Non-Magmatic, Volcanic Ash Fragments and Condensates From the Passively Degassing Plume of Popocatepetl

Volcano, Mexico - Indicators of Active Contact Metamorphism and Related Degassing ?

Obenholzner, J.H. (+), Schroettner, H. (*), Golob, P. (*), Delgado, H. (**)

(+) Naturhistorisches Museum/Mineralogie, Postfach 417, A-1014 Vienna, Austria (+436642535783@a1plus.at)

(*) ZFE, TU-Graz, Steyrerg. 17, A-8010 Austria (hartmuth.schroettner@felmi-zfe.at)

(**) Instituto de Geofisica, UNAM, Coyoacan 04510, Mexico D.F., Mexico (hugo@tonatiuh.igeofcu.unam.mx)

Abstract: Magma-wall rock interaction can contribute gases to the magmatic system as well as absorb volatiles to the

country rock. These processes are happening at depth being far away from direct observation. Besides isotopic

composition of gases, if it is possible to sample gases, the micro-analytical investigation of xenoliths is a potential

instrument to study the mineralogical products of related chemical reactions. Plume-derived aerosols can be another

source to obtain information about these processes. At Popocatepetl volcano FESEM/EDS could analyse contact

metamorphosed particles as part of the volcanic ash and aerosol particles from the passively degassing plume.

Wollastonite, hercynite and glass of contact metamorphism-related origin is present in the ash. Condensates from the

passively degassing plume are rich in phosphorus, indicating a possible non-magmatic source for this element.

Keywords: SEM, FESEM; EDS, contact metamorphosed particles, volcanic ash, wollastonite, ferrobustamite,

hercynite, buchites, volcanic aerosol, gas, recent activity of Popocatepetl volcano, Mexico; carbonate platforms,

evaporites.

Micro- and nanometer-sized particles of volcanic origin are transporting various chemical elements into

different layers of the atmosphere, according to the energy of an eruption. These particles can act as a nucleus for

further condensation processes or can trigger heterogeneous chemical reactions in the atmosphere. The abundance of

chlorine in the studied particles, especially in crystallites on spherical particles, can be an underevaluated contribution

to volcanogenic ozone destruction and climate change. A review on observed and possible interdependence of

volcanism and climate change is summarized by Robock (2000).

This study presents new data of plume-derived aerosols, a chemically very distinct environment. The

analytical possibilities of state-of-the-art SEM techniques could provide a unique opportunity to cooperate between the

analyst and the modeler. Results could also be fertile for better modeling of stratospheric chemistry.

1

The size of volcanic fragments, associated wall rock fragments and volcanic aerosols varies from meter-sized

blocks to smaller 1/16 mm for fine volcanic ash (1/16 mm equals 62.5 µm). Volcanic aerosol particles (fluids and

solids) are micro- to nanometer sized. Aerosol sciences classify solid particles from fume (ca. 0.001-1µm), clay

(ca.0.01-2µm), silt (2-20µm), fine sand (20-200µm) to dust (1-10000 µm). Size ranges are according to Hinds (1982).

All particles smaller than several millimeters are candidates for scanning electron microscope (SEM) analysis.

Recent developments in SEM techniques, like the very high resolution field emission gun SEM (FESEM) have

improved morphological, mineralogical and textural examinations. Modern SEM systems are equipped with an energy

dispersive X-ray detector (EDS) for chemical analysis.

A variety of contact metamorphism (CM) and related metasomatic minerals indicate the interaction of magma

with carbonate bearing rocks or other sediments. Contact metamorphic events might occur in the region of a magma

chamber, within magma storage zones at higher levels or as wall rock-magma interaction during magma ascent. These

processes can contribute CO2 (even SO2, F, Cl and other volatiles) to the original gas content of the magmatic system.

Fluffy and spherical aerosols are interpreted to be the product of complex condensation processes, leading to

the formation of fluffy, maybe semi-solid or spherical, solid particles. The latter ones are sometimes coated by

crystallites of an S-bearing Mg-chloride. The high contents of Mg are unusual as the average row of abundant cations

in a volcanic gas from a calc-alkaline magma is Na>K>Al>Fe>Ca>Zn>Mg. Usually Mg is 10-20 times lower than Na

or K. Spherical particles occur as single spheres or as coagulated spheres.

The size of studied spherical particles varies about 1 µm in diameter, coagulated spheres can show diameters

of up to 10 µm, fluffy particles vary from 1.5 to 10 µm. Smaller particles are abundant but have not been analysed yet,

according to limited FESEM time.

According to the limits of EDS (interaction volume depending on acceleration voltage and the density of the

sample), the C- and F-content of the teflon filter and the C-coating) we could characterize the chemical composition of

individual aerosols. Elements detected in fluffy aerosols are Si, Al, Ca, Na, Mg, K, Fe, Ti, P, Cu, Zn, Bi, Pb, Mo, Sn,

S, Cl, O. Most of the spherical particles do not show internal mixing (particle inside a particle) as the surfaces appear

smooth and none of the studied particles showed breakage. Only one particle contains Ti-, Pb- and Cr-rich crystallites

inside the sphere. Many of the coagulated or aggregated particles are internally mixed according to different

chemical/mineralogical components.

2

Elements detected in spherical particles are Si, Al, Ca, Na, K, Mg, Fe, Ti, P, Mn, Cu, Zn, S, V, Ni, O, (Pb, Cr).

Individual spheres show also individual chemical composition. The variety of chemical composition indicate different

micro-environments in the plume and/or at the zone of mixing with the surrounding atmosphere.

Part 1: Indicators of contact metamorphism asssociated with the eruptions at Popocatepetl volcano.

Stratigraphy:

According to Fries (1965a+b) the basement of Popocatepetl volcano comprises Cretaceous carbonates (ca. 3

km) and Tertiary sediments with intercalated evaporites (ca. 500 m). The stratigraphy of the younger history of

Popocatepetl volcano is described in Siebe et al. (1996). In relation to the archeological time scale for central Mexico

the major Holocene and younger Pleistocene volcanic events at Popocatepetl are:

Present Eruption (1994-?) [CM]

Upper Ceramic Plinian Eruptive Sequence (675-1095 AD)

Intermediate Ceramic Eruptive Sequence (125-255 AD)

Nealtican andesite flow (ca. 2300 BP) [CM]

Lower Ceramic Plinian Eruptive Sequence (215-800 BC)

Upper Pre-Ceramic Plinian Eruptive Sequence (2830-3195 BC)

Plinian eruption (ca. 14.000 BP) [CM]

All age data, except for present eruptions, are bracketed by 14C analysis. [CM] refers to contact metamorphosed

xenoliths or particles observed in lava flows, pyroclastic deposits or recent volcanic ash.

Contact metamorphism indicated in past eruptions:

Past eruptions of Popocatepetl (plinian eruption, ca. 14.000 years B.P.) contain blocks of Ca-silicate rocks

with diopside, grossularite as the predominant mineral, or dolomite containing clinohumite ((Mg, Fe++) 9(SiO4)4(OH,

F)2). As the availability of F in most sediments is rather low, clinohumite probably indicates the migration of F from the

magma into the wall rock. Major explosive eruptions caused the fragmentation of the host or country rock and the

sedimentation of polymict fall and flow deposits in proximal facies .

Mm-sized inclusions in pumice of the Tocuila lahar (plinian eruption of ca. 14.000 BP, redeposited 11.000-

12.000 years B.P.) contain grossularite with droplet-shaped haüyne (Na, Ca)4-8[Al6Si6 O24](SO4,S)1-2, diopside, K-

3

feldspar, quartz and sylvite. The inclusions document a homogeneous distribution of CM-altered wall rock fragments

within the vesiculating magma.

Only one sample (P 97-02) from the “ocre surge” beneath the “pink pumice” (Upper Ceramic Plinian Eruptive

Sequence, ca. 800 A.D.) contains abundant euhedral to anhedral anhydrite. Anhydrite could be formed by

hydrothermal alteration of limestone/dolomite or hydrothermal deposition inside the crater. However, a magmatic

origin of the anhydrite is possible but not proven yet.

Contact metamorphism indicated in on-going eruptions:

Ash from the eruptions of winter 1995 (01-02-1995), spring 1995 (03-20-1995 and winter 1996 (11-28-1996)

accidentially contains a variety of contact-metamorphosed rock fragments. The 11-28-1996 ash is characterized by

dome fragments and contains pseudo-vesiculated or mossy shaped aggregates of a Ca-silicate whose composition is

close to CaSiO3. The Ca-silicate is probably wollastonite. High-temperature wollastonite (pseudowollastonite: -

CaSiO3) is known in nature only in pyrometamorphosed rocks. There are two low-temperature modifications of

wollastonite (-CaSiO3 ,wo-Tc and wo-2M). A wollastonite-2M is described from highly metamorphosed ejecta of

volcanoes like Monte Somma, Vesuvius. Modern crystallographic studies could demonstrate that Fe-rich wollastonite

is identical with ferrobustamite. EDS data of analysed wollastonites document traces of manganese, which is present in

minor amounts in ferrobustamite. Only wo-15 has a MnO content (1.21 %) similar to typical ferrobustamite (see table

1.1).

Most of the observed Ca-silicates are highly herterogeneous at very small scale. Electrom microprobe analysis

would barely produce better results than EDS. For these reasons we try to model the P-T-t conditions of Ca-silicate

formation following the well studied reaction quartz + calcite <=> wollastonite and carbon dioxide. Vesicles might be

indicative for fast mineral reactions and CO2-release. High pressure experiments (1GPa) performed at the

GeoForschungsZentrum Potsdam (GDR) document CO2-filled porespace around wollastonite rims grown on quartz

(Lauterjung et al., 2000).The wollastonite particles are unusually rich in Fe; the BSE images do not show the typical

fibrous features known from light microscopy. There is no volcanic glass associated with the wollastonite particles, it

therefore might have formed without direct interaction with the melt. One microclast appears like a vesiculated glassy

fragment, but mineralogically it is SiO2, Ca-feldspar and Ca-silicate. Wollastonite aggregates contain randomly

distributed pyrite (FeS) and rare molybdenite (MoS2) occurs.

Textural and compositional differences between wollastonite types

Micro-clasts are texturally very different. Type 1 appears like an aureole-derived wollastonite (fig. 1). Veinlets

(width ca. 20µm) of massive, chemically homogeneous, non-vesiculated wollastonite are intergrown with K-feldspar

4

which has inclusions of quartz. This assemblage indicates sanidinite facies of contact metamorphism. Sanidinite facies

is defined by low pressure and increased temperature. Water escapes at these temperatures and is not available for

facilitating reactions or crystallizations. The mineralogical consequences of this combination of conditions are:

“1. Chemical and thermal equilibrium are rarely attained. The number of associated minerals therefore is likely to

exceed that demand by the mineralogical phase rule.

2. High-temperature minerals appear in the mineral assemblages of the sanidinite facies.

3. Sanidine, often with a high content of soda, is a critical mineral of this facies. Whether stable or metastable at the

time of crystallization, its presence in a mineral assemblage indicates rapid cooling from an unusual high

temperature of metamorphism.

4. As a result of partial or complete fusion, glass is sometimes present in rocks of the sanidinite facies.” (Turner et al.

1960).

Insert tab. 1.1

Insert fig.1

One micro clast appears like a vesiculated glassy fragment. Mineralogically it is Si-rich, glassy(?) material,

Ca-feldspar and Ca-silicate (type 2). The wollastonite is massive, with some inclusions of quartz and rare sulfides.

Maximal width of the wollastonite is ca. 40 µm (fig. 2). One vesicle-like vug (diameter is ca. 5 µm) is surrounded by

massive, chemically homogeneous wollastonite. Other vesicle-like vugs of the glassy (?) part of the microclast are open

or filled with sulfide.

Insert fig. 2

Type 3 wollastonite is chemically heterogeneous according to Fe-content (fig. 3, 4). This type shows patchy

intergrowth of Fe-rich (FeO ca. 12%) and Fe-poor (FeO ca. 3%) wollastonite which contains randomly distributed

pyrite (FeS) and rare molybdenite (MoS2). The size of patches ranges from 1 to 10 µm in diameter. Type 3

wollastonites are pseudo-vesiculated indicating fast kinetics of the wollastonite-forming reaction with vigorous gas

release. Vesicle diameters range from 1 to 10 µm. The composition of the gas could have been CO 2 or SO3 according

to the reaction calcite + quartz = wollastonite + CO2 or anhydrite + quartz = wollastonite + SO3 (Wood 1994). Both

reactions do not explain the high Fe-content of wollastonite. An Fe-Ca-carbonate component of the Cretaceous

5

carbonate protolith is more likely than a Fe-/Ca-sulfate component of an evaporitic succession or of hydrothermally

altered volcanic rocks.

Insert fig. 3

Insert fig, 4

Another type of wollastonite (type 4) is Fe-poor and occurs in a paragenesis with quartz (fig. 5). These patches

are strongly intergrown with SiO2, are chemically homogeneous, non-vesiculated and have diameters of 10 to 20 µm.

Insert fig. 5

Adjectives (A, B, C) of wollastonite refer to grain size of Nov. 96 volcanic ash: A: <65µm,>32µm; B: >20

µm; C:>20 µm (floats on water). Wollastonite (B1) is vesiculated and rounded, like the grain was tumbled. The grain

shows a reaction rim at the outermost edge, parallel to the grain boundary. Chemically this zone is slightly Mg-rich

relative to the core, maybe indicating an interaction with a fluid while tumbled. Size of the grain is 20x40 µm, vesicles

are heterogeneously distributed, max. length is 10µm. One vesicle intruding the grain from the outer rim is partially

filled with a glassy material rich in P (15-28% P2O5). Wollastonite (B2) is a vesiculated fragment with a glassy rim

(SiO2: 80-95%). At the interface between wo and glass, sphene crystals are abundant. The size of the fragment is 20x50

µm, the maximal. length of vesicles is 5 µm. Wollastonite (B3) is a subangular clast comprising vesiculated wo with

inclusions of Cl-rich glass (SiO2: 75-80%, Cl: ca. 3%), plagioclase-like inclusions and irregularly distributed Fe-oxides,

and a glass (SiO2: 60-65%) vesiculated towards the boundary with wo and a pyx inclusion. Vesicles sometimes

resemble the shape of crystals, maybe remnants of dissolved minerals. The particles rich in Ca-silicates (11-28-96 ash:)

can also contain fluorite crystals (diameter: 1µm), documenting another F-phase except of clinohumite found in

dolostone clasts of the ca. 14.000 BP eruption.

Other particles from the 11-28.96 event are Si-rich glasses of contact metamorphic origin („buchites“)

containing anhydrite crystals and inherited zircons. The vesiculation of this glass is typical bimodal with coalescing

vesicles (~10µm) and a sponge-type vesiculation (~1µm). EDS analysis of the glass is Na 2O=1.93, MgO=0.19,

Al2O3=4.37, SiO2=85.55, K2O=0.30, CaO=0.35, TiO2=0.48, MnO=0.16, FeO=0.53, P2O5=0.95, SO3=4.41, Cl=0.13.

None of the different wollastonite types show alteration (exception: B1) or indications of incipient retrograde

metamorphism like growth of hydrous lime silicates. This observation is interpretated as an indicator for the young age

of wollastonite. Rounded grains refer to tumbling in the conduit. We could detect rounded wollastonite particles (B1)

6

with alteration rims due to interaction with a liquid or gas phase and rounded grains of microliths-containing volcanic

glass, also with alteration rims.

Temperatur indication for wollastonite formation

If Fe-wollastonite occurs with hedenbergite it can serve as a geothermometer indicating 500-600° C. The

formation temperature for MoS2 is reported as 480 to 570 C from fumarole experiments (Bernard 1985) and from

fumarole-wall rock alteration at about 700 C (Getahun et al. 1996). Molybdenite from fumarole incrustations can

contain minor amounts of Fe and Re (Bernard 1985). These data are in accordance with the data reported by Tilley

(1948) for the formation of Fe-wollastonite. Wollastonite is also reported as a sublimate phase from Merapi volcano

(Symonds et al. 1987).

Reported occurrences of wollastonite-bearing xenoliths from volcanoes and intrusions (selected literature)

Wollastonite had been studied extensively from country rocks of intrusions. The few reports of wollastonite

from xenoliths of volcanoes are in the best case mineralogical descriptions, ignoring the volcanological context.

Wollastonite is reported from xenoliths in lava flows of Pacaya volcano (March 1989 and July 1990 eruption – Janik et

al. 1992), Guatemala and of Popocatepetl volcano (2000 years old andesite flow; Goff et al. 2001). Wollastonite-

bearing xenoliths from volcanoes are reported from lavas of Tarumai volcano, Hokkaido (Ishikawa 1953), from

pyroclastic rocks of the 1902 eruption of La Soufrier, St. Vincent (Devine et al, 1980), from pyrometamorphic ejecta of

White Island volcano and Ruapehu, NZ (Wood 1994; Wood et al. 1996). Aramaki (1961) decribes wollastonite from

xenoliths of Asama volcano. Fe-wollastonite is reported from a xenolith of Kanpu volcano, Japan (Isshiki 1954) and

from aureoles of granite intrusions at Beinn an Dubhaich, Skye, dolerite intrusions at Scawt Hill, Co. Antrim and alkali

gabbro intrusions from Muck (Tilley 1948).

Spinel, buchites and other indicators of contact metamorphism

Hercynite (Fe-Al-spinel)-bearing particles are also recognized. Euhedral hercynite (clast 1: H1, H2; clast 2:

H3; tab 1.2) is scattered in a BSE homogeneous matrix of maybe glassy composition (clast 1: M1-4; tab. 1.2).

Insert tab. 1.2

7

The sizes of individual crystals range from <1 µm to 2-3 micrometers. The spinel-shaped crystals from

clusters. Mg might partially replace Fe. The SiO2 and other element contents of H1, H2 and H3 probably derive from

background or interstitial matrix. Hercynite can be the product of a metamorphosed argillaceous rock. It is a rare

mineral because it requires an extremely Fe-Al-rich environment to grow. Baldwin (1955) has investigated the rate of

formation of hercynite from mixtures of FeO and Al2O3 at between 800° and 1300°C. Hercynite was determined from

the ca. 900°C section of a silica tube experiment from Merapi (Le Guern et al. 1982A).

Other fragments are composed of phlogopite and sphene. In between euhedral sphene crystalls very small magnetite

crystalls are widespread. Possible reactions to form sphene and phlogopite are:

TiO2 + calcite + quartz <=> sphene + CO2

2 dolomite + K-feldspar + H2O <=> phlogopite + 3 calcite + 3 CO2

In the case of active contact metamorphism, both reactions can contribute CO 2 to the fumarole gases. As most

marine limestones have 13C values close to zero, one would expect that decarbonation of the limestones would give off

CO2 with 13C-values of about +5.0 per mille (Shieh et al. 1969).

The 20 March 1995 ash contains lithic fragments. They consist mostly of euhedral pyroxenes (tab. 1.3). The crystals

are liberated from adhering melt, but there are relicts of a Si-rich homogeneous, interstitial matrix (tab.1.3). Cavities

between the pyroxene crystals host round Ni-rich minerals.

Insert tab. 1.3

The Ni-bearing minerals are heterogeneous in the BSE image. The ratios of Fe:Cu:Ni:S equal ca. 3:3:3:12

(maximum diameter is 5 µm). Very bright (in BSE image), smaller crystals are embedded in a bright matrix. The

diameter of the smaller crystals is ca. 1 µm. The relative high P2O5 contents of the matrix indicate a CM origin. Fe-Ni-

Sulfides are widespread in hornfels-type xenoliths (#95/73). On the surface of this xenolith pseudo-concentric patches

of Mn-Fe-(Ni?)-coatings are observed, showing dehydration cracks. The genesis of these patches might be related to

processes after deposition.

Further examinations of the recent ash utilizing the SEM and FE SEM could document a variety of contact

metamorphism-related particles. The 01-02-95 volcanic ash contains fragments of anhydrite crystals in a matrix rich in

Al, P and S. This matrix is heterogeneous chemically, Fe-„rich“ and Fe-„poor“ parts are intergrown in a Haring texture

8

(named after Keith Haring, US artist, known for a graphical style of drawing). This texture can be interpretated as a

spinodal decomposition of an original homogeneous glass or as the product of a combustion-like process, frozen in the

state of fingering reaction-products. The size of chemically homogeneous parts of the Haring texture is 1-5µm.

Chemical analysis of this material is as precise as analytical technique (EDS): Na2O=1.93, MgO=0.69, Al2O3=48,86,

SiO2=5.35, K20=1.89, Ca0=6.95, FeO=1.32, P2O5=14.20, SO3=17.78, Cl=0.76 (oxides of Ti, Mn and Ni are 0.00).

Discussion:

As we do not have data of CM-indicators in ash emissions of the past, an interpretation of such minerals as

signals for major eruptions would be highly speculative. At this stage of investigation the occurrence of CM-related

minerals co-documents – together with dome-related fragments - a change in the recent eruptive process. Xenoliths

might be present in the extruding dome. The interpretation of the age of the CM-minerals seems to be crucial: Does

Popocatepetl erupt old CM-minerals or is CM an on-going process when magma is sitting at shallow levels?

At Vesuvius, another volcano situated on massive carbonates, at the end of some of the shorter periods of

activity wo-bearing xenoliths are reported (Zambonini 1935). For example: 1.1796-1822 (wo-bearing ejected block

1822), 2.1870-1872 (wo-bearing xenolith in lava flow of 1872) and 3.1875-1906 (wo-bearing ejected block from the

1906 eruption).This might be an effect of non-systematical sampling or of peaked occurrence of wo-bearing xenoliths

at the end of an eruption periode which terminates with a more energetic, CM clast-clearing eruption .

Wollastonite usually forms between 2 kb (=7-8 km) at >700 C and only 500 bars (=2 km) at 600 C. The

maximum of hypocenters of volcano-tectonic earthquakes between December 1994 and November 1996 had been

between 0 and -3.5 km beneath the volcanic edifice which is based at +2.5 km (Valdes et a. 1995; Arciniega-Ceballos

et al. 2000). Dacitic lavas have temperatures between 750 and 950 C. At the plutonic or magma chamber-stage

temperatures should be lower. Volcanotectonic earthquakes occur in the brittle rocks around the magma reservoire,

conduits and may involve shear linked to stress induced by magma movement or tensile failure of rocks. This can lead

to the fragmentation of surrounded rocks which may be added to the ascending magma or are being draged from wall

rocks by the magma movement (Arcienega-Ceballos, pers. comm.). All decarbonation reactions decrease the volume of

the country rock, maybe creating a network of fractures. The long lasting volcanic activity at Popocatepetl might have

formed a magmatic “karst” system. Fenitization could have progressed at rising temperature and pressure, causing

thermodynamic shattering of the micro-fabric of the surrounding country rock. Mechanical stress, however, which

generally preceeds metasomatism, can be widespread and locally intensive, and brecciation is a common feature.

Well studied aureoles, like the Ballachulish (diameter: 5.8 km) or the Monzoni complex (diameter: 2.5 km),

demonstrate that a ca. 1000°C hot Si-rich intrusions forced the limestone-rich environment to react and forming

9

isogrades of decarbonation reations as far away from the contact as ca. 500 m (Masch et al. 1991; Heuss-Aßbichler et

al. 1991). The average width of hypocenters of volcano-tectonic earthquakes is ca. 10 km at a depth of –2 km.

Applying these data to the subsurface structure of Popocatepetl, a gigantic reservoire for CO2 release would be

available. The formation of wollastonite is normally approximating the highest grades of CM against the

pluton/magma, but is not next to the immediate contact with the melt. Fluids can transport ions deep into the country

rock. Examples are known like wollastonitized cherts intercalated between limestones. These cherts operated as

metamorphic aquifers (Romer et al. 1998).

The time of formation of wollastonite in aureoles can be very low. 10 mm thick aggregates of wollastonite can

grow in ca. 500 years at >800°C (Kridelbaugh 1973). The 20 µm thick aureole-like wollastonite (type 1) could grow in

ca. 1-2 years. Kalinin (1967) could synthetizes wollastonite at temperatures as low as 350 C over periods of 2 to 10

days at pressures of ca. 150-1700 bars Harker et al. (1956) could grow wollastonite in ½ hour at 660°C and ca. 340 bars

in the presence of water. Tanner et al. (1985) report replacement of calcite by wollastonite (100 µm) in 1.1 years at

850°C and 1 kb. A study to evaluate the residence time of the magma beneath Popocatepetl as suggested by Pyle

(1992) and Albarede (1993) would be an important step for further conclusions.

Goff et al. (2001) could measure CO2/SO2 excursions of normal CO2/SO2 flux, roughly 30 times higher than

typical values. These events had been interpretated as active assimilation of limestones into the magma chamber.

Contact metamorphism of evaporites is exceptionally well documented at the Werra-Fulda minig district

(GDR). Sulfur enrichment around basaltic dikes and rare CO2 explosions are described by Knipping (1989). Many

evaporite minerals start decripitating at low temperatures, like carnallite, releasing HCl. Anhydrite reacts at 1193° C to

CaO + SO2 + O. Most of the evaporite minerals melt at magmatic temperatures (Herrmann 1980).

Part 2: Volcanic aerosol particles

Aerosols are a suspension of solid or liquid particles in a gas. Aerosols are usually stable for at least a few

seconds and in some cases may last a year or more. The term aerosol includes both the particle and the suspension gas,

which is usually air. In the plume environment magmatic, non-magmatic (CM-derived) gases and entrained air are

blended together.

Aerosol particle sizes range from 0.001 to over 100 µm. The term fine particle describes a morphologically

well defined object (size < 2µm), coarse particles are larger than 2 µm in diameter (Hinds 1982). A secondary particle

is a particle formed in the air, usually by gas-to-particle conversion; also sometimes used to describe agglomerated or

redispersed particles. Dust are solid particles formed by crushing or other mechanical breakage of a parent material.

10

These particles have irregular shapes and are larger than about 0.5 µm. Agglomerate is a group of particles held

together by van der Waals forces or surface tension. Aggregate is a heterogeneous particle in which the various

components are not easily broken apart. Particles can be or can contain minerals or amorphous substances. The

parameters to describe aerosols or aggregates are size, geometry, primary particle etc., surface and if possible, local

chemistry (Willeke et al. 1993, p.18).

Methods to analyse aerosols are described in Willeke et al. (1993). The authors compare the SEM/EDS with

other instruments, but do not discuss the modern, high resolution FE SEM, which had not been available at this time.

The term condensate is used in this paper to descibe the product of a gas-to-particle conversion, regardless if the

particle is solid or liquid. Solid particles formed from a gas are also called sublimates.

Sampling Procedure (H. Delgado)

Volcanic aerosols from Popocatépetl volcano were obtained by flying across the plume. This was carried out

by using a fixed-wing two engine airplane (Cessna 421), departing from Atizapán Airport (nearly 75 km northwest of

the volcano) at 2,600 masl. A couple of containers with Teflon filters (1 µm) impregnated with 3M LiOH/20 %

glycerin were set out of the lateral window (co-pilot’s window) of the craft, separated about 10 cm from the plane’s

fuselage connected by Teflon piping to an air pump controlled from inside the plane. The impregnation had been

designed to adsorb acid gases. The holders had a rigid plastic cover and the pump was off at the time of take-off. The

holders were uncovered above the volcano at about 19,000 ft (5790 masl) and the pump was turned on with an average

flow rate of 30 ml/min.

The cross section of the plume was set at the southeastern side of the volcano with an azimuth of 200º the

volcano’s plume was moving outwards with a rough direction of 290º. At this altitude the ambient temperature was - 3º

C as recorded from the aircarft’s instrumentation. Ten cross sections were made between 18500 ft and 12100 ft (5640

and 3690 masl) spaced every 400 ft in altitude (122 m). Every line had a distance of 12.7 km in preserving a direction

normal to the plume’s direction. At 18500 ft (5640 masl) the volcanic plume was detected by the correlation

spectrometer (COSPEC) yielding sulfur dioxide burdens as high as 236 ppm•m, and maximum volcanic carbon dioxide

concentrations of 0.5 µmol/mol. 1 ppm•m is the unit of SO2 measured by COSPEC, it means the total SO2 concentrated

along a line of a meter since the instrument integrates it along the optical trajectory between the instrument and the

light source.

The maximum burden of sulfur dioxide was 445 ppm•m, and maximum concentration of carbon dioxide was 8

µmol/mol, both detected at 16900 ft (5213 masl). The total output of sulfur dioxide calculated by integration of the

cross sections was 6530 ± 790 tons/day. Carbon dioxide emission rate was calculated at 13630 ± 1000 tons/day.

11

Traditional analysis of volcanic aerosols (selected literature):

The production of aerosols, dust and gases is difficult to assess, because of the scarcity of data. In addition to

primary particles in the form of ash, volcanic activity produces significant quantities of reactive gases such as H2S, SO2,

or HCl. The Mt. St. Helens eruption in 1980 had given a first opportunity to carry out scientific investigation (Hidy

1984). Sampling is performed as stratospheric aerosol collection (Oberbeck et al. 1992) or tropospheric aerosol

collection, or by aircraft around non-eruptive plumes (Phlean et al. 1992). The size distribution of particles of the Mt.

St. Helens plume ranges between 10-2 and 10+2 µm in diameter (Hobbs et al. 1981). Most of the analytical studies

performed on volcanic aerosols utilized bulk chemical analysis, standard SEM analysis of solid particles (Meeker et al.

1991), or TEM analysis. Liquid aerosols from the plume of Pacaya volcano (1978) and solid/liquid aerosols from the

plume of Fuego volcano (1978) had been studied by TEM (Rose , 1980; Ammann et al., 1993). Indirect measurements

of aerosols are widely used, like spectral optical depth (Valero et al., 1992), LIDAR (Winker et al., 1992) and a variety

of satellite-hosted instruments.

Several publications are devoted to individual eruptions. For example: Mount St. Helens (Newell et al. 1982),

climatic effects of the eruption of El Chichon 1982 (several papers in Geophysical Research Letters, 10/11, 1983, the

Pinatubo eruption of 1991 (Sheridan et al. 1992, Solomon et al. 1993, Hamill et al. 1996), Kilauea: Zoller et al.,

(1983). Mt. Erebus: Meeker et al. (1991), Mount Etna: Andres et al. (1987) or Heimaey: Mroz et al. (1975). Global

effects are discussed by Handler (1989), Nriagu (1989) and Mills (2000).

Results from standard SEM analysis:

The filters had been analysed by wet chemical techniques at the Los Alamos National Laboratory (EES-1).

Only a few gases were sufficiently concentrated in spring 1997 to get results from the filter. CO 2 (2400 ppm), S (2ppm)

and Hg (1ppb) could be analysed (Counce, pers. comm.). Particles show a wide range of composition. Solid aerosols

are volcanic glass, magmatic silicates, Cl-,P- and S-bearing minerals/particles. We could identify an S-bearing mineral,

mercallite: KHSO4, known from fumaroles of Vesuvius (Imbo 1965); mercallite indicates formation temperature less

than 650 C. A Ca-chloride is present (antarcticite: CaCl2.6H2O or sinjarite: CaCl2.2H2O), also a Ca-rich particle (CaO

or Ca(OH)2). Cu- and Zn-bearing particles could be detected. Most of them also show a sulfur peak. Cu and Zn-bearing

condensates are reported by Taylor et al. (1973) from Central American volcanoes.

Galindo et al. (1998) analysed solid aerosols from an impactor filter collected at Popocatepetl. This study

utilizes standard SEM and XRF for aerosol characterization. The particles described are spherical or blocky,

fragmental, dust-like interpreted as pulverized rock. XRF analysis documents: Si, Al, Ca, S, P, Cl, K, Ni, Fe, Ti, Sc,

Cu, Zn, Mn, Sr, Cr, Co, Y, Br, Se, Ga, Rb, Hg and Pb.

12

Realizing the much better resolution of FESEM than SEM we did not continue on standard SEM

investigations. Detection limits for different elements are different for XRF and EDS, nevertheless we can add more

elements to the elements as detected by XRF and can relate elements to certain particles. Important to note is that

Galindo et al. (1998) and we worked on 2 different filters collected at different altitudes and different dates.

Results from field emission gun SEM analysis:

In a first attempt field emission gun (FE) SEM was tested to elucidate the composition of very small particles

(<5µm) on the impactor filter. The advanced resolution of the FESEM permits new insights in micro-morphology and

micro-chemistry of solid aerosols. Analytical limits are given as the filter utilized is made from teflon. EDS spectra do

not permit characterization of C- and F-peaks generated by particles. First studies taught that chemical characterization

of particles does not necessarily define minerals. As a preliminary tool to describe what can be analysed, we would like

to introduce a micromorphological classification, rather than a chemical/mineralogical. Often tiny spheres (S?) are

adhering to the surface of particles. Explanations to the tables 2.1 to 2.4: Diameter (d) and length (l) are measured in

µm. Elements listed after the “&”-sign are present in minor amounts. Elements listed are detected by a peak in the

spectrum. Further FESEM images than the presented ones can be obtained from Obenholzner and Schroettner.

Tables 2.1 to 2.4 document particles analysed by FE SEM. Most of the particles are imagined by a composite

image of BSE and SE images. Several particles are not documented as images. The category, single or aggregated

particle, describes the overall appearence according to chemical composition and morphology. Larger aggregated

particles consisting of dozens or hundreds of fragments are not mentioned in the tables. For these particles EDS

analysis is problematical as X-rays would derive from too many individual, smaller 1-5 µm fragments. Size,

morphology and chemical elements detected by EDS are listed for all other particles. Interpretation of the detected

elements refers to known minerals and their chemical composition. For most of the particles such an interpretation is

highly speculative. The morphology does not show typical crystal structures. This has to be kept in mind for all

particles categorized as fluffy, spherical or complex.

13

Complex or simple particles can be observed as aggregates or clusters.. The overall appearance is blocky or

fragmental (table 2.1).

insert tab. 2.1

As examples some particles are described in more detail. A pseudohexagonal particle shows network-like

structure in composite image. (fig. 9 c & d). Others are irregularly formed particles with rough surfaces deriving from

coagulation of thousands of small spheres building bumps and horn-like extrusions (fig. Fig. 10 c). Pop9 and Popoc18

are aggregated particles, slightly elongated (length is ca. 20 µm); these particles consists of hundreds of mineral or

glassy fragments. Popoc16 consists of ca. 2 dozens of smaller fragments, lenght is ca. 10 µm. Similar appearence is

documented by Popoc28 (diameter ca. 10 µm)). These particles are not listed in tables 2.1 to 2.4.

Further studies could demonstrate that individual particles often appear chemically homogeneous, like

minerals would do, but morphologically they are fluffy, amoeboidal, botryoidal, moss-like, flakey, spongey or gel-like.

They might be the remnants of collapsed or decrepited, short living particles.

Insert tab. 2.2

14

As examples some particles listet in table 2.2 are described in the following section. One amoeboidal particle

shows a rough surface, diameter is ca. 3 µm. The overall appearance of this particle refers to a process which

coagulated smaller spheres in a string-like manner. Individual spheres show diameters of ca. 50-100 nm. Al, Si and Fe

are the dominant elements with minor aounts of P, K, Ca and Ti. The outline of this particle shows edges, like it was

broken away from a larger particle. The particle might have been a part of a fumarole-related deposition (fig. 10 a &

b)). Popoc17 has a cloud-shape with a core surrounded by a ring. Popoc22 has a feather-like appearance (fig. 6 a).

Lenght of the particle is ca. 10 µm. Popoc25 shows 2 particles side by side. One is defined by edges the other is fluffy.

Diameter of each particle is ca. 1 µm. Popoc26 shows 2 fluffy particles side by side (fig. 6 d). Diameter of each particle

is ca. 10 µm. Popoc27 (fig. 6 b) is a fluffy particle, diameter is ca. 7 µm. Popoc10 is another fluffy particle (fig. 6 c).

Insert fig. 6

Spherical particles can have a variety of chemical compositions and might be coated with a chemically distinct

population of more or less well crystallized minerals (tab. 2.3). We do not know yet if the chemical different spheres

are hollow or massive. Processes forming spherical particles are known, but are not attributed to volcanic plumes.

Nucleation, adsorption, desorption, condensation, coagulation, condensation, evaporation or diffusion (deposition

across a boundary layer) are well modeled (Whitby et al. 1997). At this point of data collection it is difficult to imagine

processes like sintering (Seto et al. 1997) or spray pyrolysis (Jain et al. 1997) happening in plumes. Wohletz et al.

(1984) could produce spherical particles in water-melt interaction experiments. They found hollow spheres

(cenospheres) and filled hollow spheres (plerospheres). Spherical particles are also known from phreatomagmatic ash

deposits around Kilbourne Hole, New Mexico. Spherical particles from phreatomagmatic processes should be easily

characterized by similar chemical composition.

15

Fall times of spherical particles can be calculated according to Stoke´s law. As an average the time of fall for a

50µm sphere would be ca. 16-24 hours for 14.000 m. For finer grains, fall times are uncertain. Clumping due to

electrostatic forces and moisture can enhance fall rates. Fine grains that do not clump may have lower fall rates due to

lift resulting from turbulences (Riehle et al. 1994).

Le Guern et al. (1982B) could collect spherical aerosols utilizing the silica tube to condensate gases from a

lava flow of Mt. Etna. SEM investigations could demonstrate that these spherical aerosols had been droplets which

crystallized as aphthitalite (K2Na(SO4)2). The shape of the aphtitalite aggregates still resembles the droplet (diameter is

ca. 300 µm), from which fluid evaporated. Internal structures show also droplet-like features. According to the much

better resolution of the FE SEM we could detect many types of spherical particles. Important to note is that none of the

described particles in this study shows features of desiccated droplets remaining crystallized substances.

Insert tab. 2.3

One type of particle has been observed several times. Popo11 (fig. 7 a) consists of Ca, P, Mg, and Cl with

minor amounts of K, S, Si, Al, Na. A very similar morphology and composition shows popo10 (fig. 7 b). On the

surface of a Ca-phosphate (apatite or collophane = crypto-crystalline apatite-like mineral) sphere chlorides are

homogeneously scattered. The anhedral crystallites are a complex S-bearing Mg-chloride (i.e. bischofite: MgCl 2.6H2O;

carnallite: KMgCl3.6H2O; chloromagnesite: MgCl2; dansite: Na21Mg(SO4)10Cl3; kainite: MgSO4.KCl.3H2O). According

to EDS spectra a CO3-component might be present in the formula. The origin of Mg-chlorides might be related to the

reaction: MgSiO3+2HCl=MgCl2+H2O+SiO2 (Stoiber et al. 1970). Similar reactions can be assumed for the formation of

CaCl2 (CaSiO3+2HCl=CaCl2+H2O+SiO2; CaAl2Si2O8+8HCl+8H2O=CaCl2+2SiO2+2AlCl3.6H2O?). Chloraluminate is

known as sublimate at Cerro Negro volcanic fumarole (Stoiber et al. 1973).

The source for phosphorus in apatite might be related to contact metamorphism, as mobilisation of P is

documented in P-rich buchite particles. Red clayey soils deriving from volcanic detritus can be highly phosphatic

(Blackburn et al. 1969). These types of red soils can contain P2O5 between 7 and 20% and might be intercalated

between the limestones underneath Popocatepetl. It is worth mentioning that Ivlev et al. (1995) found relatively high

concentrations of S (4300-114000 ng/m³), Cl (1000-4100 ng/m³) and P (700-<4000 ng/m³), analyzing the ash collected

between Dec. 1994 and Jan. 1995) at the Hnos. Serdan airport (Mexico, D.F.).

Particles with low P-contents had also been detected at the degassing lava flow of Mt. Etna (2001), Italy. Silica

tube experiments from the fumaroles of Colima volcano, Mexico document crystallization of euhedral apatite at

temperature ranges from 600 to 650 C (Taran, pers. comm.). Data of P in volcanic gases are rare. Taran et al. (1995)

16

report 2000-30800 ppb from Kudryavy volcano, Kurile Islands, Russia. Carmichael et al. (1974) mention P from the

Showa-shinzan fumarole (Usu volcano, Japan) recalculated as P4O10 (0,0008-0,0053%). Acid gases are suspected to

leach P from older volcanic rocks or from the continental crust. The average content of P in the upper continental crust

is 665 ppm according to Wedepohl (1995).

The overall appearence of these spherical particles refers to condensation in the plume without essential

subsequent particle-particle-collision. As a model for particle formation we assume as a first step the condensation of

the Ca-phosphate, also trapping minor amounts of Si and Al. Later this sphere became the nucleus of further

condensation of a fluid phase coating the micro-sphere. The evaporation of the fluid caused the crystallization of the

Mg-chloride. These three steps of particle formation document the changes of the chemical environment inside the

plume. Changing parameters along the path of this particle had been temperature, gas chemistry, dilution, velocity of

particle and pressure. At this point of time we do not know if the three steps of formation document a descending or

ascending path of the particle in the plume.

Repeated FESEM analysis of this type of particle in 1999 by Heinrichs (Univ. Goettingen, Germany)

documented the stability of spheres and Mg-Cl-crystallites. The filter had been stored in a sealed plastic box under

prevaccum conditions in an exciccator. Electron backscatter diffraction (EBSD) on spherical particles did not provide

results due to surface roughness, shape and/or missing crystallinity.

Insert fig. 7

Other spherical particles are compositionally Fe, Si, Cl with a few Ca sulfate crystalls on top of the sphere

(popoc01, close-up: popoc02). The sphere is slightly acentrical, partially exhibiting a subsurface layer or core (Si?).

The surface layer might be an Fe-chloride. Compositionally different is a Zn-bearing sphere irregularly covered with

clustering Mg-chlorides (popoc07, close-up: popoc08; fig. 7 c and d)). This particle could be interpreted as an

apatite+anhydrite-phase sphere, trapping Si, Al and Zn during condensation.

Spherical particle (diameter ca. 1.5 µm) can enclose small crystals rich in Pb, Cr or Ti with minor amounts of

Cu, Al, Si. The particle itself seems to consist of C, also traces of Al and Si are present (pop14-1). This particle

encloses ca. 12 crystalls (fig. 8 d). Very rare CaSO4 crystals are growing on top of spherical particles (popoc01/02 – fig.

8 a and b)). This could be interpreted as a reaction between acid sulfates droplets and Ca-bearing particles (Ca-oxides, -

hydroxides, -carbonates or silicates) brought together by in-plume processes. Fig. 8 c shows another spherical particle,

composed of Na, Mg, S, and minor amounts of Ca, the diameter is 1,5 µm (pop3-1). The surface of this tiny sphere is

scattered with droplet-like objects, organized in a ring-like structure. In the center of the ring larger droplet-like objects

17

are visible (diameter ca. 100 nm). The composite image of the host-sphere refers to a homogeneous material. The

droplet objects are totally embedded in this material, some parts seem to stick out of the skin of the host-sphere.

Insert fig 8

Coagulated spheres and collapsed (?) particles are imaged in popo09 (fig. 9 b). Spheres are rich in Ca and P.

Popoc20 documents several coagulated spheres enbedded in a matrix. Lenght of the particle is ca. 1.5 µm (Fe, Zn, Si,

Mg, Al, Ca, S, Cl with minor amounts of K, Mn). Similar morphological features shows Popoc21. Spheres seem to

have coagulated to worm-like structures. Lenght of the particle is ca. 4 µm (Fe, Si, Ca, S, Cl also minor amounts of.

Mg). Popo 12 is an aggregate of tiny spheres (<2µm) and blade-like crystals (fig. 9 a). The bulk particle consists of Si,

Al, Na, S, Cl, Ca and minor amounts of Mg, P, K, Ti, Fe, Cu, Zn. Larger spherical particle can be coagulated with

smaller particles. The diameter is ca. 3 µm, it consists of P, K and minor amounts of Cu, Si, S, Cl (pop8-1). The shape

of particles can be deformed as seen in a pseudo-spherical particle, diameter is ca. 3 µm. The surface seems to be

wrinkled, maybe as an effect of desiccation (pop13 – fig. 10 d).

Insert fig. 9

Insert fig. 10

Particles resembling crystalline material. These particles appear compact, crystalline without euhedral

morphology (tab. 2.4). According to spectra they show mono- or polymineralogical composition. It was possible to

identify feldspar fragments, some are highly corroded and Fe-Ti oxides (?). Ca-rich particles are present.

18

Insert tab. 2.4

Other analysed particles are fragments of vesiculated glass with adhering dust particles. The following images

document vesiculated glass (pop16, pop17 – close-ups: pop18 and pop19, popoc19).

Metals in volcanic gas are postulated as halogen, sulphate and oxide compounds. Chloride and fluoride

compounds are considered to be the primary transporters of metals in volcanic gas. Significant correlations between

bromine content and metal concentration suggest that bromide compounds play a role in the transportation of metals in

volcanic gas (Gemmell 1987). Copper is known for its extreme partitioning into a magmatic vapor phase in

pantellerites (Italy) (Lowenstern et al. 1995). Cu, Zn and other metallic trace elements are known in fumarolic

condensates and as soluble material on ash from Central American volcanoes (Gemmell 1987, Taylor et al. 1973).

Vanadates are rare but known from sublimates (Hughes et al. 1987). Vanadium containing Na-K sulfates and V-W-Co

phases are described from Colima volcano, Mexico (Taran et al. 2000, 2001). Pb and Bi can easily be volatilized as

outlined by Stimac et al. (1996) and Symonds et al. (1987). Galena (?) and PbO (massicot) might be present in aerosol

particles, indicating temperatures <897 °C (melting temperature of massicot). Cotunnite (PbCl2) is known from recent

sublimates of Vesuvius (Houtermans et al. 1964). The Pb-contents of several particles should be kept in mind when

applying decay-series disequilibria methods.

Discussion:

Various microphysical processes affect the aerosols, for example nucleation, coagulation, growth by

condensation and sedimentation. Theoretical models of nucleation of stratospheric aerosols after a large volcanic

eruption exists on the assumption that newly formed particles are droplets of condensed sulfuric acid. The spontaneous

formation of sulfuric acid water solutions not involving any other species or any foreign surfaces is modeled by the

theory of homogeneous heteromolecular nucleation. Problems in these assumptions are discussed by Hamill et al.

(1996). The spectrum of observed particles and their chemical composition can barely be the product of such processes,

although homogeneous (gas-to-particle conversion) nucleation must have played a major role in the formation of

spherical particles. Assuming a constant growth rate (several nm per hour) for particle formation, nucleation must have

started several times to produce different size and chemical categories of particles, reflecting changing chemical and

19

physical parameters in the plume. According to limited FESEM time we could not analyze statistically size and/or

related chemical composition.

The occurrence of S- and Cl-bearing minerals/particles should be considered as a contribution to the bulk S-

and Cl-vapour budget of Popocatepetl volcano. For future studies the use of field emission gun (FE) SEM, imaging ion

microprobe or time-of-flight secondary ion mass spectrometry (TOF SIMS) is recommended. These studies should

focus on pressure and temperature conditions of mineral/particle formation to estimate the origin (magmatic or

sedimentary) of S, Cl and C (?).

Several elements observed as chemical components of analysed particles have also a crucial importance as

chemical components of volcanic gases, like S and Cl. These preliminary results do not permit a quantification of how

much S and Cl cannot be measured by open- or closed-path gas analysers. Correlation spectrometers (COSPEC) or

Fourier-transform infrared (FTIR) spectroscopy can detect only gaseous phases. The latter system is able to perform an

aerosol correction but the chemical composition of aerosols is usually not determined. Many volcanoes release gases

like SO2 at a dimension of tons/day (Bluth et al. 1993, Stoiber et al. 1987). Such estimates neglected sulfur in any other

form than gaseous SO2 (Malinconico 1987). If S in solid aerosols contributes to the total S budget of the atmosphere,

this contribution is bracketed by sedimentation of particles and chemical reactions taking place during fall time.

Goff et al. (1998) calculated that Popocatepetl released from December 1994 to November 1996 3.9 Mt SO2,

16 Mt CO2, 0.75 Mt HCl, 0.075 HF, 260 t As, 2.6 t Hg and 200 Mt H2O. Data had been obtained as a combination of

“volatile” trap chemistry (500 ml of 4N KOH as collector fluid) and SO 2 flux. Delgado et al. (2001) report 9 Mt of SO2

for the pre-eruptive periode in 1994 to Jannuary 1st 1998. These data are mainly obtained by COSPEC. Love et al.

(1998) document SiF4 detected by FTIR.

None of the analysed particles yielded a Hg or As signal. Fluorine could not be detected by EDS as the filter

utilized was made from teflon. Allmost all particles show F in their spectra, probably as a result of X-ray collection

from the filter.

EDS is not the ultimate tool for chemical analysis of single particles in the submicron region. Mass

spectrometric methods (Peter 1996), like time-of-flight secondary mass spectrometry (TOF SIMS), can analyse

isotopes or particle-induced X-ray emission (PIXE) for trace elements, if the optical resolution provides the

identification of single particles. According to the detection limits, instrumental neutron activation analysis (INAA)

would be the best instrument for chemical analysis. The disadvantage is the impossibility to focus on single particles.

However Allard et al. (2000) had been successfully analyzing bulk elements combining INAA and ICP-MS on a filter

from the plume of Stromboli. Analytical transmission electron microscopy (TEM) would be another approach, but

sample preparation is difficult. TEM analysis should be performed on several spherical particles to define the physical

20

state (crystallized or amorphous). Glass created by condensation of vapor is theoretically possible (Vogel 1994).

Progress in applying atomic force microscopy (AFM) to insulators (Pethica et al. 2001) might open new doors to study

physical and chemical properties of aerosol particles.

At Popocatépetl we discussed the possibility of P-contamination of the plume by P-fertilizer used in Mexico.

An FESEM study of the fertilizer could demonstrate, that even very small particles are always blocky. The P containing

volcanic particles are fluffy or spherical, showing smooth surfaces. A coagulation or aggregation of wind-blown,

blocky fertilizer particles can be excluded. A contamination at the molecular scale is possible, but would not lead to the

necessary concentration detected by EDS.

The fraction of elements, like S and Cl, in solid particles compared to these elements in volcanic gases is not

determined yet. A quantification of S and Cl in particles by automatic EDS analysis is possible, if the surface of the

filter is very plain. The ZFE recommends nuclepore filters. The efforts of plume and fumarole monitoring could

become veiled by ignoring the chemical composition of condensing particles. Chemical analysis of aerosol collection

on filters compared with analysis of lava leads to enrichment factors for elements in the plume. Normalizations against

highly volatile elements, like bromine, helps distinguish elements hosted in glass shards and magmatic crystals, from

elements hosted in condensates. This FESEM/EDS study detected i.e. Si and Al, typical for silicate melts and a variety

of magmatic crystals, as constituents of condensed particles, indicating the existence of SiF4 and AlF2O(?) in the plume.

CONCLUSIONS:

Recent models of the geochemical C and S cycle (Arthur 2000) do not differentiate between volcanic gases,

derived from mantle and/or subducted rocks, and gases derived from the basement of volcanoes. Volcanoes situated on

massive carbonates and/or intercalated evaporites should be treated separately in model calculations. The importance of

volcanic recycling of ancient evaporites is outlined by Risacher et al. (2001) for the 1993 Lascar eruption.

Keeping in mind that 1 kg of natural limestone contains 0, 4395 kg CO2 (reference limestone from Maria

Trost, Austria; Schoklitsch 1935), an important contribution of CO2 from decarbonation reactions can be expected. An

average density of limestone is 2.6 g/cm³. In a cylindrical model, a Monzoni-like intrusion (diameter ca. 2.5 km),

would release ca. 105 Mt of CO2, assuming a 20 m wide zone of total decarbonation around a 500 m high intrusion.

Comparing this model with the known subsurface structure of Vesuvius or Popocatepetl, these assumptions are minima

according to intrusion or magma chamber hights. These assumptions are exaggerated according to known outcrops at

Monzoni or Ballachulish, where 20 m wide, tube-like zones of total decarbonation do not exist, but decarbonation

reactions are documented as far as 500 m away from the contact. Time is another parameter limiting the value of such a

21

model, is magma stored long enough to drive the necessary decarbonation reactions? At prevailing pressures in the

realm of contact metamorphism gases might enter the magmatic system, but also can be dissolved in secondary melts

and fluids. Melt and fluid inclusion studies of skarn xenoliths are an important approach to get insights into the

complex interaction between melt and carbonate host roks, as documented from Vesuvius (Gilg et al. 2001; Fulignati et

al. 2001). Skarn lithics from the plinian Avellino, and Pompei, and the sub-plinian Pollena eruption had been studied,

but no age data of the skarn-forming process could be obtained. Although these microanalytical techniques provide

precious data, a link to the eruption history of a carbonate hosted volcano is still missing.

The critical temperature for CO2 is only 31.1°C, already attained at a normal geothermal gradient at ca. 1 km

depth. CO2-rich fluid inclusions originating from deeper or hotter environments must have been trapped in a

supercritical state, making it difficult to read the depth of formation. Density measurements might bridge the gap, but

leakage during sometimes long lasting exposure of xenoliths to atmospheric pressure conditions cannot be excluded as

mineralogical encapsulation does not provide perfect sealing, especially after being incorporated in volcanic eruptions.

The calculated and measured 16 Mt of CO2 from Popocatepetl (Dec. 1994-Nov. 1996) can be a mixture of

magmatic and basement-derived CO2. Remaining questions are, how does the input of gases from metamorphosed

country rock effect the style and the dynamics of eruptions of volcanoes situated on massive carbonate

platforms? .Could the monitoring value of CO2 become reduced by such assumptions? What is the frequency of

volcanoes to be siituated on massive carbonate platforms in the history of the plate tectonic puzzle of Earth? The

Mesozoic carbonate platforms are in a way dominant according to the now-a-days observer of the Earth‘ s surface, but

burried and/or metamorphosed carbonate and evaporite sequences could play/have played a role in contributing gases

to volcanic eruptions and as a consequence contributing gases to the atmosphere throughout Earth´s history.

„Whereas direct experimental investigations in magmatic gases, as well as extensive thermodynamic modeling

have revealed a large set of possible reaction mechanisms, very little is known about the actual plume formation

processes where the complexity of the fully heterogeneous system must be considered“ (Amann et al. 1993). More

FESEM/EDS data of particles from fumaroles, degassing lava flows and plumes are needed. Automatic sizing and

chemical analysis can reduce FESEM time. The disadvantage is a loss of information of individual particles, especially

morphological features which cannot be characterized fully by available sizing parameters. To develope SEM analysis

of volcanic ash particles and aerosol as an instrument to assess volcanic hazards and monitoring, it would be necessary

to sample in a continous, time-resolved way and set up data bases.

Phosphorus is known to exist as a compound of a variety of gaseous molecules. Remote sensing methods and

direct sampling techniques and subsequent analysis should focus on the detection of these molecules. This would help

to explain the high abundance of P in observed aerosol particles.

22

Satellite-borne aerosol data should be correlated with FESEM data on the same object. Analysis should also be

correlated according to the time of observation and sampling. Airborne sampling might be too dangerous for pilots and

planes. Small meteorological rockets could be utilized for sampling. It is necessary to develope an impactor which can

be opened and shut according to flight operation to minimize possible contamination by non-volcanic aerosol. To avoid

risks to pilots and scientists a cooperation with small rocket engineers is proposed. Rocket-borne in situ aerosol

measurements have been tested positively to study noctilucent clouds (Rapp et al. 1999). Rocket-borne data from

plumes could be an important contribution to atmospheric chemistry and volcanism and climate change.

Acknowledgements:

SEM and FE SEM studies of volcanic ash and aerosols from Popocatepetl volcano, Mexico had been sponsored by the

Naturhistorisches Museum (NHM)/Mineralogie, Vienna (part 1) and the Zentrum für Elektronenmikroskopie (ZFE),

TU Graz (part 1 and 2). Data collection was possible due to cooperation with S. Hughes, C. Siebe (UNAM), G. Kurat,

F. Brandstätter (NHM, Vienna, Austria). We are also grateful to Y. Taran (UNAM) and F. Goff (LANL), who critically

reviewed earlier versions of the manuscript.

Appendix 1:

Methodology

The field emission scanning electron microscope (FESEM) laborartory at the Zentrum für Elektronenmikroskopie

(Technische Universität Graz, Austria) – P. Golob .

FE SEM techniques had been applied partially for part 1. All aerosol studies utilized the FESEM (part 2).

Scanning electron microscopic images are generated by a thin electron beam which scans one sample line by another.

Primary electrons can go into the sample and leave the studied material as high-energetic back scattered electrons or are

emitted from near-surface layers as low-energetic secondary electrons. Detectors collect the emitted electrons and are

responsible for the brightness of a synchronous scanning beam of a screen.

The lateral resolution of a microscope is defined by the minimum distance of two objects documented as

separated identities. The capacity of resolution of an SEM is also limited by the diameter of the scanning electron beam

at the level of the sample. Conventional SEMs emit electrons from a heated cathode. The anode field accelerates the

electrons to obtain the desired (primary) energy. Inside the electron gun there is a cross-over of the electron trajectories

showing a diameter of ca. 50 µm. This cross-over is modified by electron-optical lenses to an even smaller diameter

which provides a highly focused scanning electron beam at the level of the sample.

23

Electron-optical lenses, like light-optical lenses, can show defects. These defects are responsible for the

limited availability of the beam diameter and the beam current which governs the intensity of the electron beam. If

imaging and analysis need higher beam currents the diameter of the beam is enlarged and resolution is diminished.

According to the chemical properties of the electron source the increase in beam diameter becomes more dramatic

when a low accelaertion energy is applied.. Modern instruments can reach electron-optical resolution of 5 nm utilizing

high primary energy and low beam current. High primary energy causes the primary electrons to go deeper into the

sample and resulting information is diluted by a larger sample volume from which the signals are obtained.. Therefore

resolution is diminished again, if a sample is heterogeneous at very small scale.

Field emission cathodes are characterized by a very fine tungsten tip, which emits electrons under the

influence of electrical fields. Utilizing such a device extremely small beam diameters can be produced on the surface of

a sample, therefore a high electron-optical resolution is provided by applying a low primary energy. As a result

informations for imaging and analysis can be obtained from near-surface areas. The aerosols and selected clasts of the

Popocatepetl ash had been analysed by the LEO DSM 982 Gemini microscope. This FE SEM provides lateral

resolution of ca. 1 nm at 25 keV and ca. 4 nm at 1 keV.

Applying high energy the primary electrons can interact with the atoms of the sample. The ionisation of the

electron shells of the atoms generates X-rays with characteristic energy due to the atomic source . This phenomenon

can be used to analyse the chemical composition of the source area from which X-rays are emitted. Energy-dispersive

X-ray spectrometer (EDS) are splitting the X-ray spectrum electronically. Monochromatic X-ray lines are the sources

for profiles with half-life width of ca. 150 keV, permitting a reasonable separation of lines of different elements. The

LEO DSM 982 Gemini is equipped with a Noran Voyager System, utilizing an ultra-thin window. Elements from B to

U can be analysed; resoltuion limits are in the best case at ca. 0,1 weight percents. Increasing energy of accelerating

electrons increases the intensity of the X-ray lines, leading also to a larger area of acceleration and therefore to a larger

analysed volume of the sample. This is very important for the analysis of particles. As soon as the area of acceleration

becomes larger than the particle, the substratum or the filter material is integrated in the result. After optimizing the

working conditions areas of acceleration can be obtained at a sub-µm-scale.

Correction models are needed to calculate quantitative data of the composition of a sample. X-ray intensity

and the concentration of individual elements are in a proportional relation. This relation and other parameters like the

accelerating energy are the most important factors for the correction models. These models are developed for

homogeneous samples with a flat surface. Particle analysis causes therefore a problem, simulations, like Monte-Carlo-

simulation or reference particles would be necessary. Providing these premises automated analysis is possible and

statistical data of size and composition can be obtained.

24

Except for the composition of a particle an examination of the crystal structure is needed sometimes.

Transmission electron microscopes can be used to produce diffraction patterns of selected particles. New technologies

like the electron backscattered diffraction (EBSD) permit crystal structures to be analysed in-situ in the SEM. However

analytical restrictions of this type of phase identification have to be examined by experiments. It is well known that it

can be applied to flat surfaces.

The impact of the primary electrons causes charging of the sample. To avoid this phenomenon a very thin

layer of a high-conductivity material is sputtered on top of the non- or low-conductive surface of particles and filter

material. This technique provides the charging electrons to escape and improves the imaging conditions. The teflon

filter had been coated with carbon. Carbon does not produce misleading X-ray lines and does not suppress emitting X-

rays. For imaging only Au is the best material as a coating substance as it highly improves contrast. For the aerosol

study we utilized the coating and sputtering equipment EPA 101, Leybold.

Traditional SEMs work under high vacuum. This limits the application to dry objects. New techologies like

low-vacuum systems can by-pass these problems. In these systems the sample chamber is not directly connected to the

beam-producing electron-optical equipment. The sample chamber can be flooded by a mixture of gases under different

pressures. Such a system is going to be installed at the ZFE Graz. Up to now wet samples can be analysed only in a

frozen state, utilizing cryo-preparation and cryo-transfer, which permits deep freezing, breaking and coating without

contamination (icing). Cooling of the sample and transfer to a precooled sample chamber is reducing damage to

destructable samples by primary electrons. Further Reading on electron microscopy-related literature applied to

geosciences can be found in Obenholzner et al. (1999).

REFERENCES:

Albarede, F. 1993. Residence time analysis of geochemical fluctuations in volcanic series. Geochimica et

Cosmochimica Acta, 57, 615-621.

Allard, P., Aiuppa, A., Loyer, H., Carrot, F., Gaudry, A., Pinte, G., Michel, A., Dongarra, G. 2000. Acid gas and metal

emission rates during long-lived basalt degassing st Stromboli volcano. Geophysical Research Letters, 27, 8, 1207-

1210.

Ammann, M., & Burtscher, H. 1993. Aerosol dynamics and light-scattering properties of a volcanic plume. Journal of

Geophysical Research, 98. B11, 19.705-19.711.

Andres, R.J., Kyle, P.R. & Chuan, R.L. 1993. Sulfur dioxide, particle and elemental emissions from Mount Etna, Italy

during July 1987. Geologische Rundschau, 82, 687-695.

25

Aramaki, S. 1961. Sillimanite and cordierite from volcanic xenoliths. American Mineralogist, 46, 1154-1165.

Arciniega-Ceballos, A., Valdes-Gonzales & Dawson, P. 2000. Temporal and spectral characteristics of seismicity

observed at Popocatepetl volcano, central Mexico. Journal of Volcanology and Geothermal Research, 102, 207-216.

Arthur, M.J. 2000. Volcanic contributions to the carbon and sulfur geochemnical cycles and global change. In:

Sigurdsson, H. (Ed.) Encyclopedia of Volcanoes. 1045-1056.

Baldwin, B.G. 1955. The formation and decomposition of hercynite (FeO.Al2O3). Journal of the Iron & Steel Research

Institute, 179, p.142.

Bernard, A. 1985. Les mechansimes de condensation des gaz volcaniques (chimie, mineralogie et equilibres des phases

condensees majeures et mineures). These, Universite Libre de Bruxelles,Tome 1+2,1-412.

Blackburn, G. & Taylor, R.M. 1969. Limestones and red soils of Bermuda. Geological Society America Bulletin, 80,

1595-1598.

Bluth, G.J.S., Schnetzler, C.C., Krueger, A.J. & Walter, L.S. 1993. The contribution of explosive volcanism to global

atmospheric sulphur dioxide concentrations. Nature, 366, 327-329.

Carmichael, I.S.E., Turner, F.J. & Verhoogen, J. 1974. Igneous Petrology. 715 Pp., McGraw-Hill Book Company, New

York.

Delgado-Granados, H., Cardenas Gonzales, L. & Piedad Sanchez, N. 2001. Sulfur dioxide emissions from Popocatepetl

volcano (Mexico): case study of a high-emission rate, passively degassing erupting volcano. Journal of Volcanology

and Geothermal Researchj, 108, 1-4, 107-120.

Devine, J.D. & Sigurdson, H. 1980. Garnet-fassaite calc-silicate nodule from La Soufriere, St. Vincent. American

Mineralogist, 65:302-305.

Fries, K. 1965 a. Resumen de la geologia de la hoja Cuernavaca, Estados Morelos, Mexico, Guerrero y Puebla. UNAM,

Instituto de Geologia, Mexico, D.F..

Fries. K. 1965 b. Hojo Cuernavaca 14Q-h (8). Carta geologica de Mexico, Serie de 1:100 000. UNAM, Instituto de

Geologia, Mexico, D.F..

Fulignati, P., Kamenetsky, V. S., Marianelli, P., Sbrana, A. & Mernagh, T.P. 2001. Melt inclusion record of

immiscibility between silicate, hydrosaline, and carbonate melts: Application to skarn genesis at Mount Vesuvius.

Geology, 29, 11, 1043-1046.

Galindo, I., Ivlev, L.S., Gonzalez, A. & Ayala, R. 1998. Airborne measurements of particle and gas emissions from the

December 1994-January 1995 eruption of Popocatepetl volcano, Mexico. Journal of Volcanology and Geothermal

Research, 83:197-217.

26

Gemmel, J.B. 1987. Geochemistry of metallic trace elements in fumarolic condensates from Nicaraguan and Costa

Rican volcanoes. Journal of Volcanology and Geothermal Research, 33, 161-181.

Getahun, A, Reed, M.H., Symonds, R. 1996. Mount St. Augustin volcano fumarole wall rock alteration: mineralogy,

zoning, composition and numerical models of its formation process. Journal of Volcanology and Geothermal Research.,

71: 73-107.

Gilg, H.A., Lima, A., Somma, R., Belkin, H.E., De Vivo, B. & Ayuso, R.A. 2001. Isotope geochemistry and fluid

inclusion study of skarns from Vesuvius. Mineralogy and Petrology, 73, 145-176.

Goff, F., Janik, C.J., Delgado, H., Werner, C., Counce, D., Stimac, J.A., Siebe, C., Love, S.P., Williams, S. N., Fischer,

T. & Johnson, L. 1998. Geochemical surveillance of magmatic volatiles at Popocatepetl volcano, Mexico. Geological

Society America Bulletin, 110, 6, 695-710.

Goff, F., Love, S.P., Warren, R., Counce, D., Obenholzner, J., Siebe, C. & Schmidt, S. 2001. Passive infrared remote

sensing evidence for large, intermittent CO2 emissions at Popocatepetl volcano, Mexico. Chemical Geology, 177, 1-2,

133-156.

Hamill, P., Houben, H., Young, R., Turco, R. & Zhao, J. 1996. Microphysical processes affecting the Pinatubo plume.

In: Fiocco, G., Fua, D. & Visconti, G. (eds.): The Mount Pinatubo Eruption. NATO ASI Series I; Global

Environmental Change, Vol. 42, 49-59.

Handler, P. 1989. The effect of volcanic aerosols on global climate. Journal of Volcanology and Geothermal Research,

37, 233-249.

Harker, R.I. & Tuttle, O.F. 1956. Experimental data on the PCO2-T curve for the reaction: Calcite + Quartz

Wollastonite + Carbon Dioxide. American Journal of Science., 254, 239-256.

Herrmann, A.G. 1980. Geochemische Prozesse in marinen Salzablagerungen: Bedeutung und Konsequenzen fuer die

Endlagerung radioaktiver Substanzen in Salzdiapiren. Zeitschrift der deutschen geologischen Gesellschaft, 131, 433-

459.

Heus-Aßbichler, S. & Masch, L. 1991. Microtextures and reaction mechanisms of carbonate rocks: A comparison

between the thermoaureoles of Ballachulish and Monzoni (N. Italy). In: Voll, G., Töpel, J., Pattision, D.R.M & Seifert,

F.: Equilibrium and Kinetics in Contact Metamorphism. 229-250.

Hinds, W.C. 1982. Aerosol Techology. Pp. 424. John Wiley & Sons, NY.

Hidy, G.M. 1984. Aerosols – An Industrial and Environmental Science. Pp. 774. Academic Press, NY.

Hobbs, P.V., Radke, L., Eltgroth, M. & Hegg, D. 1981. Science, 211: 816.

Houtermans, F.G., Eberhardt, A. & Ferrara, G. 1964. Lead of volcanic origin. In: Craig, H., Miller, S.L. & Wasserburg,

G.J. (eds.): Isotopic and Cosmic Chemistry. 233-243.

27

Hughes, J.M., Christian, B.S., Finger, L.W., & Maliconico, L.L. 1987. McBirneyite, Cu 3(VO4)2, a new sublimate

mineral from the fumaroles of Izalco Volcano, El Salvador. Journal of Volcanology and Geothermal Research, 33:183-

190.

Imbo, G. 1965. Catalogue of the active volcanoes of the world including solfatra fields. Part XVIII, Italy.6-19.

Ishikawa, T. 1953. Xenoliths included in the lavas from volcano Tarumai, Hokkaido, Japan. Jour. Fac. Sci., Hokkaido

University, Ser. 4, 8, 225-244.

Isshiki, N. 1954. On iron-wollastonite from Kanpu volcano, Japan. Proc. Japan. Acad., 30:869-872.

Ivlev, L.S., Galindo, I. & Kudryashov, V.I. 1995. Estudio de aerosoles y cenizas. In: Volcan Popocatepetl-Estudios

realizados durante la crisis de 1994-1995, CENAPRED.

Jain, S., Skamser, D.J., & Kodas, T.T. 1997. Morphology of single-component particles produced by spray pyrolysis.

Aerosol Science and Technology, 27, 575-590.

Janik, C.J., Goff, F., Fahlquist, L., Adams, A.I., Roldan-M., A., Chipera, S.J., Trujillo, P.E. & Counce, D. 1992.

Hydrogeochemical exploration of geothermal prospects in the Tecuamburro volcano region, Guatemala. Geothermics,

21, 4, 447-481.

Kalinin, D.V. 1967. Lower temperature boundaries for the formation of tremolite, diopside and wollastonite under

hydrothermal conditions; experimental data. Geochem. Intern., 4, 836-839.

Knipping, B. 1989. Basalt Intrusions in Evaporites. Lecture Notes in Earth Sciences, 24. Pp.131.

Kridelbaugh, S.J. 1973. The kinetics of the reaction: calcite + quartz wollastonite + CO2 at elevated temperatures

and pressures. American Journal of Science, 273, 757-777.

Lauterjung, J. & Ossing, 2000. Zweijahresbericht – GeoForschungsZentrum Potsdam - 1998/1999. P.271

Le Guern, F. & Bernard, A. 1982 A. A new method for sampling and analyzing volcanic sublimates-application to

Merapi Volcano, Java. Journal of Volcanology and Geothermal Research, 12:133-146.

Le Guern, F. & Bernard, A. 1982 B. Etude des mecanismes de condensation des gaz magmatiques - Example de l‘Etna.

Bulletin of Volcanology, 45, 3:161-166.

Love, S.P., Goff, F., Counce, D., Siebe, C. & Delgado, H. 1998. Passive infrared spectroscopy of the eruption plume at

Popocatepetl volcano, Mexico. Nature, 396, 563-567.

Lowenstern, J.B., Mahood, G.A., Rivers, M.L. & Sutton, S.R. 1991. Evidence for extreme partitioning of copper into a

magmatic vapor phase. Science, 1405, 7 June 1991.

Malinconico, L.L. 1987. On the variation of SO2 emission from volcanoes. Journal of Volcanology and Geothermal

Research, 33, 231-237.

28

Mash, L. & Heuss-Aßbichler, S. 1991. Decarbonation reactions in siliceous dolomite and impure limestone. In: Voll,

G., Töpel, J., Pattision, D.R.M & Seifert, F.: Equilibrium and Kinetics in Contact Metamorphism. 211-228.

Mroz, E.J. & Zoller, W.H. 1975. Composition of atmospheric particulate matter from the eruption of Heimaey, Iceland.

Science, 190, 461-463.

Meeker, K.A., Chuan, R.L., Kyle, P.R. & Palais, J.M. 1991. Emission of elemental gold particles from Mount Erebus,

Ross Island, Antarctica. Geophysical Research Letters, 18/8:1405-1408.

Mills, M.J. 2000. Volcanic Aerosol and Global Atmospheric Effects. In: Sigurdsson, H. (Ed.) Encyclopedia of

Volcanoes. 931-943.

Newell, R.E. & Deepak, A. (eds.), 1982. Mount St. Helens eruption of 1980 – Atmospheric effects and potential

climatic impact – NASA SP-458.

Nriagu, J.O. 1989. A global assessment of natural sources of atmospheric trace metals. Nature, 338, 47-49.

Obenholzner, J.H. & Heiken, G. 1999. Textural studies of altered/metamorphosed tuffs and secondary REE-, Zr- and

Y-minerals: scanning electron microscopic examinations. Annalen des Naturhistorischen Museums in Wien. 100A, 13-

38.

Oberbeck, V.R., Farlow, N.H., Fong, W., Snetsinger, K.G., Ferry, G.V. & Haynes, D.M. 1982. Mount St. Helens

aerosol evolution. Geophysical Research Letters, 9/9:1089-1092.

Peter, T. 1996. Airborne particle analysis for climate studies. Science, 273, 1352-1353.

Pethica, J.B. & Egdell, R. 2001. The insulator uncovered. Nature, 414, 27-28.

Phlean, J.M., Finnegan, D.L., Ballantine, D.S., Zoller, W.H., Hart, M.A. & Moyers, J.L. 1982. Airborne aerosol

measurements in the quiescent plume of Mount S. Helens: September, 1980. Geophysical Research Letters, 9/9:1093-

1096.

Pyle, D.M. 1992. The volume and residence time of magma beneath active volcanoes determined by decay-series

disequilibria methods. Earth and Planetary Science Letters, 112, 61-73.

Rapp, M. & Lübken, F.-J. 1999. Modelling of positively charged aerosols in the polar summer mesopause region. Earth

Planets Space, 51, 799-807.

Riehle, J.R., Rose, W.I., Schneider, D.J., Casadevall, T.J. & Langford, J.S. 1994. Unmanned aerial sampling of a

volcanic ash cloud. Earth in Space, Vol. 6, 8, 10-11.

Risacher, F. & Alonso, H. 2001. Geochemistry of ash leachates from the 1993 Lascar eruption, northern Chile.

Implications for recycling of ancient evaporites. Journal of Volcanology and Geothermal Research, 109, 319-337.3xsw

Robock, A. 2000. Volcanic eruptions and climate. Reviews of Geophysics, 38, 2, 191-219.

29

Romer, R. L. & Heinrich, W. 1998. Transport of Pb and Sr in leaky aquifers of the Bufa del Diente contact

metamorphic aureole, north-east Mexico. Contribution Mineralogy Petrology, 131, 155-170.

Rose, W., I., Jr., Chuan, R.L., Cadle, R.D. & Woods, D.C. 1980. Small particles in volcanic eruption clouds. American

Journal of Science., 280:671-696.

Schoklitsch, K. 1935. Beitrag zur Physiographie steirischer Karbonspaete (Gitterkonstanten, physikalische Angaben

und chemische Zusammensetzung). Zeitschrift fuer Kristallographie, vol.90, no.4, pp.433-445.

Seto, T., Hirota, A., Fujimoto, T., Shimada, M. & Okuyama, K. 1997. Sintering of polydisperse nanometer-sized

agglomerates. Aerosol Science and Technology, 27:422-438.

Siebe, C., Abrams, M., Macias, J.L. & Obenholzner, J.H. 1996. Repeated volcanic disasters in Prehispanic time at

Popocatepetl, central Mexico: Past key to the future? Geology, 24, 399-402.

Sheridan, P.J., Schnell, R.C., Hofmann, D. J., Deshler, T., 1992. Electron microscope studies of Mt. Pinatubo aerosol

layers over Laramie, Wyoming during summer 1991.Geophysical Research Letters, 19, 2, 203-206.

Sieh, Y.N. & Taylor, H.P. 1969. Oxygen and carbon isotope studies of contact metamorphism of carbonate rocks.

Journal of Petrology, 10, 2, 307-331.

Stimac, J., Hickmott, D., Abell, R., Larocque, A. C. L., Broxton, D., Gardner, J., Chipera, S., Wolff, J. & Gauerke, E.

1996. Redistribution of Pb and other volatile trace metals during eruption, devitrification, and vapor-phase

crystallization of the Bandelier Tuff, New Mexico. Journal of Volcanology and Geothermal Research, 73, 245-266.

Stoiber,R.E. & Rose, W.I. 1970. The geochemistry of Central American volcanic gas condensates. Geological Society

America Bulletin, 81:2891-2912.

Stoiber, R.E. & Rose, W.I. 1973. Sublimates at volcanic fumaroles of Cerro Negro volcano, Nicaragua. Publ. Geol.del

ICATI,4, 63-68.

Stoiber, R.E., Williams, S.N. & Huebert, B. 1987. Annual contribution of sulphur dioxide to the atmosphere by

volcanoes. Journal of Volcanology and Geothermal Research, 33, 1-8.

Solomon, S., Sanders, R.W., Garcia, R.R. & Keys, J.G. 1993. Increased chlorine dioxide over Antarctica caused by

volcanic aerosols from Mount Pinatubo. Nature, 363, 245.

Symonds, R.B., Rose, W.I., Reed, M.H., Lichte, F.E.& Finnegan, D.L. 1987. Volatilization, transport and sublimation

of metallic and non-metallic elements in high temperature gases at Merapi Volcano, Indonesia. Geochimica et

Cosmochimica Acta, 51:2083-2101.

Tanner, S.B., Kerrick, D.M. & Lasaga, A.C. 1985. Experimental kinetic study of the reaction: Calcite + Quartz

Wollastonite + Carbon Dioxide, from 1 to 3 kilobars and 500° to 850°C. American Journal of Science, 285, 577-620.

30

Taran, Y.A., Hedenquist, J.W., Korzhinsky, M.A., Tkachenko, S.I. & Shmulovich, K.I. 1995. Geochemistry of

magmatic gases from Kyudryavy volcano, Iturup, Kuril Islands. Geochimica et Cosmochimica Acta, 59, 9, 1749-1761.

Taran, Y.A., Bernard, A., Gavilanes, J.-C., Lunezheva, E., Cortes, A. & Armienta, M.A. 2001. Chemistry and

Mineralogy of high-temperature gas discharges from Colima volcano, Mexico. Implications for magmatic gas-

atmosphere interaction. Journal of Volcanology and Geothermal Research, 108, 245-264.

Taran, Y.A., Bernard, A., Gavilanes, J.-C., & Africano, F. 2000. Native gold in mineral precipitates from high-

temperature volcanic gases of Colima volcano, Mexico. Applied Geochemistry, 15, 337-346.

Taylor, P.S. & Stoiber, R.E. 1973. Soluble Material on ash from active Central American volcanoes. Geological

Society America Bulletin, 84, 1031-1042.

Tilley, C.E. 1948. On iron-wollastonite in contact skarns: an example from Skye. American Mineralogist, 33:736-738.

Turner, F.J. & Verhoogen, J., 1960. Igneous and Metamorphic Petrology. 2nd ed. McGraw-Hill Book Co., New York-

London.

Valdes C., Gonzales, G., Arciniega, A, Guzman, M., Nava, E, Gutierrez, C. & Santoyo, M. 1995. Sismicidad del

volcan Popocatepetl a partir del 21 de diciembre de 1994 al Marzo de 1995. In: Volcan Popocatepetl-Estudios

realizados durante la crisis de 1994-1995, CENAPRED.

Valero, F.P.J. & Pilewskie, P. 1992. Latitudinal survey of spectral optical depth of the Pinatubo volcanic cloud-derived

particle sizes, columnar mass loading, and effects on planetary albedo. Geophysical Research Letters, 19/2:163-166.

Vogel, W. 1994. Glass Chemistry. Pp. 464, Springer-Verlag.

Wedepohl, K.H. 1995. The composition of the continental crust. Geochimica et Cosmochimica Acta, 59, 7, 1217-1232.

Winker, D.M. & Osborn, M.T. 1992. Airborne LIDAR observations of the Pinatubo volcanic plume. Geophysical

Research Letters, 19/2:167-170.

Willeke, K. & Baron, P.A. (eds.), 1993. Aerosol measurement: Principles, techniques and Applications. 1-876. Van

Nostrand Reinhold, NY.

Whitby, E. R., & McMurry, P.H. 1997. Modal Aerosol dynamics modeling. Aerosol Science and Technology, 27:673-

688.

Wohletz, K.H. & McQueen, R.G. 1984. Volcanic and stratospheric dustlike particles produced by experimental water-

melt interactions. Geology, 12, 591-594.

Wood, C.P. 1994. Mineralogy at the magma-hydrothermal system interface in andesite volcanoes, New Zealand.

Geology, 22:75-78.

Wood, C.P. & Browne, P.R.L. 1996. Chlorine-rich pyrometamorphic magma at White Island volcano, New Zealand.

Journal of Volcanology and Geothermal Research, 72:21-35.

31

Zambonini, F. 1935. Mineralogia vesuviana ed appendice alla mineralogia vesuviana. Atti. Acc. Sc. Fis. e Mat. di

Napoli, XIV, 2a edizione, a cura di Quercigh E., Napoli, S.I.E.M..

Zoller, W.H., Parrington, J.R. & Kotra, J. M. P. 1983. Iridium enrichment in airborne particles from Kilauea volcano:

January 1983. Science, 222, 1118-1119,

32