazomethane proved to be the only method to

form [7]tBuPh. Because of the high toxicity of trimethylsilyldiazomethane and the low

reaction yields, the search for alternatives was continued. Thus, bistrimethylsilyltriyne 2.19

was used as a starting material (Scheme 2.9). This triyne, a stable white solid, was synthesized

using a known procedure.19–21

The reaction of 2.19 with MeLi (>1 equiv) was initiated at –78 °C (route a, Scheme

2.9). The temperature was increased to –5 °C, and ketone 2.2 was added. Finally, quenching

of the reaction via addition of MeI gave triyne 2.18, i.e., the precursor to [7]tBuPh. The

reaction afforded many byproducts and unconverted ketone 2.2, and triyne 2.18 could not be

isolated. It has been noteworthy that low temperatures (<–15 °C) were important for lithiation

reactions of bistrimethylsilyltriyne 2.19 to avoid polymerization of the in situ formed

dilithiated triynes.22

46

Route b was performed at lower temperature: Triyne 2.19 was treated with MeLi

(1 equiv) at –78 °C, and ketone 2.2 was added at –25 °C (Scheme 2.9). After quenching of the

reaction with MeI, the reaction outcome included a mixture of ketone 2.2, the triyne 2.18, and

the terminal triyne 2.20 that could not be separated. At this point, it seemed that all reactions

performed using MeI addition in the final step showed bad reproducibility aside from low

yields and isolation problems.

An analogous reaction without MeI addition was attempted as shown by route c in

Scheme 2.9. Triyne 2.19 was treated with MeLi at –20 °C, followed by addition of ketone 2.2.

Aqueous work-up gave terminal triyne 2.21 as the main product in yields of 30–50%. The

second product, triyne 2.22, the precursor to [7]tBuPh, was obtained in 20% yield. An

interesting observation of this reaction was that no TMS protected triyne was observed in the

product mixture, i.e., the bistrimethylsilyltriyne appeared to be lithiated on both ends using

only one equivalent of MeLi or the remaining trimethylsilyl groups were inadvertently

removed during work-up. To conclude, an alternative non-toxic synthetic pathway to the

precursor to [7]tBuPh, triyne 2.22 (containing OH groups instead of OMe groups), was

developed. This route, however, had disadvantages such as low reaction yields and poor

reproducibility.

47

Scheme 2.9 Synthetic approaches to triynes 2.18 and 2.22 (precursors to [7]tBuPh) using

bistrimethylsilyltriyne 2.19.


The synthesis of [9]tBuPh was straightforward starting with diyne 2.23, which was

formed through addition of lithiated trimethylsilylbutadiyne to ketone 2.2 (Scheme 2.10).

Desilylation of diyne 2.23 with K2CO3 gave terminal diyne 2.24. Oxidative homocoupling of

2.24 under Hay conditions (CuCl and TMEDA) gave tetrayne diol 2.25, the precursor to

[9]tBuPh, in 96% yield. Reductive elimination of 2.25 afforded a blue solution of [9]tBuPh.

Unfortunately, [9]tBuPh could not be obtained in solid form and was unstable in solution

decomposing after a period of hours/days even in deoxygenated solvents and shielded from

light.

48

Scheme 2.10 Synthesis of [9]tBuPh.

Cumulene [9]tBuPh was successfully synthesized and purified through filtration,

although handling was only possible in solution. While [7]tBuPh could reproducibly be

crystallized in form of stable needles, all crystallization attempts to afford [9]tBuPh as a

stable solid failed. The stability of [9]tBuPh was limited to several days when kept under

inert conditions. After several days, or in the presence of O2/hν, decomposition occured,

indicated by a color change from blue to brown or decolorization. It was noted during my

diploma research14 that the color of [9]tBuPh persisted for only several hours or a maximum

of ca. one day. This “time of survival” could be increased by changing several conditions.

The purity of [9]tBuPh in the reaction solution played an important role, and the product

purity could be increased by the use of an argon atmosphere instead of a nitrogen atmosphere

during the reductive elimination step. Additionally, neutral alumina was replaced by basic

alumina in the filtration step to efficiently trap the acidic aqueous residue. Finally, intense

attention was paid to maintain strictly anhydrous conditions, e.g., with dried anhydrous SnCl2

as reductant instead of SnCl2·2H2O. These optimization attempts increased the time before

decolorization of [9]tBuPh occurred, but unfortunately did not enable the isolation of

[9]tBuPh as a solid. Therefore, another precursor to [9]tBuPh was targeted, namely tetrayne

2.14 containing OMe groups instead of OH groups (Scheme 2.11). The idea was that the

replacement of elimination of H2O by an elimination of MeOH during the reductive

elimination step might positively affect the stability of [9]tBuPh. The reaction started with

49

bistrimethylsilyldiyne 2.26, which was lithiated by MeLi and treated with ketone 2.2. Final

addition of MeI gave compound 2.27, which could not be purified and thus, was directly

desilylated with K2CO3 to give diyne 2.12 in an overall yield of 66% (Scheme 2.11).

Compound 2.12 was converted to tetrayne 2.14 in 87% yield via a homocoupling reaction

under Hay conditions (CuCl and TMEDA). Although the 1H NMR- and 13C NMR spectra

confirmed the formation of very pure 2.14, TLC analysis showed a second spot in addition to

the product. The solvent mixture hexanes/CH2Cl2 (1:1) showed very good spot separation on

the TLC plate, however, column chromatography failed using this mixture. Recrystallization

in hexanes led to an off-white crystalline compound, which was also contaminated with the

byproduct. Another solvent mixture, namely hexanes/ethyl acetate (20:1) was used for column

chromatography, and pure 2.14 could be obtained in 27% yield.

Scheme 2.11 Synthesis of tetrayne 2.14.

Conversion of precursor 2.14 to [9]tBuPh was accomplished under inert conditions at

0 °C. After 1.5 h, tetrayne 2.14 was still present in the reaction mixture while a brown residue

had already appeared. After 3 h, TLC analysis showed two additional byproducts in addition

to the baseline fraction, precursor 2.14, and [9]tBuPh. The reaction mixture seemed to

decompose gradually without complete conversion of the starting material. Regarding the

synthesis of [9]tBuPh, tetrayne 2.14 seemed to be less suitable as precursor to [9]tBuPh than

tetrayne 2.25 (Scheme 2.10).

50

Several other test reactions were employed for the synthesis of [9]tBuPh by variation

of different aspects of the conversion, such as reactants, stoichiometry, temperature, and

solvent (Table 2.1). The first test reaction was performed in CH2Cl2 instead of Et2O under an

argon atmosphere at 0 °C (entry 1, Table 2.1). Tetrayne 2.25 was treated with anhydrous

SnCl2 and HCl (1 M in Et2O). The color of the reaction mixture turned to blue indicating the

formation of [9]tBuPh. After few seconds, the color turned to pink/violet affording certainly a

different product, although the identity of this product was not determined (entry 1, Table

2.1).23

The reaction was performed in CH2Cl2 once again (entry 2), but the amount of HCl

(1 M in Et2O) was reduced to only one drop. A blue solution was observed which turned to

grey after several minutes. TLC analysis showed several byproducts aside from tetrayne 2.25

and [9]tBuPh. After some minutes, an excess of HCl was added, and the solution turned to

pink/violet again. TLC analysis illustrated a pink spot and a violet spot right below indicating

two different products. After one week, decomposition of the products (dissolved in MeOH)

was observed.

The third reaction was done in CH2Cl2 at –78 °C instead of 0 °C (entry 3). After 15

min, a blue solution was observed. After additional 30 min, the blue solution became more

intense, and the TLC analysis confirmed formation of [9]tBuPh aside from unconverted 2.25.

The reaction mixture was stirred for 10 min and filtered through a plug of silica (bottom) and

alumina (top). No single crystals could be obtained by several crystallization methods.

Overlaying or diffusion methods using MeOH led to decomposition, while hexanes showed a

longer “time of survival” with decomposition occurring after 3 d when kept at ca. –20 °C.

Next, the synthesis of [9]tBuPh was tried in hexanes (entry 4) since this solvent

showed the most delayed decomposition of a solution of [9]tBuPh. The reaction was

monitored via TLC analysis, which illustrated several byproducts. After 3 h, the solvent was

reduced by bubbling nitrogen through the reaction mixture. During solvent evaporation, the

solution became black indicating decomposition of [9]tBuPh.

The synthesis of [9]tBuPh was performed at rt using only anhydrous SnCl2 without

additional HCl (entries 5 and 6 for Et2O and CH2Cl2, respectively). After 30 min, the color of

the Et2O solution did not change, while the CH2Cl2 solution became bluish. The TLC analysis

of the CH2Cl2 solution showed the greatest amount of byproducts, turning to one big

decomposition spot overnight. After 1 d, the color of the Et2O solution also darkened, and

after 3 d, a brown precipitate was observed.

51

In conclusion, none of the test reactions mentioned above improved the outcome for

the synthesis of [9]tBuPh. While the reaction in CH2Cl2 showed a new pink product after

initial formation of [9]tBuPh, the reaction done in hexanes showed no improvement, and only

decomposition of [9]tBuPh was observed. In addition, in combination with SnCl2, HCl

appeared to be essential for the synthesis of [9]tBuPh.

52

Table 2.1 Reductive elimination of 2.25 under various conditions.

entry solvent temp. SnCl2 (anhydr) HCl (1 M in Et2O) result

1 CH2Cl2 0 °C 3 equiv 4 equiv initial blue solution

([9]tBuPh) turning to

pink/violet, new product

was not completely stable

in solution

2 CH2Cl2 0 °C 3 equiv one drop;

after some

minutes,

excess

bluish solution turning to

grey; after excess of HCl,

pink/violet solution

showing a pink and a violet

spot in TLC

3 CH2Cl2 –78 °C 3 equiv 4 equiv blue solution after 30 min

(2.25 still present)

4 hexanes 0 °C 3 equiv 4 equiv decomposition during

solvent evaporation by

bubbling N2 into the

reaction mixture

5 CH2Cl2 rt 3 equiv none after 30 min blue solution;

many byproducts

6 Et2O rt 3 equiv none after 1 h dark solution

turning to brown residue

53

2.1.2.6 Synthetic approaches to [11]cumulene [11]tBuPh and [13]cumulene [13]tBuPh

Even though [9]tBuPh showed poor stability and difficult product isolation, the

synthesis of higher [n]tBuPh cumulenes with n = 11 and 13 was targeted. Initially, the

synthetic route to [11]tBuPh was pursued via synthesis of secondary alcohol 2.28 in order to

form ketone 2.29 and finally the pentayne 2.30 through a FBW rearrangement (Scheme 2.12).

Compound 2.12 was treated with EtMgBr, and the mixture was stirred for 20 min. Ethyl

formate was added, and the reaction was monitored by TLC analysis. After 10 min, starting

material and several new spots were observed, while one main spot could be assigned to the

product. Stirring of the reaction mixture was continued for 50 min to achieve full conversion

of the starting material. The presumptive product spot by TLC, however, vanished affording

several (>8) new spots. Thus, it seemed that 2.28 was not stable, and no further efforts were

made following this approach.

Scheme 2.12 Synthetic approach to [11]tBuPh.

54

The synthesis of [13]tBuPh was attempted from terminal triyne 2.21 (Scheme 2.13),

which had been successfully formed during the synthesis of 2.18 and 2.22 as outlined in

Scheme 2.9. Compound 2.21 was converted in a homocoupling reaction under Hay conditions

(CuCl and TMEDA) to hexayne 2.31, the precursor to [13]tBuPh (Scheme 2.13). The

reaction mixture showed many spots, as monitored by TLC analysis, and after aqueous work-

up and purification by column chromatography, one main product could be obtained as a

brownish solid in 31% yield. 1H- and 13C NMR spectra of the brownish solid fit very well to

the structure of hexayne 2.31. After repeated dissolution of 2.31, however, it was observed

that 2.31 was not stable and tended to decompose as confirmed by TLC analysis showing

plenty of new spots while the main product spot had vanished.

Scheme 2.13 Synthetic approach to [13]tBuPh.

Attempts were made to convert 2.31 to the [13]tBuPh without using work-up methods

such as filtration over alumina and solvent evaporation, since during this, relatively speaking,

time-consuming work-up, decomposition occured giving no possibility for subsequent

identification or characterization. The conversion of precursor 2.31 to [13]tBuPh was carried

out in Et2O (Figure 2.3), and reaction progress was monitored directly by UV/vis

spectroscopy. Initially, the UV/vis spectrum of pure 2.31 showed no significant absorption in

the visible region >350 nm (black curve, Figure 2.3). After addition of SnCl2 and HCl,

however, an absorption band at 361 nm appeared (red curve), which broadened during the

course of the reaction forming a shoulder (light and dark blue curves). After several hours and

slight evaporation of solvent via bubbling argon through the reaction mixture, several weakly

55

visible absorption signals between 350 and 450 nm were observed. After stirring the mixture

overnight, the UV/vis spectra showed three distinctive absorption signals at 342, 368, and 398

nm resembling the vibronic fine structure of a polyyne with values of ν = 2066 cm–1 and ν =

2048 cm–1. Additionally, a shoulder absorption at about 500 nm was observed (orange curve,

Figure 2.3)

Regarding the color of the reaction progress, the dissolved precursor was colorless,

while the reaction mixture became more and more orange after SnCl2 was added, from pale

apricot-orange to darker orange over time. Hence, there was no intense color change as

usually observed for longer cumulenes ([7]cumulenes: deep violet, [9]cumulenes: deep blue).

Additionally, no absorption bands were present in the lower energy region (>600 nm) that

would hint to cumulene formation. Consequently, no evidence for successful formation of

[13]tBuPh was observed.

56

Figure 2.3 UV/vis spectra taken during attempted conversion of precursor 2.31 to [13]tBuPh

(in Et2O).

2.1.3 Synthesis of the [n]Mes cumulene series

2.1.3.1 Limitations of “common” synthetic pathways: Toward the synthesis of

precursors to [n]Mes

The synthesis of [n]Mes cumulenes deviated significantly from those previously

described for other [n]cumulenes because of the bulkiness of the mesityl endgroup,

particularly with respect to ketone 2.32 (Scheme 2.14). The formation of the precursors to

[n]Mes required either a terminal monoyne with a structure of 2.33 (with m = 1, for [3]Mes,

57

[5]Mes, and [7]Mes) or an analogous terminal diyne 2.33 (with m = 2, for [9]Mes) as

described in the [n]tBuPh series of cumulenes in Section 2.1.2. It was possible to form the

ketone 2.32, but the addition of acetylides to form the acetylenes 2.33 failed, probably due to

steric hindrance from the ortho-methyl groups of the aryl ring (Scheme 2.14). Thus,

alternative synthetic routes had to be developed to circumvent this problem and have been

presented in the following sections starting with the longest representative of the [n]Mes

series, i.e., the [9]cumulene [9]Mes.

Scheme 2.14 Unsuccessful approach to mesityl acetylene 2.33.

2.1.3.2 Synthesis of [9]cumulene [9]Mes

The synthesis of [9]Mes started with mesityl acid chloride, which was treated with

bistrimethylsilyldiyne 2.26 and AlCl3 to give ketone 2.34 (Scheme 2.15).24–26 Addition of

lithiated mesitylene gave diyne 2.35. Desilylation of 2.35 afforded terminal diyne 2.36.

Compound 2.36 was converted to tetrayne 2.37 in a homocoupling reaction under Hay

conditions (CuCl and TMEDA), which was then subjected to reductive elimination to give

[9]cumulene [9]Mes, i.e., using SnCl2 and HCl (Scheme 2.15).

58

Scheme 2.15 Synthesis of [9]Mes.27


Formation of [7]Mes started with treatment of mesityl aldehyde with ethynyl

Grignard, followed by conversion to the alkynyl bromide 2.38 with NBS/AgNO3.28–30

Reaction of 2.38 with diyne 2.36 using the Cadiot-Chodkiewicz reaction and further oxidation

afforded ketone 2.39. Addition of lithiated mesitylene gave triyne 2.40, and reductive

elimination with SnCl2 and HCl afforded [7]Mes.

59


The synthesis of [7]Mes via a FBW rearrangement was targeted (see Scheme 2.7 for

synthesis of [7]tBuPh via FBW rearrangement reaction). This reaction required the methyl

ether 2.41 to enable the formation of the Li-acetylide and addition to ethyl formate (Table

2.2). The OH leaving group of terminal acetylene 2.42,31 however, has been sterically

shielded by the methyl groups of the mesityl moiety, leading to synthetic challenges in the

conversion of OH to OMe. The following methylation studies were done toward the synthesis

of 2.41 (Table 2.2).

60

Table 2.2 Methylation reaction of 2.42 under various conditions.

entry base

(1 equiv)

time

before MeI

addition

solvent temperature resulting

product(s)

1 EtMgBr 15 min THF rt 2.42

2 EtMgBr 2 h THF rt 2.42

3 n-BuLi 30 min THF –78 °C 2.32, 2.41

(ca. 20%)

4 n-BuLi 3 min THF –78 °C 2.42, 2.32, 2.41

(ca. 22%)

5 n-BuLi 1 min THF –95 °C 2.32, 2.41

(41%)

6 n-BuLi 1 min hexanes/Et2O –78 °C 2.42, 2.32

7 n-BuLi 1 min toluene –78 °C 2.32, 2.41

(ca. 46%)

8 n-BuLi 1 min toluene – 90 °C 2.42, 2.32

Reaction conditions such as solvent, temperature, and time before addition of MeI

were varied. TLC analysis from these reactions showed several products including starting

material 2.42, dimesitylmethanone 2.32, and the desired product 2.41. The best yields for the

synthesis of 2.41 were achieved using toluene as solvent at –78 °C (entry 7, Table 2.2) or

THF at –95 °C (entry 5, Table 2.2) with 46% and 41%, respectively. Despite the moderate

results achieved for some of the methylation studies, the synthesis of 2.41 was abandoned due

to poor reproducibility.

The FBW rearrangement approach was then targeted using propargylic alcohol 2.42

(Scheme 2.17). The acetylenic proton of 2.42 could not be converted to a Grignard reagent by

the use of EtMgBr and thus, the reaction did not afford the desired alcohol 2.43. Two other

61

products, however, were formed depending on the stoichiometry of the reactants. Addition of

one equivalent of EtMgBr (route a, Scheme 2.17) formed a product that was obtained by

column chromatography as a light yellow solid. The IR band at 1930 cm–1 as well as the

signals at 6.11 ppm and 203.6 ppm observed in the 1H- and 13C NMR spectra, respectively,

indicated an allene. The signal at 6.11 ppm in the 1H NMR spectrum, however, integrated to

only one proton. A crystallographic analysis identified the product as allene 2.44, which

contained a bromine atom instead of one allenic proton. In contrast, the use of two equivalents

of EtMgBr afforded a different product, i.e., the alcohol 2.45 as an off-white solid in 11%

yield (route b, Scheme 2.17). This compound was identified by NMR spectroscopy. From the

mechanistic point-of-view, the reaction was assumed to proceed via the aldehyde intermediate

2.46 that was initially formed through addition of the in situ formed acetylide of 2.42 to ethyl

formate. Due to the excess of EtMgBr in the reaction mixture, EtMgBr could then add to

aldehyde 2.46 to give alcohol 2.45.

62

OH

HO

2.45

HHO

2.42

O

EtO H

1. EtMgBr, THF, rt

2.

OH

HO OH

2.43

Br

H

2.44

O

EtO H

1. 1 equiv EtMgBr, THF, rt

2.

16%

O

EtO H

1. 2 equiv EtMgBr, THF, rt

2.

11%

O

HO

route a)

route b)

2.46

H

3. NH4Cl, H2O

3. NH4Cl, H2O

Scheme 2.17 Reaction of terminal acetylene 2.42 with EtMgBr and ethyl formate.


The triisopropylsilyl-terminated acetylene 2.47 was obtained through a Grignard

reaction using mesitylbromide and ester 2.48 (Scheme 2.18). Desilylation of 2.47 with TBAF

gave the terminal alkyne 2.42. A homocoupling reaction of compound 2.42 under Hay

conditions (CuCl and TMEDA) gave diyne 2.49, which was converted to [5]cumulene [5]Mes

via reductive elimination reaction using SnCl2 and HCl.

63


The homocoupling reaction of compound 2.42 under Hay conditions

(CuCl and TMEDA) showed limited reproducibility and purity of the final product 2.49. Only

impure product was obtained that could not be further purified by either column

chromatography or recrystallization. The crude product was carried forward to the [5]Mes,

although this led to formation of even more byproducts. To circumvent these problems, two

palladium-catalyzed homocoupling reactions were attempted (route a and b, Scheme 2.19),

using Pd(PPh3)2Cl2 and CuI as catalysts.32,33 Route a was performed with Et3N as base and

ethyl bromoacetate as oxidant affording 2.49 in 12% yield (after recrystallization) along with

slight impurity (<10%) of the starting material. Route b used i-Pr2NH as base and I2 as

oxidant affording pure 2.49 in 26% yield. Both reactions were favored compared to the

standard homocoupling reaction under Hay conditions in spite of the low yields, since

purification and characterization of 2.49 was accomplished more easily.

64

Scheme 2.19 Pd-catalyzed homocoupling reactions giving precursor 2.49.


The synthesis of [3]Mes was not possible, because the necessary precursors could not

be formed. Specifically, terminal acetylene 2.41 would be used for the synthesis of [3]Mes.

The synthesis of compound 2.41, however, was problematic as described in Section 2.1.3.3

and Table 2.2.

2.2 Summary and conclusion regarding the stability of [n]cumulenes

In summary, the synthesis of two series of [n]cumulenes, [n]tBuPh and [n]Mes was

accomplished. For the lower representatives of [n]tBuPh, i.e., with n = 3 and 5, the developed

synthetic routes were straightforward giving good to very good isolated yields. An alternative

reaction to form [7]cumulene [7]tBuPh, as shown in Scheme 2.9, was developed in order to

avoid the use of the highly toxic trimethylsilyldiazomethane that was initially used to form

[7]tBuPh (Scheme 2.8). Both approaches gave low yields, although the approach using

bistrimethylsilyltriyne was favored because of easier formation of intermediates and

precursors. Regarding the synthesis of [n]Mes, it was not possible to proceed using protocols

that were successful for [n]tBuPh because of either the inability to add acetylides to the

mesityl ketone 2.32 or the unsuccessful formation of building blocks containing methyl ether

units (Section 2.1.3.3). During the course of the doctoral research, methylation reactions

affording methyl ether derivatives of propargylic alcohols could finally be realized.

Unfortunately, the methylation reactions showed low yields and bad reproducibility (Table

2.2) and meanwhile, alternative routes had already been developed to enable the synthesis of

[n]Mes.§

65

[n]Cumulenes with n ≤ 5 that have been studied in this thesis are infinitely stable at

ambient conditions either in solution or as solid. In contrast, longer [n]cumulenes with n = 7

or 9 are not particularly stable in solution under ambient conditions, but some derivatives

show sufficient stability in the solid state so that X-ray crystallographic analysis could be

accomplished for all derivatives except [9]tBuPh (see Chapter IV). This includes the first

crystallographic analysis of [7]- and [9]cumulenes reported to date. Attempts to isolate higher

cumulenes via precipitation result in decomposition. The pure cumulene [7]tBuPh can be

isolated by slow crystallization from a solution of CH2Cl2/MeOH, and the crystalline solid is

indefinitely stable in the absence of O2 (glove box).34 Compound [7]Mes is stable in solution

for weeks when kept at ca. –20 °C in a deoxygenated solution, and crystals of [7]Mes from

CH2Cl2/hexane are stable for at least one week when stored at ca. –20 °C in the absence of

light and oxygen.34 Cumulenes [9]tBuPh and [9]Mes decompose rapidly in solution when

exposed to oxygen, but they can be handled for hours ([9]tBuPh) or days ([9]Mes) when kept

in a cold (–20 °C), deoxygenated solution of Et2O that is shielded from ambient light. Crystals

of [9]Mes grown from a C2D2Cl4 solution at –20 °C are stable for at least one week when

stored in the absence of light and O2.34

Changing several aspects of the synthetic conditions improved the stability of higher

cumulenes and their precursors in terms of handling, purity, persistence, and increased

reaction yields. More specifically, nitrogen atmosphere was changed to argon atmosphere for

all syntheses that were accomplished under inert conditions (intermediates, precursors, and

cumulenes). In addition, attention was paid to exclude water using dried anhydrous reactants.

Furthermore, basic alumina was used for cumulene purification instead of neutral alumina.35

Finally, attempts to synthesize even longer [n]tBuPh cumulenes with n = 11 and 13

have failed mainly due to instability at the stage of the precursor syntheses. Hence, it cannot

be ruled out that higher cumulenes might be sufficiently stable for characterization via e.g.,

UV/vis spectroscopy. It is, however, necessary to search for alternative routes for the

synthesis of the precursors to facilitate the synthesis of longer cumulenes.

66

2.3 Experimental part

2.3.1 General procedures and methods

Reagents were purchased reagent grade from commercial suppliers and used without further

purification. THF and Et2O were distilled from sodium/benzophenone. CH2Cl2 was distilled

from CaH2, and MeOH was distilled from magnesium.

Na2SO4 were used as standard drying reagents after aqueous work-up.

1H and 13C NMR spectra were recorded on a Bruker Avance 300 operating at 300 MHz (1H

NMR) and 75 MHz (13C NMR), a Bruker Avance 400 operating at 400 MHz (1H NMR) and

100 MHz (13C NMR), or a Jeol Alpha 500 operating at 500 MHz (1H NMR) and 126 Hz (13C

NMR). NMR spectra were referenced to the residual solvent signal (1H: CDCl3, 7.24 ppm; 13C: CDCl3, 77.0 ppm; 1H: CD2Cl2, 5.32 ppm; 13C: CD2Cl2, 53.8 ppm) and recorded at

ambient probe temperature. CDCl3 (99.8%, Deutero GmbH) was stored over 4 Å molecular

sieves. CD2Cl2 (99.6%, Deutero GmbH) was used as received.

UV/vis spectroscopic measurements were carried out on a Varian Cary 5000 UV/vis-NIR

spectrophotometer or an Agilent Cary 60 UV/Vis spectrophotometer at rt.

Mass spectra were obtained from a Kratos MS50G (EI), Micromass Zabspec (EI), Bruker

9.4T Apex-Qe FTICR (MALDI, Matrix: DCTB), Agilent Technologies 6220 oaTOF (ESI),

Bruker micro TOF II focus, and Bruker maxis 4G (APPI, ESI, in MeOH/ACN) instruments.

IR spectra were recorded on a Varian 660-IR spectrometer as solids in ATR-mode.

Differential scanning calorimetry (DSC) measurements were made on a Mettler Toledo TGA/

STDA 851e/1100/SF.

Cyclic voltammetry was performed using 0.1 M n-Bu4NPF6 as supporting electrolyte. Pt wire

was used as counter electrode, Ag/AgNO3 was used as reference electrode, and a glassy

carbon disc was used as working electrode. Ferrocene was added to the samples serving as

internal standard.

Melting points were measured with an Electrothermal 9100 instrument.

67

TLC analyses were carried out on TLC plates from Macherey-Nagel (ALUGRAM® SIL

G/UV254) and visualized via UV-light (264/364 nm) or standard coloring reagents. Column

chromatography was performed using Silica Gel 60M (Merck).

Several compounds mentioned herein have already been synthesized during my diploma

thesis. These reactions, however, have been modified, optimized, or improved during my

doctoral research, leading to better yields or higher purity. Furthermore, some of the

characterization data missing in my diploma thesis is provided in following experimental

data.35

2.3.2 Experimental data and compound characterization

Bis-(3,5-di-t-butyl-phenyl)-methanol 2.1.14 To a mixture of Mg (2.27 g, 93.3 mmol) in THF

(15 mL) under a N2 atmosphere was added a small amount (5 mL) of a solution of 3,5-di-t-

butylphenylbromide (22.1 g, 82.1 mmol) in THF (30 mL).36 A small crystal of I2 was added,

and the mixture was heated to promote the Grignard formation. The rest of the solution of 3,5-

di-t-butylphenylbromide in THF was added, and the reaction mixture was stirred for 30 min at

a gentle reflux while slowly diluting with THF (150 mL). The reaction was cooled to rt, a

solution of ethyl formate (2.76 g, 3.00 mL, 37.3 mmol) in dry THF (20 mL) was slowly

added, and the mixture was stirred at rt for 3 d. The reaction was quenched via the addition of

saturated aq NH4Cl (150 mL), and Et2O (50 mL) was added. The layers were separated, the

organic phase washed with saturated aq NH4Cl (100 mL) and saturated aq NaCl (100 mL),

dried (Na2SO4), and filtered. Solvent removal and purification by column chromatography

(silica gel, hexanes/EtOAc 20:1) afforded 2.1 (13.6 g, 89%) as a colorless solid. Mp 138–140

°C. Rf = 0.56 (hexanes/EtOAc 5:1). IR 3571 (m), 3058 (vw), 2954 (s), 2900 (m), 2866 (m),

1594 (m), 1469 (m), 1360 (m), 1063 (m), 730 (m) cm−1; 1H NMR (300 MHz, CDCl3) δ 7.34

(t, J = 1.7 Hz, 2H), 7.27 (d, J = 1.7 Hz, 4H), 5.85 (d, J = 3.4 Hz, 1H), 2.26 (d, J = 3.4 Hz, 1H),

1.31 (s, 36H); 13C NMR (75 MHz, CDCl3) δ 150.7, 143.0, 121.4, 121.0, 77.5, 34.9, 31.5. EI

68

HRMS m/z calcd. for C29H44O (M+) 408.33920, found 408.33867. Anal. calcd. for C29H44O:

C, 85.23; H, 10.85. Found: C, 85.22; H, 10.90.

Bis-(3,5-di-t-butyl-phenyl)-methanol 2.1. To a solution of 3,5-di-t-butylphenylbromide (15.4

g, 0.0572 mol) in THF (145 mL) was added n-BuLi (2.5 M in hexanes, 23 mL, 0.058 mol)

under a N2 atmosphere at –78 °C via a syringe.37 After stirring for 30 min, ethyl formate

(2.1 g, 2.3 mL, 0.028 mol) was slowly added, and the mixture was stirred for 30 min. The

cooling bath was removed, and the reaction was stirred overnight. The reaction was quenched

via the addition of saturated aq NH4Cl (150 mL), and Et2O (100 mL) was added. The layers

were separated, the organic phase washed with saturated aq NH4Cl (150 mL) and saturated aq

NaCl (150 mL), dried (Na2SO4), and filtered. Solvent removal and purification by column

chromatography (silica gel, hexanes/EtOAc 20:1) afforded 2.1 (6.64 g, 57%) as a light yellow

solid. Spectral data are consistent with that described above.

Bis-(3,5-di-t-butyl-phenyl)-methanone 2.2.14,38 To a solution of 2.1 (3.00 g, 7.33 mmol) in

CH2Cl2 (70 mL) was added PCC (2.31 g, 10.7 mmol), celite (2 g), and molecular sieves (4 Å,

2 g). After 3 h, the reaction mixture was passed through a plug of silica gel to remove the

chromium waste. Solvent removal afforded 2.2 (2.96 g, 99%) as a colorless solid. Mp 115–

118 °C. Rf = 0.61 (hexanes/EtOAc 5:1). IR 3063 (vw), 2954 (s), 2903 (m), 2867 (m), 1657

(s), 1592 (m), 1462 (m), 1231 (s) cm−1; 1H NMR (400 MHz, CDCl3) δ 7.65 (d, J = 1.8 Hz,

2H), 7.63 (d, J = 1.8 Hz, 4H), 1.34 (s, 36H); 13C NMR (100 MHz, CDCl3) δ 197.9, 150.7,

69

137.3, 126.3, 124.7, 35.0, 31.4. EI MS m/z 406.3 (M+, 25), 391.3 ([M − CH3]+, 100). EI

HRMS m/z calcd. for C29H42O (M+) 406.32358, found 406.32378.

Bis-(3,5-di-t-butyl-phenyl)-methanone 2.2. To a solution of 3,5-di-t-butylphenylbromide

(15.4 g, 0.0572 mol) in THF (145 mL) was added n-BuLi (2.5 M in hexanes, 23 mL,

0.058 mol) under a N2 atmosphere at –78 °C via a syringe. After stirring for 30 min, ethyl

formate (2.1 g, 2.3 mL, 0.028 mol) was slowly added, and the mixture was stirred for 30 min.

The cooling bath was removed, and the reaction was stirred overnight. The reaction was

quenched via the addition of saturated aq NH4Cl (150 mL), and Et2O (100 mL) was added.

The layers were separated, the organic phase washed with saturated aq NH4Cl (150 mL) and

saturated aq NaCl (150 mL), dried (Na2SO4), and filtered. Solvent removal and purification by

column chromatography (silica gel, hexanes/EtOAc 20:1) afforded 2.2 (1.68 g, 14%) as an

off-white solid. Spectral data are consistent with the above mentioned.

3,3-Bis(3,5-di-t-butyl-phenyl)-3-methoxy-1-(trimethylsilyl)prop-1-yne 2.3. To a solution

of trimethylsilylacetylene (2.2 g, 3.1 mL, 22 mmol) in THF (25 mL) at −78 °C was added n-

BuLi (2.5 M in hexanes, 6.5 mL, 16 mmol) under a N2 atmosphere via a syringe. The reaction

mixture was stirred for 40 min, and a solution of 2.2 (6.0 g, 15 mmol) in THF (60 mL) was

added. The cooling bath was removed, the reaction was stirred for 4 h, and MeI (20 g, 9.0 mL,

0.14 mol) was added. The reaction mixture was stirred for 2 h. The solution was quenched via

70

the addition of saturated aq NH4Cl (80 mL), and Et2O (60 mL) was added. The layers were

separated, the organic phase was washed with saturated aq NaCl (60 mL), dried (Na2SO4), and

filtered. Solvent removal and recrystallization from MeOH afforded 2.3 (6.7 g, 86%) as a

colorless solid. Mp 93–95 °C. Rf = 0.64 (hexanes/CH2Cl2 1:1). IR 3076 (vw), 2955 (m), 2901

(m), 2866 (w), 2827 (w), 2161 (w), 1596 (m), 1086 (s), 843 (s) cm−1; 1H NMR (300 MHz,

CDCl3) δ 7.42 (d, J = 1.8 Hz, 4H), 7.28 (t, J = 1.8 Hz, 2H), 3.42 (s, 3H), 1.27 (s, 36H), 0.25

(s, 9H); 13C NMR (75 MHz, CDCl3) δ 150.1, 141.8, 121.3, 121.2, 105.7, 93.7, 82.6, 52.9,

34.9, 31.4, 0.0. ESI MS m/z 487.4 ([M − MeO]+, 100); ESI HRMS m/z calcd. for

C35H54NaOSi ([M + Na]+) 541.38361, found 541.38246.

3,3-Bis(3,5-di-t-butyl-phenyl)-3-methoxyprop-1-yne 2.4. To a solution of 2.3 (6.7 g,

13 mmol) in MeOH/THF (105 mL, 20:1) was added K2CO3 (2.0 g, 14 mmol). The reaction

mixture was stirred for 3 h and quenched via the addition of saturated aq NH4Cl (60 mL). The

layers were separated, and the aqueous phase was extracted with CH2Cl2 (2 x 50 mL). The

organic phase was washed with saturated aq NH4Cl (60 mL) and saturated aq NaCl (60 mL),

dried (Na2SO4), and filtered. Solvent removal afforded 2.4 (5.6 g, 97%) as a colorless solid.

Mp 106–109 °C. Rf = 0.60 (hexanes/CH2Cl2 1:1). IR 3298 (w), 3272 (w), 3070 (vw), 2954 (s),

2902 (m), 2865 (m), 2828 (w), 2156 (w), 1594 (m), 1084 (s) cm−1; 1H NMR (400 MHz,

CDCl3) δ 7.35 (d, J = 1.8 Hz, 4H), 7.29 (t, J = 1.8 Hz, 2H), 3.41 (s, 3H), 2.84 (s, 1H), 1.26 (s,

36H); 13C NMR (100 MHz, CDCl3) δ 150.2, 141.7, 121.40, 121.36, 84.1, 82.1, 76.8, 52.8,

34.9, 31.4. ESI MS m/z 415.3 ([M − MeO]+, 100); ESI HRMS m/z calcd. for C32H46NaO

([M + Na]+) 469.34409, found 469.34511.

71

1,1,4,4-Tetrakis(3,5-di-t-butylphenyl)-4-methoxybut-2-yn-1-ol 2.5. To a solution of 2.4

(1.27 g, 2.84 mmol) in THF (15 mL) at −78 °C was added n-BuLi (2.5 M in hexanes,

1.14 mL, 2.85 mmol) under a N2 atmosphere via a syringe. The reaction mixture was stirred

for 1 h, and a solution of 2.2 (1.15 g, 2.83 mmol) in THF (15 mL) was added. The cooling

bath was removed, and the reaction was stirred for 3 h. The solution was quenched via the

addition of saturated aq NH4Cl (30 mL), and Et2O (30 mL) was added. The layers were

separated, and the organic phase was washed with saturated aq NaCl (30 mL), dried

(Na2SO4), and filtered. Solvent removal and purification by column chromatography (silica

gel, hexanes/EtOAc 30:1) afforded 2.5 (1.9 g, 78%) as a colorless solid. Mp 117−120 °C.

Rf = 0.36 (hexanes/EtOAc 20:1). IR 3464 (br), 3069 (vw), 2955 (s), 2903 (m), 2866 (m), 1596

(m), 1458 (m), 1066 (m), 716 (s) cm−1; 1H NMR (300 MHz, CDCl3) δ 7.49 (br d, J = 1.3 Hz,

4H), 7.37 (br d, J = 1.1 Hz, 4H), 7.34 (br s, 2H), 7.31 (br s, 2H), 3.49 (s, 3H), 2.84 (s, 1H),

1.27 (s, 36H), 1.25 (s, 36H); 13C NMR (75 MHz, CDCl3) δ 150.3, 150.0, 144.1, 142.1, 121.6,

121.4, 121.3, 120.7, 93.2, 87.5, 82.3, 75.8, 52.9, 34.9, 34.8, 31.5 (one signal coincident or not

observed). ESI HRMS m/z calcd. for C61H88NaO2 ([M + Na]+) 875.6677, found 875.6673.

1,1,4,4-Tetrakis(3,5-di-t-butyl-phenyl)buta-1,2,3-triene ([3]tBuPh). To a solution of 2.5

(1.0 g, 1.2 mmol) in Et2O (15 mL) was added anhydrous SnCl2 (0.67 g, 3.5 mmol) and HCl

(1 M in Et2O, 4.7 mL, 4.7 mmol) at 0 °C under an Ar atmosphere. After stirring for 1 h, the

72

solution was filtered through a plug of basic alumina oxide. Solvent removal afforded

[3]tBuPh (0.79 g, 84%) as a yellow solid. Mp 206–208°C. Rf = 0.26 (hexanes/EtOAc 20:1).

UV/vis (CHCl3) λmax (ε) 240 (12600), 277 (26200), 426 (37900) nm. UV/vis (Et2O) λmax 278,

323, 424 nm; IR 3060 (vw), 2952 (s), 2864 (m), 1587 (m), 1244 (s) cm−1; 1H NMR (400

MHz, CD2Cl2) δ 7.43 (s, 4H), 7.37 (d, J = 1.6 Hz, 8H), 1.31 (s, 72H); 13C NMR (100 MHz,

CD2Cl2) δ 151.9, 151.0, 139.0, 124.1, 124.0, 122.4, 35.2, 31.7. MALDI HRMS m/z calcd. for

C60H84 (M+) 804.65675, found 804.65640.

Crystal data for [3]tBuPh: C60H84, M = 805.27; monoclinic crystal system; space group C2/c,

a = 51.7261(10), b = 10.9049(2), c = 19.0153(4) Å; β = 95.00(3)°; V = 10685.2(4) Å3; Z = 8;

ρcalcd = 1.001 g cm–3; µ(MoKα) = 0.056 mm–1; λ = 0.71073 Å; 173.15 K; 2θ max = 54.98°;

total data collected = 22006; R1 = 0.0873 [6868 observed reflections with F ≥ 4σ(F)]; wR2 =

0.2878 for 541 variables, 12113 unique reflections, and 66 restraints; residual electron density

= 0.76 and –0.53 e Å–3. Several t-butyl groups showed disorder, which have been resolved

and refined to the following occupation factors: C15b/C15e = 40:50%; C13b,c,d/C13e,f,g =

76:24%; C23b,c,d/C23e,f,g = 63:37%; C43b,c,d/C43e,f,g = 68:32%; C45b,c,d/C45e,f,g =

85:15 %. CCDC 903382.

1,1-Bis-(3,5-di-t-butyl-phenyl)-3-trimethylsilyl-prop-2-yn-1-ol 2.6. To a solution of

trimethylsilylacetylene (1.1 g, 1.6 mL, 12 mmol) in THF (25 mL) at −78 °C under a N2

atmosphere was added n-BuLi (2.5 M in hexanes, 3.4 mL, 8.5 mmol) via a syringe. The

reaction mixture was stirred for 1 h, and a solution of 2.2 (3.0 g, 7.4 mmol) in THF (25 mL)

was added. The cooling bath was removed, and the reaction was stirred for 1 h. The solution

was quenched via the addition of saturated aq NH4Cl (30 mL), and Et2O (30 mL) was added.

The layers were separated, the organic phase dried (Na2SO4), and filtered. Solvent removal

afforded 2.6 (3.71 g, quant.) as a colorless solid. Mp 150–152 °C. Rf = 0.67 (hexanes/EtOAc

5:1). IR 3523 (w), 2956 (m), 2901 (m), 2865 (m), 2169 (w), 1596 (m), 843 (s) cm−1; 1H NMR

73

(300 MHz, CDCl3) δ 7.51 (d, J = 1.8 Hz, 4H), 7.30 (t, J = 1.8 Hz, 2H), 2.74 (s, 1H), 1.28 (s,

36H), 0.23 (s, 9H); 13C NMR (100 MHz, CDCl3) δ 150.3, 143.8, 121.4, 120.6, 109.0, 91.2,

75.8, 35.0, 31.4, -0.1. EI MS m/z 504.4 (M+, 7), 57.1 ([t-butyl]+, 100). EI HRMS m/z calcd.

for C34H52SiO (M+) 504.37875, found 504.37688. Anal. calcd for C34H52SiO: C, 80.89; H,

10.38. Found: C, 80.51; H, 10.57.

1,1-Bis-(3,5-di-t-butyl-phenyl)-prop-2-yn-1-ol 2.7. To a solution of 2.6 (3.7 g, 7.3 mmol) in

MeOH (40 mL) was added K2CO3 (1.0 g, 7.2 mmol). After stirring for 5 h, the reaction

mixture was quenched via the addition of saturated aq NH4Cl (30 mL), and CH2Cl2 (30 mL)

was added. The layers were separated, the organic phase was washed with saturated aq NH4Cl

(40 mL) and saturated aq NaCl (40 mL), dried (Na2SO4), and filtered. Solvent removal

afforded 2.7 (2.8 g, 89%) as an off-white solid. Mp 101–104 °C. Rf = 0.43 (hexanes/EtOAc

20:1). IR 3543 (m), 3294 (w), 3065 (vw), 2956 (s), 2902 (m), 2866 (m), 1594 (m), 717 (s)

cm−1; 1H NMR (300 MHz, CDCl3) δ 7.46 (d, J = 1.8 Hz, 4H), 7.34 (t, J = 1.8 Hz, 2H), 2.85 (s,

1H), 2.80 (s, 1H), 1.29 (s, 36H); 13C NMR (75 MHz, CDCl3) δ 150.4, 143.4, 121.6, 120.5,

87.4, 75.4, 74.7, 34.9, 31.4. EI HRMS m/z calcd. for C31H44O (M+) 432.33920, found

432.33958.

74

1,1,6,6-Tetrakis(3,5-di-t-butylphenyl)hexa-2,4-diyne-1,6-diol 2.8. To a solution of 2.7

(2.8 g, 6.5 mmol) in CH2Cl2 (30 mL) was added a solution of Hay catalyst [CuCl (0.64 g,

6.5 mmol) and TMEDA (1.5 g, 1.9 mL, 13 mmol) in CH2Cl2 (10 mL)]. The reaction mixture

was stirred for 1 d, saturated aq NH4Cl (50 mL) was added, and the resulting mixture was

extracted with Et2O (50 mL). The organic phase was washed with saturated aq NH4Cl


afforded 2.8 (2.73 g, 98%) as a white solid. Mp 190–191 °C. Rf = 0.31 (hexanes/EtOAc 10:1).

IR 3595 (vw), 3520 (w), 3440 (br), 3070 (vw), 2955 (s), 2903 (m), 2866 (m), 1595 (m), 715

(s) cm−1; 1H NMR (300 MHz, CDCl3) δ 7.42 (d, J = 1.8 Hz, 8H), 7.34 (t, J = 1.8 Hz, 4H),


76.0, 70.7, 35.0, 31.4. ESI HRMS m/z calcd. for C62H86NaO2 ([M + Na]+) 885.65200, found

885.65186.

Crystal data for 2.8: C64H94O4, M = 927.39, triclinic crystal system; space group P–1, a =

10.1043(2), b = 10.2676(1), c = 16.7900(4) Å, α = 97.703(1)°, β = 102.907(1)°, γ =

112.992(1)°; V = 1514.74(5) Å3, Z = 1; ρcalcd = 1.017 g cm–3; µ(MoKα) = 0.061 mm–1; λ =

0.71073 Å; 173(2) K; 2θ max = 55.06°; total data collected = 12974; R1 = 0.0726 [5614

observed reflections with F ≥ 4σ(F)]; wR2 = 0.2249 for 307 variables, 6905 unique

reflections, and 0 restraints; residual electron density = 0.894 and –0.431 e Å–3. One MeOH

per asymmetric unit.

1,1,6,6-Tetrakis(3,5-di-t-butyl-phenyl)hexa-1,2,3,4,5-pentaene ([5]tBuPh). To a solution of

2.8 (6.9 g, 8.0 mmol) in Et2O (60 mL) was added anhydrous SnCl2 (2.8 g, 15 mmol) and HCl

(1 M in Et2O, 16 mL, 16 mmol) at 0 °C under an Ar atmosphere. After stirring for 1 h, the

solution was filtered through a plug of basic alumina oxide. Solvent removal and purification

by column chromatography (silica gel, hexanes/CH2Cl2 4:1) afforded [5]tBuPh (4.89 g, 74%)

as crystalline red solid. Mp ~ 234 °C. Rf = 0.84 (hexanes/EtOAc 30:1). UV/vis (CHCl3) λmax

75

(ε) 255 (35000), 275 (25500), 296 (25800), 373 (9900), 437 (14900), 510 (66700) nm. UV/vis

(Et2O) λmax 253, 274, 292, 371, 500 nm; IR 3064 (w), 2955 (s), 2903 (m), 2866 (m), 1995

(w), 1585 (m), 1241 (s), 706 (s) cm−1; 1H NMR (300 MHz, CDCl3) δ 7.48 (d, J = 1.7 Hz, 8H),

7.39 (d, J = 1.7 Hz, 4H), 1.33 (s, 72H); 13C NMR (75 MHz, CDCl3) δ 150.7, 148.3, 137.4,

126.3, 125.6, 123.9, 122.6, 35.0, 31.5. MALDI HRMS m/z calcd. for C62H84 (M+) 828.65675,

found 828.65633.

Crystal data for [5]tBuPh: C62H84, M = 829.29, monoclinic crystal system; space group P21/c,

a = 14.1531(2), b = 18.8068(4), c = 11.0114(2) Å, β = 108.030(1)°; V = 2787.03(9) Å3, Z = 2;

ρcalcd = 0.988 g cm–3; µ(MoKα) = 0.055 mm–1; λ = 0.71073 Å; 173.15 K; 2θ max = 55°; total

data collected = 12488; R1 = 0.0661 [4591 observed reflections with F ≥ 4σ(F)]; wR2 =

0.1984 for 293 variables, 6395 unique reflections, and 15 restraints; residual electron density

= 0.29 and –0.33 e Å–3. One t-butyl group showed disorder, which was resolved and refined to

the following occupation factor: C25b,c,d/C25e,f,g = 62:38%. CCDC 903379.

1,1,7,7-Tetrakis(3,5-di-t-butyl-phenyl)-1,7-dimethoxyhepta-2,5-diyn-4-ol 2.15. To a

solution of EtMgBr (1.0 M in THF, 12.1 mL, 12.1 mmol) in THF (10 mL) was added

compound 2.4 (5.42 g, 12.1 mmol) in THF (50 mL) under a N2 atmosphere via a syringe. The

reaction mixture was stirred for 20 min, and ethyl formate (0.45 g, 0.49 mL 6.1 mmol) was

added. The reaction was stirred for 30 min and quenched via the addition of saturated aq

NH4Cl (50 mL), and Et2O (50 mL) was added. The layers were separated, the organic phase

was washed with water (50 mL) and saturated aq NaCl (50 mL), dried (Na2SO4), and filtered.

Solvent removal and purification by column chromatography (silica gel, hexanes/EtOAc 20:1)

afforded 2.15 (4.74 g, 85%) as a colorless solid. Mp 75−80 °C (from EtOAc). Rf = 0.12

(hexanes/EtOAc 20:1). IR 3442 (br), 3071 (vw), 2956 (s), 2903 (m), 2867 (m), 1597 (m),

1072 (s) cm−1; 1H NMR (300 MHz, CDCl3) δ 7.35 (d, J = 1.7 Hz, 4H), 7.33−7.31 (m, 4H),

76

5.53 (d, J = 8.1 Hz, 1H), 3.42 (s, 6H), 2.39 (d, J = 8.1 Hz, 1H), 1.27 (s, 36H), 1.26 (s, 36H); 13C NMR (75 MHz, CDCl3) δ 150.2, 141.5, 141.4, 121.43, 121.40, 85.8, 85.3, 82.1, 53.0,

52.7, 34.9, 31.4. ESI HRMS m/z calcd. for C65H92NaO3 ([M + Na]+) 943.6939 found

943.6928.

1,1,7,7-Tetrakis(3,5-di-t-butyl-phenyl)-1,7-dimethoxyhepta-2,5-diyn-4-one 2.16. To a

solution of 2.15 (4.74 g, 5.15 mmol) in CH2Cl2 (50 mL) were added PCC (1.7 g, 7.9 mmol),

celite (2.0 g), and molecular sieves (4 Å, 2.0 g). The reaction mixture was stirred for 3 h and

passed through a plug of silica gel to remove the chromium waste. Solvent removal and

recrystallization from MeOH afforded 2.16 (3.158 g, 67%) as an off-white solid. Mp 161–164

°C. Rf = 0.62 (hexanes/EtOAc 10:1). IR 3067 (vw), 2954 (s), 2903 (m), 2867 (m), 2208 (w),

1656 (m), 1632 (m), 1594 (m), 1224 (s), 1072 (m), 712 (m) cm−1; 1H NMR (300 MHz,

CDCl3) δ 7.32 (t, J = 1.8 Hz, 4H), 7.26 (d, J = 1.8 Hz, 8H), 3.41 (s, 6H), 1.23 (s, 72H); 13C

NMR (75 MHz, CDCl3) δ 150.6, 140.1, 122.0, 121.4, 92.9, 82.4, 53.4, 34.9, 31.4 (two signals

coincident or not observed). ESI HRMS m/z calcd. for C65H90NaO3 ([M + Na]+) 941.6782,

found 941.6806.

77

1,1,8,8-Tetrakis(3,5-di-t-butyl-phenyl)-1,8-dimethoxyocta-2,4,6-triyne 2.18. To a solution

of trimethylsilyldiazomethane (2 M in Et2O, 3 mL, 6 mmol) in THF (100 mL) was added n-

BuLi (2.5 M in hexanes, 2.4 mL, 5.9 mmol) at –78 °C under a N2 atmosphere. Caution!

Trimethylsilyldiazomethane should be regarded as extremely toxic and should only be

handled by individuals trained in its proper and safe use. All operations must be carried

out in a well-ventilated fume hood and all skin contact should be avoided. The solution

was stirred for 30 min and transferred via a cannula to a solution of ketone 2.16 (2.73 g,

2.97 mmol) in THF (200 mL) cooled to –78 °C under a N2 atmosphere. The reaction was

stirred for 3 h, quenched via the addition of saturated aq NH4Cl (150 mL), and Et2O (100 mL)

was added. The layers were separated, the organic phase washed with saturated aq NaHCO3

(100 mL), water (100 mL), and saturated aq NaCl (100 mL), dried (Na2SO4), and filtered.

Solvent removal and recrystallization from hexanes afforded 2.18 (0.564 g, 21%) as a light

yellow solid. Mp 210 °C. Rf = 0.52 (hexanes/EtOAc 20:1). IR 3068 (vw), 2955 (s), 2903 (m),

2867 (m), 2177 (w), 1593 (m) cm−1; 1H NMR (300 MHz, CDCl3) δ 7.32 (t, J = 1.7 Hz, 4H),

7.26 (d, J = 1.7 Hz, 8H), 3.41 (s, 6H), 1.27 (s, 72H); 13C NMR (75 MHz, CDCl3) δ 150.4,

140.9, 121.8, 121.4, 82.9, 79.5, 73.4, 63.6, 53.3, 34.9, 31.4. ESI MS m/z 883.7 ([M – OMe]+,

100); ESI HRMS m/z calcd. for C66H90NaO2 ([M + Na]+) 937.6833 found 937.6840.

1,1,8,8-Tetrakis(3,5-di-t-butylphenyl)octa-2,4,6-triyne-1,8-diol 2.22. To a solution of 1,6-

bis(trimethylsilyl)hexa-1,3,5-triyne 2.19 (78 mg, 0.36 mmol) in THF (12 mL) was added

MeLi·LiBr complex (1.5 M in Et2O, 0.25 mL, 0.38 mmol) at –20 °C under an Ar atmosphere

via a syringe. The reaction mixture was stirred for 1 h, and a solution of 2.2 (0.16 g, 0.39

mmol) in THF (5 mL) was added. After stirring for 1 h, the cooling bath was removed, and

the reaction was stirred for 2 h at rt. The solution was quenched via the addition of saturated

aq NH4Cl (30 mL), and Et2O (30 mL) was added. The layers were separated, the organic

phase was dried (Na2SO4), and filtered. Solvent removal and purification by column

78

chromatography (silica gel, hexanes/EtOAc 20:1) afforded pure 2.22 (65 mg, 20%) as a

colorless solid. Mp 224–226 °C. Rf = 0.08 (hexanes/EtOAc 20:1). IR 3588 (w), 3440 (bw),

3070 (vw), 2955 (s), 2903 (m), 2866 (m), 1592 (m) cm−1; 1H NMR (300 MHz, CDCl3) δ 7.39

(d, J = 1.7 Hz, 8H), 7.37 (t, J = 1.7 Hz, 4H), 2.87 (s, 2H), 1.31 (s, 72H); 13C NMR (75 MHz,

CDCl3) δ 150.7, 142.7, 122.0, 120.6, 81.7, 76.0, 71.3, 64.1, 35.0, 31.4. ESI HRMS m/z calcd.

for C64H85O ([M – OH]+) 869.6595 found 869.6580.

1,1,6,6-Tetrakis(3,5-di-t-butyl-phenyl)octa-1,2,3,4,5,6,7-heptaene ([7]tBuPh). To a

solution of 2.18 (0.125 g, 0.137 mmol) in Et2O (16 mL) was added anhydrous SnCl2 (78 mg,

0.41 mmol) and HCl (1 M in Et2O, 0.55 mL, 0.55 mmol) at 0 °C under an Ar atmosphere.

After 1 h, the solution was filtered through a plug of basic alumina oxide and eluted with

CH2Cl2 affording the purified [7]tBuPh. Since the cumulene is not stable as amorphous solid,

crystalline [7]tBuPh was obtained as violet/green needles (51 mg, 44%) by overlaying a

CH2Cl2 solution with MeOH. Mp 160–162 °C (decolorization). Rf = 0.61 (hexanes/EtOAc

20:1). UV/vis (CHCl3) λmax (ε) 297 (49800), 315 (66200), 459 (23300), 546 (40200), 573

(65200) nm. UV/vis (Et2O) λmax 294, 313, 453, 542, 565 nm; IR 3066 (vw), 2957 (s), 2904

(s), 2865 (m), 2055 (w) 1586 (m) cm−1; 1H NMR (400 MHz, CDCl3) δ 7.42 (d, J = 1.6 Hz,

8H), 7.40 (t, J = 1.6 Hz, 4H), 1.32 (s, 72H); 13C NMR (100 MHz, CDCl3) δ 150.8, 146.9,

137.2, 126.9, 124.9, 124.1, 123.1, 122.4, 35.0, 31.5. ESI HRMS m/z calcd. for C64H84 (M+)

852.6568 found 852.6535.

Crystal data for [7]tBuPh: C64H84, M = 853.31, triclinic crystal system; space group P–1, a =

13.8993(8), b = 14.0696(9), c = 15.1248(10) Å, α = 81.231(5)°, β = 89.873(5)°, γ =

78.628(5)°, V = 2864.6(3) Å3, Z = 2, ρcalcd = 0.989 g cm–3; µ(CuKα) = 0.406 mm–1, λ =

1.5418 Å; 173.0(8) K; 2θ max = 141.82°; total data collected = 17423; R1 = 0.0698 [7006


79

reflections, and 5 restraints; residual electron density = 0.35 and –0.30 e Å–3. Two t-butyl

group showed disorder, which was resolved and refined to the following occupation factor:

C22,23,24/C22a,23a,24a = 87:13%; C82,83,84/C82a,83a,84a = 61:39%. CCDC 903381.

1,1-Bis(3,5-di-t-butyl-phenyl)-5-(trimethylsilyl)penta-2,4-diyn-1-ol 2.23. To a solution of

1,4-bis(trimethylsilyl)buta-1,3-diyne (2.00 g, 10.3 mmol) in THF (20 mL) was added

MeLi·LiBr complex (1.5 M in Et2O, 7.0 mL, 11 mmol) at 0 °C under a N2 atmosphere via a

syringe. The cooling bath was removed, and the red-brown mixture was stirred for 0.5 h

before it was cooled again to 0 °C. A solution of 2.2 (4.18 g, 10.3 mmol) in THF (40 mL) was

added. The cooling bath was removed, and the reaction was stirred overnight. The solution

was quenched via the addition of saturated aq NH4Cl (70 mL), and Et2O (50 mL) was added.

The layers were separated, the organic phase was dried (Na2SO4), and filtered. Solvent

removal and purification by column chromatography (silica gel, hexanes/EtOAc 20:1)

afforded pure 2.23 (3.87 g, 71%) as a light yellow solid. Mp 114–116 °C. Rf = 0.35

(hexanes/EtOAc 20:1). IR 3547 (w), 3069 (vw), 2956 (m), 2901 (w), 2867 (w), 2218 (w),

2094 (w), 1595 (m), 1248 (m), 845 (s) cm−1; 1H NMR (300 MHz, CDCl3) δ 7.37 (d, J = 1.7

Hz, 4H), 7.33 (t, J = 1.7 Hz, 2H), 2.78 (s, 1H), 1.29 (s, 36H), 0.20 (s, 9H); 13C NMR (75

MHz, CDCl3) δ 150.5, 143.0, 121.8, 120.6, 88.6, 87.5, 80.6, 75.9, 71.4, 34.9, 31.4, –0.4. ESI

MS m/z 511.4 ([M – OH]+, 100); ESI HRMS m/z calcd. for C36H52NaOSi ([M + Na]+)

551.36796, found 551.36687.

80

1,1-Bis(3,5-di-t-butyl-phenyl)penta-2,4-diyn-1-ol 2.24. To a solution of 2.23 (3.35 g,

6.34 mmol) in MeOH (40 mL) was added K2CO3 (1.0 g, 7.2 mmol). The reaction mixture was

stirred overnight and quenched via the addition of saturated aq NH4Cl (40 mL). The layers

were separated, and the aqueous phase was extracted with CH2Cl2 (2 x 20 mL). The organic

phases were combined and washed with saturated aq NH4Cl (30 mL) and saturated aq NaCl

(30 mL), dried (Na2SO4), and filtered. Solvent removal afforded 2.24 (2.73 g, 94%) as an off-

white solid. Mp 120−124 °C. Rf = 0.40 (hexanes/EtOAc 10:1). IR 3591 (m), 3437 (br), 3276

(m), 3068 (vw), 2954 (s), 2902 (m), 2864 (m), 2057 (vw), 2012 (vw), 1597 (m), 1360 (m),

1247 (m), 879 (s), 628 (s) cm−1; 1H NMR (300 MHz, CDCl3) δ 7.38 (d, J = 1.7 Hz, 4H), 7.34

(t, J = 1.7 Hz, 2H), 2.79 (s, 1H), 2.26 (s, 1H), 1.28 (s, 36H); 13C NMR (75 MHz, CDCl3) δ

150.6, 142.8, 121.9, 120.6, 79.2, 75.8, 70.6, 69.4, 67.6, 35.0, 31.4. ESI MS m/z 479.3

([M + Na]+, 62), 439 ([M – OH]+, 100). ESI HRMS m/z calcd. for C33H44NaO ([M + Na]+)

479.32844, found 479.32740.

1,1,10,10-Tetrakis(3,5-di-t-butyl-phenyl)deca-2,4,6,8-tetrayne-1,10-diol 2.25. To a

solution of 2.24 (2.73 g, 5.98 mmol) in CH2Cl2 (30 mL) was added a solution of Hay catalyst

[CuCl (0.600 g, 5.98 mmol) and TMEDA (1.4 g, 1.8 mL, 12 mmol) in CH2Cl2 (20 mL)]. The

reaction mixture was stirred for 1 d, saturated aq NH4Cl (50 mL) was added, and the mixture

was extracted with Et2O (50 mL). The organic phase was washed with saturated aq NH4Cl

(100 mL) and saturated aq NaCl (100 mL), dried (Na2SO4), and filtered. Solvent removal and

81

twofold recrystallization from hexanes afforded 2.25 (2.62 g, 96%) as a light yellow solid. Mp

185 °C. Rf = 0.36 (hexanes/EtOAc 10:1). IR 3588 (vw), 3447 (br), 3070 (vw), 2956 (s), 2904

(m), 2866 (m), 2217 (w), 1595 (m), 715 (s) cm−1; 1H NMR (400 MHz, CDCl3) δ 7.34 (s, 12

H), 2.81 (s, 2H), 1.27 (s, 72H); 13C NMR (100 MHz, CDCl3) δ 150.8, 142.5, 122.1, 120.6,

81.4, 76.1, 71.2, 65.0, 62.3, 35.0, 31.4. ESI MS m/z 893.7 ([M – OH]+, 100); ESI HRMS m/z

calcd. for C66H86NaO2 ([M + Na]+) 933.6520, found 933.6517.

1,1,6,6-Tetrakis(3,5-di-t-butyl-phenyl)deca-1,2,3,4,5,6,7,8,9-nonaene ([9]tBuPh). To a

solution of 2.25 (34 mg, 0.037 mmol) in Et2O (4 mL) was added anhydrous SnCl2 (22 mg,

0.12 mmol) and HCl (1 M in Et2O, 0.15 mL, 0.15 mmol) at 0 °C under an Ar atmosphere.

After 30−60 min, the solution was filtered through a plug of basic alumina oxide and eluted

with Et2O to afford the purified [9]tBuPh as a blue solution in Et2O. The yield could not be

determined due to instability of this compound. Rf = 0.62 (hexanes/EtOAc 20:1). UV/vis

(Et2O) λmax 316, 338, 366, 528, 574, 610, 664 nm.

5,5-Bis(3,5-di-t-butylphenyl)-5-methoxy-1-(trimethylsilyl)penta-1,3-diyne 2.27. To a

solution of 1,4-bis(trimethylsilyl)buta-1,3-diyne 2.26 (1.46 g, 7.51 mmol) in THF (30 mL)

was added MeLi·LiBr complex (1.5 M in Et2O, 5.0 mL, 7.5 mmol) at 0 °C under an Ar

atmosphere via a syringe. The cooling bath was removed, and the red-brown mixture was

82

stirred for 0.5 h before it was cooled again to 0 °C. A solution of 2.2 (3.05 g, 7.50 mmol) in

THF (30 mL) was added. After stirring for 20 min, the cooling bath was removed, and the

reaction mixture was stirred overnight. Saturated aq NH4Cl (70 mL) and Et2O (70 mL) were

added. The layers were separated, the organic phase was washed with saturated aq NH4Cl (70

mL) and saturated aq NaCl (70 mL), dried (Na2SO4), and filtered. Solvent removal afforded

the crude product 2.27 as a yellow solid (3.94 g, 97%) that could not be purified. It was then

carried on to the next step as formed.

5,5-Bis(3,5-di-t-butylphenyl)-5-methoxypenta-1,3-diyne 2.12. To a solution of 2.27 (3.94 g,

7.26 mmol) in MeOH/THF (4:1, 100 mL) was added K2CO3 (1.0 g, 7.2 mmol). The reaction

mixture was stirred for 2 h, and saturated aq NH4Cl (100 mL) was added. The layers were

separated, and the aqueous phase was extracted with CH2Cl2 (2 x 50 mL). The organic phases

were combined and washed with saturated aq NH4Cl (50 mL) and saturated aq NaCl (50 mL),


(silica gel, hexanes/EtOAc 30:1) afforded pure 2.12 (2.35 g, 66% over two steps based on 2.2)

as a colorless solid. Mp 110−113 °C. Rf = 0.42 (hexanes/CH2Cl2 4:1). IR 3260 (m), 3068

(vw), 2959 (s), 2904 (m), 2868 (m), 2826 (w), 2163 (w), 1592 (m), 1087 (s), 877 (m) cm−1; 1H NMR (300 MHz, CDCl3) δ 7.39 (t, J = 1.7 Hz, 2H), 7.36 (d, J = 1.7 Hz, 4H), 3.49 (s, 3H),


76.8, 72.7, 68.8, 67.7, 53.2, 34.9, 31.4. ESI HRMS m/z calcd. for C34H46NaO ([M + Na]+)

493.34409, found 493.34459.

MeO OMeMeO H

2.12

CuCl, TMEDA

CH2Cl2

2.14

83

1,1,10,10-Tetrakis(3,5-di-t-butylphenyl)-1,10-dimethoxydeca-2,4,6,8-tetrayne 2.14. To a

solution of 2.12 (1.02 g, 2.17 mmol) in CH2Cl2 (10 mL) was added a solution of Hay catalyst

[CuCl (0.22 g, 2.2 mmol) and TMEDA (0.50 g, 0.65 mL, 4.3 mmol) in CH2Cl2 (5 mL)]. The

reaction mixture was stirred for 1 h, saturated aq NH4Cl (40 mL) was added, and the mixture

was extracted with Et2O (50 mL). The organic phase was washed with saturated aq NH4Cl


afforded 2.14 (0.89 g, 87%) as an off-white solid. Mp 78–80 °C. Rf = 0.75 (hexanes/CH2Cl2

1:1). IR 3069 (vw), 2955 (s), 2903 (m), 2867 (m), 2825 (v), 2215 (w), 1594 (m) cm−1; 1H

NMR (400 MHz, CDCl3) δ 7.31 (t, J = 1.8 Hz, 4H), 7.23 (d, J = 1.8 Hz, 8H), 3.39 (s, 6H),

1.25 (s, 72H); 13C NMR (101 MHz, CDCl3) δ 150.5, 140.8, 121.9, 121.4, 82.9, 79.4, 73.3,

64.5, 62.2, 53.3, 34.9, 31.4. APPI HRMS m/z calcd. for C67H87O ([M – OMe]+) 907.67514,

found 907.67536.

1,1-Bis(3,5-di-t-butylphenyl)-7-(trimethylsilyl)hepta-2,4,6-triyn-1-ol S1. To a solution of

1,6-bis(trimethylsilyl)hexa-1,3,5-triyne 2.19 (0.10 g, 0.46 mmol) in THF (30 mL) was added

MeLi·LiBr complex (1.5 M in Et2O, 0.30 mL, 0.45 mmol) at –78 °C under an Ar atmosphere

via a syringe. The reaction mixture was stirred for 1 h at –20 °C, and a solution of 2.2 (0.21 g,

0.52 mmol) in THF (20 mL) was added. The reaction mixture was stirred for 1.5 h, and

saturated aq NH4Cl (70 mL) and Et2O (50 mL) were added. The layers were separated, the



(silica gel, hexanes/CH2Cl2 1:1) afforded pure S1 (0.104 g, 41%) as a red oil. Rf = 0.38

(hexanes/ CH2Cl2 1:1). 1H NMR (300 MHz, CDCl3) δ 7.36–7.34 (m, 6H), 2.80 (s, 1H), 1.27

(s, 36H), 0.20 (s, 9H); 13C NMR (100 MHz, CDCl3) δ 150.7, 142.7, 122.0, 120.6, 87.9, 80.9,

76.1, 71.3, 64.6, 61.2, 35.0, 31.4, –0.6 (one signal coincident or not observed). MALDI MS

84

m/z 535 ([M − OH]+, 100); ESI HRMS m/z calcd. for C38H52NaOSi ([M + Na]+) 575.3680,

found 575.3677.

1,1-Bis(3,5-di-t-butylphenyl)hepta-2,4,6-triyn-1-ol 2.21. To a solution of 1,6-

bis(trimethylsilyl)hexa-1,3,5-triyne 2.19 (0.10 g, 0.46 mmol) in THF (30 mL) was added

MeLi·LiBr complex (1.5 M in Et2O, 0.30 mL, 0.45 mmol) at –78 °C under a N2 atmosphere

via a syringe. The reaction mixture was stirred for 1 h at –20 °C, and a solution of 2.2 (0.21 g,

0.52 mmol) in THF (20 mL) was added. The reaction mixture was stirred for 1.5 h and

saturated aq NH4Cl (70 mL) and Et2O (50 mL) were added. The layers were separated, the



(silica gel, hexanes/CH2Cl2 1:1) afforded pure 2.21 (0.029 g, 13%) as a brown solid. Mp >115

°C (decomp.). Rf = 0.31 (hexanes/ CH2Cl2 1:1). IR 3587 (w), 3456 (br), 3230 (m), 3069 (vw),

2957 (s), 2903 (m), 2865 (m), 2038 (vw), 1595 (m) cm−1; 1H NMR (300 MHz, CDCl3) δ 7.35

(s, 6H), 2.80 (s, 1H), 2.16 (s, 1H), 1.27 (s, 36H); 13C NMR (100 MHz, CDCl3) δ 150.8, 142.6,

122.0, 120.6, 80.2, 76.0, 71.0, 68.4, 68.0, 64.3, 60.0, 35.0, 31.4. ESI HRMS m/z calcd. for

C35H44NaO ([M + Na]+) 503.3284, found 503.3288.

Improved synthesis of 2.21:

To a solution of 1,6-bis(trimethylsilyl)hexa-1,3,5-triyne 2.9 (0.655 g, 3.00 mmol) in THF

(100 mL) was added MeLi·LiBr complex (1.5 M in Et2O, 2.0 mL, 3.0 mmol) at –20 °C under

an Ar atmosphere via a syringe. The reaction mixture was stirred for 1.5 h at –20 °C, and a

solution of 2.2 (1.28 g, 3.15 mmol) in THF (20 mL) was added. The reaction mixture was

stirred for 1 h at –20 °C and for further 2 h at rt. Saturated aq NH4Cl (100 mL) and Et2O (100

85

mL) were added. The layers were separated, the organic phase was washed with saturated aq

NH4Cl (100 mL) and saturated aq NaCl (100 mL), dried (Na2SO4), and filtered. Solvent

removal and purification by column chromatography (silica gel, hexanes/EtOAc 20:1)

afforded pure 2.21 (0.689 g, 48%) as an off-white solid.

1,1,14,14-Tetrakis(3,5-di-t-butylphenyl)tetradeca-2,4,6,8,10,12-hexayne-1,14-diol 2.31.

To a solution of 2.21 (0.029 g, 0.060 mmol) in CH2Cl2 (2 mL) was added a solution of Hay

catalyst [CuCl (0.20 g, 2.0 mmol) and TMEDA (0.046 g, 0.060 mL, 0.40 mmol) in CH2Cl2

(1 mL)]. The reaction mixture was stirred for 1.5 h, saturated aq NH4Cl (10 mL) was added,

and the mixture was extracted with Et2O (10 mL). The organic phase was washed with

saturated aq NH4Cl (20 mL) and saturated aq NaCl (20 mL), dried (Na2SO4), and filtered.

Solvent removal and purification by column chromatography (silica gel, hexanes/CH2Cl2 1:1)

afforded pure 2.31 (0.009 g, 31%) as a brown solid. Mp >85 °C (darkening), >95 °C (viscous

black). Rf = 0.22 (hexanes/ CH2Cl2 1:1). IR 3426 (br), 3071 (vw), 2955 (s), 2904 (m), 2866

(m), 2168 (m), 2074 (vw), 1594 (m) cm−1; 1H NMR (400 MHz, CDCl3) δ 7.34 (t, J = 1.8 Hz,

4H), 7.32 (d, J = 1.8 Hz, 8H), 2.82 (s, 2H), 1.27 (s, 72H); 13C NMR (100 MHz, CDCl3) δ

150.8, 142.3, 122.2, 120.6, 81.6, 76.1, 71.1, 65.1, 63.4, 62.9, 61.9, 35.0, 31.4. ESI HRMS m/z

calcd. for C70H86NaO2 ([M + Na]+) 981.6520, found 981.6536.

1,1,6,6-Tetramesitylhexa-2,4-diyne-1,6-diol 2.49. To a solution of 2.42 (50 mg, 0.17 mmol)

in CH2Cl2 (3 mL) was added a solution of Hay catalyst [CuCl (16.9 mg, 0.17 mmol) and

86

TMEDA (0.04 g, 0.05 mL, 0.3 mmol) in CH2Cl2 (1 mL)]. The reaction mixture was stirred for

4 d. Saturated aq NH4Cl (15 mL) and Et2O (15 mL) were added. The organic phase was

washed with saturated aq NH4Cl (30 mL), saturated aq NaCl (30 mL), water (30 mL), dried

(Na2SO4), and filtered. Solvent removal and purification by column chromatography (silica

gel, hexanes/CH2Cl2 4:1) afforded 2.49 as a yellow solid that could not be further purified.

NMR spectra of unpurified material have been included in the Appendix Section.

HO OHHHOPd(PPh3)2Cl2, CuI, Et3N

ethyl bromoacetate, THF

2.42

2.49

1,1,6,6-Tetramesitylhexa-2,4-diyne-1,6-diol 2.49. To a solution of Pd(PPh3)2Cl2 (7.2 mg,

0.010 mmol), CuI (3.9 mg, 0.020 mmol), and Et3N (60 µL) in THF (3 mL) was added 2.42

(50 mg, 0.17 mmol) in THF (3 mL). Finally, ethyl bromoacetate (0.023 g, 0.015 mL, 0.14

mmol) was added, and the reaction mixture was stirred overnight. Saturated aq NH4Cl (15

mL) and Et2O (15 mL) were added. The organic phase was washed with saturated aq NH4Cl

(20 mL), saturated aq NaCl (20 mL), dried (Na2SO4), and filtered. Solvent removal,

purification by column chromatography (silica gel, hexanes/EtOAc 20:1), and further

recrystallization from hexanes afforded 2.49 (6 mg, 12%) as a light brown solid. Rf = 0.21

(hexanes/CH2Cl2 1:4). 1H NMR (400 MHz, CDCl3) δ 6.71 (s, 8H), 2.43 (s, 2H), 2.20 (s, 12H),

2.19 (s, 24H); 13C NMR (100 MHz, CDCl3) δ 139.3, 136.6, 136.1, 131.6, 84.4, 74.7, 23.2,

20.5.

87

1,1,6,6-Tetramesitylhexa-2,4-diyne-1,6-diol 2.49. To a solution of Pd(PPh3)2Cl2 (1.6 mg,

0.0023 mmol), CuI (1.6 mg, 0.0084 mmol), I2 (22 mg, 0.087 mmol), and 2.42 (50 mg,

0.17 mmol) were added i-Pr2NH (2 mL) and THF (2 mL) under a N2 atmosphere. The

reaction mixture was stirred overnight, and saturated aq NH4Cl (10 mL) and Et2O (10 mL)

were added. The organic phase was washed with saturated aq NH4Cl (15 mL), saturated aq

Na2SO3 (15 mL), saturated aq NaCl (15 mL), dried (Na2SO4), and filtered. Solvent removal,

filtration through a plug of silica gel with hexanes/CH2Cl2 = 1:1, and recrystallization from

hexanes afforded 2.49 (13 mg, 26%) as a brownish solid. Spectral data are consistent with that

described above.

(2-Bromo-1,1’-dimesityl)allene 2.44. To a solution of 2.42 (0.298 g, 1.01 mmol) in THF

(5 mL) was added EtMgBr (1 M in THF, 1 mL, 1 mmol) at rt under a N2 atmosphere via a

syringe. The reaction mixture was stirred for 20 min, and ethyl formate (0.039 g, 42 µl, 0.52

mmol) was added. The cooling bath was removed, and the reaction was stirred for 30 min.

Saturated aq NH4Cl (10 mL) and Et2O (10 mL) were added. The layers were separated, the

organic phase washed with brine (10 mL), dried over Na2SO4, and filtered. The solvent was

removed in vacuo to yield the crude product as a viscous oil which was purified by column

chromatography (hexanes/ethyl acetate 20:1) to afford 2.44 (0.05 g, 14%) as a light yellow

solid. Rf = 0.57 (hexanes/EtOAc 20:1). IR 3033 (w), 2956 (m), 2915 (m), 2855 (m), 2730 (w),

1931 (m), 1607 (m), 1442 (s) cm−1; 1H NMR (400 MHz, CDCl3) δ 6.85 (s, 4H), 6.11 (s, 1H),


112.4, 72.9, 21.2, 20.9.

Crystal data for allene 2.44: C21H23Br, M = 355.30, monoclinic crystal system; space group

P21/n, a = 8.7221(2), b = 8.8492(4), c = 22.8496(8) Å, β = 95.017(2)°, V = 1756.86(11) Å3, Z

= 4, ρcalcd = 1.343 g cm–3; µ(MoKα) = 2.335 mm–1, λ = 0.71073 Å; 173.15 K; 2θ max =

54.94°; total data collected = 5797; R1 = 0.0396 [2905 observed reflections with F ≥ 4σ(F)];

88

wR2 = 0.1143 for 205 variables, 4038 unique reflections, and 0 restraints; residual electron

density = 0.342 and –0.635 e Å–3.

1,1-Dimesitylhex-2-yne-1,4-diol 2.45. To a solution of 2.42 (0.10 g, 0.34 mmol) in THF

(5 mL) was added EtMgBr (1 M in THF, 0.7 mL, 0.7 mmol) at rt under a N2 atmosphere via a

syringe. The reaction mixture was stirred for 30 min, and ethyl formate (0.013 g, 14 µl, 0.18

mmol) was added. The cooling bath was removed, and the reaction was stirred for 40 min.

Saturated aq NH4Cl (10 mL) and Et2O (10 mL) were added. The layers were separated, the

organic phase washed with brine (10 mL), dried over Na2SO4, and filtered. Solvent removal

and recrystallization from hexanes afforded 2.45 (0.013 g, 11%) as an off-white solid. Rf =

0.10 (hexanes/EtOAc 20:1). 1H NMR (300 MHz, CDCl3) δ 6.70 (s, 4H), 4.32 (br s, 1H), 2.47

(s, 1H), 2.21 (s, 12H), 2.20 (s, 6H), 1.72–1.63 (m, 2H), 0.91 (t, J = 7.4 Hz, 3H); 13C NMR (75

MHz, CDCl3) δ 140.22, 140.18, 136.25, 136.23, 135.92, 135.89, 131.6, 131.5, 90.0, 89.1,

76.5, 64.1, 30.4, 23.2, 20.5, 9.4.

2.4 References

1 Early work on cumulenes, see: a) H. Hopf, Classics in Hydrocarbon Chemistry,

Wiley-VCH, Weinheim, 2000, Chapter 9; b) H. Fischer, in The chemistry of alkenes

(Ed.: S. Patai), John Wiley & Sons, New York, 1964, pp. 1025–1159.

2 For two notable exceptions, see: a) Y. Kuwatani, G. Yamamoto, M. Oda, M. Iyoda,

Bull. Chem. Soc. Jpn. 2005, 78, 2188–2208; b) W. Skibar, H. Kopacka, K. Wurst, C.

Salzmann, K.-H. Ongania, F. F. de Biani, P. Zanello, B. Bildstein, Organometallics

2004, 23, 1024–1041.

3 R. Kuhn, H. Krauch, Chem. Ber. 1955, 88, 309–315.

89

4 [n]Ph: a) R. Kuhn, K. Wallenfels, Chem. Ber. 1938, 71, 783–790; b) R. Kuhn, H.

Zahn, Chem. Ber. 1951, 84, 566–570.

5 [n]Ph: F. Bohlmann, K. Kieslich, Abh. Braunsch. Wiss. Ges. 1957, 9, 147–160.

6 [n]Cy: F. Bohlmann, K. Kieslich, Chem. Ber. 1954, 87, 1363–1372.

7 The electronic structure of [n]cumulenes is fundamentally different between two

classes of molecules, i.e., when n is even (n = 2, 4, 6...) or odd (n = 3, 5, 7...). Only for

n = odd, π-conjugation between the endgroups via the cumulene framework is

possible. Only odd cumulenes are considered for the discussion in this chapter.

8 F. Innocenti, A. Milani, C. Castiglioni, J. Raman Spectrosc. 2010, 41, 226–236.


10 U. Mölder, P. Burk, I. A. Koppel, J. Mol. Struct. THEOCHEM 2004, 712, 81–89.

11 R. Hoffmann, Angew. Chem. Int. Ed. 1987, 26, 846–878; Angew. Chem. 1987, 99,

871–906.

12 Based on a search of WebCSD, see http://webcsd.ccdc.cam.ac.uk/ on 11/07/14, for

equally substituted [n]cumulenes with n = odd and alkyl or aryl endgroups. X-ray

crystallographic structures obtained in this thesis or in publications of the Tykwinski

group have not been considered.


14 Diploma thesis “Carbon in One Dimension – Synthesis of [n]Cumulenes”, Johanna A.

Januszewski, Friedrich-Alexander-Universität Erlangen-Nürnberg, May 2010.

15 E. J. Corey, P. L. Fuchs, Tetrahedron Lett. 1972, 36, 3769–3772.

16 S. Eisler, R. R. Tykwinski, in Acetylene Chemistry: Chemistry, Biology, and Material


2005, chapter 7.

17 W. A. Chalifoux, R. R. Tykwinski, C. R. Chimie 2009, 12, 341–358.

18 W. A. Chalifoux, R. R. Tykwinski, Chem. Rec. 2006, 6, 169–182.

19 H. Hauptmann, M. Mader, Synthesis 1978, 307–309.

20 J. Anthony, A. M. Boldi, Y. Rubin, M. Hobi, V. Gramlich, C. B. Knobler, P. Seiler, F.

Diederich, Helv. Chim. Acta 1995, 78, 13–45.

21 S. Eisler, N. Chahal, R. McDonald, R. R. Tykwinski, Chem. Eur. J. 2003, 9, 2542–

2550.

22 J. Hlavatý, L. Kavan, K. Okabe, A. Oya, Carbon 2002, 40, 1131–1150.

90

23 This reaction has been studied more intensely, and the resulting outcome is presented

in Section 5.3.1.

24 D. R. M. Walton, F. Waugh, J. Organomet. Chem. 1972, 37, 45–56.

25 A. L. K. Shi Shun, E. T. Chernick, S. Eisler, R. R. Tykwinski, J. Org. Chem. 2003, 68,

1339–1347.

26 T. Luu, E. Elliott, A. D. Slepkov, S. Eisler, R. McDonald, F. A. Hegmann, R. R.

Tykwinski, Org. Lett. 2005, 7, 51–54.

27 The synthesis is adapted from the work of Dominik Wendinger, see also J. A.

Januszewski, D. Wendinger, C. D. Methfessel, F. Hampel, R. R. Tykwinski, Angew.

Chem. Int. Ed. 2013, 52, 1817–1821; Angew. Chem. 2013, 125, 1862–1867.

28 J. P. Mario, H. N. Nguyen, J. Org. Chem. 2002, 67, 6841–6844.

29 L. Schmiech, C. Alayrac, B. Witulski, T. Hofmann, J. Agric. Food Chem. 2009, 57,

11030–11040.

30 T. Lee, H. R. Kang, S. Kim, S. Kim, Tetrahedron 2006, 62, 4081–4085.

31 The synthesis of acetylene 2.42 is presented in Section 2.1.3.4.

32 A. Lei, M. Srivastava, X. Zhang, J. Org. Chem. 2002, 67, 1969–1971.

33 Q. Liu, D. J. Burton, Tetrahedron Lett. 1997, 38, 4371–4374.


Angew. Chem. Int. Ed. 2013, 52, 1817–1821.

35 Several compounds that have been synthesized during my diploma thesis also appear

in the experimental part of this thesis since the procedures have been repeated several

times under different conditions in order to improve purity and/or yield of the product.

36 a) K. Yamada, Y. Matsumoto, K. B. Selim, Y. Yamamoto, K. Tomioka, Tetrahedron

2012, 68, 4159–4165. b) K. Schreiner, H. Oehling, H. E. Zieger, I. Angres, J. Am.

Chem. Soc. 1977, 99, 2638–2641.

37 a) S. R. Ditto, R. J. Card, P. D. Davis, D. C. Neckers, J. Org. Chem. 1979, 44,

894−896. b) P. D. Bartlett, M. Roha, R. M. Stiles, J. Am. Chem. Soc. 1954, 76, 2349–

2353.

38 A. Ceccon, C. Corvaja, G. Giacometti, A. Venzo, J. Chem. Soc., Perkin Trans. 2

1978, 283–288.

91

3. Chapter III. Cumulene rotaxanes – Synthesis and stability of [n]tBuPh

rotaxanes

3.1 General introduction to rotaxanes

Rotaxanes are compounds that combine two different molecular components. A linear

molecule containing two endgroups, i.e., a dumbbell-shaped molecule, is one part of a

rotaxane and is often called the axle. The axle is surrounded by at least one macrocycle

defining the second component of the rotaxane (Figure 3.1).1 The large endgroups on the

dumbbell, also called stoppers, have to be sufficiently bulky to prevent the loss of the

macrocycle from its position around the linear molecule, the axle. These stoppers are

necessary, since the macrocyle is not bound covalently to the chain, but only via non-covalent

interactions.2 Pseudorotaxanes represents a special subclass of rotaxanes which do not possess

stoppers, and the axle and macrocycle are kept together only by non-covalent interactions.

Rotaxane nomenclature is based on the number of components, which is put in brackets as a

prefix, i.e., one dumbbell-shaped molecule and one macrocycle unit give a [2]rotaxane. In this

thesis, only [2]rotaxanes will be discussed.

Figure 3.1 Definition of a [2]rotaxane.

92

The interesting dynamics of rotaxanes offer potential applications in molecular

electronics, e.g., as sensors, switches, and molecular machines or rotary motors.3,4 Aside from

the above mentioned usage of these interlocked molecules, rotaxanes can also act as

“insulated molecular wires” which have been reviewed by Anderson and Frampton.3 Besides,

rotaxane formation serves for stabilization of the axle compound.5–7 The current chapter

describes the synthesis of rotaxanes with a cumulene as the axle component targeting

stabilization enhancement of the cumulene chain.

Rotaxane formation can be accomplished via several conventional methods including

capping, clipping, and slipping that belong to the passive strategies for template formation of

rotaxanes (Figure 3.2).1 In the case of capping (Figure 3.2a), first the axle and the macrocycle

are connected together via non-covalent interactions forming a pseudorotaxane, which is

finally converted to the rotaxane by a reaction with the stopper molecules. In contrast, the

clipping method (Figure 3.2b) uses an already formed dumbbell-shaped molecule, and the

final addition of an incomplete “open” macrocycle gives the rotaxane via a ring closing

reaction. The slipping method (Figure 3.2c) is based on the kinetic stability of the rotaxane.

Herein, higher temperatures are used to enable the slipping of the macrocycle onto the

dumbbell-shaped chain, where it gets “trapped” by decreasing the temperature.

Figure 3.2 Three common methods for rotaxane formation: a) capping, b) clipping, and c)

slipping.

In contrast to the passive methods described in Figure 3.2, there is an active template

synthesis of rotaxanes (Figure 3.3). This is the most attractive method for rotaxane formation

and has been applied in this thesis. In this active template method, two “half-dumbbell-

shaped” units (e.g., acetylene units containing a bulky endgroup) and the macrocycle are

93

connected via a metal ion to form the rotaxane. This method has been reported by Leigh and

coworkers for the first time8–11 utilizing a Cu(I)-catalyzed reaction. Specifically, a 1,3-

cycloaddition of azides with terminal alkynes, i.e., a CuAAC- or “click” reaction has been

carried out to assemble a [2]rotaxane.8 The idea behind this strategy is that the metal ion has a

dual function acting on one hand as a template for the rotaxane formation and on the other

hand as catalyst that facilitates the bond formation of two half-dumbbell-shaped units. In one

step, the thread is generated, while at the same time the inclusion of the macrocycle via

coordination occurs.

Figure 3.3 Active template method for rotaxane formation.

3.2 Polyyne rotaxanes as motivation for cumulene rotaxane formation

Long polyynes are not particularly stable, upon reaching a certain number of acetylene

units in the chain, polyynes can often only be handled in a dilute solution.12 Recently, several

methods have been developed to increase the stability of longer polyynes or to encapsulate the

polyyne chain. Complexation offers one opportunity for increased stabilization of polyynes.

For example, Taylor and Gabbai13 have described the complexation of the polyyne chain to

tridentate mercury-containing Lewis acids via weak supramolecular interactions.

Encapsulation of polyynes into carbon nanotubes can also shield the sp-hybridized carbon

chain, as demonstrated by Zhao et al.14 Briefly, fusion reactions of linear polyyne units in

double wall carbon nanotubes (DWCNTs) afforded long linear carbon chains. Another

encapsulation method has been described by Ben Shir et al.15 using guest-host complex

formation. Herein, the polyyne forms a molecular rotary motor via insertion of a cucurbituril

host around the carbon chain.

Introduction of macrocycles to form polyyne rotaxanes has been recently reported by

several groups2,4,16 presumably to determine if this strategy can be used for stability

enhancement of a sp-hybridized carbon chain. It is important to also consider that the

macrocycle in a rotaxane does interact strongly with the linear component, i.e., the polyyne

94

chain, and thus the investigation of structural and physical properties should not be hampered

by rotaxane formation. In 2012, Gladysz17 has demonstrated a polyyne rotaxane formation for

a tetrayne with a dimetallic polyynediyl axle and a phenanthroline based macrocycle (Figure

3.4a). Anderson and Tykwinski18 have constructed even longer polyyne rotaxanes with four,

six, and ten acetylene units (Figure 3.4b) using the same phenanthroline-based macrocycle as

Gladysz has done, but with supertrityl endcapped polyynes.

Figure 3.4 Polyyne rotaxanes reported by a) Gladysz17 and b) Anderson/Tykwinski.18

The synthesis of the known polyyne rotaxanes was based on the active metal template

approach and included a complexation as well as a homocoupling reaction. Initially, the

copper-macrocycle complex was formed by treatment of naked macrocycle with CuI in

CH2Cl2/CH3CN at rt.16 Afterwards, addition of the terminal acetylene, K2CO3, and an oxidant

95

in THF gave the desired polyyne rotaxane. Gladysz initially used iodine as oxidant and 55 °C

as reaction temperature resulting in low yields and high impurity, and he switched later to the

use of oxygen as oxidant and decreased the temperature to 50 °C, which gave the rotaxane in

9% yield (Scheme 3.1). Anderson and Tykwinski maintained iodine as oxidant and were able

to achieve rotaxanes in yields of 15–34% using polyynes with supertrityl endgroups (Scheme

3.1). The comparison of the yields for supertrityl endcapped [n]polyynes (34, 32, and 15% for

n = 4, 6, and 10, respectively) demonstrated that the increase of chain length correlated with

the decrease of reaction yields.

Scheme 3.1 Synthesis of polyyne rotaxanes.

3.3 Introduction to cumulene rotaxanes: Motivation and target

During the period of this doctoral thesis, higher [n]cumulenes with n > 5 were

synthesized by the incorporation of bulky endgroups for stabilization. This strategy had been

proven successful for polyyne formation but showed limitations in the synthesis of

cumulenes. Thus far, a [7]cumulene, [7]tBuPh, had been successfully synthesized that has

been infinitely stable as crystalline solid. With respect to the next higher analogue, [9]tBuPh,

the synthesis was successful as well, but the limit was reached at this stage for stabilization of

the cumulene series [n]tBuPh under normal laboratory conditions. Namely, [9]tBuPh was

only stable for several days when kept in Et2O solution at ca. –20 °C under water- and

oxygen-free conditions. Furthermore, it was not possible to obtain this compound as a

crystalline solid in order to probe its stability in the solid state. An alternative strategy to

96

enhance the stability of longer cumulenes had to be found. In the following section, the

formation of cumulene rotaxanes has been presented.

3.4 Synthesis of rotaxane precursors and the appropriate cumulene rotaxanes

([9]tBuPh rotaxanes) using three different macrocycles**

For the synthesis of cumulene rotaxanes, three different macrocycles, 3.1, 3.2, and 3.3,

have been chosen (Figure 3.5). All three macrocycles are based on a phenanthroline unit

which has been reliable for polyyne rotaxane formation as reported by several groups, such as

Saito,2,16 Gladysz,17 and Anderson.3,4,18 The macrocycles possess the same modified

phenanthroline backbone (dotted box in Figure 3.5) with variation of the macrocycle size due

to different linkers. Macrocycle 3.119 shows a relatively rigid structure using a diphenyl ether

group as linker. Compound 3.1 is the smallest homologue, followed by compound 3.220 with

an alkyl linker, i.e., a (CH2)10 unit. The last representative is macrocycle 3.32,16 which uses a

combination of a resorcinol aromatic ring linked via (CH2)6 fragments to the phenanthroline

backbone. Compound 3.3 represents the largest of the macrocycles.

Figure 3.5 Three macrocycles, compounds 3.1, 3.2, and 3.3, used in the synthesis of

cumulene rotaxanes.

** The synthesis of all macrocycles as well as the precursors to rotaxanes (oligoyne diethers) has been performed

by Michael Franz and Levon Movsisyan who are working on polyyne rotaxanes in the Tykwinski and Anderson

groups, respectively.

97

Regarding the synthesis of cumulene rotaxanes, the precursors, i.e., polyyne rotaxanes

3.4, 3.5, and 3.6 had to be synthesized (Scheme 3.2). Thus, using standard methods, two

equivalents of the terminal diyne 2.12 were subjected to rotaxination using K2CO3 as base, I2

as oxidant, and the macrocyclic copper complexes 3.1·CuI, 3.2·CuI, and 3.3·CuI,

respectively.21,22 Unfortunately, this synthetic route was only mildly successful, resulting in

unsatisfying yields of 5% and 6% using macrocycles 3.1 and 3.2, respectively. Macrocycle

3.3, however, showed the highest yield of 68% for the synthesis of 3.6 and thus, 3.6 seemed to

represent the most suitable macrocycle for rotaxane formation of [9]tBuPh cumulenes.

Scheme 3.2 Formation of rotaxanes 3.4, 3.5, and 3.6 via an active metal templated

homocoupling reaction.22

98

The yield of polyyne rotaxane 3.6 was further increased via an alternative pathway,

namely a Cadiot-Chodkiewicz heterocoupling (Scheme 3.3).22 Terminal diyne 2.12 and

bromodiyne 3.7 were reacted in the presence of K2CO3 and macrocyclic complex 3.3·CuI in

THF at 60 °C. Additional oxidant was not necessary, and the rotaxane 3.6 could be

synthesized with an increased yield of 74%. This reaction was also performed using complex

3.1·CuI giving 3.4 in a low yield of 5%.

Scheme 3.3 Formation of rotaxanes 3.4 and 3.6 via an active metal templated Cadiot-

Chodkiewicz heterocoupling reaction.22

With precursors 3.4 and 3.6 in hand, further conversion to the cumulene rotaxanes 3.8

and 3.9, respectively, was accomplished via reductive elimination (Scheme 3.4). The

precursors were dissolved in dry Et2O and treated with anhydrous SnCl2 and HCl

(1 M in Et2O) at rt under an argon atmosphere. After ca. 30 min, the cumulene rotaxane

formation was judged complete via TLC analysis, and the reaction mixture was filtered

through a plug of basic alumina. Conventional crystallization attempts by overlaying a

99

CH2Cl2 solution containing the cumulene rotaxanes with MeOH (at ca. –20 °C) afforded

crystalline precipitates of [9]cumulene rotaxanes 3.8 and 3.9 in 29% and 33% yield,

respectively. This was the first time that a 3,5-di-t-butylphenyl substituted [9]cumulene could

be isolated in the solid state. Cumulene rotaxane 3.10 was also successfully synthesized,

however, with much lower yields that appeared already in the formation of the appropriate

precursor 3.5 (Scheme 3.4). The limited amount of precursor 3.5 and cumulene rotaxane 3.10

prevented full characterization and thus, no further comparisons or discussions of its

properties were made.

Scheme 3.4 Synthesis of cumulene rotaxanes 3.8, 3.9, and 3.10.

100

3.5 Stability of [9]cumulene rotaxanes and comparison to [9]tBuPh

The stability of [9]cumulene rotaxane 3.9 was qualitatively compared to the “naked”

[9]cumulene [9]tBuPh by observing optical changes in e.g., color changes. These studies

were performed in Et2O, kept under an argon atmosphere at rt – exposed to light and kept in

the dark. Results have been summarized in Table 3.1. The results showed that a solution of

[9]tBuPh decolorized already after 3 h under ambient light or overnight when kept in the

dark, whereas for 3.9, decolorization was observed after one week under ambient light or after

several weeks when kept in dark. In addition, the color change of the TLC spots of [9]tBuPh

and the cumulene rotaxane 3.9 was compared. The color of the TLC spot changed from blue

to orange/yellow after seconds for [9]tBuPh or within 4 h for 3.9. Regarding [9]cumulenes in

the solid state, the “naked” [9]tBuPh could not be isolated as solid, while crystalline

precipitates of [9]cumulene rotaxanes were obtained reproducibly from solutions of

CH2Cl2/MeOH. All synthesized [9]cumulene rotaxanes were stable for at least weeks to

months in the crystalline solid state.

101

Table 3.1 Comparison of the qualitative stability of [9]tBuPh and [9]cumulene rotaxane 3.9

when kept under an argon atmosphere at rt.

Et2O solution kept at ambient light kept in the dark TLC spot

[9]tBuPh decolorization from

blue to grey/green to

orange within 3

hours

decolorization to

orange overnight

decolorization from

blue to an orange

spot after seconds

and already during

spotting

3.9 decolorization after

one week

decolorization after

several weeks

decolorization from

blue to a light yellow

spot within 4 h

Additional methods for monitoring stability based on decomposition, such as

differential scanning calorimetry (DSC) or melting/decomposition point analysis have been

rare for longer [n]cumulenes with n ≥ 5. For example, Iyoda and coworkers23 reported melting

points of a variety of tetraryl[5]cumulene derivatives that were measured as >250 °C. The

increased stability of the [9]cumulene rotaxanes enabled studies describing more closely the

thermal stability of these rotaxanes. Thus, DSC measurements for [9]cumulene rotaxane 3.9

were performed (Figure 3.6). The results showed that no melting point was observed for

rotaxane 3.9, but rather an onset of decomposition at ca. 170 °C, with a maximum at 176 °C.

102

Figure 3.6 DSC scan of [9]cumulene rotaxane 3.9.

While it was not possible to directly compare the thermal stability of rotaxane 3.9 with

cumulene [9]tBuPh, the DSC analysis of [9]cumulene rotaxane 3.9 was compared to the DSC

scan of [7]cumulene [7]tBuPh (Figure 3.7). Similar to the [9]cumulene rotaxane 3.9,

[7]tBuPh showed no melting point, but instead, an onset of decomposition at ca. 187 °C with

a maximum at 215 °C. Thus, the [7]cumulene [7]tBuPh appeared to possess slightly a higher

thermal stability compared to the [9]cumulene rotaxane 3.9. If cumulenes followed the same

trend as polyynes,24 decomposition of the “naked” [9]cumulene [9]tBuPh should be lower

than 187 °C, the onset of decomposition of [7]tBuPh. If the decomposition of [9]tBuPh

started between 160 and 187 °C, no stabilization of the cumulene via rotaxane formation

would be observed. In contrast, if the decomposition already started below 160 °C, rotaxane

formation would result in stabilization. Consequently, at this point, no conclusive remarks

regarding DSC analysis could be made.

103

Figure 3.7 DSC scan of [7]tBuPh.

Since the [9]cumulene rotaxanes showed an increased stability compared to the

“naked” analogues, especially as microcrystalline solids, extensive characterization was

possible for the first time based on 1D and 2D NMR-, UV/vis-, and IR spectroscopy, as well

as mass spectrometry and cyclic voltammetry. Discussions and comparisons of the

[9]cumulene rotaxanes with the series of [n]tBuPh cumulenes have been described in Chapter

IV.

3.6 Synthetic approach to higher [n]cumulene rotaxanes (n > 9)

3.6.1 Synthetic approach to [11]cumulene rotaxane

Since an enhancement of stability of longer cumulenes via rotaxane formation was

observed through the formation of 3.9, the synthesis of the [11]cumulene rotaxane 3.11 was

attempted (Scheme 3.5). Precursor 3.12 was obtained in a very low yield via a heterocoupling

reaction of terminal triyne 2.21 and bromodiyne 3.7.22 Nevertheless, with a small amount of

104

compound 3.12 in hand, reductive elimination was conducted at rt. TLC analysis confirmed

the formation of [11]cumulene rotaxane 3.11, indicated by a blue-greenish spot, which

showed behavior typical for cumulenes. The blue-greenish reaction mixture, however, started

to decompose within minutes as observed by a color change of the solution to brown and

definitely decomposed when filtered over a plug of basic alumina. Nonetheless, this was the

first time where a [11]cumulene could be synthesized, although, disappointingly, no concrete

confirmation of this statement could be obtained by traditional characterization methods.

105

Scheme 3.5 Synthetic approach to [11]cumulene rotaxane 3.11.

106

3.6.2 Synthetic approach to [13]cumulene rotaxanes including UV/vis spectroscopy

studies

With the terminal triyne 2.21 in hand, the synthesis of the [13]cumulene rotaxane 3.13

was also attempted (Scheme 3.6). Terminal triyne 2.21 was homocoupled in the presence of

3.3·CuI to afford the hexayne rotaxane 3.14 in 17% yield.22 This precursor for the

[13]cumulene rotaxane was further converted in a reductive elimination reaction using

anhydrous SnCl2. The light yellow Et2O solution of 3.14 turned to a brighter yellow color

after SnCl2 and HCl addition before a darkening to orange occurred after several minutes.

TLC analysis, however, showed only a baseline spot aside from the starting material spot.

After additional SnCl2 and HCl were added, more side products were observed without any

evidence for rotaxane formation. In conclusion, TLC analysis and unsuccessful crystallization

attempts indicated that either rotaxane 3.13 was not formed or it was too unstable and

decomposed already in the reaction mixture.


Further attempts to convert 3.14 to the rotaxane 3.13 were carried out without using

work-up methods such as filtration over alumina and further solvent evaporation as described

107

for the synthesis of [13]tBuPh in Section 2.1.2.6. The reaction of precursor 3.14 to form

[13]cumulene rotaxane 3.13 was monitored via UV/vis spectroscopy (Figure 3.8). Precursor

3.14 showed a weak shoulder absorption at ca. 360 nm in the UV/vis spectrum (dark blue

curve). After addition of SnCl2 and HCl, this shoulder disappeared, and a new absorption

band at 364 nm was observed (yellow curve, Figure 3.8) similar to the UV/vis spectrum of the

synthesis of [13]tBuPh (red curve, Figure 2.3). After stirring and evaporation via bubbling

argon through the reaction mixture, three signals (e.g., pink curve) were formed at 345, 371,

and 400 nm, analogous to UV/vis spectra of [13]tBuPh in Figure 2.3 (see Section 2.1.2.6).

The values of vibrations were slightly lower with 2032 and 1954 cm–1. After second addition

of reactants (in order to potentially accelerate the reaction) and several hours, only one

absorption band, at 363 nm, remained (green curve), which persisted after stirring overnight

(black curve). The color of the reaction mixture appeared also to be apricot-orange but

slightly brighter than in the case of conversion to the “naked” [13]cumulene [13]tBuPh.

Again, however, no hint for cumulene formation could be observed based on lack of

absorption bands in the lower energy region.

108


[13]cumulene rotaxane 3.13 (in Et2O).

Finally, a new and potentially promising macrocycle, compound 3.15, has been

synthesized (Scheme 3.7).25 This macrocycle possesses a similar structure to macrocycle 3.3,

however, the resorcinol aromatic ring contains two additional t-butyl groups in the ortho

position to the oxygen substituents. Molecular modeling26 of 3,5-di-t-butylphenyl-substituted

[n]cumulene rotaxanes (n = 9 and 11) with macrocycles 3.3 and 3.15 has suggested a better

shielding of the cumulene chain in the case of macrocycle 3.15. Thus, direct synthesis of the

109

[13]cumulene rotaxane has been carried out, starting with the formation of 3.16 in 12%

yield.22


110

With compound 3.16 in hand, the synthesis of [13]cumulene rotaxane 3.17 was

attempted. The reaction was monitored via UV/vis absorption spectroscopy before and after

addition of the reactants. Standard reaction methods, such as TLC analysis and work-up via

filtration were not carried out to circumvent potential decomposition. Furthermore, in the case

of decomposition, UV/vis measurements that were recorded directly from the reaction

mixture could show at least potential evidence for the successful formation of the rotaxane

before complete decomposition. The conversion of 3.16 to [13]cumulene rotaxane 3.17 was

performed in Et2O at rt (Figure 3.9). Pure precursor 3.16 showed a shoulder absorption

between 350 and 360 nm in the UV/vis spectrum (black curve). After addition of SnCl2, the

frequently observed absorption band at 366 nm occurred (red curve). Stirring the reaction

mixture for ca. 1 h revealed the formation of three vague signals, rather shoulders, between

350 and 450 nm (green curve). The interesting feature was that by time passing, the intensity

of these absorption decreased, in contrast to both synthetic approaches to [13]tBuPh and

[13]cumulene rotaxane 3.13 in Figure 2.3 and Figure 3.8, respectively, where the intensity of

the three signals was increasing by proceeded time. In addition, two new absorption bands in

the region of 475–550 nm, at 487 and 528 nm were observed (green curve), which showed a

debut compared to the synthetic approaches to [13]tBuPh and [13]cumulene rotaxane 3.13 in

Figure 2.3 and Figure 3.8, respectively, even though two potential absorption bands could be

observed in the case of conversion of precursor 3.14 to rotaxane 3.13 in Figure 3.8. The

UV/vis spectrum that was recorded after stirring overnight revealed again only one absorption

band at ca. 359 nm (violet curve). The two absorption bands in the lower energy region also

disappeared hereby. The color of the reaction mixture was again apricot-orange, and no hint

for cumulene formation by color or by absorption in the low energy region was observed.

111


[13]cumulene rotaxane 3.17 (in Et2O).

3.7 Summary and conclusion

In summary, the strategy for rotaxane formation that has recently been successfully

implemented for polyynes, also succeeded when applied to cumulenes. To the best of my

knowledge, this was the first time that cumulene rotaxanes had been synthesized. The

synthesis of precursors to cumulene rotaxanes (i.e., polyyne rotaxanes) was optimized to the

112

stage that reaction yields reached an outstanding value of 74%. This result was achieved using

macrocycle 3.3, which was the most suitable of the three macrocycles that were tested. The

reaction yields for the synthesis of cumulene rotaxanes 3.8 and 3.9 (29% and

33%, respectively) were also quite good comparing the rather low yields of known polyyne

rotaxane syntheses.17 The introduction of a macrocycle, which ideally has offered protection

to the instable cumulene chain led to the desired stability enhancement of longer cumulenes

and hence, the possibility to afford [9]cumulenes as stable solids. Unfortunately, the limit to

stability enhancement by rotaxane formation appeared to be reached at the stage of a

[13]cumulene as observed by UV/vis spectroscopic studies. More specifically, according to

the similar colors of the reaction outcome in Figures 3.7 and 3.8, and the lack of absorption

bands in the low energy region that have been characteristic for longer cumulenes, no

[13]cumulene rotaxanes, 3.13 and 3.17, were formed in the two synthetic approaches. It

seemed, however, that some new compounds were formed instead, e.g., an acetylene-like

vibronic fine structure was observed for [13]cumulene rotaxane 3.13. In addition, overnight,

the three absorptions signals that were observed in both reactions converted to only one broad

absorption band at ca. 360 nm indicating a potential decomposition of the initially formed yet

unknown and uncharacterized compounds.

As an outlook or future directions to go with this project, it is worth to again approach

the synthesis of a [11]cumulene rotaxane to determine if this cumulene can be stabilized by

encapsulation with a macrocycle, which would represent the first [11]cumulene synthesized to

date. Another optimization that could be done in the case of cumulene rotaxanes is to improve

the reaction yields for the polyyne rotaxanes since further formations of cumulene rotaxanes

appeared much less problematic. Finally, regarding the macrocycles, optimizations, e.g., cycle

size and functionalization issues could be varied to develop the most suitable macrocycle for

cumulene rotaxanes.



The general procedures and methods are analogous to that in Section 2.3.1.

113


1,1,6,6-Tetrakis(3,5-di-t-butylphenyl)deca-1,2,3,4,5,6,7,8,9-nonaene rotaxane 3.8. To a

solution of 3.4 (10 mg, 6.7 µmol) in Et2O (4 mL) was added anhydrous SnCl2 (3.8 mg,

20 µmol) and HCl (1 M in Et2O, 0.03 mL, 30 µmol) at rt under an Ar atmosphere. After 25

min, the solution was filtered through a plug of basic alumina oxide and eluted with CH2Cl2

affording the purified 3.8. Since the cumulene is not stable as amorphous solid and in

solution, “microcrystalline” 3.8 was obtained as blue solid with red shining (2.8 mg, 29%) by

overlaying a CH2Cl2 solution with dry MeOH at ca. –20 °C for ca. 2 days being stable for an

infinite time kept under an Ar atmosphere. Rf = 0.23 (hexanes/EtOAc 10:1). UV/vis (Et2O)

λmax 278, 319, 341, 369, 530, 578, 610, 670 nm; IR 3065 (vw), 2957 (m), 2901 (m), 2862 (w),

1920 (vw) 1584 (m) 1257 (s) cm−1; 1H NMR (400 MHz, CD2Cl2) δ 8.21 (d, J = 8.4 Hz, 2H),

7.94 (d, J = 8.8 Hz, 4H), 7.87 (d, J = 8.4 Hz, 2H), 7.74 (s, 2H), 7.38 (t, J = 1.6 Hz, 4H), 7.28

(d, J = 1.6 Hz, 8H), 7.24−7.17 (m, 8H), 6.88 (d, J = 8.8 Hz, 4H), 5.21 (s, 4H), 1.22 (s, 72H); 13C NMR (100 MHz, CD2Cl2) δ 159.7, 158.3, 158.0, 151.4, 147.0, 144.5, 137.3, 136.4, 134.3,

134.2, 130.0, 127.6, 127.4, 127.3, 125.8, 124.3, 123.7, 123.2, 122.6, 122.2, 122.0, 120.5,

116.5, 70.6, 35.1, 31.5. ESI HRMS m/z calcd. for C104H110N2NaO3 ([M + Na]+) 1457.84087

found 1457.83717.

114


solution of 3.5 (12.5 mg, 7.90 µmol) in Et2O (4 mL) was added anhydrous SnCl2 (4.5 mg,

0.024 mmol) and HCl (1 M in Et2O, 0.035 mL, 0.035 mmol) at rt under an Ar atmosphere.

After 20 min, the solution was filtered through a plug of basic alumina oxide and eluted with

CH2Cl2 affording the purified 3.9. Since the cumulene is not stable as amorphous solid and in

solution, “microcrystalline” 3.9 was obtained as blue needles with green shining (4.0 mg,

33%) by overlaying a CH2Cl2 solution with dry MeOH at ca. –20 °C for ca. 2 days being

stable for an infinite time kept under an Ar atmosphere. The isolated product was

contaminated with ca. 5% precursor 3.6. Rf = 0.46 (CH2Cl2). UV/vis (Et2O) λmax (ε) 287

(78200), 319 (65600), 341 (102700), 368 (158700), 530 (56700), 576 (50000), 615 (74500),

666 (38600) nm; IR 3069 (vw), 3036 (vw), 2955 (s), 2902 (m), 2863 (m), 2168 (vw), 2055

(w), 2038 (w), 1919 (w), 1590 (s), 1241 (s) cm−1; 1H NMR (300 MHz, CD2Cl2) δ 8.46 (d, J =

8.9 Hz, 4H), 8.26 (d, J = 8.5 Hz, 2H), 8.08 (d, J = 8.5 Hz, 2H), 7.74 (s, 2H), 7.37 (t, J = 1.6

Hz, 4H), 7.32 (d, J = 1.6 Hz, 8H), 7.16 (d, J = 8.9 Hz, 4H), 7.00 (t, J = 8.2 Hz, 1H), 6.61 (t, J

= 2.3 Hz, 1H), 6.37 (dd, J = 8.2 Hz, J = 2.3 Hz, 2H), 4.08 (t, J = 7.3 Hz, 4H), 3.96 (t, J = 6.6

Hz, 4H), 1.82 (t, J = 6.7 Hz, 4H), 1.75 (t, J = 6.5 Hz, 4H), 1.54 (s, 8H), 1.20 (s, 72H); 13C

NMR (75 MHz, CD2Cl2) δ 160.92, 160.87, 156.1, 151.4, 146.4, 144.8, 137.4, 136.7, 131.9,

129.7, 129.1, 128.7, 127.7, 125.7, 124.5, 124.1, 123.2, 120.3, 119.8, 119.0, 115.2, 107.8,

100.0, 68.4, 68.0, 35.1, 31.4, 29.8, 29.4, 26.13, 26.08. ESI HRMS m/z calcd. for

C108H126N2NaO4 ([M + Na]+) 1537.96098, found 1537.95928, for C108H127N2O4 ([M + H]+)

1515.97904, found 1515.97866.

115


solution of 3.5 (10 mg, 6.9 µmol) in Et2O (4 mL) was added anhydrous SnCl2 (9.0 mg, 0.047

mmol) and HCl (1 M in Et2O, 0.1 mL, 0.1 mmol) at 0 °C under an Ar atmosphere. After 15

min, the solution was filtered through a plug of basic alumina oxide and eluted with CH2Cl2

affording the purified 3.10. Since the cumulene is not stable as amorphous solid and in

solution, “microcrystalline” 3.10 was obtained as blue precipitate by overlaying a CH2Cl2

solution with dry MeOH at ca. –20 °C for ca. 2 days being stable for an infinite time kept

under an Ar atmosphere. Due to the small amount of 3.10 that could be isolated, no yield was

determined. Rf = 0.12 (CH2Cl2). UV/vis (Et2O) λmax 286, 320, 341, 369, 532, 578, 616, 668

nm; IR 3069 (vw), 2953 (s), 2920 (s), 2852 (s), 2050 (vw), 1929 (w), 1585 (m) cm−1; 1H

NMR (400 MHz, CDCl3) δ 8.47 (d, J = 8.9 Hz, 4H), 8.17 (d, J = 8.4 Hz, 2H), 7.99 (d, J = 8.4

Hz, 2H), 7.67 (s, 2H), 7.31–7.30 (m, 4H), 7.27 (d, J = 1.6 Hz, 8H), 7.18 (d, J = 8.9 Hz, 4H),

4.18–4.14 (m, 4H), 1.91–1.89 (m, 4H), 1.46 (br s, 12H), 1.17 (s, 72H). Insufficient material

was available to obtain a meaningful 13C spectrum. ESI HRMS m/z calcd. for C100H119N2O2

([M + H]+) 1379.92661 found 1379.92568.

5,5'-(5-Bromo-1-methoxypenta-2,4-diyne-1,1-diyl)bis(1,3-di-t-butylbenzene) 3.7. To a

solution of 2.12 (0.243 g, 0.516 mmol) in acetone (5 mL) was added N-bromosuccinimide

(NBS) (0.138 g, 0.775 mmol) and AgNO3 (8.8 mg, 0.052 mmol) at rt. The reaction mixture

was covered with alumina foil and stirred overnight. n-Pentane (15 mL) and saturated aq

116

Na2S2O5 (15 mL) were added. The layers were separated, and the organic phase was washed

with brine (20 mL), dried over Na2SO4, and filtered. Solvent removal afforded 3.7 (0.28 g,

99%) as a white solid. Mp 48−52 °C. Rf = 0.60 (hexanes/EtOAc 10:1). IR 3070 (vw), 2955

(s), 2903 (m), 2866 (m), 2824 (w), 1594 (m) cm−1; 1H NMR (300 MHz, CDCl3) δ 7.33 (t, J =

1.8 Hz, 2H), 7.29 (d, J = 1.8 Hz, 4H), 3.42 (s, 3H), 1.29 (s, 36H); 13C NMR (75 MHz, CDCl3)

δ 150.4, 141.1, 121.7, 121.4, 82.6, 75.6, 73.4, 65.1, 53.1, 42.0, 34.9, 31.4. ESI HRMS m/z

calcd. for C34H4579BrNaO ([M + Na]+) 571.25460, found 571.25464; for C33H42

79Br ([M –

OMe]+) 517.24644, found 517.24587.

3.9 References

1 F. Aricó, J. D. Badjic, S. J. Cantrill, A. H. Flood, K. C.-F. Leung, Y. Liu, J. F.

Stoddart, Top. Curr. Chem. 2005, 249, 203–259.

2 S. Saito, K. Nakazono, E. Takahashi, J. Org. Chem. 2006, 71, 7477–7480.

3 M. J. Frampton, H. L. Anderson, Angew. Chem. Int. Ed. 2007, 46, 1028–1064; Angew.

Chem. 2007, 119, 1046–1083.

4 M. J. Langton, J. D. Matichak, A. L. Thompson, H. L. Anderson, Chem. Sci. 2011, 2,

1897–1901.

5 Y. Kohsaka, K. Nakazono, Y. Koyama, S. Asai, T. Takata, Angew. Chem. Int. Ed.

2011, 50, 4872–4875.

6 R. Eelkerna, K. Maeda, B. Odell, H. L. Anderson, J. Am. Chem. Soc. 2007, 129,

12384–12385.

7 G. Wenz, B.-H. Han, A. Müller, Chem. Rev. 2006, 106, 782–817.

8 V. Aucagne, K. D. Hänni, D. A. Leigh, P. J. Lusby, D. B. Walker, J. Am. Chem. Soc.

2006, 128, 2186–2187.

9 J. D. Crowley, S. M. Goldup, A.-L. Lee, D. A. Leigh, R. T. McBurney, Chem. Soc.

Rev. 2009, 38, 1530–1541.

10 V. Aucagne, J. Berná, J. D. Crowley, S. M. Goldup, K. D. Hänni, D. A. Leigh, P. J.

Lusby, V. E. Ronaldson, A. M. Slawin, A. Viterisi, D. B. Walker, J. Am. Chem. Soc.

2007, 129, 11950–11963.

11 The active template method using Cu(I) has already been applied to the formation of

catenanes, see: C. O. Dietrich-Buchecker, J. P. Sauvage, J. P. Kintzinger, Tetrahedron

Lett. 1983, 24, 5095–5098; J.-P. Sauvage, Acc. Chem. Res. 1990, 23, 319–327.

117

12 R. Eastmond, T. R. Johnson, D. R. M. Walton, Tetrahedron 1972, 28, 4601–4616.

13 T. J. Taylor, F. P. Gabbai, Organometallics 2006, 25, 2143–2147.

14 C. Zhao, R. Kitaura, H. Hara, S. Irle, H. Shinohara, J. Phys. Chem. C 2011, 115,

13166–13170.

15 I. Ben Shir, S. Sasmal, T. Mejuch, M. K. Sinha, M. Kapon, E. Keinan, J. Org. Chem.

2008, 73, 8772–8779.

16 S. Saito, E. Takahashi, K. Nakazono, Org. Lett. 2006, 8, 5133–5136.

17 N. Weisbach, Z. Baranová, S. Gauthier, J. H. Reibenspies, J. A. Gladysz, Chem.

Commun. 2012, 48, 7562–7564.

18 L. D. Movsisyan, D. V. Kondratuk, M. Franz, A. L. Thompson, R. R. Tykwinski, H.

L. Anderson, Org. Lett. 2012, 14, 3424–3426.

19 A. M. Blanco-Rodríguez, M. Towrie, J.-P. Collin, S. Záliš, A. Vlček Jr, Dalton Trans.

2009, 3941–3949.

20 F. Eggers, U. Lüning, Eur. J. Org. Chem. 2009, 2328–2341.

21 The leaving group of the terminal acetylene plays a huge role in this reaction

manifesting in higher reaction yields for OMe leaving groups than for OH groups. As

an example, the precursor rotaxane formation have been accomplished using 3.2 as

macrocycle and the terminal diyne 2.24. The yields have been lousy, and this reaction

type has not continued.

22 The reaction was performed by Michael Franz from the Tykwinski group.


2188–2208.


25 Compound 3.15 was developed and synthesized by Michael Franz.

26 3D models have been calculated by Michael Franz using the software “Chem3D” and

“MM2 geometry optimization” as method.

118

4. Chapter IV. Characterization of [3]-, [5]-, [7]-, and [9]tBuPh including

[9]tBuPh rotaxanes and comparison to different series of [n]cumulenes††

4.1 UV/vis spectroscopy

4.1.1 Introduction

UV/vis spectroscopy has been the most common characterization method for long

[n]cumulenes and in early studies it has been the essential tool to confirm formation of [7]-

and [9]cumulenes.1 Using homologous series of [n]cumulenes, UV/vis spectroscopy allows

analysis of electronic trends, i.e., optical band gap, as a function of cumulene length and

substitution. Similar to that demonstrated for polyynes,2 UV/vis spectroscopic analysis versus

molecular length might also allow extrapolation to infinite chain length, which would offer a

prediction of the band gap of the cumulenic version of carbyne. To date, however, this has not

been possible due to the limited number of model compounds that are currently available.

In principle, three major factors govern trends typically observed in the UV/vis spectra

of [n]cumulenes, including (1) structure of the [n]cumulene (odd- versus even-numbered

[n]cumulenes), (2) molecular length, and (3) the nature of the terminal substitution, i.e.,

endgroup effects (aryl versus alkyl and mesomeric versus inductive effects, respectively).

The influence of odd- versus even-numbered [n]cumulenes is delineated schematically

in Figure 4.1, demonstrating the potential mesomeric or inductive contribution from the

endgroups. For even-numbered cumulenes, there are two π-systems that are degenerate and

spatially orthogonal (Figure 4.1a). In the case of aryl endcapping groups, each orthogonal π-

system can conjugate with substituents at one end of the cumulenic framework, but not both.

The situation is distinctly different for odd-numbered cumulenes (Figure 4.1b), where the two

π-systems of the sp-carbon framework are no longer degenerate. In this case, one π-system

spans the length of the cumulene skeleton and can conjugate with both sets of endgroups

(Figure 4.1b, in red). The other π-system (in blue) does not communicate directly with the

endgroups (i.e., via resonance) and is thus considerably shortened, although hyperconjugation

†† Portions of this chapter have been published: J. A. Januszewski, R. R. Tykwinski, Chem. Soc. Rev. 2014, 43, 3184–3203, see http://dx.doi.org/10.1039/C4CS00022F - Reproduced by permission of The Royal Society of Chemistry;

119

with terminal groups is easily envisioned. It is worth noting that the influence of odd- versus

even-numbered [n]cumulene structure should also be observed in the bond length alternation

(BLA) of cumulenes. As shown by the mesomeric structures in Figure 4.1b, increased BLA is

expected for odd [n]cumulenes and should be further enhanced by groups able to conjugate to

the cumulene core.

Figure 4.1 Electronic effects based on odd- and even-numbered, as well as alkyl- and aryl

endcapped [n]cumulenes, demonstrated schematically with canonical structures for [4]- and

[5]cumulenes.

Within this section, characterization of all representatives of the cumulene series of

[n]tBuPh including [9]cumulene rotaxanes 3.8–3.10 based on qualitative and quantitative

UV/vis spectroscopy has been performed. Initially, the UV/vis data of [n]tBuPh and [n]Mes

will be presented. Comparisons of both series to each other as well as to other cumulene

120

derivatives will be discussed. Finally, the spectroscopic results of [9]cumulene rotaxanes and

comparison to the “naked” [9]cumulene [9]tBuPh as well as to the series of [n]tBuPh will be

given. A discussion of the band gap of cumulenes and a following conclusion will complete

the UV/vis spectroscopic section.

4.1.2 UV/vis spectroscopy of [3]-, [5]-, [7]-, and [9]tBuPh

4.1.2.1 General observations

Figures 4.2a and 4.2b present the measured UV/vis spectra of [n]tBuPh (n = 3, 5, 7, 9)

and [n]Mes (n = 5, 7, 9), respectively. A distinct signal pattern is immediately observed,

showing two regions of absorption bands for both series of tetraaryl[n]cumulenes with the

most intense absorptions mainly found at higher energy in the UV region (<400 nm).

Noteworthy is the observation that the fine structure of the absorption in the UV region

becomes more distinct as a function of molecular length, i.e., when conjugation is extended.

The second set of absorption bands is found in the visible region from 400–700 nm. With

increasing conjugation in the cumulene chain, the lowest absorption λmax values become red-

shifted to lower energy values. In the following sections, the spectroscopic results of

[n]tBuPh and [n]Mes will be compared to two [n]cumulene series, that are known from

literature, [n]Ph and [n]Cy with n = 3, 5, 7, and 9 (Figures 4.2c and 4.2d, respectively).1

These two series offer the most complete analyses reported to date (aside from [n]tBuPh and

[n]Mes). The [n]Ph shows a very similar signal pattern compared to [n]tBuPh and [n]Mes.

The alkyl-substituted [n]Cy also possesses two regions of absorption bands, however, with

the absorption intensities being much more pronounced in the high energy region (<400 nm)

than in the visible region.

121

Figure 4.2 UV/vis spectra of [n]cumulenes: UV/vis spectra of a) [n]tBuPh and b) [n]Mes.

Both sets of spectra were measured in Et2O and normalized to the most intense low energy

absorption. Spectra of c) [n]Ph (in benzene) and d) [n]Cy (in Et2O). Spectra of [n]Ph and

[n]Cy were adapted with permission from reference 1. Copyright 1964 John Wiley & Sons.

4.1.2.2 Influence of cumulene chain length

The most obvious consequence of π-electron conjugation is observed in the lowest

energy absorption values, λmax, as a function of cumulene length chain, as demonstrated by

values of the two [n]cumulene series, [n]tBuPh and [n]Mes that are listed in Table 4.1. A

monotonic red-shift in λmax is clearly visible as cumulene length is increased within each

series, indicating a decreasing HOMO-LUMO energy gap. In the case of [n]tBuPh, the

smallest cumulene homologue, [3]tBuPh, shows a λmax value of 424 nm reaching a value of

664 nm for [9]tBuPh. Similar effects can be observed for [n]Mes with n increasing from 5 to

9 giving 460 and 666 nm, respectively.

As expected, the comparison of tetraaryl[n]cumulenes with tetraalkyl[n]cumulene [n]Cy

showed that λmax values of tetraaryl[n]cumulenes are found at much lower energy than those

122

of the tetraalkyl[n]cumulenes, e.g., 664 nm versus 465 nm for [9]tBuPh and [9]Cy,

respectively (Table 4.1).8 This is explained by the decreased conjugation in the case of the

alkyl endgroups in contrast to tetraaryl[n]cumulenes that possess aryl endgroups that are

contributing to the π-conjugation of the cumulene (Figure 4.1).

Table 4.1 Lowest energy absorption λmax (in nm) and energy values Eg (in eV)[a] of [n]tBuPh,

[n]Mes, [n]Ph, and [n]Cy (in Et2O).

[n]cumulene [3] [5] [7] [9] refs

[n]tBuPh 424 (2.92) 500 (2.48) 564 (2.20) 664 (1.87)

[n]Mes - 460 (2.70) 560 (2.21) 666 (1.86)

[n]Ph 420 (2.95)[b] 489 (2.54)[b] 557 (2.23)[b] 663 (1.87) 3–7

[n]Cy 272 (4.56) 339 (3.66) 401 (3.09) 465 (2.67) 8

[a] Determined via common conversion tools from http://halas.rice.edu/conversions (nm to

eV). [b] Measured in benzene.

4.1.2.3 Influence of endgroups

While λmax values of the tetraaryl[5]cumulenes [5]tBuPh and [5]Mes (500 and 460

nm, respectively), vary rather significantly (Table 4.1), the λmax values of the longer

analogues [7]tBuPh/[7]Mes and [9]tBuPh/[9]Mes are nearly identical, with values of about

560 nm and 665 nm, respectively. Thus, lower [n]cumulenes, e.g., [5]cumulenes have a

pronounced endgroup effect compared to longer [n]cumulenes. Furthermore, the experimental

data shows that the influence of the endgroups decreases rapidly as a function of cumulene

length. In comparison with [5]Ph, which is electronically more or less neutral (i.e., no

endgroup effects are present), the endgroups of [5]cumulenes [5]tBuPh and [5]Mes are

expected to provide positive inductive effects due to the alkyl substituents (t-butyl and

methyl, respectively). The λmax absorption of [5]tBuPh is red-shifted to 500 nm compared to

λmax = 489 nm for [5]Ph4,5 (Table 4.1), while the λmax value of 460 nm for [5]Mes is

decreased compared to [5]Ph (λmax = 489 nm). This correlates well with the diminished

conjugation between the mesityl endgroups and the cumulene chain displayed by the

increased aryl twist angle relative to [5]tBuPh. This effect will be explained below in the X-

ray crystallography section (see Section 4.2).

123

Obviously, the endgroups also affect the λmax values of [n]cumulenes, aside from the

influence of chain length (increase of conjugation) as mentioned above. Therefore, a

comparison of endgroup effects of several [5]cumulenes has been done, and the appropriate

λmax values are summarized below in Table 4.2. The biggest shift of λmax values to lower

energy is observed for cumulenes in which the pendent aryl rings are forced to be coplanar to

the cumulene framework, namely [5]An with λmax = 555 nm9 and [5]Fl with λmax = 540 nm5

(for the structure of the cumulenes, see Figure 1.6). Thus, planarity of the aryl endgroups

appeared to be a dominant factor.

p-Methoxyphenyl-substituted [5]cumulene [5]MeOPh shows red-shifted λmax values

of 517 nm10 as reported by Cadiot relative to [5]Ph (λmax = 488 nm).5 This feature also occurs

in the case of [5]tBuPh and is readily explained by mesomeric and inductive effects of the

OMe endgroups, which donate electron density to the electron deficient sp-hybridized carbon

framework yielding a donor-acceptor-donor (D-A-D) type conjugated system.

Finally, the greatest blue-shift of λmax is found for tetraalkyl[n]cumulenes, such as

[5]Cy and [5]tBu (λmax = 339 and 337 nm, respectively)8,11 which differ by about 150 nm

compared to [5]Ph (λmax = 488 nm). There is, obviously, no mesomeric contribution from the

alkyl endgroups to the conjugated structure of [5]Cy and [5]tBu, but hyperconjugation with

the four methyl units in the cyclohexyl group or t-butyl groups, respectively, appears to be

present based on the comparison to [5]Me with λmax = 320 nm.12

Table 4.2 UV/vis spectroscopic data (λmax in nm) of selected [5]cumulenes with different

endgroups.

[5]An [5]Fl [5]MeOPh [5]tBuPh [5]Ph [5]Mes [5]Cy [5]tBu [5]Me

λmax 555[a] 540[a]

517[a] 510[a] 500[b]

488[a] 460[b] 339[b] 337[b]

320[c]

refs 9 5 10 4,5 8 11 12

[a] Measured in CHCl3. [b] Measured in Et2O. [c] Measured in EtOH.

124

4.1.2.4 Conclusion including comparison of the band gap of cumulenes

UV/vis results that have been studied for polyynes suggest that at some length, the

lowering of the λmax values achieves saturation and reaches a maximum and constant value.

This limiting value would then represent an estimate of the energy gap (Eg) of the material

“polyyne” carbyne. The same should be true for cumulenes. Comparison of the optical band

gap of cumulenes, i.e., the energy values Eg,13 shows a decrease from ca. 2.95 eV to 1.86 eV

by increasing chain length of aryl-substituted cumulenes (Table 4.1). Thus, with values at ca.

1.86 eV for [9]cumulenes, saturation and hence a “carbyne-like” cumulene is still not reached.

Several methods have been used for polyynes to describe the relationship between Eg, λmax,

and n, which also might be applied to cumulenes, for example an empirical power-law14,15 (Eg

= 1/λmax ~ n–x) or the exponential function proposed by Meier and coworkers.16

Unfortunately, attempts to apply these protocols to the cumulenes reported in Table 4.1

provide inconclusive estimates for Eg, and longer cumulenes (n > 9) are needed to complete

this analysis. Thus, no experimental estimate for the cumulenic form of carbyne is currently

available. At this point, a “semiconducting level” is present rather than a “metallic level” as

predicted by theoretical calculations for (infinite) long cumulenes.

In conclusion, evidence of saturation still needs to be established to make a prediction

of the HOMO-LUMO gap of carbyne. As mentioned before in Section 4.1.1, this estimation

was not possible experimentally due to the limited number of model compounds.

4.1.3 UV/vis spectroscopy of [9]cumulene rotaxanes and comparison to [9]tBuPh

The UV/vis spectra of [9]cumulene rotaxanes 3.8–3.10 show the same basic

features as the spectrum of the “naked” [9]cumulene [9]tBuPh including two absorption

regions, a similar fine structure pattern, as well as almost identical λmax values (Figure 4.3 and

Table 4.3). In detail, the λmax values of all three rotaxanes 3.9, 3.10, and 3.8 are slightly red-

shifted with values of 665, 668, and 670 nm, respectively, compared to λmax of 664 nm for

[9]tBuPh. Cumulene rotaxane 3.9 possesses the biggest macrocycle surrounding the

cumulene chain and the closest value of λmax compared to [9]tBuPh. The remaining rotaxanes

3.10 and 3.8 have smaller macrocycles that probably interact slightly with the cumulene

framework, resulting in red-shifted values. While the macrocycle of rotaxane 3.10 possesses

alkyl linkers, that of 3.8 contains an aryl ether that could also contribute to even greater

125

interactions with the cumulene chain, and thus resulting in a higher λmax value. The absorption

values in the high-energy region are similar for all three rotaxanes (Figure 4.3 and Table 4.3).

Compared to the “naked" [9]cumulene [9]tBuPh, the signals of the rotaxanes are slightly red-

shifted about ca. 3 nm.

Figure 4.3 Qualitative UV/vis spectra (in Et2O) of the [9]cumulene rotaxanes 3.8, 3.9, and

3.10 as well as the “naked” [9]cumulene [9]tBuPh (a quantitative spectrum was recorded for

3.9, see right axis).

Due to the increased stability of the cumulene rotaxanes, molar absorptivities of a

[9]cumulene have been measured and are discussed for the first time. A quantitative UV/vis

spectrum of cumulene rotaxane 3.9 measured in Et2O shows intense absorptions at both high

and low energy, e.g., 368 nm (ε = 159,000 M–1cm–1) and 615 nm (ε = 75,000 M–1cm–1). The

ε values at high energy resemble those of polyynes, which are well-known to show very

intense absorptions, such as, for example, precursor to rotaxane 3.9, tetrayne 3.6 (245 nm,

ε = 125,000 M–1cm–1, hexanes),17 supertrityl-substituted tetrayne (268 nm, ε = 149,000 M–

1cm–1, hexanes),18 and t-butyl-substituted tetrayne (λmax = 240 nm, ε = 277,000 M–1cm–1,

hexanes).19

126

Table 4.3 UV/vis spectroscopic data (absorption wavelengths in nm) of [9]tBuPh and

[9]cumulene rotaxanes 3.9, 3.10, and 3.8 (in Et2O).

compound [9]tBuPh 3.9[a] 3.10 3.8

Absorption

bands

316

338

366

528

574

610

664

321 (319)

341

368

531 (530)

577 (576)

615

665 (666)

320

341

369

532

578

616

668

319

341

369

530

578

610

670

[a] Absorption bands for qualitative and quantitative UV/vis spectra. If qualitative and

quantitative data differ, the quantitative values are given in parentheses.

The comparison of molar absorptivity ε values of [n]tBuPh (n = 3, 5, 7) measured in

CHCl3 and [9]cumulene rotaxane 3.9 (measured in Et2O) shows that ε (mainly in the case of

absorptions of the fine structure in the high energy region) increases as the chain length of the

cumulene increases (Figure 4.4). This feature mirrors the effect already known for

polyynes2,18 as well as several reported cumulene series, such as [n]Ph and [n]Cy (Figure

4.2c and Figure 4.2d, respectively),1 as well as the [n]Fc series with n = 1, 3, and 5.20

127

Figure 4.4 Quantitative UV/vis spectra of [n]tBuPh (n = 3, 5, 9, in CHCl3) and [9]cumulene

rotaxane 3.9 (in Et2O).

4.2 X-ray crystallography of [n]cumulenes and discussion of bond length

alternation (BLA)

4.2.1 Introduction

X-ray crystallographic analysis is relatively uncommon for [n]cumulenes (n ≥ 5) due

to the instability of longer cumulenes under ambient conditions and limited synthetic

accessibility. X-ray crystallography, however, offers profound insight into both the physical

and electronic structure of cumulenes, especially via the analysis of bond length alternation as

a function of molecular length (BLA, defined as the bond length difference between the two

central-most double bonds of the cumulene chain). As discussed above, UV/vis spectroscopy

nicely shows that there is a relationship between the optical HOMO-LUMO gap and the

length of the cumulene chain. In principle, this change in HOMO-LUMO gap should coincide

with structural changes of the cumulene framework. More specifically, BLA should diminish

for longer cumulenes and eventually reach a constant value. Several theoretical calculations

128

for cumulenes suggest that BLA cumulenes should approach a value of nearly zero,21–23

although at the time of these reports experimental validation of these predictions does not

exist.

4.2.2 General observations

Regarding reports of X-ray crystallography of cumulenes to date, analysis of

[3]cumulenes is common (>15 structures), while data for [5]cumulenes is rare (three

structures), and no data exists for longer cumulenes to the best of my knowledge.24

In the following section, new X-ray crystallographic structures, [3]tBuPh, [5]tBuPh,

[7]tBuPh, [5]Mes, [7]Mes, and [9]Mes are presented.25 All single crystals have been obtained

in this research group26 with [7]tBuPh, [7]Mes, and [9]Mes representing the longest

[n]cumulenes yet known to be analyzed via X-ray crystallography. These series of molecules

enables a detailed discussion of results and structural trends and of the BLA analysis that is

required to answer a key question concerning cumulenes as possible model compounds for

carbyne: Do cumulenes show experimental evidence of reduced BLA as a function of

increasing length and can thus reach a “carbyne-like” status? Initially, general trends

regarding bond angles and bond lengths will be presented. The effects of twist angles (also

named as torsional angles) that are derived from aryl endgroups and the cumulene core are

discussed. Finally, a discussion of BLA of the cumulenes will complete this section including

a comparison to cumulenes known from literature. More specifically, only series of

cumulenes with at least three reported analogous X-ray structures are considered for

comparison, i.e., the two cumulene series, [n]Ph27–29 and [n]Cy27,30 with n = 3, 4, 5. As well,

X-ray crystallographic analysis is included for even-numbered cumulenes in addition to the

odd-numbered cumulenes that are the main focus in this thesis. For a better understanding of

following discussions, Figure 4.5 shows a [9]cumulene including the carbon atom labeling of

the cumulene chain as well as the origin of the twist angle calculations via intersection of

planes formed from the aryl rings (C11–C16, grey-colored plane) and the cumulene skeleton

(constructed from C11, C21, and C1–C5, as well as the appropriate symmetric atoms, blue-

colored plane).

129

Figure 4.5 Description of bond lengths and twist angles using [9]Mes (ORTEP

drawings with 20% probability level): a) structure of [9]Mes including carbon labeling, b)

planes defining the twist angle (front view), and c) planes defining the twist angle (side view).

Carbons C11–C16 define the grey-colored plane, while carbons C11, C21, and C1–C5, as

well as the appropriate symmetric atoms define the blue-colored plane.

130

4.2.3 Bond angles

Figure 4.6 shows the bond angles of the cumulenic chain of [3]tBuPh, [5]tBuPh,

[7]tBuPh, [5]Mes, [7]Mes, and [9]Mes. While X-ray crystallographic structures of polyynes

regularly show a bending (“bow” or “S” shape) of the sp-carbon framework in the solid

state,31,32 cumulenes appear to maintain a more linear structure. An examination of the bond

angles of the two series [n]tBuPh and [n]Mes shows that they rarely vary by more than a few

degrees from the ideal of 180°. This feature fits to other known cumulenes reported in

literature.33

Figure 4.6 Bond angles of the cumulene chain in [n]tBuPh and [n]Mes with n = 3, 5, 7 and

n = 5, 7, 9, respectively.

The bond angles of the cumulene chain in [n]tBuPh and [n]Mes show values between

177.1° and 179.8° (Figure 4.6). An exception to the almost linear cumulene chains described

so far is the structure of [3]tBuPh with cumulenic angles of 169.4° (C1–C2–C3) and 168.4°

(C2–C3–C4). In addition to the likely influence of undefined crystal-packing forces, the bent

shape of the [3]tBuPh might arise from favorable, intramolecular C–H/π-interactions between

hydrogen atoms of a t-butyl unit of the aryl group with the aromatic ring at the opposite

terminus of the cumulene chain (Figure 4.7).25 This premise is supported by short C–H/π

distances of 2.81–3.27 Å,34 and these interactions are reminiscent of a recent study by

Grimme and Schreiner highlighting dispersive forces in hexaphenylethane derivatives.25,35

131

Figure 4.7 Illustration of possible intramolecular C–H/π-interactions of [3]tBuPh.

In conclusion, the bond angle values of [n]tBuPh and [n]Mes show no indication for

trends, such as increase or decrease of linearity by increasing the chain length. In contrast,

comparison of the bond angles in the series of [n]Ph27–29 and [n]Cy27,30 shows that by

increasing the chain length, the bond angles resemble more and more 180° with values

reaching almost 179° for both [5]cumulenes, [5]Ph and [5]Cy starting with values of ca. 176°

for [3]Ph and [3]Cy. Comparison of the bond angles regarding the difference of substitution

(aryl versus alkyl) in [n]Ph and [n]Cy shows no significant change. Consequently, the

substitution pattern does not appear to influence the linearity of [n]cumulenes.

4.2.4 Bond lengths

Figure 4.8 shows the crystallographic structures of [3]tBuPh, [5]tBuPh, [7]tBuPh,

[5]Mes, [7]Mes, and [9]Mes including the labeling of carbon atoms of the cumulene chain.

Appropriate data for further discussion of the bond lengths of the cumulene chain can be

gathered from Figure 4.9 and Table 4.5.

The terminal double bonds, C1-C2 (α-bonds) of the cumulene chain are always the

longest at 1.33–1.35 Å. The second outermost double bonds, C2-C3 (β-bonds) of cumulenes

are the shortest with values of 1.25–1.26 Å resembling almost acetylenic bonds (Figure 4.9).

Herein, one exception needs to be mentioned, i.e., in the case of the [7]cumulenes, the central

γ-bonds with 1.252 and 1.257 Å are slightly shorter than the β-bonds lengths of 1.254 and

132

1.260 Å for [7]tBuPh and [7]Mes, respectively. Finally, the central double bonds (γ-bonds)

are intermediate to those of α- and β-double bonds (Figure 4.9). Hence, a significant

alternation in bond length is maintaining in the cumulene chain and will be discussed in

Section 4.2.6.

Figure 4.8 X-ray crystallographic structures of [3]tBuPh, [5]tBuPh, [7]tBuPh, [5]Mes,

[7]Mes, and [9]Mes (ORTEP drawings with 20% probability level).

A comparison between even-numbered and odd-numbered [n]cumulenes shows that

the α-bonds in even-numbered cumulenes are much shorter than in the odd-numbered analogs

with a decrease of almost 0.02 Å and 0.015 Å for [4]Ph and [4]Cy, respectively, compared to

the α bonds of [n]Ph and [n]Cy with n = 3, 5 (Figure 4.9). This feature can be explained by

the mesomeric structures of odd-numbered [n]cumulenes that differ to the even-numbered

[n]cumulenes as already described in Section 4.1 (Figure 4.1). The α-bonds in even-numbered

cumulenes show more double bond character in the canonical structures than in odd-

numbered cumulenes that contain more single bond character. While even-numbered

cumulenes show shorter α-bonds than odd-numbered cumulenes, the β-double bonds show

the opposite trend, namely they are longer in even-numbered- than in odd-numbered

[n]cumulenes, however, in lower extent. The difference between β-bond lengths and α-bond

lengths is lower for alkyl-substituted [n]cumulenes, with ca. 0.03–0.07 Å for [n]Cy compared

to the difference in bond lengths of aryl-substituted [n]cumulenes giving values of 0.06–0.10

Å for [n]tBuPh, [n]Mes, and [n]Ph.

133

Interestingly, the comparison of [5]cumulenes [5]tBuPh, [5]Mes, [5]Ph, and [5]Cy

shows that the bond lengths of [5]tBuPh resemble those of [5]Ph, while the bond lengths of

[5]Mes are similar to those of the alkyl-substituted [5]Cy (Figure 4.9). Consequently, it

seemed that the structures of mesityl-substituted cumulenes more closely resemble those of

alkyl-substituted cumulenes than aryl-substituted cumulenes (when comparing to [n]Ph). This

fact might be explained by the significant torsional angle of the endgroups of [n]Mes, which

prevents communication of these endgroups with the cumulene chain as it is the case for

alkyl-substituted cumulenes.

Figure 4.9 Bond lengths (in Å) of the cumulene chain of the series [n]tBuPh, [n]Mes,

[n]Ph,27–29 and [n]Cy.27,30

4.2.5 Torsional angles

For a cumulene chain to conjugate efficiently with the endgroups, these endgroups

should be coplanar to the π-orbitals of the chain (red π-system in Figure 4.1b). In both

cumulene series [n]tBuPh and [n]Mes, however, a twisting of the endgroups to the chain is

134

observed (Figure 4.5, Figure 4.8, and Table 4.4). [n]tBuPh cumulenes show two pairs of

twisted endgroups having different torsional angles and being centrosymmetric to each other.

One pair of endgroups is more coplanar to the red π-system of the cumulene chain (see Figure

4.1b) with torsional angles between 14° and 21° (Table 4.4, aryl rings A and C).36 The other

pair, however, is twisted to a much higher extent with values of 44° to 55° (Table 4.4, aryl

rings B and D). In contrast, in the [n]Mes series, all four endgroups are twisted significantly

out of the plane of the cumulene framework, in a range of 45° to 52° (Table 4.4). In

conclusion, the resulting steric congestion prevents coplanarity between the endgroups and

the cumulene chain for both cumulenes, however, in a lower extent for the [n]tBuPh series

since one pair of endgroups is in a position to enable conjugation with the sp-hybridized

carbon chain. Nevertheless, this feature affects the BLA of cumulenes that will be covered in

the following section.

Table 4.4. Aryl twist angles of aromatic ring relative to cumulenic framework.[a]

cumulene ring[b] [3]tBuPh [5]tBuPh [7]tBuPh [5]Mes [7]Mes [9]Mes

aryl twist angle (°)

A 30.9 14.4 16.6 46.2 45.4 48.8

B 43.1 47.9 54.9 51.4 52.1 54.4

C 26.6 20.6

D 40.4 43.6

[a] Aryl twist angles have been calculated as the difference between planes generated from (i)

the six carbons of the aryl ring and (ii) the carbons of the cumulene skeleton, along with the

four ipso-carbons of the aryl rings (see Figure 4.5). [b] See Figure 4.8 for labeling of aryl

rings.

4.2.6 Bond length alternation

The values for bond lengths in a cumulene chain confirm definitely BLA in the

cumulene chain. Table 4.5 summarizes all relevant bond lengths and the appropriate BLA

values of the cumulene chain for four series of cumulenes, [n]Ph,27–29 [n]Cy,27,30 [n]tBuPh,

and [n]Mes, as well as selected [n]H21–23 cumulenes from theoretical studies. The BLA values

for all cumulenes decrease as a function of length, e.g., from 0.086 Å ([3]tBuPh) to 0.052 Å

([7]tBuPh, Table 4.5, entries 7–9) or from 0.048 Å ([5]Mes) to 0.038 Å ([9]Mes, Table 4.5,

entries 10–12). An interesting fact is observed when BLA values have been compared

135

between even- and odd-numbered [n]cumulenes. More specifically, BLA values for the [4]-

and [5]cumulenes in both series, [n]Ph and [n]Cy, are approximately identical and vary only

by 0.001–0.002 Å (Table 4.5, entries 2/3 and 5/6, respectively).

BLA values also reflect endgroup effects, as shown for [5]cumulenes [5]Cy, [5]Mes,

[5]tBuPh, and [5]Ph with 0.028/0.040, 0.048, 0.054, and 0.058 Å, respectively (Table 4.5,

entries 6, 10, 8, and 3, respectively). In contrast to aryl-substituted cumulenes, alkyl-

substituted cumulenes, e.g., [n]Cy, show the lowest BLA values so far known for cumulenes.

This might be explained by the absence of a significant endgroup influence on the cumulene

chain via conjugation (see Figure 4.1). The aryl-substituted cumulenes show similar BLA

values although the [n]Mes series shows slight deviations, probably due to the limited

conjugation between the endgroups and the sp-hybridized cumulene chain as a result of the

twisted mesityl endgroups.

Theoretical studies of BLA in long cumulenes that have been reported in literature

suggest that BLA values converge quite rapidly to a value of nearly zero (Table 4.5).21–23

Castiglioni and coworkers, for example, have calculated the BLA for a series of [n]H,

showing already very low BLA values with 0.0136 and 0.010 Å for [7]H and [9]H,

respectively (Table 4.5, entries 14 and 16). Calculations for even longer [n]H cumulenes with

n = 29 result in a BLA value of 0.004 Å (Table 4.5, entry 19). In addition, Shakibazadeh and

coworkers have reported that BLA values for [7]H and [9]H are almost the same, 0.014 and

0.010 Å, respectively (Table 4.5, entries 13 and 15). Furthermore, Görling and coworkers

have calculated BLA values of 0.009 Å and 0.006 Å for [11]H and [n]H (n = 19–39),

respectively, (Table 4.5, entries 17 and 18) approaching BLA = 0, i.e., the central double

bonds in the cumulene chain would be equal length.

136

Table 4.5 Selected bond lengths (Å) for cumulene series [n]Ph, [n]Cy, [n]tBuPh and

[n]Mes, as well as theoretically calculated values for [n]H including BLA data.[a]

Entry Cumulene C1−C2 C2−C3 C3−C4 C4−C5 C5−C5’ BLA[b] Ref

1 [3]Ph 1.344(3)[c]

1.346(2)[d]

1.246(3)[c]

1.260(2)[d]

1.345(3)[c]

1.349(2)[d] – –

0.099

0.088 28

2 [4]Ph[e] 1.327 1.270 1.271 1.326 – 0.056 27

3 [5]Ph 1.3453(17) 1.2503(18) 1.3091(19) 1.2515(18) 1.3456(17) 0.058 29

4 [3]Cy[e] 1.328[g]

1.332

1.256[g,h]

1.261[h] – – –

0.072

0.071 27,30

5 [4]Cy[e] 1.317 1.273 1.279 1.313 – 0.039 27

6 [5]Cy[e] 1.329[i]

1.332[d]

1.260[i]

1.267[d]

1.300[f,i]

1.295[d,f]

–

–

0.040

0.028 30

7 [3]tBuPh 1.334(3)

1.336(3)[j]

1.249(3)

– – 0.086 25

8 [5]tBuPh 1.342(2) 1.255(2) 1.309(3)[f] - – 0.054 25

9 [7]tBuPh 1.345(3)

1.347(3)[k]

1.254(3)

1.252(3)[l]

1.302(3)

1.306(3)[m] 1.252(3) - 0.052 25

10 [5]Mes 1.339(2) 1.255(2) 1.303(3)[f] – – 0.048 25

11 [7]Mes 1.334(3) 1.260(3) 1.299(3) 1.257(4)[n] – 0.042 25

12 [9]Mes 1.330(3) 1.255(3) 1.298(3) 1.260(4) 1.298(5) 0.038 25

13[o] [7]H 1.319 1.274 1.289 1.275 - 0.014 22

14[p] [7]H 1.310342 1.266802 1.281530 1.267928 - 0.0136 21

15[o] [9]H 1.319 1.274 1.289 1.277 1.287 0.010 22

16[p] [9]H 1.310424 1.267176 1.281012 1.268959 1.279270 0.010 21

17[q] [11]H - - - - - 0.009[r] 23

18[q] [19]H–[39]H - - - - - 0.006[s] 23

19[p] [29]H - - - - - 0.004[t] 21

[a] See Figure 4.5 for atomic numbering scheme. [b] Calculated as difference in bond length

between the two central-most bonds. For non-centrosymmetric structures, BLA has been

calculated using the average of positionally equivalent bonds. [c] Structure determination at

20 °C. [d] Structure determination at –160 °C. [e] ESDs not reported in refs[27,30]. [f]

C3−C3’. [g] Averaged in case of multiple determination (see ref[27]). [h] C2−C2’. [i]

Structure determination at 22 °C. [j] C3−C4. [k] C7−C8. [l] C6−C7. [m] C5−C6. [n] C4−C4’.

[o] Geometry optimization at B3LYP/6-31G* level. [p] Geometry optimization at

PBE1PBE/cc-pVTZ level. [q] Geometry optimization at B3LYP/TZVPP level. [r] Bond

length values of 1.280 Å and 1.271 Å. [s] Bond length values of 1.278 Å and 1.272 Å. [t]

Estimated (bond lengths are not given for this value in ref[21]).

137

A plot of cumulene BLA values versus the number of double bonds n is depicted in

Figure 4.10 for [n]Cy, [n]Ph, [n]Mes, and [n]tBuPh. The graph suggests a limiting BLA

value of 0.03–0.05 Å for [n]Mes and [n]tBuPh (the inserted lines are only a guide for the

eye). For [n]Cy and [n]Ph, only two data points for each are available hindering extrapolation

to afford an suggested asymptotic limit. Noticeable, however, is that the BLA values of the

[n]Ph series with n = 3, 5 are slightly increased compared to [n]Mes and [n]tBuPh. These

predictions are, however, still relatively higher than that predicted by theory for the “parent”

series [n]H. While computational results can differ depending on the method of analysis, the

trend appears clear that BLA ≤ 0.01 Å by the length of [9]H. The difference between

experiment and theory likely arises from endgroup effects, although confirmation of this

hypothesis is still necessary.

Figure 4.10 BLA values (inset) versus chain length n (lines are only a guide for the eye).

138

4.3 Theoretical studies including comparison to UV/vis spectroscopy and BLA

analysis

In a collaborative project, theoretical studies have been performed for [9]cumulenes by

Görling and coworkers,37 dealing with BLA, twist angles, and electronic absorptions. In the

case of [9]tBuPh, a simplified structure, i.e., [9]MePh is used since [9]tBuPh is too large for

SCS-MP2 and CC2 calculations (Figure 4.11). The substitution change from t-Bu to Me,

however, is expected to have a negligible influence on the molecular structure and low-lying

excitation energies.

Figure 4.11 Optimized geometries of [9]tBuPh and [9]MePh cumulenes.

4.3.1 Influence of twist angles on BLA and electronic absorption energy

The influence of the twist angle between the aryl endgroups and the cumulene chain

on the BLA is depicted in Figure 4.12. The two lines describe two different calculation

methods, that have been performed, i.e., B3LYP (red) and SCS-MP2 (blue). The same trend is

predicted by both methods, namely a decrease of BLA by increasing the twist angle values

reaching a BLA between 0.01 and 0.02 Å for twist angles of 90°. In contrast, twist angle

139

values of 10° give a much higher BLA with ca. 0.034 Å using B3LYP and ca. 0.062 Å using

SCS-MP2.38 These results confirm that conjugation between the aryl endgroups and the

cumulene chain does indeed influence the BLA values of cumulenes.

Figure 4.12 BLA calculation of [9]Ph versus the aryl twist angle.

HOMO–1, HOMO, LUMO, and LUMO+1 for [9]tBuPh and [9]Mes, with twist

angles fixed to 32° and 49°, respectively, calculated via the B3LYP method are outlined in

Figure 4.13. The HOMO and LUMO of both [9]cumulenes show that conjugation of aryl

endgroups with the cumulene chain is present. In contrast, HOMO and LUMO of [9]Ph with

a twist angle of exact 90° (Figure 4.14) show that, as expected, the endgroups do not

conjugate to the cumulene core.

140

Figure 4.13 Left: [9]tBuPh cumulene: B3LYP-MOs for equilibrium geometry (aryl twist

angle of 32°). Right: [9]Mes cumulene: B3LYP-MOs for SCS-MP2/def2-TZVPP equilibrium

geometry (aryl twist angle of 49°).

141

Figure 4.14 Hartree-Fock-MOs for [9]Ph with an aryl twist angle of 90°.

The influence of the twist angles also affects the lowest energy absorption λmax for

[n]cumulenes (Figure 4.15). A red-shift of λmax with values ranging from ca. 500 nm � 550

nm � 630 nm � 680 nm39 for twist angle values of 90°, 60°, 30°, and 10°, respectively, is

observed for one possible conformer of [9]Ph (with D2 symmetry).40 In contrast, the shift of

absorption values in the high energy region are not dramatically affected by the magnitude

twist angles. The intensities of the higher energy absorptions, however, seem to increase as a

function of the twist angle value, while the intensities of the λmax values decrease by

increasing twist angle.

142

Figure 4.15 UV/vis spectra of one possible conformer of [9]Ph (D2 symmetry)40 in

dependence of the phenyl twist angle calculated at CC2/def2-TZVPP//SCS-MP2/def2-TZVPP

level.

4.3.2 UV/vis spectroscopy – Theory and experiment

Experimentally obtained UV/vis spectra of [9]tBuPh and [9]Mes, and the

corresponding theoretical spectra (calculated at the CC2/def2-TZVPP//SCS-MP2/def2-

TZVPP level of theory) are depicted in Figure 4.16. Herein, [9]MePh is used instead of

[9]tBuPh for theoretical comparisons (see Figure 4.11). Both, the lower and the higher energy

region of the theoretical spectra fit well to the experimentally measured UV/vis spectra of the

[9]cumulenes. Furthermore, theory suggests that the high-energy absorptions between 300

and 400 nm are dominated by HOMO–1 to LUMO+1 transition of the cumulene chain (see

Figure 4.13). The absorption bands in the lower energy region correspond mainly to HOMO

to LUMO transitions, which are also influenced by the endgroups, as shown in Figure 4.13.

Much stronger coupling of the cumulene π-system to the aryl rings is observed for the

HOMO-LUMO transitions compared to the high energy region, and the resulting fine

structure in the experimental spectrum can be attributed to several cumulene chain coupled

143

vibrations such as, for example, C-C stretch vibrations of the aryl rings coupled to C-H

deformation and C-C stretch vibrations of the cumulene chain.

Figure 4.16 Calculated and experimental UV/vis spectra of [9]MePh/[9]tBuPh (top) and

[9]Mes (bottom) cumulenes. The twist angles are 31° for [9]MePh, as well as 49° for [9]Mes.

All theoretical UV/vis spectra have been computed at the CC2/def2-TZVPP//SCS-MP2/def2-

TZVPP level of theory.

144

4.4 Electrochemistry (cyclic voltammetry) including comparison of the

electronic band gap (Eele) to the optical band gap (Eopt)

4.4.1 Introduction

To date, very little is known about electrochemical analysis of carbyne. While

polyynes have been sporadically analyzed via cyclic voltammetry (CV), e.g., by Tykwinski,41

Gladysz,42–44 and Hirsch,45 relatively little has been reported for cumulenes. To date, several

reports on CV studies on lower [n]cumulenes (n ≤ 3) have been published, for example, by

Kemula and Kornacki46 as well as recently by Zhu and coworkers.47 Furthermore, CV data

have been briefly discussed for six higher [n]cumulenes, i.e., [5]cumulenes with ferrocenyl

([5]Fc),20,48 ferrocenyl/phenyl ([5]Fc/Ph),49 phenyl ([5]Ph),20,50 t-butyl/phenyl endcapping

groups ([5]tBu/Ph),51 [5]tBu,51 and [5]EtPh.51

4.4.2 Cyclic voltammetry of [3]tBuPh, [5]tBuPh, and [7]tBuPh

The synthetic accessibility and stability of the cumulene series [n]tBuPh with n = 3, 5,

and even 7 has facilitated electrochemical studies. The cyclic voltammetry experiments have

been carried out in CH2Cl2 as solvent with 0.1 M Bu4NPF6 as supporting electrolyte using a

standard three-electrode cell. A glassy carbon disc has been used as working electrode, and

ferrocene has been added to the samples serving as internal standard. Figures 4.17–4.20

present the cyclic voltammograms of [3]tBuPh, [5]tBuPh, and [7]tBuPh, respectively, and

the according data for oxidation and reduction potentials is summarized in Table 4.7. For all

three [n]tBuPh cumulenes (n = 3, 5, 7), two separate oxidation steps and at least one

reduction process are observed. For [3]tBuPh, the first oxidation process (E1/2 = 0.49 V) is

reversible52 while the second oxidation process (E1/2 = 0.95 V) shows a quasireversible level

(Figure 4.17). In the case of [5]tBuPh (Figure 4.18), oxidation processes are similar giving a

reversible first oxidation step (E1/2 = 0.43 V) and a quasireversible second oxidation step (E1/2

= 0.80 V). For [7]tBuPh, different results occur, i.e., the first oxidation process (E1/2 = 0.42 V)

is quasireversible (Figure 4.19), while the second oxidation process (E1/2 = 0.68 V) appears to

be reversible.

145


Ferrocene (Fc) has been used as the internal standard. Scan rate: 150 mV/s.



146



A single, quasireversible reduction step (E1/2 = –2.18 V) is found for [3]tBuPh (Figure

4.17). In contrast, [5]tBuPh shows two reduction processes, the first one (E1/2 = –1.72 V) is

reversible and the second reduction step (E1/2 = –2.19 V) is quasireversible step (Figure 4.18).

For [7]tBuPh, two reversible reduction processes (E1/2 = –1.37 and –1.72 V) are observed

(Figure 4.19). For the reduction processes of [n]tBuPh, a trend is observed showing an

increase of reduction steps accompanied by an increase of reversibility with increasing chain

length.

4.4.3 Comparison to known cumulene systems

Comparisons of the electrochemical results to known cumulene systems are limited

due to few and poorly investigated CV studies of cumulenes reported to date. Kemula and

Kornacki, for example have studied the reduction of the [3]Ph cumulene and present CV

voltammograms of the resulting reduction processes at –1.30, –1.63, –2.22, and –2.46 (in

DMF, referred to S.C.E.).46 Furthermore, Hoijtink and van der Meij have also reported about

147

the reduction of [3]Ph as well as [5]Ph, which show two reductions – at 1.20 and 1.03 V, as

well as, 1.55 and 1.30 V, respectively (in 1,2-dimethoxyethane, relative to the reduction

potential of biphenyl).50,53 In both cases, no oxidation processes are discussed. Bildstein and

coworkers have reported the synthesis and characterization of ferrocenyl-substituted

cumulenes [n]Fc also including CV studies.20,49 The redox properties of these cumulenes

depend on the length of the cumulenic chain as well as on the number of double bonds (odd-

versus even-numbered cumulenes). Thus, strong electronic communication between the

electron-donating ferrocenyl groups through the bridging cumulene chain offers interesting

CV characteristics including both oxidation and reduction processes.20,49 The group of

Bildstein20 has performed cyclic voltammetry of [5]Ph that serves as a reference to the

ferrocenyl endcapped cumulenes (Figure 4.20). Herein, two reduction steps, one reversible

two-electron (E1/2 = –1.11 V) and one quasireversible two-electron (E1/2 = –1.43 V) process,

are observed. Furthermore, a partially reversible two-electron oxidation (E1/2 = 1.02 V) is

found, followed by a poorly resolved second oxidation step (Figure 4.20). The cyclic

voltammograms of [5]Ph and [5]tBuPh resemble each other (Figure 4.20 and Figure 4.18,

respectively), both showing two oxidation and reduction events, respectively. The calculated

energy values of the HOMO-LUMO gap of both [5]cumulenes are almost identic with 2.13

eV and 2.15 eV for [5]Ph and [5]tBuPh, respectively. No further comparisons, however, have

been done, since both CV measurements have been recorded under different conditions, such

as concentration, scan rate, temperature, etc.

Figure 4.20 Cyclic voltammogram of 0.001 M [5]Ph. 0.2 M electrolyte (Bu4NPF6) in

CH2Cl2, referenced to SCE. Scan rate 200 mV/s. The graphic is adapted from ref[20].

148

Suzuki and coworkers have reported CV studies on [5]tBu/Ph and compared these

results to the two [5]cumulenes [5]tBu and [5]EtPh (Table 4.6).51 The voltammograms of

[5]tBu show no reduction (in the region reaching –2.2 V) but an irreversible oxidation

process. Replacement of two t-butyl groups by two phenyl groups yielding [5]tBu/Ph leads to

a reversible reduction step as well as two reversible oxidation steps. [5]EtPh containing only

aryl endgroups, shows two reduction events with the first one being reversible and the second

irreversible. Two oxidation steps are observed, with the first one being irreversible and second

reversible. The redox potentials of [5]tBu/Ph are positioned between those of the [5]tBu and

the aryl-substituted [5]EtPh or [5]tBuPh, which is logical due to the respective substituent

effects. A comparison of the [5]cumulenes [5]tBu, [5]tBu/Ph, [5]EtPh, and [5]tBuPh shows

an increased number of reduction events with increasing number of aryl rings as endgroups.

In addition, a facilitated reduction is observed. The increase of alkyl substituents on the aryl

rings results in a more difficult first reduction comparing [5]EtPh and [5]tBuPh (see Table

4.6). A similar behavior is observed when comparing the first oxidation processes of [5]tBu,

[5]tBu/Ph, [5]EtPh, and [5]tBuPh. By increasing the number of aryl endgroups, the oxidation

processes become easier. The alkyl groups on the aryl rings seem to influence the oxidation

events in an opposite way compared to the reduction events. The increase of alkyl substituents

on the aryl rings results in an easier first oxidation resulting in [5]tBuPh as the [5]cumulene

with the lowest oxidation potential (Table 4.6).

149

Table 4.6 CV data of [5]cumulenes [5]tBu, [5]tBu/Ph, [5]EtPh,[a] and [5]tBuPh.[b]

reduction potential (V) oxidation potential (V)

[5]tBu51 n.d. (< –2.2) +0.84irr

[5]tBu/Ph51 –1.93 +0.78, +1.00

[5]EtPh51 –1.92irr, –1.58 +0.51irr, +0.72

[5]tBuPh –2.19qurev, –1.72 +0.43, +0.80 [a] Conditions: CH2Cl2, –20 °C, 0.1 M Et4NPF6, 0.001 M [5]cumulene, scan rate = 100 mV/s,

irr = irreversible (referenced to ferrocene). [b] Conditions: CH2Cl2, rt, 0.1 M Bu4NPF6,

0.002 M [5]cumulene scan rate = 150 mV/s, qurev = quasireversible (referenced to ferrocene).

4.4.4 Cyclic voltammetry of a [9]cumulene rotaxane

A cyclic voltammogram has been recorded for the [9]cumulene rotaxane 3.9 (Figure

4.21). Herein, only an irreversible oxidation (Epeak = 0.43 V) is observed aside from two

reversible reduction steps (E1/2 = –1.20 and –1.56 V). The lack of a reversible oxidation

process for [9]cumulene rotaxane 3.9 might be explained by the presence of the macrocycle,

which on the one hand could primary be oxidized compared to the cumulene framework. On

the other hand, the reason for the irreversibility could rely on a reaction between the

macrocycle with the resulting cation of the cumulene. Due to insufficient amount of 3.9, no

repeated or concentration-varied measurement has been made. Regarding the reduction

behavior, the CV studies of 3.9 show that the [9]cumulene rotaxane fits well to the series of

[3]tBuPh, [5]tBuPh, and [7]tBuPh cumulenes giving the longest cumulene with the lowest

reduction potential, i.e., the reduction of [n]cumulenes becomes facilitated as a function of

cumulene chain length.

150

Figure 4.21 Cyclic voltammogram of rotaxane 3.9. 0.1 M electrolyte (Bu4NPF6) in CH2Cl2.


4.4.5 Electronic and optical band gap of [n]tBuPh

Table 4.7 presents CV data including absorption energies and absorption cutoff (taken

from the UV/vis spectra), the optical (Eopt) and electronic (Eele) band gap values in eV, as well

as the potentials for oxidation and reduction processes of [3]tBuPh, [5]tBuPh, [7]tBuPh, and

[9]cumulene rotaxane 3.9. The comparison of the first oxidation steps reveals that there is a

slightly facilitated first oxidation process with increasing chain length of the cumulenes with

potentials of 0.49, 0.43, and 0.42 V for [3]tBuPh, [5]tBuPh, and [7]tBuPh, respectively. The

second oxidation event, however, appears significantly facilitated when the chain is

lengthened (0.95, 0.80, and 0.68 V for [3]tBuPh, [5]tBuPh, and [7]tBuPh, respectively). The

first reduction process for all four cumulene compounds shows that the cumulene can be

reduced easier when the cumulene chain is elongated, i.e., –2.18, –1.72, –1.37, and –1.20 V

for [3]tBuPh, [5]tBuPh, [7]tBuPh, and [9]cumulene rotaxane 3.9, respectively. This feature

is also common for reported polyyne structures.44,45 Furthermore, the second reduction

151

process for [5]tBuPh, [7]tBuPh, and [9]cumulene rotaxane 3.9 shows the same trend, i.e.,

facilitated reduction with increasing chain length.

Comparison of the optical band gap values (Eopt) of [3]tBuPh, [5]tBuPh, and

[7]tBuPh determined from the UV/vis spectra and the electronic band gap values (Eele) from

CV measurements shows a high accordance (Table 4.7). The biggest difference is observed

for [7]tBuPh with 1.97 eV (Eopt) versus 1.79 eV (Eele).

Table 4.7 Selected UV/vis spectroscopic and electrochemical details including optical (Eopt)

and electronic (Eele) band gap values for [3]tBuPh, [5]tBuPh, [7]tBuPh, and 3.9.

compound [3]tBuPh [5]tBuPh [7]tBuPh 3.9

λmax abs. energies (nm),

(ε in M–1cm–1) 426 (37,900) 510 (66,700) 573 (65,200) 666 (38,600)

λmax abs. cutoff (nm),

(Eopt in eV)[a] 470 (2.64) 550 (2.25) 628 (1.97) 711 (1.74)

Eele (eV)[b] 2.67 2.15 1.79 1.63[c]

E1/2 ox2 0.95 0.80 0.68 -

Epeak ox1 - - - 0.43

E1/2 ox1 0.49 0.43 0.42 -

E1/2 red1 –2.18 –1.72 –1.37 –1.20

E1/2 red2 - - –1.72 –1.56

Epeak red2 - –2.19 - -

[a] As estimated from the intercept of a tangent line to the lowest energy absorption of the

UV/vis spectrum with the x-axis. [b] As estimated from cyclic voltammetry, Eox1 – Ered1. [c]

estimated value using the potentials of Epeak ox1 and E1/2 red1.

Figure 4.22 shows the plot of the electronic (Eele) and optical (Eopt) band gaps versus

1/n of the cumulene series [n]tBuPh with n = 3, 5, and 7 and 3.9. Regarding endgroup

influences in the shorter cumulene, [3]tBuPh, only the higher cumulenes with n = 5, 7 and

[9]cumulene rotaxane 3.9 have been considered for the extrapolation of data in Figure 4.22. In

both cases, a linear correlation between the HOMO-LUMO gaps and 1/n is observed. A linear

dependence is common in previous studies of conjugated oligomers, as described, for

example, by Hirsch,45 Chance,54 Diederich,55,56 and Bäuerle.57 The extrapolation of the data to

infinite cumulene chain length predicts a HOMO-LUMO gap of 0.96 eV and 1.13 eV for the

electronic (Eele) and optical (Eopt) band gaps, respectively, which gives λmax values of 1287

152

and 1097 nm, respectively, for an infinite cumulenic chain. In contrast to the applied

extrapolation method using the plot of E versus 1/n, the exponential function proposed by

Meier and coworkers16 represents a more significant extrapolation method giving a better

estimate for infinite cumulenes. In order to make this estimate, however, data for longer

cumulenes (n > 9) are necessary.

Figure 4.22 Plots of electronic (left) and optical (right) band gaps (Eele and Eopt, respectively)

versus 1/n for the cumulenes [3]tBuPh, [5]tBuPh, [7]tBuPh, and 3.9 (band gap data taken

from Table 4.7).

4.4.6 Comparison of electrochemical properties of cumulenes and polyynes

Regarding the electrochemical behavior of polyynes, no oxidation processes are

expected for aryl-substituted polyynes because of their electron deficient nature. For example,

Hirsch’s aryl-endcapped dendrameric polyynes show only reduction processes without any

hint for oxidation.45 The endcapping groups show that there is a strong influence or

interaction of the endgroups with the polyyne chain and thus its π-system. Tykwinski and

coworkers, hence, have suggested that by the use of alkyl-endcapping groups, CV

measurements of polyynes should give more suitable redox characteristics of the “pure”

polyyne chain due to lack of interactions between the endgroups and the chain framework.41

Thus, polyynes with adamantyl endgroups have been examined by CV including the tetrayne,

hexayne, and octayne.41 The results vary showing no redox event, oxidation and reduction

events, as well as only a reduction event for the tetrayne, hexayne, and octayne, respectively.

Thus, no concrete statements have yet been done concerning the influence of aryl- or alkyl

endgroups on the redox activity of polyynes.

153

Regarding the electrochemical behavior of cumulenes, however, both, oxidation and

reduction events occur contrary to polyynes. A preliminary trend can be drawn for the series

of [n]tBuPh with n = 3, 5, and 7, including [9]cumulene rotaxane 3.9, as well as for various

[5]cumulenes: With increasing chain length and increasing number of aryl endgroups attached

to the cumulene chain, reduction and oxidation processes become facilitated and more

distinct, and also the number of redox events increases.

4.5 NMR spectroscopy of [n]cumulenes

4.5.1 Introduction

With a rigid framework of sp-hybridized carbons that can be unstable for longer

derivatives, 13C NMR spectroscopy of both cumulenes and polyynes can be challenging due

to long T1 relaxation times, insolubility of the sample, and decomposition of the material

during acquisition. This is particularly true for the analysis of cumulenes. Whereas several

very recent studies provide more details for polyynes,32,58–61 no 13C NMR spectroscopic data

are reported for cumulenes beyond the length of a [5]cumulene, and even data describing

[5]cumulenes are scarce. Iyoda has described the characterization of several [5]cumulene

derivatives by 13C NMR spectroscopy, however, comparisons to longer or shorter derivatives

are not available.62,63

During the doctoral thesis research, a 13C NMR spectrum of one longer [n]cumulene

with n > 5, namely [7]tBuPh has been recorded for the first time.25 The next higher odd-

numbered cumulene, the [9]cumulene [9]tBuPh is already too unstable for NMR

characterization. Thus, the stabilization of this [9]cumulene by rotaxane formation to yield 3.8

and 3.9 offers the first opportunity to explore 13C NMR spectroscopy for longer derivatives.

Since the 1H NMR spectrum is not significant for the discussion of the chemical shift of the

sp-hybridized cumulene chain, only the 13C NMR spectra of [9]cumulene rotaxanes 3.8 and

3.9 will be discussed herein.

The second part of this section deals with 13C-1H correlation NMR spectroscopy, e.g.,

HMBC spectra of the series of [n]tBuPh including [9]cumulene rotaxanes. 13C-1H correlation

studies have been done to assign the shifts of several carbon atoms of the cumulenes including

also the shifts of some carbon atoms within the cumulene chain that have been identified via

this method.

154

4.5.2 13C NMR spectroscopy of [9]cumulene rotaxanes and their precursors

Figure 4.23 shows the carbon atom labeling of precursors 3.4 and 3.6 and the

[9]cumulene rotaxanes 3.8 and 3.9 that will be used for the NMR spectroscopic discussion.

Figure 4.24 shows 13C NMR spectra of [9]cumulene rotaxane 3.9 and precursor 3.6 with the

carbon atoms of the sp-hybridized carbon chain (i.e., C1–C5) and the appropriate chemical

shifts marked in green. The signals for the carbon atoms C1–C5 of the cumulene chain of

[9]cumulene rotaxane 3.9 (top, Figure 4.24) are shifted downfield compared to the carbon

atoms C1–C5 of the acetylene units of precursor 3.6 (bottom, Figure 4.24), from 83–63 ppm

for 3.6 (acetylenic region) to 145–120 ppm for 3.9 (cumulenic region). The downfield region

is also common for the cumulenic carbon atoms of lower [n]cumulenes with n ≤ 7.25 The

appearance of the cumulene and polyyne resonances in completely different regions seems

rather surprising, since both carbon chains consists of sp-hybridized carbon atoms.

Figure 4.23 Carbon atom labeling of precursors and [9]cumulene rotaxanes for NMR

spectroscopic discussion.

155

Figure 4.24 Comparison of 13C NMR spectra of [9]cumulene rotaxane 3.9 and precursor 3.6

(dotted lines highlight the assignment of signals in the spectrum of [9]cumulene rotaxane 3.9

that result from precursor 3.6, which is present as an impurity).

Comparison of [9]cumulene rotaxanes 3.9 and 3.8 shows that the chemical shift of

cumulenic carbon atoms C1–C5 reveals similarities as outlined in Figure 4.25. The cumulenic

carbon signals for both rotaxanes can be divided into three different regions. On the one hand,

one set of three carbon atoms can be assigned into the most upfield region of 125–119 ppm,

whereas, on the other hand, the fourth carbon atoms appear shifted downfield at 128.7 ppm

and 127.4 ppm for 3.9 and 3.8, respectively. The last remaining carbon atom signals are

notably deshielded to 144.8 ppm and 144.5 ppm for 3.9 and 3.8, respectively.

156

Figure 4.25 13C NMR spectra (165–100 ppm region) of [9]cumulene rotaxanes 3.9 (top) and

3.8 (bottom).

4.5.3 13C- and correlation NMR spectroscopy of [n]tBuPh (n = 3, 5, and 7) and

[9]cumulene rotaxanes

The following three Figures 4.26–4.28 present decoupled and coupled 13C NMR

spectra of [n]tBuPh (n = 3, 5, and 7) as well as the two-dimensional heteronuclear multiple

bond coherence (HMBC) spectra. The conventional decoupled spectra as well as the coupled

spectra have been performed to assign scalar couplings between H and C atoms in order to

assign the terminal cumulenic carbon atoms. The cumulenic carbon atoms are marked in red,

and the terminal carbon atom of the cumulene chain is labeled as C1. In all three figures, there

is only one carbon atom of the cumulene chain that shows a splitting pattern in the coupled

spectra defining a triplet, while all remaining carbon atoms describe singlets with the same

chemical shift as in the appropriate decoupled spectra. This feature leads to the assumption

that only 3JCH couplings are observed, while 4

J couplings are probably too small to be

detected for such systems. Using this same analysis and logic, carbon atoms C1 of all three

cumulenes, [3]tBuPh, [5]tBuPh, and [7]tBuPh can be assigned to 123.4, 125.6, and 127.5

157

ppm, respectively. The recorded HMBC spectra that are outlined in Figures 4.26–4.28

confirm the chemical shifts of C1 via a correlation to the aryl protons H2 via 3JCH couplings.

158


corresponding HMBC NMR spectrum (aryl region).

159



160



161

[9]Cumulene rotaxanes 3.9 and 3.8 have also been examined via HMBC experiments,

and the corresponding spectra are depicted in Figure 4.29 and Figure 4.30, respectively. With

the help of the HMBC spectra, the position of outermost carbon atoms C1 can be identified

and shows chemical shifts of 128.7 and 127.4 ppm for 3.9 and 3.8, respectively. These signal

positions have also been determined through 3JCH couplings between the protons H2 and

carbon C1 as mentioned before. To date, the C1 remains the only carbon atom of the

cumulenic chain that could be assigned using the HMBC experiment.


correlation signals between H2 and C1).

162


correlation signals between H2 and C1).

4.5.4 Discussion and comparison

In the case of polyynes, the chemical shifts of C1 and C2 (first and second sp-

hybridized carbon atoms of the chain, respectively) follow two interesting trends, which have

been reported on both, an experimental and theoretical basis.32,58–60 On the one hand, TIPS-

substituted polyynes, for example, show that the signals for C1 steadily shift downfield with

increasing chain length while the signals for C2 show the exact opposite behaviour, namely

shifting slightly upfield, indicating differently polarized carbon atoms.32 On the other hand,

the opposite trend has been observed in phenyl-58 and t-butyl-,61 as well as donor- and

acceptor-60substituted polyynes, where the signals for C1 steadily shift upfield with increasing

chain length while the signals for C2 remain almost the same shift without significant trend.

In the case of TIPS- and phenyl substitution, the remaining carbon atoms of the chain are

163

positioned in the most upfield region. The described phenomenon, however, is yet not

completely understood, and the trend of this effect seems to depend on the substituents of the

sp-carbon chain.

In Figure 4.31, the 13C-NMR signals of the cumulene chain in the series of [n]tBuPh

cumulenes (n = 3, 5, and 7) as well as [9]cumulene rotaxanes 3.8 and 3.9 are shown. As

discussed before for the rotaxanes 3.8 and 3.9, “naked” cumulenes show also three regions

with cumulenic 13C signals, one, that is the most downfield shifted region (between 145 and

152 ppm, region I), the second one describes a range between 123 and 129 ppm (region II),

and the last one is shifted upfield between 126 and 120 ppm (region III). The 13C signals of

the first two regions seem to converge with increasing length (n) of the cumulene chain

(Figure 4.31).

Figure 4.31 Plot of 13C NMR carbon chemical shifts versus the number of double bonds n for

[n]tBuPh (n = 3, 5, 7) and [9]cumulene rotaxanes 3.8 and 3.9.

The 13C signals of the second region might be assigned to the outermost carbon C1 of

the cumulene chain by HMBC- as well as coupled 13C NMR spectroscopic measurements

(Figure 4.31). It has been assumed that the 13C signals from region I originates from the

second outermost carbon C2, as it is also the case in the 13C NMR study of TIPS-endcapped

164

polyynes showing a similar convergence of carbon signals in which the carbon positions are

determined by 13C labeling.32 The assumption that region I describes C2 resonances is not

random since these two outermost carbons reflects most the effect of endgroups. The

cumulenes, especially longer cumulenes, however, have never been intensely characterized

using 13C NMR spectroscopy so far. Therefore, the origin of these interesting observations

needs to be further investigated.

4.6 Summary and conclusion

In summary, a variety of characterization methods for the series of [n]tBuPh have

been performed, including UV/vis spectroscopy and structural studies via X-ray

crystallographic analysis. In addition, theoretical studies have been done by Görling and

coworkers to help to explain the observed trend in the performed analyses. Furthermore,

cyclic voltammetry and 13C NMR spectroscopy including C-H correlation studies have been

done, to give unprecedented insight into structural and electronic properties of cumulenes.

UV/vis spectroscopy and X-ray crystallography of [n]cumulenes show a correlation

between cumulene chain length and properties, such as HOMO-LUMO gap and BLA, and

both decrease with increasing chain length, i.e., increasing n. An asymptotic or saturated

value for n = 9 in order to achieve a “carbyne-like” molecule, however, has not yet been

reached. By means of some characterization tools, it is confirmed that the structure of

cumulenes (odd-numbered versus even-numbered) as well as the endgroups affect the

properties of cumulenes in predictable trends. Theoretical studies of cumulenes show good

agreement to experimental results, such as UV/vis data (e.g., optical band gap), bond lengths,

and BLA.

Cyclic voltammetry reveals oxidation and reduction events for the cumulene series

[n]tBuPh including rotaxane 3.9. Trends in both, oxidation and reduction values depend on

cumulene chain length. The influence of chain length on the HOMO-LUMO gap, estimated

from cyclic voltammetry, is also observed to be similar to UV/vis spectroscopic studies.

Finally, NMR spectroscopy has been used to determine the chemical shifts of the first

carbon atoms of the cumulene chain of [n]tBuPh and [9]cumulene rotaxanes using the C-H

correlation technique HMBC. A similar trend of carbon chemical shifts versus n as in the case

of polyynes has been determined giving a converging trend of the shifts of the two outermost

carbons in the chain by increasing the chain length.

165

4.7 References

1 H. Fischer, in The Chemistry of Alkenes (Ed.: S. Patai), John Wiley & Sons, New

York, 1964, pp. 1025–1159.

2 R. R. Tykwinski, W. Chalifoux, S. Eisler, A. Lucotti, M. Tommasini, D. Fazzi, M. Del

Zoppo, G. Zerbi, Pure Appl. Chem. 2010, 82, 891–904.

3 F. Bohlmann, K. Kieslich, Abh. Braunschweig. Wiss. Ges. 1957, 9, 147–166.

4 R. Kuhn, G. Platzer, Ber. Dt. Chem. Ges. 1940, 73B, 1410–1417.


6 K. Brand, A. Busse-Sundermann, Chem. Ber. 1950, 83, 119–128.

7 R. Kuhn, H. Zahn, Chem. Ber. 1951, 84, 566–570.

8 F. Bohlmann, K. Kieslich, Chem. Ber. 1954, 87, 1363–1372.

9 H. Fischer, W. D. Hell, Angew. Chem., Int. Ed. Engl. 1967, 6, 954–955; Angew. Chem.

1967, 79, 931–932.

10 P. Cadiot, Ann. Chim. [Paris] 1956, 13, 214–272.

11 T. Negi, T. Kaneda, H. Mizuno, T. Toyoda, Y. Sakata, S. Misumi, Bull. Chem. Soc.

Jpn. 1974, 47, 2398–2405.

12 L. T. Scott, G. J. DeCicco, Tetrahedron Lett. 1976, 31, 2663–2666.

13 Determined via common conversion tools from http://halas.rice.edu/conversions (nm

to eV).

14 C. Bubeck, in Electronic Materials: The Oligomer Approach (Eds.: K. Müllen, G.

Wegner), Wiley-VCH, Weinheim, 1998, Chapter 8.

15 G. N. Lewis, M. Calvin, Chem. Rev. 1939, 25, 273–328.

16 H. Meier, U. Stalmach, H. Kolshorn, Acta Polym. 1997, 48, 379–384.

17 Determined by doctoral student Michael Franz from the Tykwinski group.

18 W. A. Chalifoux, R. R. Tykwinski, Nat. Chem. 2010, 2, 967–971.

19 W. A. Chalifoux, R. McDonald, M. J. Ferguson, R. R. Tykwinski, Angew. Chem. Int.

Ed. 2009, 48, 7915–7919; Angew. Chem. 2009, 121, 8056–8060.



21 F. Innocenti, A. Milani, C. Castiglioni, J. Raman Spectrosc. 2010, 41, 226–236.

22 D. Nori-Shargh, F. Deyhimi, J. E. Boggs, S. Jameh-Bozorghi, R. Shakibazadeh, J.

Phys. Org. Chem. 2007, 20, 355–364.


166

24 Based on a search of WebCSD, see http://webcsd.ccdc.cam.ac.uk/ on 11/07/14, for

equally substituted [n]cumulenes with n = odd and alkyl or aryl endgroups. Results do

not include [5]Ph (for the structure of [5]Ph, see M. M. Woolfson, Acta. Cryst. 1953,

6, 838–841.)


Angew. Chem. Int. Ed. 2013, 52, 1817–1821; Angew. Chem. 2013, 125, 1862–1867.

26 [3]tBuPh, [5]tBuPh, [7]tBuPh, and [5]Mes have been prepared by myself or under

my supervision (as in the case of [5]Mes) while [7]Mes and [9]Mes have been

prepared by Dominik Wendinger and used in this thesis for better comprehension,

discussion, and comparison.

27 H. Irngartinger, W. Götzmann, Angew. Chem. Int. Ed. Engl. 1986, 25, 340–342;

Angew. Chem. 1986, 98, 359–361.

28 Z. Berkovitch-Yellin, L. Leiserowitz, Acta Crystallogr. Sect. B 1977, 33, 3657–3669.

29 The structure of [5]Ph has been reported without bond length or angle values see: M.

M. Woolfson, Acta. Cryst. 1953, 6, 838–841. Bond lengths and angles from J. A.

Januszewski, F. Hampel, R. R. Tykwinski, unpublished, CCDC 817302.

30 H. Irngartinger, H.-U. Jäger, Angew. Chem. Int. Ed. Engl. 1976, 15, 562–563; Angew.

Chem. 1976, 88, 615–616.

31 S. Szafert, J. A. Gladysz, Chem. Rev. 2006, 106, PR1–PR33.

32 S. Eisler, A. D. Slepkov, E. Elliott, T. Luu, R. McDonald, F. A. Hegmann, R. R.

Tykwinski, J. Am. Chem. Soc. 2005, 127, 2666–2676.

33 Exceptions include: a) T. Kawase, N. Nishigaki, H. Kurata, M. Oda, Eur. J. Org.

Chem. 2004, 3090–3096. b) Y. Kuwatani, G. Yamamoto, M. Iyoda, Org. Lett. 2003, 5,

3371–3374.

34 Calculated as the distance from H-atom to a plane generated from the six atoms of the

aromatic ring.

35 S. Grimme, P. R. Schreiner, Angew. Chem. Int. Ed. 2011, 50, 12639–12642.

36 Herein, the twist angles of [3]tBuPh are not included due to the specific bending of

the cumulene chain that might be caused by intramolecular C–H/π-interactions.

37 Unpublished results by A. Görling and coworkers in collaboration with R. R.

Tykwinski and coworkers.

38 Values have been estimated from the graph in Figure 4.12.

39 Values have been estimated from the graph in Figure 4.15.

167

40 Two conformers for [9]Ph (D2 and C2h symmetry, respectively) that differ in the

orientation of the phenyl groups have been used for calculations performed by Görling

and coworkers. Both conformers show a similar minimum potential energy (difference

of only 0.13 kJ/mol). The theoretical results show no difference between properties of

the two conformers.

41 W. A. Chalifoux, M. J. Ferguson, R. McDonald, F. Melin, L. Echegoyen, R. R.

Tykwinski, J. Phys. Org. Chem. 2012, 25, 69–76.

42 R. Dembinski, T. Bartik, B. Bartik, M. Jaeger, J. A. Gladysz, J. Am. Chem. Soc. 2002,

122, 810–822.

43 W. Mohr, J. Stahl, F. Hampel, J. A. Gladysz, Chem. Eur. J. 2003, 9, 3324–3340.

44 Q. Zheng, J. C. Bohling, T. B. Peters, A. C. Frisch, F. Hampel, J. A. Gladysz, Chem.

Eur. J. 2006, 12, 6486–6505.

45 T. Gibtner, F. Hampel, J.-P. Gisselbrecht, A. Hirsch, Chem. Eur. J. 2002, 8, 408–432.

46 W. Kemula, J. Kornacki, Rocz. Chem.: Ann. Soc. Chim. Pol. 1962, 36, 1835–1874.

47 Y. Li, K. Chandra Mondal, P. P. Samuel, H. Zhu, C. M. Orben, S. Panneerselvam, B.

Dittrich, B. Schwederski, W. Kaim, T. Mondal, D. Koley, H. W. Roesky, Angew.

Chem. Int. Ed. 2014, 53, 4168−4172; Angew. Chem. 2014, 126, 4252−4256.

48 B. Bildstein, Coord. Chem. Rev. 2000, 206–207, 369−394.

49 W. Skibar, H. Kopacka, K. Wurst, C. Salzmann, K.-H. Ongania, F. Fabrizi de Biani, P.

Zanello, B. Bildstein, Organometallics 2004, 23, 1024−1041.

50 G. J. Hoijtink, P. H. van der Meij, Z. physik. Chem. 1959, 20, 1−14.

51 N. Suzuki, N. Ohara, K. Nishimura, Y. Sakaguchi, S. Nanbu, S. Fukui, H. Nagao, Y.

Masuyama, Organometallics 2011, 30, 3544–3548.

52 Reversibility has been investigated using equation ipa/ipc ≈ 1 for reversible systems.

Redox processes with values of ipa/ipc in a range of 0.7–1.2 and peak widths similar to

that of Fc+/Fc (110−120 mV) have been considered reversible. Otherwise, redox

processes have been considered quasireversible or irreversible (only the oxidation

process of compound 3.9 has been considered irreversible due to the missing cathodic

current peak). This is a simplified determination, since it has not yet been assigned if

the redox processes are one- or multielectron processes.

53 Reduction potentials have not been determined by cyclic voltammetry but by another

method using a glass apparatus for preparation of negative ions under high vacuum as

reported by van der Meij in ref[50].

168

54 J. L. Brédas, R. Silbey, D. S. Boudreaux, R. R. Chance, J. Am. Chem. Soc. 1983, 105,

6555–6559.

55 J. Anthony, C. Boudon, F. Diederich, J.-P. Gisselbrecht, V. Gramlich, M. Gross, M.

Hobi, P. Seiler, Angew. Chem. Int. Ed. Engl. 1994, 33, 763–766; Angew. Chem. 1994,

106, 794–798.

56 R. E. Martin, U. Gubler, C. Boudon, V. Gramlich, C. Bosshard, J.-P. Gisselbrecht, P.

Günter, M. Gross, F. Diederich, Chem. Eur. J. 1997, 3, 1505–1512.

57 P. Bäuerle, Adv. Mater. 1992, 4, 102–107.

58 R. R. Tykwinski, T. Luu, Synthesis 2012, 44, 1915–1922.

59 M. Haque, L. Yin, A. R. T. Nugraha, R. Saito, Carbon 2011, 49, 3340–3345.

60 M. Štefko, M. D. Tzirakis, B. Breiten, M.-O. Ebert, O. Dumele, W. B. Schweizer, J.-P.

Gisselbrecht, C. Boudon, M. T. Beels, I. Biaggio, F. Diederich, Chem. Eur. J. 2013,

19, 12693–12704.

61 F. Bohlmann, M. Brehm, Chem. Ber. 1979, 112, 1071–1073.


2188–2208.

63 This paragraph has been adapted from: S. Frankenberger, J. A. Januszewski, R. R.

Tykwinski, in Fullerenes and Other Carbon-Rich Nanostructures (Ed.: J.-F.

Nierengarten), Springer, Berlin, 2014, vol. 159, 219–256.

169

5. Chapter V. Reactions of [n]cumulenes‡‡

5.1 Addition reaction of [5]tBuPh with tetracyanoethylene (TCNE)

5.1.1 Motivation and objective

Cumulenes undergo reactions such as photo- or thermal cycloadditions, either as

dimerizations or with other reactants, such as alkynes and alkenes, e.g., tetrafluoroethylene or

tetracyanoethylene (TCNE). The reactions of shorter [n]cumulenes (n = 2, 3) with TCNE

have been explored by Kawamura and coworkers giving several novel cyclic compounds with

diverse structures.1,2 Recently, the group of Kawamura has reported results on the reaction of

several [3]cumulenes 5.1 with TCNE (5.2) showing the interesting reactivity of a cumulene

chain at consecutive double bonds (Scheme 5.1).3 The reaction gives two different

cycloadducts, 5.3 and 5.4. Cyclic compound 5.3 is the main product and is assumed to be

formed from intermediate 5.5 that derives from a [2 + 2]-cycloaddition reaction at the central

C=C bond supported by π-π-stacking of the aryl rings. Following bond cleavage of the

(NC)2C-C(CN)2 bond, rotation, and rearrangement forms the product 5.3. In addition, a

[4 + 2]-cycloaddition product, compound 5.6, has been suggested as possible intermediate.

‡‡ Portions of this chapter have been published: J. A. Januszewski, F. Hampel, C. Neiss, A. Görling, R. R.

Tykwinski, Angew. Chem. Int. Ed. 2014, 53, 3743–3747.

170

Scheme 5.1 Reaction of [3]cumulenes with TCNE in CH2Cl2 at rt.

When dissolved in MeOH or CH3CN, 5.3 is converted to the bi- and tricyclic

compounds 5.7 and 5.8, respectively (Scheme 5.2). In both conversions, one aryl group takes

part in the reaction. The synthetic route to 5.7 is assumed to run via a rearrangement reaction,

while the formation of 5.8 is part of an equilibrium that becomes irreversible during a final

hydrolysis step.

Scheme 5.2 Conversion of cyclobutane 5.3 to 5.7 and 5.8.

171

Another [2 + 2] reaction of a cumulene with TCNE (5.2) has been reported by

Bildstein and coworkers.4 [5]Cumulene [5]Fc has been treated with TCNE in benzene

showing addition at the β-bond of the cumulene chain to give cycloadduct 5.9 in yields of ca.

41% (Scheme 5.3).

Scheme 5.3 TCNE addition to [5]Fc giving cycloadduct 5.9.

Reactions of TCNE to carbon rich compounds, such as polyynes, have been

investigated in the last decades giving products with interesting optical properties, e.g., push-

pull chromophores described as potential candidates for nonlinear optics.5–8 Reactions of

acetylenes with TCNE have been developed in the group of Diederich5–7,9 and are based on a

sequence of cyclization and cycloreversion reaction steps (Scheme 5.4).10 Herein, a general

example is presented that uses an acetylenic compound containing an electron-donating group

(this group is required for the realization of the reaction), which is treated with TCNE. As

intermediate, a four-membered cyclobutene derivative is formed via cycloaddition, which is

transformed to donor-substituted tetracyanobutadiene (TCBD) derivatives via electrocyclic

ring opening.

Scheme 5.4 General cyclization/cycloreversion method giving donor-substituted TCBD

derivatives.

Diederich and coworkers,5–7 and others8 describe the [2 + 2]-cycloaddition reaction of

polyynes with TCNE as a very efficient “click reaction” to form new chromophores with

outstanding stability and unique optical properties. This successful work in the subject of

polyyne chemistry gives rise to the question whether a [5]cumulene might be converted to a

[3]cumulene (5.10) via the cyclization and cycloreversion method using TCNE (Scheme 5.5).

172

Scheme 5.5 Expected cyclization/cycloreversion reaction of a [5]cumulene with TCNE to the

unsymmetrically substituted [3]cumulene 5.10.

[3]Cumulenes, such as 5.10, containing donor and acceptor endgroups, are push-pull

systems and are assumed to possess outstanding properties.7,11 In addition, they are predicted

to constitute more potent nonlinear optical chromophores than the appropriate oligoenes and

oligoynes.12,13 To date, several push-pull [3]cumulenes have been synthesized in order to

mainly examine whether these compounds exist in a covalent form (cumulene form) or a

dipolar form (acetylene form). Once formed, the dipolar character has been examined by, e.g.,

X-ray crystallography and NMR spectroscopy studies.7,11,14–18 Furthermore, cumulenes such

as 5.10 show a “chameleonic”, rather acetylene-like reactivity including cycloaddition

reactions as well as polar additions of electrophilic and nucleophilic reactants.7,15

Reported studies of cumulene reactivity offer little guidance as to whether the

regiochemistry of a cycloaddition reaction of a [5]cumulene with TCNE would be governed

by steric or electronic factors. Thus, density-functional calculations have been performed

based on [5]cumulene [5]tBuPh.19 These calculations show that the two central-most carbons

(C3) bear a slight negative charge, while the neighboring carbons (C2) are slightly positively

charged (Table 5.1). This charge distribution may guide the highly electrophilic TCNE toward

the central γ-bond (for bond labeling see Scheme 5.5). The bulky aryl groups at termini are

expected to hinder approach of TCNE to the β-bond.

173

Table 5.1 Natural population analysis (NPA) charges of the cumulenic C-atoms in

[5]cumulene [5]tBuPh.

• • • •Ar

Ar Ar

Ar

[5]tBuPh

C1

C2

C3

C1

C2

C3

C-atom PBE PBE-D3 B3LYP B3LYP-D3

C1 –0.068 –0.068 –0.058 –0.059

C2 +0.053 +0.055 +0.048 +0.050

C3 –0.066 –0.068 –0.064 –0.066

Examination of the frontier molecular orbitals of TCNE and [5]tBuPh shows,

however, that a concerted [2 + 2] addition of TCNE at the γ-bond is forbidden by orbital

symmetry, as expected based on the Woodward-Hoffmann rules for a 4-electron system

(Figure 5.1).20 We have hypothesized that the addition of TCNE to [5]cumulene [5]tBuPh

might occur via a sterically directed stepwise mechanism toward the γ-bond of a [5]cumulene

(Scheme 5.5). A subsequent retro-[2 + 2] reaction from the cyclic intermediate might then

give an unsymmetrically substituted [3]cumulene 5.10. This overall sequence could provide a

useful metathesis reaction to form polarized cumulenes.21

Figure 5.1 HOMO and LUMO of TCNE and [5]tBuPh.

174

5.1.2 Target, synthetic pathway, and test reactions

The reaction of [5]tBuPh and TCNE in CH2Cl2 at rt was explored to form the

unsymmetrically substituted [3]cumulene 5.10a (Figure 5.2). This compound should be

obtained via the cyclobutane derivative 5.11, which has been expected to be formed via

[2 + 2]-cyclization reaction of TCNE and the central double bond of [5]tBuPh, followed by

cycloreversion.

[5]Cumulene [5]tBuPh was treated with TCNE in CH2Cl2 at rt under an argon

atmosphere (Figure 5.2). The reaction mixture was additionally covered with alumina foil,

and it was monitored by TLC analysis resulting in a “virtual rainbow” of different colored

spots. The concentration of several products, however, seemed to be negligible and three

predominant products A, B, and C were observed, which have been discussed in this section

(Figure 5.2). The orange colored spot A seemed to be the main product in this reaction, spot B

was initially orange/brown and turned to green after some minutes. Spot C was initially only

visible under UV light, converting to orange over time. Purification by chromatography

(silica gel, CH2Cl2/hexanes 1:1 or CHCl3) allowed for the isolation of all three fractions A, B,

and C.

175

Figure 5.2 Performed reaction of [5]tBuPh with TCNE, monitored by TLC analysis.

The identity of product A was tackled first. High resolution ESI MS analysis (positive

mode) offered a signal at m/z 979.6557, which was consistent with [[5]tBuPh + TCNE + Na]+

(C68H84N4Na). The aryl region of the 1H NMR spectrum showed triplets at 7.61, 7.53, and

7.49 ppm (integration ratio 1:1:2) and doublets at 7.29, 7.19, and 7.12 ppm (integration ratio

4:2:2, see Figure 5.3). These signals were consistent with four 3,5-di-t-butylphenyl groups,

but also confirmed a loss of symmetry during the course of the reaction. In the 13C NMR

spectrum, there was a noteworthy resonance at 203.4 ppm (Figure 5.4), while the IR spectrum

revealed a weak signal at 1920 cm–1; both suggestive of an allene moiety.

176

Figure 5.3 1H NMR spectrum of product A (inset: expansion of the aryl region).

Figure 5.4 13C NMR spectrum of product A, allene signal highlighted; * = Et2O.

177

Finally, overlaying a solution of A in CH2Cl2 with either EtOH or MeOH afforded

crystals, and structural analysis (Figure 5.5) showed the ethyl and methyl enol ethers 5.12 and

5.13, respectively. Thus, product A could be confidently assigned as cyclic [3]dendralene 5.14

(Figure 5.6).22

MeOH

NC CN

NC CN

O

EtOH

NC CN

NC CN

O

product A

5.12 5.13

Figure 5.5 Top: Overlaying of product A (in CH2Cl2) with EtOH and MeOH giving adducts

5.12 and 5.13, respectively. Bottom: ORTEP drawings (20% probability level) for compounds

5.12 and 5.13.

178

A

C

B

•

NC CN

CNNC

5.14

CH2Cl2/hexanes

1/1

Figure 5.6 Identification of product A as cyclic [3]dendralene 5.14.

Secondly, the identity of product B was investigated. ESI MS analysis showed a

parent peak at m/z 979.6576 ([[5]tBuPh + TCNE + Na]+), analogous to that observed for

product A (i.e., compound 5.14). The aryl region of 1H NMR spectrum for product B,

however, showed two sets of signals, i.e., two triplets and two broad singlets, integrating to 2

and 4, respectively. This corresponded to a structure with only two unique aryl groups and a

two-fold symmetry. Furthermore, no resonance, which was consistent with an allenic sp-

carbon was found in the 13C NMR spectrum. Ultimately, X-ray crystallography identified the

structure of product B as radialene 5.15 (Figure 5.7).

Ar

Ar

Ar

Ar

NCCN

CNNC

Ar =

A

C

B

5.15

CH2Cl2/hexanes

1/1

Figure 5.7 Left: Identification of product B as [4]radialene 5.15. Right: ORTEP drawing

(20% probability level) for compound 5.15.

179

The identification of 5.14 and 5.15 (products A and B) was, in principle, consistent

with intermediate 5.10a during the reaction of [5]tBuPh with TCNE. As the study progressed,

it became clear that product C (Figure 5.8) was not stable, and it transformed to products A

(compound 5.14) and B (compound 5.15) in a ratio of ca. 9:1 over time, by warming, or via

concentration of the reaction mixture. Thus, attention turned to identification of product C.

Figure 5.8 Conversion of product C to A (compound 5.14) and B (compound 5.15), and the

associated TLC analysis.

The work of Viehe and coworkers15 has shown that polarized [3]cumulenes could be

efficiently trapped via reaction with EtOH or Br2 (Scheme 5.6). The resulting butadiene

adducts are obtained through addition of EtOH and Br2 to the central double bond of the

cumulene chain in the push-pull [3]cumulene.

N

N CN

CNEtOH

CN

CN

Br

BrN

N

CN

CN

EtO

HN

N

Br2

Scheme 5.6 Reaction of a push-pull [3]cumulene with EtOH and Br2, respectively.15

In order to test if product C was consistent with the desired polarized [3]cumulene

5.10a (Figure 5.2), product C was isolated at low temperature and treated with EtOH and Br2,

180

respectively, at 0 °C in a solution of CHCl3 (Figure 5.9). The addition of EtOH showed no

productive results, and the butadiene derivative 5.16 was not observed. In contrast, the

outcome of the reaction with Br2 showed two new products D and E by TLC analysis. Product

E was unstable and decomposed. In contrast, the less polar brown product D could be

separated via column chromatography and was identified as the symmetrical and stable

[4]dendralene 5.17, as established by X-ray crystallography (Figure 5.9). Compound 5.17

was, unfortunately, not a product easily linked to the presence of the [3]cumulene 5.10a

during the reaction of [5]tBuPh with TCNE. Besides, no hint for formation of expected

butadiene 5.18 was observed in this reaction.

Figure 5.9 Conversion of product C with Br2 affording [4]dendralene 5.17; ORTEP drawing

(20% probability level) for compound 5.17.

It was hypothesized that dendralene 5.17 could be formed from the reaction of Br2

with radialene 5.15, which might be produced in situ from product C. Therefore, a test

reaction was carried out in which radialene 5.15 was treated directly with an excess of Br2

(Scheme 5.7). This reaction did not give the dibromoadduct 5.17, but rather, [4]dendralene

5.19, an isomeric analog of 5.17 (bromination of one aryl ring is observed). The structure of

5.19 was confirmed by X-ray crystallography. The use of less Br2 also gave a single product

with almost identical Rf value as for 5.19 (as observed via TLC analysis) but differently

181

colored than compound 5.19 (i.e., brown instead of red-brown). Unfortunately, this structure

could not yet be confirmed by X-ray crystallographic analysis but was identified tentatively as

[4]dendralene 5.20 in which no aromatic substitution occured.23

Scheme 5.7 Conversion of product B (compound 5.15) to [4]dendralenes 5.19 and 5.20;

ORTEP drawing (20% probability level) for compound 5.19.

With the identity of product C not yet confirmed, direct isolation and characterization

was attempted. The reaction of [5]tBuPh and TCNE was conducted over 24 h at –25 °C in

CH2Cl2. The desired product C was purified by column chromatography with CDCl3 with the

column temperature maintained between –20 to 0 °C using a jacketed column. The fractions

containing C were stored at low temperature (i.e., on dry ice) to prevent conversion to A and

B. The combined CDCl3 fractions (ca. 250 mL) were reduced to less than 1 mL under

vacuum, and NMR spectra were acquired.

The 1H NMR spectrum showed only one set of signals representative of the 3,5-di-t-

butylphenyl group, namely a broad singlet (suggesting a triplet) at 7.39 ppm and a doublet at

7.16 ppm. This indicated a highly symmetrical structure for product C. More significantly, the 13C NMR spectrum showed the signal of an allene at 203.6 ppm, as well as that for a sp3-

hybridized carbon atom at 42.9 ppm. Combined with ESI MS analysis, which showed a signal

at m/z 979.6575 ([[5]tBuPh + TCNE + Na]+), these data supported that product C could be

assigned as compound 5.11 based on a cyclobutane derivative with two exocyclic allene units

(Figure 5.10).24

182

Figure 5.10 Identification of product C as cyclobutane 5.11.

5.1.3 Mechanistic studies and characterization by UV/vis spectroscopy and X-ray

crystallography

As described in Section 5.1.2, the reaction of [5]tBuPh with TCNE, as well as further

addition reactions using EtOH, MeOH, and Br2, afford a variety of interesting compounds

with potentially unusual properties. This section of the thesis presents the mechanistic studies

that have been performed with the help of theoretical calculations by Görling and coworkers.

It is still necessary to find a fundamental basis of cycloaddition reactions of cumulenes since

to date, no general guidelines concerning the reactivity pattern exist. Thus, we have hoped to

get further insight into the reaction behavior of cumulenes through the help of theoretical

investigations. Furthermore, the results of UV/vis spectroscopic measurements of some

products obtained during the TCNE reaction will be discussed. Based on X-ray

crystallographic analysis, the structural properties of the formed cyclobutanes, radialenes, and

dendralenes will be outlined and compared to known structures. For simplification concerning

further discussions, Scheme 5.8 shows an overview of products identified from the reaction of

[5]tBuPh with TCNE and further addition reactions.

183

Scheme 5.8 Overview of compounds identified from the reaction of [5]tBuPh with TCNE

including further addition reactions.

Kawamura and coworkers have described a possible mechanism for the reaction of

[3]cumulenes with TCNE affording compound 5.3 as mentioned in Section 5.1.1 (see Scheme

5.1).1,3 In their studies, cyclobutane 5.5 (with two exocyclic ethylene units) has been

described as an intermediate that resembles the cyclobutane 5.11 (see Scheme 5.8)

synthesized herein. Kawamura has assumed that the single C-C-bond connecting four cyano

groups of 5.5 is broken and results in compound 5.3 via bond rotation and cyclization

(Scheme 5.9).

Scheme 5.9 Summary of a mechanism of the reaction of a [3]cumulene with TCNE as

suggested by Kawamura and coworkers.1,3

184

Based on Kawamuras’s work, we have postulated a mechanism for the reaction of

[5]tBuPh with TCNE (Scheme 5.10). Homolytic cleavage of the central (NC)2C-C(CN)2 bond

of 5.11 gives intermediate 5.21, in which the resulting radicals are stabilized by allylic

delocalization. Cyclization of rotamer 5.21’ then affords [3]dendralene 5.14. Radialene 5.15

can also be formed from 5.11 through a stepwise mechanism via intermediate 5.21.

Alternatively, neither a concerted reaction that converts 5.11 directly to 5.15 nor a two-step

4e-electrocyclic ring closing/opening via 5.21 can be ruled out.25

NCNC CN

CN

Ar

Ar Ar

Ar

Ar

Ar

NCNC

CN

NC

Ar

Ar

Ar

Ar

Ar

Ar

CNCN

CNNCNCNC CN

CN

Ar

Ar Ar

Ar

NCNC CN

CN

Ar

Ar Ar

Ar

NCNC

CN

CNAr

Ar

Ar

Ar

•

•

5.145.15

5.11

5.11

• •

5.21

concerted

stepwise

5.21'

Ar

Ar Ar

ArNC

NC CN

CN

[5]tBuPh

Ar =

Scheme 5.10 Proposed mechanism for the conversion of 5.11 to 5.14 and 5.15.

Scheme 5.11 shows a possible concerted mechanism for the addition reaction that

have been applied to 5.11 (i.e., a product from the reaction of [5]tBuPh and TCNE) using Br2

as well as a two-step mechanism for addition reactions applied to 5.14 using ROH.

Compound 5.11, for example, adds Br2 onto the allenic carbon atoms. A simultaneous bond

cleavage of the single bond containing the cyano groups would be necessary to obtain

[4]dendralene 5.17. Aside from the concerted mechanism, also a radical containing

transformation can be considered based on the biradical intermediate 5.21 in Scheme 5.10.

Finally, additions of ROH to cyclobutane 5.14 take place on the allenic bond giving adducts

5.12 and 5.13 for R = Et and Me, respectively.

185

Scheme 5.11 Proposed concerted mechanism for the addition reaction of bromine to 5.11 and

the stepwise addition of ROH to 5.14.

DFT calculations show that the reaction of [5]tBuPh with TCNE to 5.11 followed by a

rearrangement reaction to 5.14 and 5.15 is in excellent agreement with predictions based on

stability (Figure 5.11), i.e., compounds 5.14 and 5.15 are clearly thermodynamically more

stable than 5.11. Interestingly, the inclusion of van-der-Waals corrections to the DFT

calculations gives enhanced stabilization of products 5.11, 5.14, and 5.15 when compared to

calculations without this correction. This observation is easily rationalized by the fact that the

products are stabilized by intramolecular dispersive interactions between the aryl and alkyl

groups. The computed energies also explain why 5.10a is not observed, given that metathesis

reaction from 5.11 to 5.10a is significantly endothermic when dispersive interactions are

included. In the absence of dispersion corrections, the formation of 5.14, and 5.15 is still

preferred over 5.10a, albeit the differences are less pronounced.

186

Figure 5.11 Computed relative energies (kcal/mol) of the products 5.11 (product C), 5.14

(product A), and 5.15 (product B), as well as the hypothesized product 5.10a, in comparison

to the reactants TCNE and [5]tBuPh. Calculations based on DFT including (red), and without

(blue) dispersion interaction corrections.

Regarding the properties of all synthesized cyclobutane, radialene, and dendralene

derivatives, radialene 5.15 shows the most interesting features. The most obvious

characteristic of 5.15 is its color, which changes from orange, when adsorbed on silica gel to

green in the solid state.26 Compound 5.15 also shows solvatochromism by UV/vis

spectroscopy. Specifically, the UV/vis spectrum shows a broad low energy absorption with

λmax > 700 nm (Figure 5.12) that ranges from λmax = 720 nm (cyclohexane) to λmax = 771 nm

(CHCl3), which is characteristic for an intramolecular charge transfer absorption.27

187

Figure 5.12 Quantitative UV/vis spectrum of radialene 5.15 (in CHCl3); Inset: λmax values for

5.15 as a function of solvent.

The rectangular structure of 5.1528,29 suggests a donor-acceptor (push-pull) interaction

between the dicyanovinyl acceptor and the electron-rich dialkylaryl groups, with shortened

C1–C2 bond lengths of 1.469(2) Å and longer

C1–C1’ and C2–C2’ bonds (1.494(3) and

1.504(3) Å, respectively). This is concurrent

with elongation of the alkylidene bonds C1–C4

and C2–C3 (1.36–1.37 Å) compared to several

equally substituted symmetrical [4]radialenes

as described in the following paragraph.

Table 5.2 presents several known

[4]radialenes 5.22–5.26 that are substituted by

a variety of endgroups giving symmetrical and unsymmetrical compounds.7,29–32 The bond

lengths of the cyclobutane unit (a–d in red, Table 5.2), as well as the exocyclic double bonds

(e–h in green, Table 5.2) are compared to radialene 5.15. Radialenes 5.22 and 5.23 show

cyclobutane bond lengths a–d of 1.502(6)–1.507(6) Å and 1.484(2)–1.492(2) Å,

respectively.30,31 The exocyclic double bonds e–h show values of 1.343(6)–1.351(6) Å and ca.

188

1.326(2) Å for 5.22 and 5.23, respectively. The bond lengths a and b of radialene 5.15

(1.469(2) Å) are shorter compared to a and b for symmetrical radialenes 5.22

(1.502(6)/1.507(6) Å) and 5.23 (1.484(2) Å) probably caused by push-pull effects in 5.15. In

contrast, the bond lengths e–h of the exocyclic double bonds in 5.15 are slightly longer than

those of radialenes 5.22 and 5.23. This might be derived from the increased single bond

character due to the push-pull effect of the endgroups. Radialene 5.24, reported by Diederich,

possesses similar push-pull endgroups as radialene 5.15, namely dimethylaniline and cyano

endgroups.7 Unlike 5.15, however, compound 5.24 is centrosymmetric showing no effect on

bond lengths of the cyclobutane ring having similar bond lengths a–d of 1.473(3)–1.474(3) Å.

A minor difference in bond length of the exocyclic double bonds can be observed giving

1.383(3) Å for e and h and 1.402(3) Å for f and g, probably due to endgroup effects.

Radialene 5.25 shows similar effects in bond lengths as radialene 5.15 with shortened a and b

and elongated c and d bonds (compared to 5.22 and 5.23), as well as similar values for e–h

(Table 5.2).32 The last example is shown by radialene 5.26 describing similar structural

properties as radialenes 5.25 and 5.15.29 In conclusion, endgroup effects show a significant

influence on the bond lengths of radialenes resulting in shortening and elongation of bond

lengths, thus, enabling the possibility of tuning the properties of radialenes.

189

Table 5.2 Bond lengths (Å) of radialene 5.15 and selected radialenes known from literature.

radialene 5.15 5.22 5.23 5.24 5.25 5.26

a 1.469(2) 1.502(6) 1.484(2) 1.474(3) 1.4279(15) 1.440(3)

b 1.469(2) 1.507(6) 1.484(2) 1.474(3) 1.4275(15) 1.440(3)

c 1.494(3) 1.504(6) -[a] 1.473(3) 1.5156(15) 1.491(3)

d 1.504(3) 1.502(6) 1.492(2) 1.473(3) 1.5157(16) 1.488(3)

e 1.370(2) -[a] 1.326(2) 1.383(3) 1.3747(15) 1.361(3)

f 1.370(2) 1.343(6) 1.326(2) 1.402(3) 1.3722(15) 1.361(3)

g 1.362(3) 1.351(6) 1.326(1) 1.402(3) 1.3735(16) 1.343(3)

h 1.362(3) -[a] 1.326(1) 1.383(3) 1.3753(16) 1.345(3)

ref 30 31 7 32 29

[a] Value not given in literature.

5.2 Addition reactions of other cumulenes with TCNE

5.2.1 [7]tBuPh cumulene

A reaction of TCNE with the [7]cumulene [7]tBuPh was performed (Figure 5.13). A

purified solution of the [7]cumulene [7]tBuPh in CH2Cl2 was treated with TCNE at rt. After

stirring overnight, the color of the reaction mixture turned from violet to bluish. The TLC

analysis showed a wide variety of colorful spots similar to the reaction of [5]tBuPh with

TCNE, and five major products, A–E, were formed:

190

1. The first product A was initially blue and changed on the TLC plate to violet after

few seconds, and finally, complete decolorization occurred. By spotting the

purified product A on the TLC plate and putting it immediately into the TLC

chamber, only the blue spot was visible. Leaving the plate with product A for some

seconds on air before starting TLC analysis, however, resulted in an additional

pink baseline spot.

2. On the TLC plate, product B was initially only visible under UV light changing

into pale pink after some seconds under exposure to UV light.

3. The color of the purified product C was pale yellow, and after spotting on a silica

gel TLC plate, it turned to an intense orange. Using alumina TLC plates, no color

change was observed upon spotting. Based on the observations of the TLC

analysis using silica gel TLC plates, that could be considered as slightly acidic,

addition of HCl to a solution of C was performed also resulting in a color change

from pale yellow to orange.

4. Product D and baseline product E remained yellow and orange, respectively.

Figure 5.13 Reaction of [7]tBuPh with TCNE, and the associated TLC analysis.

After separation of products A–E via column chromatography (hexanes/CH2Cl2 1:1,

silica gel), several characterization methods were applied. A qualitative UV/vis spectrum was

recorded for product A as shown in Figure 5.14. The value for the lowest energy absorption

λmax was at 566 nm, while a shoulder absorption band could be observed at 609 nm. It was

interesting to note, that both values were shifted into the low-energy absorption region

191

compared to [7]tBuPh as well as [5]tBuPh. Hence, addition of TCNE to the cumulene system

affording product A could not be excluded, and the resulting red-shift of λmax could be

explained by the strong electron-accepting cyano groups.

Figure 5.14 Qualitative UV/vis spectrum of product A measured in hexanes/CH2Cl2 (eluent

from column chromatography).

The IR spectrum of product A given in Figure 5.15 showed evidence for the presence

of a cumulene group observed by an IR signal at 2038 cm–1. In general, the region between

1900 and 2050 cm–1 has been common for cumulenic vibrations. Unfortunately, no further

spectroscopic characterization of product A was possible due to decomposition.

192

Figure 5.15 IR spectrum of product A.

Product C (Figure 5.13) showed a color change from almost colorless to orange under

acidic conditions. Unfortunately, this product could not yet been identified, but it was possible

to record a 1H NMR spectrum of the compound in the orange solution, i.e., of product C after

addition of HCl (Figure 5.16). Based on this analysis, the signature of a single unsymmetrical

compound could be observed showing sharp and distinct signals. The upfield region showed

four singlets, each integrated to ca. 18 protons. The downfield region showed four triplets,

each integrated to one proton and three doublets with an integration ratio of 4:2:2.

Additionally, one singlet with an integration of one proton was found. This singlet was

probably caused by an additional proton that was derived from the HCl, which could add to

the structure of C.

193

Figure 5.16 1H NMR spectrum (recorded in CD2Cl2) of product C after addition of HCl.

The color change of product C under acidic conditions was also monitored via UV/vis

spectroscopic measurements. Two UV/vis absorption curves describing fraction C before

(black curve) and after (red curve) HCl addition have been shown in Figure 5.17. The high

energy region showed in both spectra three absorption bands and a shoulder. The values of

C + HCl were shifted slightly to higher energy with e.g., 273 and ca. 303 nm compared to 278

and 307 nm for product C. Noteworthy, a broad absorption band at 478 nm was present after

HCl addition.

194

Figure 5.17 Qualitative UV/vis spectrum of product C before (black curve) and after (red

curve) addition of HCl measured in hexanes/CH2Cl2 (eluent from column chromatography).

Regarding the yellow product D (Figure 5.13), a UV/vis spectrum was recorded for

characterization (Figure 5.18). The absorption region was set to 200–800 nm (default

parameters), and several absorption bands in the higher energy region (250–500 nm) were

observed. In addition, an intense absorption band appeared at 800 nm that extended into the

near-IR region. The low-energy absorption band at ca. 800 nm was reminiscent of

[4]radialene 5.15 that was formed in the reaction of [5]tBuPh and TCNE and showed a λmax

value of 771 nm in CHCl3 (Figure 5.7 and Figure 5.12).

195

Figure 5.18 Qualitative UV/vis spectrum of product D measured in hexanes/CH2Cl2 (eluent

from column chromatography).

In conclusion, initial results from the reaction of [7]tBuPh with TCNE show several

interesting products that have been briefly investigated. Three of the products seem likely to

have cumulenic or radialenic structure. The obtained products, however, show lower stability

than the products from the reaction of [5]tBuPh with TCNE, preventing more complete

characterization under the time limits of this thesis. Unfortunately, the reaction of [7]tBuPh

with TCNE has also not yet been repeated using crystalline (i.e., pure) [7]tBuPh due to time

restrictions. This reaction is certainly suited to further investigations based on the initial

results in order to gain more information about the reactivity of even higher [n]cumulenes.

196

5.2.2 [5]MeOPh cumulene

It is of interest to investigate to what extent the endgroups of [n]cumulenes influence

the outcome of the reaction with TCNE and the stability of the products. Thus, the use of

[5]cumulene [5]MeOPh containing p-methoxyphenyl endgroups has been explored in the

reaction with TCNE. Contrary to the 3,5-di-t-butylphenyl group, the p-methoxyphenyl

endgroup, in fact, shows higher potential for the formation of donor-acceptor molecules due

to its more distinctive electron-donating effect. The group of Kawamura has already been

successful in investigation of the reaction between the lower cumulene representative

[3]MeOPh and TCNE,3 which has encouraged us to proceed using the analogous

[5]cumulene [5]MeOPh.

A sample of [5]cumulene [5]MeOPh that was synthesized during my diploma thesis

was used for the reaction with TCNE.33,34 Before using [5]MeOPh for the reaction with

TCNE, TLC analysis was performed and showed a pink (Rf value of ca. 0.6 in CH2Cl2) and a

baseline spot, which have yet not been identified. By filtration through a plug of silica gel in

CH2Cl2, the baseline fraction remained at top of the plug, and the pink fraction could be

obtained pure. Concentration of the pink fraction, however, gave very little product, and only

a poorly resolved 1H NMR spectrum could be recorded. TLC analysis confirmed that the

baseline spot also contained the pink spot (repeated elution of TLC plate after rotation by 90°)

illustrating that there were likely solubility problems.

In the first reaction, the sample of [5]MeOPh (containing the pink and the baseline

spot) was treated with TCNE at rt under inert conditions, and the reaction mixture was stirred

for several days (Figure 5.19). Several products B–E were formed according to TLC analysis,

aside from unconverted starting material (A). Products B and E did not change over time

remaining yellow and reddish on the TLC plate, respectively. Product C was initially only

visible under UV light but turned to green-brownish over time, while product D was initially a

violet spot that turned to yellow.

197

Figure 5.19 Reaction of [5]MeOPh (containing pink and baseline spot) with TCNE, and the

associated TLC analysis.

Three of four compounds, i.e., products B–D could be identified by X-ray

crystallographic analysis (Figure 5.20), and each structure consisted of an HCl or Cl2 adduct.

This led to three possible explanations:

1. The reaction or crystallization solvent (CH2Cl2) reacted with the starting material or

the reaction intermediates.

2. SnCl2 or HCl could be still present in the sample of [5]MeOPh (derived from the

reductive elimination reaction leading to [5]MeOPh) and could react with the starting

material or reaction intermediates that were formed during the reaction of [5]MeOPh

with TCNE.

3. [5]MeOPh that was used as starting material was already contaminated with

impurities or in the worst case, never possessed the structure of the desired

[5]cumulene [5]MeOPh.34

198

Figure 5.20 Products B–D with appropriate ORTEP drawings (20% probability level).

The reaction of [5]MeOPh with TCNE as shown in Figure 5.19 was repeated twice in

order to compare the results with those of Figure 5.19. Each reaction (the “second” and

“third” reactions, respectively) used a different sample as starting material,35 and both

samples showed the same pink and baseline spots in the TLC analysis.

First, the pink and the baseline spots of [5]MeOPh were separated via filtration

through a plug of silica gel using CH2Cl2 as solvent, and the separated pink fraction was used

for the “second” reaction with TCNE (Figure 5.21a). This reaction was monitored via TLC

analysis and showed a slightly different spot splitting compared to the TLC analysis of the

reaction in Figure 5.19. A violet spot, product A, turning to yellow as in the first reaction (see

199

product D in Figure 5.19) was observed (Figure 5.21a). Since the concentration was too low

for separation, the reaction was not further examined.

A “third” reaction of [5]MeOPh with TCNE was done, but with the starting material

(i.e., [5]MeOPh) containing both, the pink fraction and the baseline material (Figure 5.21b).

The results were almost the same as in the second reaction (i.e., when only the pink fraction

was used, Figure 5.21a). The baseline spot, however, was slightly darker, possible due to the

baseline material that already was present before the reaction. Crystallization attempts,

however, failed giving no confirmation if the violet spot, i.e., product A from the second and

third reactions (Figure 5.21a and b, respectively) was equal to the one describing product D

from the prior reaction in Figure 5.19.

Figure 5.21 Second and third reactions of [5]MeOPh with TCNE, and the associated TLC

analysis, in comparison to the first reaction. The precursor contained a) only the pink spot and

b) the pink and baseline spot.

To try to resolve the issues described above, the synthesis of [5]MeOPh was repeated

several times during my doctoral thesis in order to obtain material for final identification and

further reactions with TCNE. The precursor to [5]MeOPh, compound 5.27, was synthesized

200

by a standard synthetic route36 starting with ketone 5.28, followed by an addition of a

trimethylsilyl-substituted acetylide giving compound 5.29 (Scheme 5.12). Desilylation under

basic conditions afforded terminal alkyne 5.30, and final homocoupling reaction under Hay

conditions (CuCl and TMEDA) gave 5.27 which was purified by column chromatography and

carried on toward [5]MeOPh using standard conditions (SnCl2 and HCl). This reaction,

however, did not provide [5]MeOPh because too many byproducts were formed, and

decomposition of the precipitated product [5]MeOPh was observed. Taking a sample of

dissolved [5]MeOPh during the reaction progress using a capillary and spotting it on a TLC

plate already led to a color change from red to dark violet (in the capillary tube and on the

plate). Nevertheless, the formed reddish precipitate from the reaction mixture that was

assumed as [5]MeOPh was filtered, washed with water, EtOH, and Et2O, and dried. 1H NMR

spectroscopy showed two pairs of aryl groups and two signals for MeO groups aside from

impurities consisting of water, grease, and solvent. [5]MeOPh, however, should show only

one pair of aryl groups and one signal defining the OMe groups in the 1H NMR spectrum.

Unfortunately, the sample amount was not sufficient for 13C NMR spectroscopy. For

unknown reasons, it was not possible to reproduce the synthesis of [5]MeOPh in order to

provide adequate amounts for the reaction with TCNE. Consequently, no further studies on

this reaction could be performed, and also no reproducible results could be obtained.

Scheme 5.12 Synthetic approach to [5]cumulene [5]MeOPh.

201

5.2.3 [5]oTol cumulene

The readily accessible [5]cumulene [5]oTol37 containing tolyl endgroups was treated

with TCNE (Figure 5.22). TLC analysis showed a yellow spot (A), a UV spot (B) that turned

to green after time, a weak yellow spot (C), a big grey-blue spot (D), and a bright yellow

baseline spot (E). A similar spot pattern was also obtained in the reaction of [5]tBuPh with

TCNE (Figure 5.2). All fractions were separated by column chromatography in CH2Cl2. The

UV product B showed a conversion to products A and D. The same behavior was observed in

the TCNE chemistry of the [5]cumulene [5]tBuPh (Figure 5.8). Consequently, it was assumed

that product B described the cyclobutane derivative consisting of two exocyclic allene groups

(analogous to compound 5.11 in Section 5.1.2) converting to A with the assumed structure of

an unsymmetrical cyclic [3]dendralene (analogous to 5.14) and D that was expected to

possess a radialene structure analogous to compound 5.15 in Section 5.1.2 (Figure 5.22).

Figure 5.22 Reaction of [5]oTol with TCNE, and the associated TLC analysis. Proposed

product structures are shown as A, B, and D.

202

Crystallization attempts for the products from the reaction of [5]oTol with TCNE were

not successful. Thus, no product from each of the fractions could be characterized by X-ray

crystallography. Product D was not infinitely stable as observed during the crystallization

attempts, and after several days, decomposition occurred in methanol and toluene as

confirmed via TLC analysis. In hexanes, D appeared to be more stable. The product D was

assumed to be the analog of radialene 5.15 in Section 5.1.2 and was treated with Br2 to

observe if the same behavior as for 5.15 (Figure 5.8) was present (see Scheme 5.7). The spot

of product D disappeared when a huge excess of Br2 was used, but no products could be

isolated or characterized. Using less Br2 did not significantly change the outcome of the

reaction, although even after heating the reaction mixture to reflux, the spot of the starting

material was still present. The TCNE reaction with [5]oTol was repeated once again in order

to establish the radialenic structure for D. Product D was isolated, and a 1H NMR spectrum

was recorded but too many and broadened signals were observed that hindered assignment to

the assumed radialenic structure.

5.3 Addition reaction of a [9]cumulene with HCl

5.3.1 Synthesis of [5]cumulene 5.31

Several optimization reactions for the synthesis of [9]tBuPh were attempted and are

outlined in Section 2.1.2.5. One reaction was the synthesis of [9]tBuPh performed in CH2Cl2

instead of Et2O to facilitate the following crystallization attempts. Hence, the reaction was

conducted under standard conditions, i.e., argon atmosphere, low temperature, and using

anhydrous SnCl2 and HCl (Scheme 5.13). After addition of HCl to the reaction mixture, the

solution color turned to blue indicating the formation of [9]cumulene [9]tBuPh, however, this

color immediately turned violet/pink. The reaction mixture then changed from violet/pink to

orange brown after several days. Initial TLC analysis of the reaction mixture indicated two

spots, a pink spot (A) and a violet spot (B) right below (Scheme 5.13). The stability of the

pink product A appeared to be much higher than the violet product B. Crystallization attempts

in CH2Cl2/MeOH were made for both products, but only in the case of the pink product A,

single crystals could be obtained. The structure of product A was assigned as 5.31 via X-ray

crystallographic analysis (Scheme 5.13 and Table 5.3). Two equivalents of HCl were added to

both β-bonds of the [9]cumulene [9]tBuPh during the reaction. Influence of solvent was

203

certainly one possible reason for the different outcome in CH2Cl2 compared to Et2O. It was,

however, not clear, how the solvent affected the formation of 5.31. Stabilization effects of the

chlorinated solvent could be one possibility.

Scheme 5.13 Synthesis of [5]cumulene 5.31 via reductive elimination of tetrayne 2.14 in

CH2Cl2, and the associated TLC analysis.

The synthesis of [5]cumulene 5.31 was reproducible as well as the crystallization with

CH2Cl2 and MeOH affording stable deep pink crystals of compound 5.31. In addition, if the

same reaction was repeated at –78 °C instead of 0 °C, HCl addition could be suppressed

giving only a blue solution indicating the [9]cumulene [9]tBuPh showing no following color

change to pink.

5.3.2 Characterization of [5]cumulene 5.31 via UV/vis spectroscopy and X-ray

crystallography

Figure 5.23 presents the qualitative UV/vis spectrum of [5]cumulene 5.31 recorded in

Et2O. The absorption pattern is very similar to other [5]cumulenes showing the most intense

204

absorption in the low energy region (541 nm) followed by a less intense band in the high

energy region (301 nm) and finally a weak absorption in between at 434 nm.

Figure 5.23 Qualitative UV/vis spectrum of [5]cumulene 5.31 (in Et2O).

The value of the lowest energy absorption λmax is 541 nm and red-shifted compared to

[5]tBuPh (500 nm in Et2O), [5]Mes (460 nm in Et2O) as well as [5]Ph (488 nm in CHCl3; for

a complete description of [5]cumulene absorption characteristics, see Section 4.1.2.3). This

red-shift is probably explained by the increased conjugation of the molecule including the

cumulene chain, the ethylene units, and the aryl groups (Figure 5.24).

205

Figure 5.24 Two canonical structures of [5]cumulene 5.31.

The X-ray structure of 5.31 is shown below in Table 5.3 including relevant carbon

atom labels. The comparison of bond angles of the cumulene chain shows that the values

deviate slightly more from the ideal value of 180° compared to other [5]cumulenes. Known

[5]cumulenes show bond angles in a range from 178.8° to 179.7° (Section 4.2). The bond

angles of 5.31 are 175.5° (C3-C4-C5) and 178.2° (C4-C5-C5‘).

The bond lengths in the cumulene chain of 5.31 are consistent with the bond lengths of

other [5]cumulenes that are discussed in Section 4.2 (Table 5.3). The terminal α-bond length

(C3-C4) is the longest with a value of 1.340(3) Å while the β-double bond (C4-C5) is the

shortest bond and resembles more a triple bond with 1.251(3) Å. The value for the central γ-

double bond (C5-C5‘) at 1.308(5) Å lies between the α- and β-bond lengths, resulting in a

BLA value of 0.057, which resembles the values of [5]Ph (0.058) and [5]tBuPh (0.054) the

most.

The bond length value of the double bond C1-C2 is 1.348(3) Å and slightly longer

than an isolated double bond (1.34 Å) while the single bond length of C2-C3 is 1.445(3) Å

and thus slightly shorter than the bond length of an isolated C-C bond that is attached to sp2-

hybridized carbons, i.e., 1.47 Å. This can probably be explained by the canonical structures

(Figure 5.24), i.e., the C1-C2 bond possesses more single bond character while the C2-C3

bond possesses more double bond character.

206

Table 5.3 Bond lengths (Å) and bond angles (°) of [5]cumulene 5.31.[a]

Cmpd C1-C2 C2-C3 C3-C4 C4-C5 C5-C5’ C3-C4-C5 C4-C5-C5’ BLA[b]

5.31 1.348(3) 1.445(3) 1.340(3) 1.251(3) 1.308(5) 175.5(3) 178.2(3) 0.057

[a] ORTEP drawing (20% probability level) of 5.31. [b] Calculated as difference in bond

length between the two central-most bonds.

5.4 Dimerization of [5]tBuPh

During my doctoral thesis, it was possible to prepare [5]tBuPh on a large scale (>5 g)

due to the straightforward synthetic pathway and the good stability of that compound (Figure

5.25). Thus, several studies could be pursued in order to examine the reactivity of [5]tBuPh.

207

Figure 5.25 Scale-up of [5]tBuPh.

Aside from addition reactions of TCNE to [5]tBuPh (see Section 5.1), the thermal

dimerization reaction was explored with [5]tBuPh. Hartzler38 and Iyoda39 reported the

dimerization of [5]cumulenes affording [4]radialenes with four exocyclic allene units

(Scheme 5.14). The [5]cumulenes they have used were substituted by alkyl groups and were

converted to the [4]radialenes by a thermal reaction in the solid state.

• •

R

R••

R

R ••

••

R

R R

R

R

R R

R

thermal

dimerization

α β γ

R

R

=

R = t-Bu

Scheme 5.14 Thermal dimerization of [5]cumulenes reported by Hartzler38 and Iyoda.39

In contrast, Kawamura40 reported about oligomerization of the phenyl-substituted

[5]cumulene [5]Ph, which was heated to reflux in toluene affording a trimer that was assumed

to be synthesized via a [4]radialene intermediate with four exocyclic allene units (Scheme

5.15). Based on these results, similar dimerization routes were investigated with [5]tBuPh.

208

Scheme 5.15 Thermal trimerization of [5]Ph reported by Kawamura.40

5.4.1 Synthesis of the dimer of [5]tBuPh

The [5]cumulene [5]tBuPh was heated to reflux in toluene (Figure 5.26). After 5 days,

new compounds were formed as indicated by TLC analysis. The orange red spot (C) of

[5]tBuPh was converted into a new pink spot (product A) with lower polarity compared to

[5]tBuPh. Product A appeared to be unstable and tended to decompose, forming an additional

weak grey baseline spot. Product A was assumed to contain the expected dimer, i.e.,

[4]radialene 5.32 since no other significant spot was observed (Figure 5.27). In addition to A,

a green product spot B and a baseline spot D were also observed, but showed much lower

intensity. The main product A could initially not be purified by flash chromatography using

hexanes and silica gel and remained at the top of the column. Filtration with ethyl acetate led

to decomposition observed by a color change to orange brown. Using a very short plug of

silica gel, it was finally possible to purify the pink fraction by filtration with hexanes.

209

Figure 5.26 Thermal reaction of [5]tBuPh in toluene, and associated TLC analysis.

A 1H NMR spectrum of product A was recorded giving several signals in the aryl

region and the alkyl region indicating the formation of either an unsymmetrical compound or

a mixture of several compounds. 13C NMR spectroscopy of product A also showed several

signals, e.g., three signals at ca. 200 ppm were observed that were consistent with an allenic

species. Mass spectroscopy showed a signal value which fitted to [4]radialene 5.32 (center-to-

center dimer, see Figure 5.27) but also its isomers, i.e., head-to-tail-, head-to-head dimers, etc.

Figure 5.27 [4]Radialene 5.32 as the desired product from the thermal reaction of [5]tBuPh

in toluene. Additionally, the possible head-to-tail and head-to-head dimers are shown.

Crystallization attempts of product A from CH2Cl2/hexanes afforded square-like single

crystals that were investigated by X-ray crystallographic analysis confirming the structure of

[4]radialene 5.32 (Figure 5.27). The crystals were, however, colorless while the reaction

mixture was deeply pink. This result as well as the results from the NMR spectroscopy

suggested the presence of several products during the thermal dimerization of [5]tBuPh that

were either overlayed or not observed (i.e., not UV active) by TLC analysis. The thermal

210

dimerization reaction was repeated several times and showed completely different outcomes

than in the first attempt with far more byproducts. The reaction was not reproducible and

probably concentration-dependent based on the variety of resulting products. The only feature

that was at least partially reproducible in each reaction were the crystallization attempts which

afforded colorless square crystals that resembled crystals of 5.32. Unfortunately, all square

crystals obtained during these subsequent thermal reactions of [5]tBuPh were not suitable for

X-ray crystallographic analysis. Isolation and purification of the square crystals from the

reaction mixture were attempted. The crystals could not be purified by washing with different

solvents due to dissolution. Purification via column chromatography was tried using toluene

as solvent. Compound 5.32, however, could not be accurately assigned to a certain spot on the

TLC plate that significantly complicated separation. Developing agents, such as iodine or

anisaldehyde, did not show any change regarding TLC analysis. With potassium

permanganate as developing agent, one spot became brighter and more intense. After

separation of this fraction, a 1H NMR spectrum was recorded indicating that the dimer could

be present, but another compound was visible in the spectrum, and unfortunately, no 13C

NMR spectrum could be measured due to insufficient amount of the product. In conclusion,

[4]radialene 5.32 could not be identified via TLC analysis, and it was not possible to achieve

reproducibility for the thermal reaction of [5]tBuPh or the separation and purification of 5.32.

A further drawback of this reaction was the reaction time, i.e., the reaction had to be stirred

for at least one week under reflux to afford complete conversion of the starting material.

The synthesis of [4]radialene 5.32 was also attempted using the microwave under the

same thermal reaction conditions in toluene as mentioned in Figure 5.26. After 7 h, however,

no change was observed. Only starting material remained, and no evidence for formation of

other products was found. Hence, the synthesis of 5.32 using the microwave was abandoned.

Another reaction approach to afford [4]radialene 5.32 was the thermal dimerization

reaction in the solid state according to Hartzler38 and Iyoda.39 [5]Cumulene [5]tBuPh was

thus melted and stirred for 15 min at ca. 230 °C. The liquefied material was cooled to rt and

dissolved in CH2Cl2. TLC analysis showed no starting material ([5]tBuPh) but several

fluorescent spots. Unfortunately, crystallization attempts from CH2Cl2 and hexanes were not

successful. The crude material was put through a short plug of silica and washed with

hexanes. A yellow fraction was collected (filtrate). The remaining fractions were flushed out

using toluene (residue). Both the filtrate and the residue were analyzed by NMR spectroscopy,

but neither showed evidence for formation of [4]radialene 5.32. Surprisingly, it appeared that

211

the residue contained only the bis-3,5-di-t-butylphenylketone 2.2, but characterization of this

product was not further pursued.

5.4.2 Characterization of the dimer of [5]tBuPh

The characterization of [4]radialene 5.32 and pink product A, in which the presence of

5.32 is assumed (Figure 5.26) has been limited due to low substance amount and impossible

product separation, respectively. Nevertheless, a UV/vis spectrum of the crude reaction

mixture (pink fraction A in Figure 5.26) has been recorded in toluene and is shown in Figure

5.28. Two broad absorption bands are observed in the high energy region, at 277 and 321 nm,

while the low energy region shows one absorption band at 527 nm with lower intensity. This

band might be derived from the radialene structure of 5.32 or it could also arise from a

different cumulenic species that has been formed during the thermal reaction.

Figure 5.28 Qualitative UV/vis spectrum of the pink reaction mixture (A) as presented in

Figure 5.26 (in toluene).

212

Besides, the characterization of product A via NMR-, IR-, and mass spectra cannot

definitely confirm that [4]radialene 5.32 is contained in product A (as assumed and mentioned

before), although the mass spectrum reveals a signal that is consistent with [4]radialene 5.32.

This signal, however, can also be derived from possible isomers of 5.32. During the work-up

of the reaction in Figure 5.26 and after crystallization attempts, several fractions have been

obtained via separation and filtration methods, however, with none of them matching to

[4]radialene 5.32.

5.4.3 X-ray crystallographic data: Discussion and comparison

Table 5.4 shows the X-ray crystallographic structure of the [4]radialene 5.32. One

other [4]radialene with four exocyclic allenes, namely compound 5.33, is known, and has

been reported by Iyoda.39,41 The endgroups of 5.33 are aliphatic based on tetramethyl indane

(Table 5.4). Values of relevant bond lengths and angles of [4]radialenes 5.32 and 5.33 are

given in Table 5.4. While [4]radialene 5.32 is a C4 symmetric compound with the space group

P4/n, [4]radialene 5.33 shows a lower symmetry with the space group P–1. The bond angles

of the exocyclic allene units deviate from the ideal value of 180° with 176.3(4)° for 5.32 and

173.9(3)° and 177.8(3)° for 5.33. The angles of the cyclobutane are rectangular with values of

89.996(1)° for 5.32 as well as 90.2(3)° and 89.8(3)° for 5.33. The bond lengths of both

radialenes are similar except for the bond lengths C1-C2 (or C4-C5) that are longer for the

aryl-substituted radialene 5.32 (1.325(5) Å) compared to the alkyl-substituted radialene 5.33

(1.300(4) and 1.298(4) Å). The exocyclic double bonds C2-C3 and C5-C6 of both radialenes

range from 1.304 to 1.309 Å. The bond lengths of the cyclobutane ring are 1.497(5) Å for

5.32, as well as 1.499(4) and 1.510(4) Å for 5.33, which are slightly longer as the bond

lengths of single C-C bonds attached to sp2-hybridized carbons (1.47 Å), but similar to a

number of other [4]radialenes (Table 5.2).

213

Table 5.4 Bond lengths (Å) and bond angles (°) of [4]radialene 5.32 (left) and Iyoda’s

[4]radialene 5.33 (right).[a]

•

•

•

•

C1

C2C3

C4

C5

C6C7C8

5.33

Compound 5.32 5.33

C1-C2 1.325(5) 1.300(4)

C2-C3 1.308(5) 1.309(4)

C3-C3’ 1.497(5) 1.499(4)[b]

C4-C5 1.298(4)

C5-C6 1.304(4)

C6-C7 1.510(4)

C1-C2-C3 176.3(4) 173.9(3)

C4-C5-C6 177.8(3)

C3-C3’-C3’’ 89.996(1)

C8-C3-C6 90.2(3)

C3-C8-C7 89.8(3)

[a] ORTEP drawing (20% probability level) of 5.32. [b] C3-C6.

5.5 Conclusion and summary

The addition reaction of TCNE to [5]cumulene [5]tBuPh gave several unique products

that were separated and purified. One of these products, radialene 5.15, showed interesting

214

solvatochromism with a lowest energy absorption (λmax value >700 nm) approaching the

near-IR region. The reaction showed a high reproducibility, and it was possible to use the

products for further modifications by e.g., addition reactions (using Br2) to build up

interesting functionalized building blocks which could probably serve as precursors to

unprecedented conjugated structures. In addition, crystallizations of several products, such as

ROH adducts 5.12/5.13, radialene 5.15, and dendralene 5.17 were successful and

reproducible.

TCNE addition reactions to other [n]cumulenes, i.e., cumulenes [7]tBuPh, [5]MeOPh,

or [5]oTol were not successful and showed no reproducibility, high reaction times, as well as

high instability of the resulting products. Thus, characterization and identification of the

outcome of the reactions was limited.

During several optimization reactions of the reductive elimination to [9]tBuPh using

SnCl2 and HCl in CH2Cl2, an addition of two equivalents of HCl to [9]tBuPh showed the

successful and reproducible formation of a two-fold HCl adduct, i.e., [5]cumulene 5.31.

Through variation of the addition reagents and/or the nature of the [n]cumulene, such

reactions could be promising for the synthesis of substituted or functionalized cumulenes.

Finally, thermal dimerization attempts of [5]tBuPh gave one interesting new

compound, i.e., [4]radialene 5.32 with four exocyclic allene units. The thermal reaction was,

however, not reproducible. Furthermore, the reaction times were long, and the outcome

showed too many products that could not be separated, purified, or characterized.

Consequently, the [4]radialene could not be reproduced and no further investigations could be

done.

Many of the results described above show interesting and promising chemistry using

[n]cumulenes as precursors in (cyclo)addition reactions. Unfortunately, the characterization of

some of these reactions and their products have not been completely studied within the time

of this thesis. Thus, there are a number of future directions to be considered that result from

our preliminary studies of the reactivity of [n]cumulenes:

1. The reaction of a [7]cumulene with TCNE affords interesting products, one of

which shows a lowest energy absorption value in the near-IR region as observed by

UV/vis spectroscopy. Since purely organic chromophores that absorb in the IR are

rare, this reaction should be repeated, and the structure of this product should be

determined.

215

2. It is also intriguing to perform TCNE addition reactions with [n]cumulenes

containing different endgroups. Firstly, the variables concerning the reaction of

[5]MeOPh with TCNE should be determined, since there are several outstanding

questions remaining from the three reactions done during this thesis. This is

especially important, since three interesting products have already been confirmed

by X-ray crystallographic analysis.

3. The addition of HCl to the cumulene chain of a [9]tBuPh seems to show selectivity

for the second outermost double bonds, as demonstrated with the synthesis of 5.31.

The reaction conditions for this and perhaps related reactions should be established.

Also, the resulting product, a [5]cumulene with unsymmetrical substitution at the

termini, is thus interesting for the formation of functionalized cumulenes.

4. It has been established that thermal dimerization reactions of cumulenes can afford

radialene-like structures, such as 5.32, but little remains known about such products.

Iyoda and coworkers have already suggested that “such π-extended radialenes are

expected to constitute novel oligocumulene systems with interesting structures”,39,41

and the work from this thesis provides an outstanding starting point to expand on

Iyoda’s suggestion.



The general procedures and methods are analogous to that in Section 2.3.2.

Tetracyanoethylene (TCNE) was stored at low temperature (–25 °C) and weighed in the glove

box (under an argon atmosphere). Ar has been defined as 3,5-di-t-butylphenyl in the Section

5.6.2.

216


Allene 5.14. TCNE (23 mg, 0.18 mmol) was dissolved in dry CH2Cl2 (10 mL) under an Ar

atmosphere, and [5]cumulene [5]tBuPh (0.15 g, 0.18 mmol) in dry CH2Cl2 (15 mL) was

added at rt. After stirring for 6 d, the solvent was removed, and purification by column

chromatography (silica gel, hexanes/CH2Cl2 1:1) afforded product 5.14 as a yellow-orange

solid (0.117 g, 68%). Mp. 112–114 °C. Rf = 0.56 (hexanes/CH2Cl2 1:1), UV/vis (CHCl3) λmax

(ε) 292 (23800), 351 (19270), 457 (15600) nm. IR 3068 (vw), 2963 (s), 2907 (m), 2869 (w),

2221 (w), 1920 (vw, br), 1593 (m) cm−1; 1H NMR (300 MHz, CDCl3) δ 7.61 (t, J = 1.7 Hz,

1H), 7.53 (t, J = 1.7 Hz, 1H), 7.49 (t, J = 1.7 Hz, 2H), 7.29 (d, J = 1.7 Hz, 4H), 7.19 (d, J =

1.7 Hz, 2H), 7.12 (d, J = 1.7 Hz, 2H), 1.310–1.305 (m, 72H); 13C NMR (75 MHz, CDCl3) δ

203.4, 163.4, 162.3, 152.2, 151.9, 151.5, 136.8, 136.4, 131.6, 127.8, 126.8, 125.6, 125.5,

125.0, 124.1, 124.0, 122.4, 112.9, 111.7, 110.4, 104.2, 73.0, 41.2, 35.15, 35.13, 35.0, 31.4,

31.3, 31.2. MALDI MS m/z 957 ([M]+, 100); ESI HRMS m/z calcd. for C68H84N4Na ([M +

Na]+) 979.6588, found 979.6557.

Radialene 5.15. TCNE (31 mg, 0.24 mmol) was dissolved in dry CH2Cl2 (20 mL) under an

Ar atmosphere, and [5]cumulene [5]tBuPh (0.20 g, 0.24 mmol) in dry CH2Cl2 (10 mL) was

added at rt. After stirring for 3 d, the solvent was removed, and purification by column

chromatography (silica gel, hexanes/CH2Cl2 1:1) gave the radialene product 5.15 as a green

solid (17 mg, 7%). Mp. >270 °C (decomp). Rf = 0.10 (hexanes/CH2Cl2 1:1), UV/vis (CHCl3)

λmax (ε) 264 (18500), 331 (19600), 357 (25900), 381 (22500), 457 (11500), 493 (10700), 771

(3900) nm. IR (ATR) 3060 (vw), 3022 (vw), 2958 (s), 2905 (m), 2867 (m), 2217 (w), 2007

217

(vw), 1979 (vw), 1607 (m) cm−1; IR (KBr) 3063 (vw), 2961 (s), 2906 (m), 2870 (w), 2255

(w), 2218 (w), 1611 (w), 1476 (m) cm−1; 1H NMR (400 MHz, CDCl3) δ 7.52 (t, J = 1.6 Hz,

2H), 7.11 (t, J = 1.5 Hz, 2H), 7.04 (s, 4H), 6.91 (s, 4H), 1.29 (s, 36H), 1.10 (s, 36H); 13C

NMR (100 MHz, CDCl3) δ 167.4, 152.3, 151.8, 149.7, 139.4, 138.3, 135.0, 127.7, 126.7,

126.4, 125.2, 113.3, 110.2, 74.0, 34.9, 34.5, 31.33, 31.29. MALDI MS m/z 957 ([M]+, 100);

ESI HRMS m/z calcd. for C68H84N4Na ([M + Na]+) 979.6588, found 979.6576.

Crystal data for 5.15: C68H84N4, M = 957.39; monoclinic crystal system; space group C2/c, a

= 22.9875(9) Å, b = 23.8037(7) Å, c = 13.4123(6) Å, β = 119.248(5)°, V = 6403.4(4) Å3, Z =

4, ρcalcd = 0.993 mg mm–3; µ(CuKα) = 0.429 mm–1; λ = 1.54184 Å; 173.00(10) K; 2θ max =

147.02°; total data collected = 10892; R1 = 0.0498 [4651 observed reflections with

F ≥ 4σ(F)]; wR2 = 0.1460 for 350 variables, 6149 unique reflections, and 3 restraints; residual

electron density = 0.18 and –0.22 e Å–3. One t-butyl group showed disorder, which have been

resolved and refined to the following occupation factors: C18/C19/C20:C18a/C19a/C20a =

86:14%. CCDC 967293.

Ethyl vinyl ether 5.12. Compound 5.14 (8.2 mg, 0.0086 mmol) was dissolved in CH2Cl2

(2 mL) at rt, and EtOH (2 mL) was added. After stirring for 2.5 h, the solvent was removed to

give product 5.12 as a dark yellow solid (8.5 mg, quant). Rf = 0.47 (hexanes/CH2Cl2 1:1); 1H

NMR (400 MHz, CDCl3) δ 7.85 (bs, 1H), 7.61 (t, J = 1.7 Hz, 1H), 7.52 (t, J = 1.8 Hz, 1H),

7.36 (t, J = 1.8 Hz, 1H), 7.27 (t, J = 1.8 Hz, 1H), 7.17 (d, J = 1.7 Hz, 2H), 7.07 (d, J = 1.8 Hz,

2H), 6.95 (d, J = 1.5 Hz, 2H), 6.83 (bs, 1H), 4.57 (s, 1H), 3.84–3.79 (m, 1H), 3.53–3.48 (m,

1H), 1.43 (s, 9H), 1.31 (s, 18H), 1.28 (s, 18H), 1.26 (s, 18H), 1.22 (s, 9H), 1.05 (t, J = 7.0 Hz,

3H); 13C NMR (100 MHz, CDCl3) δ 170.3, 160.1, 153.0, 151.9, 151.1, 150.2, 144.8, 139.6,

138.2, 136.5, 136.1, 134.0, 126.4, 125.7, 125.5, 125.3, 124.8, 124.5, 123.7, 122.2, 121.2,

115.3, 111.0, 110.8, 109.9, 69.7, 61.6, 38.2, 35.5, 35.1, 35.0, 34.8, 34.7, 31.5, 31.38, 31.36,

31.2, 15.3. Note that there appears to be restricted rotation of one of the aryl rings of 5.12,

which leads to non-degeneracy of protons signals (at 7.85 and 6.83 ppm) as well as carbon

218

signals (an “extra” quaternary carbon signal of the t-butyl group at 35.5–34.7 ppm). MALDI

MS m/z 1003 ([M]+, 100); ESI HRMS m/z calcd. for C70H90N4NaO ([M + Na]+) 1025.7007,

found 1025.6985.

Crystal data for 5.12: C70H90N4O, M = 1003.46; triclinic crystal system; space group P–1, a =

13.6915(6) Å, b = 15.8594(10) Å, c = 15.9165(8) Å, α = 82.646(5)°, β = 76.612(4)°, γ =

86.974(5)°, V = 3333.6(3) Å3, Z = 2, ρcalcd = 1.000 mg mm–3; µ(MoKα) = 0.441 mm–1; λ =

1.54184 Å; 173.0 K; 2θ max = 148.74°; total data collected = 23364; R1 = 0.0580 [9670


reflections, and 0 restraints; residual electron density = 0.538 and –0.335 e Å–3. CCDC

967294.

Methyl vinyl ether 5.13. Compound 5.14 (9.0 mg, 0.0094 mmol) was dissolved in CH2Cl2

(2 mL) at rt, and MeOH (2 mL) was added. After stirring for 2.5 h, the solvent was removed

to give product 5.13 as a yellow-orange solid (9.3 mg, quant.). Rf = 0.46 (hexanes/CH2Cl2

1:1); 1H NMR (400 MHz, CDCl3) δ 7.69 (bs, 1H), 7.62 (t, J = 1.7 Hz, 1H), 7.52 (t, J = 1.7 Hz,

1H), 7.38 (t, J = 1.7 Hz, 1H), 7.29 (t, J = 1.7 Hz, 1H), 7.18 (d, J = 1.6 Hz, 2H), 7.13 (d,

J = 1.8 Hz, 2H), 7.00 (d, J = 1.6 Hz, 2H), 6.85 (bs, 1H), 4.58 (s, 1H), 3.53 (s, 3H), 1.40 (s,

9H), 1.32 (s, 18H), 1.30 (s, 18H), 1.27 (s, 18H), 1.24 (s, 9H); 13C NMR (100 MHz, CDCl3) δ

170.1, 161.7, 153.0, 151.8, 151.5, 151.3, 150.3, 145.1, 139.4, 137.8, 137.0, 136.0, 134.0,

126.9, 125.9, 125.5, 125.3, 125.0, 124.6, 124.3, 123.2, 122.3, 121.3, 115.0, 111.5, 111.3,

109.8, 61.4, 61.1, 38.4, 35.3, 35.12, 35.06, 34.84, 34.76, 31.4, 31.2. Note that there appears to

be restricted rotation of one of the aryl rings of 5.13, which leads to non-degeneracy of proton

signals (at 7.69 and 6.85 ppm) as well as carbon signals (an “extra” quaternary carbon signal

of the t-butyl group at 35.3–34.76 ppm). MALDI MS m/z 989 ([M]+, 100); ESI HRMS m/z

calcd. for C69H88N4NaO ([M + Na]+) 1011.6850, found 1011.6853.

Crystal data for 5.13: C69H88N4O, M = 989.43; triclinic crystal system; space group P–1, a =

14.6553(7) Å, b = 15.5967(7) Å, c = 16.2744(9) Å, α = 75.549(4)°, β = 75.159(5)°, γ =

219

72.932(4)°, V = 3375.7(3) Å3, Z = 2, ρcalcd = 0.973 mg mm–3; µ(CuKα) = 0.430 mm–1; λ =

1.54184 Å; 173.1(4) K; 2θ max = 122.86°; total data collected = 15387; R1 = 0.0578 [8483


reflections, and 30 restraints; residual electron density = 0.42 and –0.39 e Å–3. Two t-butyl

groups showed disorder, which have been resolved and refined to the following occupation

factors: C68/C70 = 58:42%; C72/C74 = 67:33%. CCDC 967295.

Cyclobutane 5.11. TCNE (15 mg, 0.12 mmol) was dissolved in dry CH2Cl2 (5 mL) under an

Ar atmosphere, and [5]cumulene [5]tBuPh (0.10 g, 0.12 mmol) in dry CH2Cl2 (10 mL) was

added at –25 °C. After stirring for 1 d at –25 °C, the solvent was removed, and the resulting

solid was purified by column chromatography. In order to prevent conversion of the

compound 5.11 into 5.14 or 5.15, the fraction was kept cold (≤0 °C), and column

chromatography (silica gel, CDCl3) was performed with a jacketed column cooled to –20 to

0 °C through the use of a constant temperature cryo-cool. Collected fractions were kept

cooled on dry ice, and the CDCl3 solvent was removed under vacuum keeping the solution at

<0 °C. This ultimately gave the desired product 5.11 dissolved in a solution of CDCl3. Rf =

0.47 (CHCl3). Rf = 0.40 (hexanes/CH2Cl2 1:1). 1H NMR (500 MHz, CDCl3) δ 7.39 (br s, 4H),

7.16 (d, J = 2.0 Hz, 8H), 1.17 (s, 72H); 13C NMR (126 MHz, CDCl3) δ 203.6, 151.2, 131.2,

128.7, 124.1, 123.7, 109.9, 99.3, 42.9, 34.7, 31.2. ESI HRMS m/z calcd. for C68H84N4Na ([M

+ Na]+) 979.6588, found 979.6575.

Dendralene 5.17. TCNE (31 mg, 0.24 mmol) was dissolved in dry CH2Cl2 (20 mL) under an

Ar atmosphere at –15 °C, and [5]cumulene [5]tBuPh (0.20 g, 0.24 mmol) in dry CH2Cl2

220

(10 mL) was added. After stirring for 1 d, the solvent was removed, and purification by

column chromatography cooled to 0 to (–10 °C) (silica gel, CHCl3) gave compound 5.11 in

CHCl3 (ca. 200 mL). To this solution was added a huge excess of bromine (2.0 mL, 6.2 g,

39 mmol) at –15 °C. The solution was stirred for 1 d, and additional bromine (1.0 mL, 3.1 g,

19 mmol) was added, and the reaction mixture was stirred for 1 d. An aq sodium bisulphate

solution (200 mL) was added. The separated organic phase was washed with saturated aq

NH4Cl (150 mL) and brine (150 mL), dried over Na2SO4, filtered, and the solvent was

removed. Purification by column chromatography (silica gel, CH2Cl2/hexanes = 1:1) afforded

the desired product 5.17 as a brown orange solid (0.070 g, 26%, over two steps). Mp. 214–

218 °C. Rf = 0.64 (hexanes/CH2Cl2 1:1), IR 3063 (vw), 2959 (s), 2906 (m), 2869 (m), 1591

(s), cm−1; 1H NMR (300 MHz, CD2Cl2) δ 7.45–7.44 (m, 4H), 7.35 (d, J = 1.8 Hz, 4H), 7.16

(bd, J = 1.3 Hz, 4H), 1.31 (s, 36H), 1.30 (s, 36H); 13C NMR (100 MHz, CD2Cl2) δ 161.7,

155.3, 151.3, 138.6, 137.3, 124.4, 123.6, 122.8, 111.0, 108.9, 104.0, 95.9, 35.3, 35.2, 31.45,

31.36. ESI HRMS m/z calcd. for C68H8479Br2N4Na ([M + Na]+) 1137.49550, found

1137.49503.

Crystal data for 5.17: C68H84Br2N4, M = 1117.21, monoclinic crystal system; space group

P21/c, a = 17.69599(15) Å, b = 27.7992(2) Å, c = 13.65440(10) Å, β = 102.2304(8)°, V =

6564.60(9) Å3, Z = 4, ρcalcd = 1.130 mg mm–3; µ(CuKα) = 1.856 mm–1; λ = 1.54184 Å;

153.00(10) K; 2θ max = 148.56°; total data collected = 22669; R1 = 0.0389 [11660 observed

reflections with F ≥ 4σ(F)]; wR2 = 0.1047 for 704 variables, 12781 unique reflections, and 3

restraints; residual electron density = 0.73 and –0.44 e Å–3. One t-butyl group showed

disorder, which have been resolved and refined to the following occupation factors:

C72/C73/C74:C72’/C73’/C74’ = 58:42%. CCDC 967296.

Dendralene 5.19. To a solution of compound 5.15 (3 mg, 0.003 mmol) in CHCl3 (ca. 1 mL)

was added a huge excess of bromine (ca. 0.6 g, 0.2 mL, 4 mmol) at 0 °C. After stirring

221

overnight, aq sodium bisulphate solution (5 mL) was added. The organic phase was separated,

washed with saturated aq NH4Cl (5 mL) and brine (5 mL), dried over Na2SO4, and filtered.

The solvent was removed to give the desired product 5.19 as a red solid (quantitative

conversion according to TLC). Mp. 185–195 °C. Rf = 0.31 (hexanes/CH2Cl2 1:1), IR 3073

(vw), 2958 (s), 2868 (m), 2219 (w), 1589 (m), 1461 (s) cm−1; ESI HRMS m/z calcd. for

C68H8379Br2N4 [M – Br]+ 1113.4979, found 1113.4975, for C68H83

79Br3N4Na ([M + Na]+)

1215.4060, found 1215.4063. 1H and 13C NMR spectra were recorded but due to the

unsymmetrical structure of 5.19, neither spectrum offered useful data toward confirming the

structure of 5.19.

Crystal data for 5.19: C68H83Br3N4, M = 1196.11; monoclinic crystal system; space group

P21/c, a = 14.0086(3) Å, b = 20.6382(4) Å, c = 23.1676(4) Å, β = 99.0353(17)°, V =




restraints; residual electron density = 0.63 and –0.69 e Å–3. One t-butyl groups showed

disorder, which have been resolved and refined to the following occupation factors:

C88/C89/C90:C88a/C89a/C90a = 87:13%. CCDC 967297.

Dendralene 5.20. To a solution of compound 5.15 (0.010 g, 0.011 mmol) in CHCl3 (5 mL)

was added an excess of bromine (32 mg, 0.010 mL, 0.20 mmol) at 0 °C. After stirring for 3 h,

additional bromine (0.16 g, 0.051 mL, 1.0 mmol) was added over 2 h in 0.010 mL portions.

After stirring overnight, aq sodium bisulphate solution (10 mL) was added. The organic phase

was separated, washed with saturated aq NH4Cl (10 mL) and brine (10 mL), dried over

Na2SO4, and filtered. The solvent was removed, and purification by flash column

chromatography (silica gel, hexanes/CH2Cl2 1:1) gave the desired product 5.20 as a dark

brown solid (quantitative conversion according to TLC). Mp. 186–187 °C. Rf = 0.32

(hexanes/CH2Cl2 1:1), IR 3065 (vw), 2958 (s), 2904 (m), 2867 (m), 2216 (m), 1587 (m), 1463

(s) cm−1; ESI HRMS m/z calcd. for C68H8479Br2N4Na ([M + Na]+) 1137.4955, found

222

1137.4981, for C68H8479BrN4 ([M – Br]+) 1035.5874, found 1035.5878. Note that 1H and 13C

NMR spectra could be obtained for compound 5.20, but signals in the resulting spectra (see

Appendix) were severely broadened and could not be interpreted.

Products B, C, and D from Figure 5.20 (the “first” reaction). TCNE (5.1 mg, 0.040 mmol)

was dissolved in dry CH2Cl2 (10 mL) under an Ar atmosphere, and [5]cumulene [5]MeOPh

(0.02 g, 0.04 mmol)* in dry CH2Cl2 (10 mL) was added at rt. After stirring for 9 d under Ar,

the solvent was removed, and the resulting product mixture separated by column

chromatography (silica gel, CH2Cl2) to afford products B, C, and D. Each product was

crystallized from CH2Cl2/hexanes at rt, which provided crystals suitable for X-ray

crystallographic analyses. Aside from X-ray crystallographic analyses, however, no further

characterization was possible due to limited amount of pure sample.

* The sample of [5]MeOPh used for this reaction was from my diploma thesis and of

unknown purity. Thus, this is an approximate value for the amount of [5]MeOPh used in this

reaction.

Crystal data for product B: C40H29N4O4Cl, M = 665.12; monoclinic crystal system; space

group C2/c, a = 22.9058(3) Å, b = 14.51631(18) Å, c = 21.6431(3) Å, β = 101.9104(12)°, V =




restraints; residual electron density = 0.28 and –0.28 e Å–3.

Crystal data for product C: C34H28Cl2O4, M = 571.46; triclinic crystal system; space group P–

1, a = 10.5303(5) Å, b = 13.0141(9) Å, c = 13.0624(8) Å, α = 116.326(6)°, β = 101.833(4)°,

γ = 92.661(5)°, V = 1550.98(15) Å3, Z = 2, ρcalcd = 1.224 mg mm–3; µ(MoKα) = 0.244 mm–1;

λ = 0.7107 Å; 173.00(10) K; 2θ max = 57.6°; total data collected = 9966; R1 = 0.0592 [5358

223


reflections, and 0 restraints; residual electron density = 0.37 and –0.94 e Å–3.

Crystal data for product D: C40H29N4O4Cl, M = 665.12; monoclinic crystal system; space

group P21/c, a = 13.6353(7) Å, b = 15.0902(6) Å, c = 17.7149(8) Å, β = 112.406(6)°, V =




restraints; residual electron density = 0.34 and –0.21 e Å–3. The Cl atom showed disorder over

two positions with an occupation of 60:40%.

[5]cumulene 5.31. To a solution of 2.14 (60 mg, 0.066 mmol) in CH2Cl2 (10 mL) was added

anhydrous SnCl2 (40 mg, 0.21 mmol) and HCl (1 M in Et2O, 0.3 mL, 0.3 mmol) at 0 °C under

an Ar atmosphere. After 10 min, the solution was filtered through a plug of basic alumina

oxide to afford the purified 5.31 as a pink solution in CH2Cl2. Since the cumulene is not stable

as amorphous solid, crystalline 5.31 was obtained as pink needles (9.4 mg, 15%) by

overlaying a CH2Cl2 solution with MeOH. Mp. ~ 85 °C (decomp. via loss of shining, 168 °C

melt). Rf = 0.75 (hexanes/EtOAc 20:1). IR 3059 (vw), 2956 (s), 2903 (m), 2866 (m), 2004

(m), 1970 (w), 1590 (m), cm−1; UV/vis (Et2O) λmax 238, 301, 434, 541 nm. 1H NMR (400

MHz, CD2Cl2) δ 7.37 (t, J = 1.7 Hz, 2H), 7.34 (t, J = 1.7 Hz, 2H), 7.16 (d, J = 1.7 Hz, 4H),

6.93 (d, J = 1.7 Hz, 4H), 6.75 (s, 2H), 1.27 (s, 36H), 1.25 (s, 36H); 13C NMR (100 MHz,

CD2Cl2) δ 151.2, 151.1, 150.4, 143.8, 139.8, 138.0, 125.5, 124.0, 123.6, 123.0, 122.2, 113.2,

35.1, 35.0, 31.5, 31.4. ESI HRMS m/z calcd. for C66H8735Cl2 ([M + H]+) 949.61793, found

949.61736.

224

Crystal data for 5.31: C68H90Cl6, M = 1120.10; triclinic crystal system; space group P–1, a =

9.5841(4) Å, b = 12.3984(5) Å, c = 15.5663(7) Å, α = 106.942(2)°, β = 99.0353(17)°, γ =

109.954(2)°, V = 1615.65(12) Å3, Z = 1, ρcalcd = 1.151 mg mm–3; µ(MoKα) = 0.304 mm–1; λ =

0.71073 Å; 173.0 K; 2θ max = 54.84°; total data collected = 13782; R1 = 0.0584 [4589


reflections, and 0 restraints; residual electron density = 0.401 and –0.701 e Å–3. Two

molecules CH2Cl2 per unit cell.

1,2,3,4-Tetrakis(2,2-bis(3,5-di-t-butylphenyl)vinylidene)cyclobutane 5.32. Cumulene

[5]tBuPh (0.030 g, 0.036 mmol) was dissolved in toluene (5 mL) and heated to reflux for 4 d

under air. Evaporation of the solvent afforded a red solid that contained more than one

compound according to TLC analysis. The solid was dissolved in CH2Cl2, and this solution

was overlayed with hexanes. This crystallization gave colorless square crystals, which were

characterized by X-ray crystallographic analysis showing that the sample was compound 5.32.

Unfortunately, the formation of 5.32 was not reproducible and no further, meaningful

characterization could be done.

Crystal data for 5.32: C124H168, M = 1658.58; tetragonal crystal system; space group P4/n, a =

19.3824(4) Å, b = 19.382(4) Å, c = 14.8604(7) Å, α = 90.00°, β = 90.00°, γ = 90.00°, V =

5582.8(3) Å3, Z = 2, ρcalcd = 0.987 mg mm–3; µ(CuKα) = 0.403 mm–1; λ = 1.5418 Å; 172.9(3)

K; 2θ max = 102.76°; total data collected = 5632; R1 = 0.0759 [2509 observed reflections

with F ≥ 4σ(F)]; wR2 = 0.2091 for 294 variables, 2972 unique reflections, and 46 restraints;

residual electron density = 0.41 and –0.32 e Å–3. Three t-butyl groups showed disorder, which

have been resolved and refined to the following occupation factors:

C18/C19/C20:C18a/C19a/C20a = 75:25%, C22/C23/C24:C22a/C23a/C24a = 75:25%,

C38/C39/C40:C38a/C39a/C40a = 75:25%.

225

5.7 References

1 N. Islam, M. Tsukayama, Y. Kawamura, Int. J. Mod. Phys. B 2006, 20, 4619–4624.

2 The work on reactions of allenes and TCNE by Kawamura and coworkers has been

presented in several poster contributions, see: The 8th International Symposium on

Functional pi-Electron Systems, Graz, Jul. 2008; 12th International Symposium on

Novel Aromatic Compounds, p.193, Osaka, Jul. 2007.

3 S. Ueta, K. Hida, M. Nishiuchi, Y. Kawamura, Org. Biomol. Chem. 2014, 12, 2784–

2791.



5 M. Kivala, F. Diederich, Acct. Chem. Res. 2009, 42, 235–248; S.-i. Kato, F. Diederich,

Chem. Commun. 2010, 46, 1994–2006.

6 M. Štefko, M. D. Tzirakis, B. Breiten, M.-O. Ebert, O. Dumele, W. B. Schweizer, J.-P.

Gisselbrecht, C. Boudon, M. T. Beels, I. Biaggio, F. Diederich, Chem. Eur. J. 2013,

19, 12693–12704.

7 Y.-L. Wu, F. Tancini, W. B. Schweizer, D. Paunescu, C. Boudon, J.-P. Gisselbrecht,

P. D. Jarowski, E. Dalcanale, F. Diederich, Chem. Asian J. 2012, 7, 1185–1190.

8 For examples, see: Y. Li, M. Ashizawa, S. Uchida, T. Michinobu, Polym. Chem. 2012,

3, 1996–2005; T. Shoji, S. Ito, T. Okujima, N. Morita, Chem. Eur. J. 2013, 19, 5721–

5730.

9 B. Breiten, Y.-L. Wu, P. D. Jarowski, J.-P. Gisselbrecht, C. Boudon, M. Griesser, C.

Onitsch, G. Gescheidt, W. B. Schweizer, N. Langer, C. Lennartz, F. Diederich, Chem.

Sci. 2011, 2, 88–93.

10 T. Michinobu, C. Boudon, J.-P. Gisselbrecht, P. Seiler, B. Frank, N. N. P. Moonen, M.

Gross, F. Diederich, Chem. Eur. J. 2006, 12, 1889–1905.

11 R. Gompper, U. Wolf, Tetrahedron Lett. 1978, 44, 4263–4264.

12 J. O. Morley, J. Phys. Chem. 1995, 99, 10166–10174.

13 W. Zhu, Y. Jiang, Phys. Chem. Chem. Phys. 1999, 1, 4169–4173.

14 B. Tinant, J.-P. Declercq, D. Bouvy, Z. Janousek, H. G. Viehe, J. Chem. Soc., Perkin

Trans. 2 1993, 911–915.

15 D. Bouvy, Z. Janousek, H. G. Viehe, B. Tinant, J.-P. Declercq, Tetrahedron Lett.

1993, 34, 1779–1782.

16 Z. Yoshida, Pure & Appl. Chem. 1982, 54, 1059–1074.

226

17 H. Fischer, H. Fischer, Chem. Ber. 1967, 100, 755–766.

18 M. Iyoda, S. Tanake, K. Nishioka, M. Oda, Tetrahedron Lett. 1983, 24, 2861–2864.

19 DFT calculations have been performed in collaboration with Prof. A. Görling (FAU).

20 R. B. Woodward, R. Hoffmann, J. Am. Chem. Soc. 1965, 87, 395–397.

21 L. Leroyer, V. Maraval, R. Chauvin, Chem. Rev. 2012, 112, 1310–1343.

22 Cyclic [3]dendralenes, trimethylenecyclobutanes, have been studied, but little is

known about this class of compounds, see for examples: J. K. Williams, W. H.

Sharkey, J. Am. Chem. Soc. 1959, 81, 4269–4272. L. Trabert, H. Hopf, D. Schomburg,

Chem. Ber. 1981, 114, 2405–2414. W. V. Dower, K. P. C. Vollhardt, Angew. Chem.

Suppl. 1982, 21, 1545–1555; Angew. Chem. 1982, 94, 712. W. T. Thorstad, N. S.

Mills, D. Q. Buckelew, L. S. Govea, J. Org. Chem. 1989, 54, 713–776.

23 The structural assignment of 5.20 rested primarily on the mass spectrum and the

analogous formation and properties to 5.19. 1H and 13C NMR spectra were acquired,

but showed severely broadened signals that limited their usefulness (spectra have been

supplied in the Appendix).

24 Formation of the head-head-dimer could not be ruled out, but no evidence for

supporting formation of that dimer was observed to date.

25 See, for example: R. Pal, R. J. Clark, M. Manoharan, I. V. Alabugin, J. Org. Chem.

2010, 75, 8689–8692.

26 Spotted onto a TLC plate from CH2Cl2, this fraction was initially orange/brown and

slowly turned to green as the solvent evaporated.

27 F. Bureš, O. Pytela, M. Kivala, F. Diederich, J. Phys. Org. Chem. 2011, 24, 274–281;

B. Strehmel, A. M. Sarker, H. Detert, ChemPhysChem 2003, 4, 249–259.

28 A. E. Learned, A. M. Arif, P. J. Stang, J. Org. Chem. 1988, 53, 3122–3123.

29 P. I. Dosa, G. D. Whitener, K. P. C. Vollhardt, A. D. Bond, S. J. Teat, Org. Lett. 2002,

4, 2075–2078.

30 M. Iyoda, H. Otani, M. Oda, Y. Kai, Y. Baba, N. Kasai, J. Am. Chem. Soc. 1986, 108,

5371–5372.

31 S. Hashmi, K. Polborn, G. Szeimies, Chem. Ber. 1989, 122, 2399–2401.

32 R. Boese, D. Bläser, R. Latz, Acta Cryst. Sec. C 1999, 55, IUC9900067.

227

33 Compound [5]MeOPh has been previously synthesized, although the synthetic

procedure was different, see: P. Cadiot, Ann. Chim. [Paris] 1956, 13, 214–272. Cadiot

reported that [5]MeOPh could be isolated as a moderately stable red solid.

34 During my diploma thesis (see ref[36]), [5]MeOPh was synthesized three times, but

could be characterized only once via 1H NMR spectroscopy; the other samples showed

broadened signals in the 1H NMR spectra or the presence of impurities. No 13C NMR

spectrum was recorded. Based on UV/vis spectroscopy and TLC analysis, however, all

three samples could be assigned to [5]MeOPh, but with different grades of purity.

35 Samples derived from two individual syntheses of [5]MeOPh done during the

doctoral thesis. 1H NMR spectra of these two samples, however, showed impurities

and/or broadened signals. TLC analysis of both samples showed similar results as

found for samples of [5]MeOPh synthesized during my diploma thesis.

36 For experimental details, see: Diploma thesis “Carbon in One Dimension – Synthesis

of [n]Cumulenes”, Johanna A. Januszewski, Friedrich-Alexander-Universität

Erlangen-Nürnberg, May 2010.

37 [5]oTol was synthesized by Dominik Wendinger.



2188–2208.

40 N. Islam, T. Ooi, T. Iwasawa, M. Nishiuchi, Y. Kawamura, Chem. Commun. 2009,

574–576.

41 M. Iyoda, M. Oda, Y. Kai, N. Kanehisa, N. Kasai, Chem. Lett. 1990, 2149–2152.

228

Appendix

Theoretical calculations

Computational details

All calculations were performed using the Turbomole quantum chemistry suite.1

Geometries were optimized adopting the PBE density-functional2 and the def2-TZVP

Gaussian basis set.3 In some cases additional calculations using the B3LYP4 density

functional were conducted, however, the results remained qualitatively unchanged. All

energies refer to the PBE functional.

Since we expected that intramolecular dispersion interaction between the aryl groups

may influence the relative energies of the investigated molecules, all calculations were re-

done applying the semi-empirical D3 dispersion correction according to Grimme et al.5

Atomic charges were calculated using the natural population analysis (NPA) of Reed,

Weinstock, and Weinhold,6 which provides physically reliable charges without significant

basis set dependence.

Computational results

Computed Energies: Energies of all computed compounds (in Hartree)

compound PBE PBE-D3 B3LYP B3LYP-D3

TCNE –447.1728148 –447.1768399 –447.4510811 –447.45789608

7 –2410.6845575 –2410.7959376 –2412.3441510 –2412.5257118

8 –2857.8748950 –2858.0126355 N.A. N.A.

9 –1428.9389218 –1428.9955777 –1429.9103034 –1430.0033091

11 –2857.9210665 –2858.0562208 N.A. N.A.

12 –2857.9290589 –2858.0831090 –2859.8465573 –2860.0889277

229

References

1 Turbomole 6.5, a development of University of Karlsruhe and Forschungszentrum

Karlsruhe GmbH, 1989-2007, Turbomole GmbH, since 2007; available from

www.turbomole.com.

2 J. P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 1996, 77, 3865–3868.

3 F. Weigend, R. Ahlrichs, Phys. Chem. Chem. Phys. 2005, 8, 3297–3305.

4 A. D. Becke, J. Chem. Phys. 1993, 98, 5648–5652.

5 S. Grimme, J. Anthony, S. Ehrlich, H. Krieg, J. Chem. Phys. 2010, 132, 154104–

154119.

6 A. E. Reed, R. B. Weinstock, F. Weinhold, J. Chem. Phys. 1985, 83, 735–746.

230

Spectra Appendix

Figure A1. 1H NMR spectrum of 2.1.

Figure A2. 13C NMR spectrum of 2.1.

231



232



233



234



235

Figure A11. 1H NMR spectrum of [3]tBuPh.

Figure A12. 13C NMR spectrum of [3]tBuPh.

236



237



238



239



240



241



242



243



244



245



246



247



248



249



250

Figure A41. 1H NMR spectrum of S1.

Figure A42. 13C NMR spectrum of S1.

251



252



253

Figure A47. 1H NMR spectrum of 2.49 (Synthesis of 2.49 under Hay conditions).

Figure A48. 13C NMR spectrum of 2.49 (Synthesis of 2.49 under Hay conditions).

254

Figure A49. 1H NMR spectrum of 2.49 (Synthesis of 2.49 using ethyl bromoacetate as

oxidant).

Figure A50. 13C NMR spectrum of 2.49 (Synthesis of 2.49 using ethyl bromoacetate as

oxidant).

255

Figure A51. 1H NMR spectrum of 2.49 (Synthesis of 2.49 using I2 as oxidant).

Figure A52. 13C NMR spectrum of 2.49 (Synthesis of 2.49 using I2 as oxidant).

256



257



258



259



260



261



262



263



264



265



266



267



268



269



270



271

List of Publications

1. Synthesis and properties of long [n]cumulenes (n ≥ 5)

J. A. Januszewski, R. R. Tykwinski

Chem. Soc. Rev. 2014, 43, 3184–3203

DOI: 10.1039/c4cs00022f

2. Unexpected Formation of a [4]Radialene and Dendralenes via Addition of TCNE to a

Tetraaryl[5]cumulene

J. A. Januszewski, F. Hampel, C. Neiss, A. Görling, R. R. Tykwinski

Angew. Chem. Int. Ed. 2014, 53, 3743–3747

DOI: 10.1002/anie.201309355

3. Unerwartete Bildung eines [4]Radialens und mehrerer Dendralene bei der Addition

von Tetracyanoethylen an ein Tetraaryl[5]cumulen

J. A. Januszewski, F. Hampel, C. Neiss, A. Görling, R. R. Tykwinski

Angew. Chem. 2014, 126, 3818–3822

DOI: 10.1002/ange.201309355

4. Synthesis and Structure of Tetraarylcumulenes: Characterization of Bond-Length

Alternation versus Molecule Length

J. A. Januszewski, D. Wendinger, C. D. Methfessel, F. Hampel, R. R. Tykwinski,

Angew. Chem. Int. Ed. 2013, 52, 1817–1821

DOI: 10.1002/anie.201208058

5. Synthese und Struktur von Tetraarylcumulenen: Charakterisierung der

Bindungslängenalternanz in Abhängigkeit der Moleküllänge

J. A. Januszewski, D. Wendinger, C. D. Methfessel, F. Hampel, R. R. Tykwinski,

Angew. Chem. 2013, 125, 1862–1867

DOI: 10.1002/ange.201208058

6. Oligomers from sp-Hybridized Carbon: Cumulenes and Polyynes

S. Frankenberger, J. A. Januszewski, R. R. Tykwinski

Series Title: Structure and Bonding, Vol. 159, Book Title: Fullerenes and Other

Carbon-Rich Nanostructures, J.-F. Nierengarten (Ed.), Springer Berlin Heidelberg,

2014, 219–256

DOI: 10.1007/430_2013_110

272

CURRICULUM VITAE

PERSÖNLICHE DATEN

Name Johanna A. Januszewski

Geburtsdaten 16.01.1985 (Bytów, Polen)

Staatsangehörigkeit Deutsch

Familienstand Ledig

AKADEMISCHE UND SCHULISCHE AUSBILDUNG

06/2010 – 08/2014 Dissertation „Synthesis and Characterization of

[n]Cumulenes“ an der Friedrich-Alexander-Universität

Erlangen-Nürnberg unter der Betreuung von Prof. Rik R.

Tykwinski

09/2004 – 05/2010 Studium der Chemie (Diplom) an der Friedrich-

Alexander-Universität Erlangen-Nürnberg

09/1995 – 06/2004 Apian-Gymnasium Ingolstadt

AUSLANDSERFAHRUNGEN

09/2011 Dipartimento di Chimica, Politecnico di Milano, Milan,

Italy, Gastwissenschaftlerin

07/2011 14th International Symposium on Novel Aromatic

Compounds (ISNA-14), Eugene, Oregon, USA

Posterpräsentation

AUSZEICHNUNGEN

07/2010 Zerweck-Preis für herausragende Abschlussarbeiten

273

Documents

Synthesis and Characterization of [n]Cumulenes