RI:NDlCONTl Socktd Iralla... ill Minno/ovlo It Pd~, 31 (U: pp. U7-1tl

Comunlcadone pracnlata al Con~ 8DlP dl Cacn&rt 11 14 ottobre 11141

CHEMICAL VARIATIONS IN PYROCLASTIC SERIES

NARA CoIlADOSSI, MARINO MARTINI

htilll!o di Minertlogi., Petrografi. e Gcochimica dell'Universidl, vi. La Pia 4, .50121 Fiff'lW'.Ccntro di studio per J. Mineraiogi. e I. Gcochimia dei Scdimc:nd (CN.R.)

ABsnACT. - The S1udy of pyroclaslic depo5.itshu bcc:n mainly bucd upon their lithoqial, gra­nulomctric and staligaphic charactcn.

It seemed possible that chemial peculiaritieshave been produced in response to the physial­chemical conditions which gave rise to Ihe diffeff'nteruptive processes, and a slUdy of pyrodastic serieswas undert.ken to check whether the presenldistribution of specific elements can provide someaddition.1 information.

Samples representing the different types of pyro­clastic deposits of Minoan eruption of Santorini andPompcian eruption of Vcwvius have been inve­stigated, as well as Iavas from the same .re-s;5Odium, pot.&SSium, iron, magnesium, fluorine .ndchlorine have been determined.

The distribution of Na, K, F and Cl, along withthe values for I«chablc Na ,nd D, .ppears of ,ettlllin interest.

Since the soJubility of volatiles in magma sc=IS

to depend mainly on their cooccntalioos in alkalies,high lcvels of F and a in pyrod..utk products canbe justified by the K .nd N. values observed atVesuvius; for Santorini this can apply to pumiccsonly, while a different explanation is needed forsamples of surge matrix, because their low contentsin alkali metals cannOt account for the observednol rqligible levels in a.

Keeping also in mind the locuion (If the volcanoan inflow (If sea water into the system appcanthe moSl likely among the natural mechanisms whichan produce .n enrichmenl of D in the magmaticproducu, in spite of the lack of Jufficicnt conditionsfor ill solubility in the melt phase.

The explosive processes which produced the basesurge deposits at Sanlorini seem then 10 be ascribedto the above mentioned inflow, while for Vesuviusthe chemical evidences point OUt to • similar roleof phrcalic w.tm.

RussuHTO. - 1.0 studio dei <kpositi piroclaslicisi ~ soprallutto buato sui 1010 aattcri Iitoqid,granulomctrici e staligafici.

.£ scmbato possibiJe cbe si sillllO prodolle pecu.liarilil chimiche in risposta .lIe condizioni chimico­fisiche che hanno dato luoge .i ditferenli processierutcivi, cd e SIaIO inlraprcso uno studio di seriepiroclastiche per verificare se I'.ttuale distribuzionedi dementi specifici pub fomire infonnuioni Jupp1c­mentari.

Sono stati considcrati depositi piroclastici den.eruzione Minoica di Santorino e di quclla Pompciana

del Vesuvio, e bve delle stessc aree, determinandosodio, potassic, ferro, maPcsio, dolO e ftuoro; mq­giore inleresse e appresentato daI.I.a distribuzioncdi N., K, F c a nelle fOC't'e totali, e di N. e anei Iisciviati delle stessc fOC't'e. Poid,e la solubililidei voladti ncl magma scmbra dipen<lere $OpranullOd.1 suo oontenulO in atali, aid livclli di FeCIpossono csscre giustificati dai valoci di Na e Kosstrvali al Vesuvio; per Santorino db pub valeresolo per le pornid, cd una dh'C:fSll spicgazionc ~

necessaria per i eampioni di • surPC'. cbe hannobasso contcnuto in .Icali e v.loci signifialivi di a.

Tcncndo conlO .nd1C dell'ubicazione del vubno,un .pporlO di acqu. marina al sistcma appaff' ilpiu plausibile fra i proccssi narurali che possonoprcxlurre un arricchimenco in Cl dei prodotri mag­rn.tid, nonostante non esistano le condizioni peruna sua maggioff' solubilitll ncl fuso.

I proccssi esplosivi che h.nno prodono qucslOtipo di deposilo • S.nlorino scmbrano quindi d&.nribuirc a11'apporto suddellO, mcntff' per it Ve­suvio le CINIneristichc chimichc indicano piullostoun ruolo simile di lICque fre-tiche.

The explosive volcanic eruptions and theirproducts have been largely investigatedduring the last ten years, and considerableresults have been obtained which besidestheir scientific meaning have a s~aJ im­portance in assessing volcanic hazards inpopulated areas (LIRER et aL, 1973; WALKER,1974,1977; BARBERI and GASPARINI, 1976;BoND and SPARKS, 1976; BARBERI et aI.,1977; :BooTH et al., 1977; BAKER, 1979;BooTH, 1979; DELlBRIAS et aI., 1979;SHEItIDAN, 1979; SHERIDAN et al., 1981).

The guickline of these researches hasconsisted in a careful reconstruction of pasteruptive events by means of the study ofpyrodastic formations, carried out followingmainly Iithological, granulometric and strati­graphic criteria. The chemical compositionof the products involved in explosive activityis normally not taken into consideration inthe frame of the above mentioned inve5ti-

288 N. CORADOSSI, M. MARTINI

Fig. I. Locations of the studied samples fromSantorini. Oplontl

SMHOR'N' N

'",

Q

o,

Fig. 2. _ Locations of the studied samples fromVesuvius area.

25 ml of water, without stirring.The analytical results are shown in table 2.

Di8CU!l8ion

The analytical data for the series hereconsidered allow to verify that, besides thebasic differences in chemical characters, atVesuvius significant compositional variationscan be found along the different deposits,while at Santorini a substantial similarity isobserved with minor exceptions.

At Vesuvius the chemical composition ofcineritic and lithic components of flowdeposits gradually approaches that of basallavas of Somma; this fact has been consideredas a cerrain degree of evidence in favour ofa phreatic event which involved lavas ofolder age with respect to those produced inthe first stages of Pompeian eruption (MAR­TINI, 1982).

Following the same logical path, thepersistence of the major features in the

gations, probably because in many instancessecondary processes of alteration obliteratedto a great extent the original characters.

On the contrary, if the emplacement settingallowed but a minor degree of secondarychanges, it seems to us that some additionalinformation could arise from the distributionof specific elements.

This condition can be fulfilled for pyro­clastic formation of recent age, characterizedby high depositional rates, considerablethickness and an early formed or con­temporary impermeable cover.

To carry on a pre1iminar essay of thiskind of investigations, samples from pyro­clastic series of Pompeian eruption of Ve·suvius (79 A.D.) and Minoan eruption ofSantorini (1470 b.C.) are here taken intoconsideration, with reference to their contentsin some major (Na, K, Fe, Mg) and volatileconstituents (F and Cl).

Experimental

In figg. I and 2 are given the locations ofthe studied samples, described in table 1.

The analyses of Na, K, Fe, Mg have beencarried out by classical procedures usingabsorption spectrophotometry; specific elec­trodes have been employed in the determi­nation of F and Cl.

The leachate solutions have been obtainedby cold extraction on 1 g of sample with

"I. ',..""","" ,..0123Krn

*Pornpel

*

CHEMICAL VARIATIONS IN PYROCLASTIC SERIES 289

TABLE 1

Description 01 the studied samplesTABLE 2

Analytical data 01 the studied samples

~ ..TO "~" ..TO "n ..~ ..~ ..~ "~ ..~ ..n"n"n"~"n"n"n"."."~"~ ..~ ..~"

~"

.,..........................0'.0'.0"

0"0'.0'.0'.~uo.~.oo

~,"'"'oo~UO,~.".

~uo,~."'"

~""o'oo

M"'o'ooM"'o' ...~""o,oo

_10. r."....,"" to".........."...«b of ...... ,D,.."

...."" toll

...<... of ~....

..."._ p' 0&. ",.roe...u~ Hoo

....""• ..,.,0< <1" "oo

....00. "'0<''''''0 fl ..

........ f.ll

....... of ,..,.

...."" 0' ._ .... 1'1"0<'''''0 fl ..

Loo. 010<'. 1'1"0<"'''0 n""....... ,.,.'0< 0 "ow

.......00. Pf'oc ~ flow

...."" foil

"""'" f.'l_"" ..... 'u,••""<l. or b••••u•••......'0........u,••""'1<'. ",.""" ••uo 0"".....'00, .lAva b.<><'. ""'<><'''''0 """

7" ,.

7" "7" 02

7" '37" , •

70 ..

~" ..~H .,

~" u~" u~" ,.n "7" "~ "rn H

~" 15

~ "~" 11

n "T" n

n "~" "~ "7Ji .l

...~'.n...., ....,"."."..,...

•. lS.. ", "."...""'.10..".."

' ••0

'. ,.,."

',0

°""...."'.J3,."

, .11,..,'.13, ..'"",".

00

"l. "1•• ,

I .• ~

I.' ,I. tl

'.121. ••

,. "".n

., '. "S1 ••~,

'.H 0.11

I U 0."l.ll D.n

o' ,.,...,. ,..,I." 0."l.n ••,... ..I.U IJ, .. .,'.10 "... .".. .." .."'I •• 03

'0 0."'" 0."" ".' I..t ~.

" Il

" 11

'll~

00'

m0<.

'"..0

."'"".mm

'",....0m

".mmm..,..0,..",m

_,I,•,Ul,...," 0'HO02.0..,~U~.

>SOSlIOO

".00'••"..,,,.lSlO

"0.l~'.,n.,...'00,,,,,,,.n '0,no

"".',",

m

'"...m

'".","'"m00.

HO

'"",'"...m

'""0,.."0

'"m'1\0

,s'."'0'",..'"

'0'0,,,.0"'"...,.,~

",....'",..0

,.o~

'0'm

00

'"'">0'm

"

chemical picture should allow to infer forSantorini a true phreatomagmatic episode,producing poorly differentiated materialsfrom the same magma chamber, If we extendthe observation to the studied volatileconstituents, however, a good uniformity IS

still present for fluorine distribution, buttwo strong peaks are shown by chlorineconcentrations (fig. 3).

The 'llbundance of volatile constituents In

magmas are normally limited by theirsolubility, which in turn seems to depend. onthe alkali metals content (KOGARKO et al.,1968; KOGARKO, 1974).

The substantial uniformity of sodium and

Total iron is expressed as FeO'; Na. and Cl.represent the results of cold extraction.

potassium content tn the whole series ofpyrodastic products at Santorini appears toinfer that the variations observed for chlorinecannot depend on different conditions forits solubility in the magma.

The results obtained by simple leachingshow that significant quantities of extractablechlorine correspond to the « highs» in totalconcentration of this element. Sodium IS

another important constituent of the same

,,,",'"",'"'""'"'","..,"""m,...."..'"...

,..",0"m...'"..,'"'"'..'"'"'"'"'00..,""'"'"'"

". 0'.- -..".","""""",.s•tt"..,.."..,.,..,." ..,,.,It"lOO<

"."..'''013'•"10'...

'1,••

I".lOll,,,.>J1O

ll"

" ..,,,,,,,."0.,..~..,.IU.III~

"",..."",,",."""""

"- -....,. ",...l. II

,.",."'."."...,."'."..".."""....

..>t........,...

,...•. o...".."....•. ll.."., "...,....,. I.., ...,.,S.•,

'.'.....'. "' ...., .... "

','~." ....,." I.'..... 1.10

'." '.1•.... ..".." ,."'.11 ,'"'.0' , ..,.•, '.1.,... ' ••0.." ...,,." ,...,.., ..,... "1. ,. "

" "" 1."" '.11.. S ..

I.O! ..,." .

P',v "0 •pp,v "".••IV ."

.... lO'.."" ,,,pr,v ,.."""'0 • "0",,.v ,U•n. '"••"" ,oopr,. >t." 0

"" 0

" 0

""""""

......0. ,.11............Pwol.. , ",,'0<'''''0 flow

Pwol... """""''''0 fl""......"". "",oe,••uo n""......... ",."",,,.u. n""..... ,...... •• 'PS""'U..... 1...... •• PT" ,.,

_ ,...... •• .... >t,_ 1...... •• • p.. , ..

fy""', o fl""

o _ ".'0 1... ""0 0.•" ..l.. 0<"0 01.. "'_0 •. , ........11_,

"""-,,,,, f.U

"-,,,,, f ...

""'"''''' f."........ f ...

"""''''' '.11

""'ooU""'o"H""'0'0<'""'on"...,1<.,,"()p'''''''-,""" ..""" ...""" ..""" ..."u ..."" ..""" ..."" ..""" ..

"""",,

pp." ,,, 0

P"" ,,,

...... • No"

..." no ~

..." '12..." '"..." ,..pp." 'npp." ,.,,ps" no..." ,.,

290 N. CORADOSSI, M. MARTINI

Na.%o CI.%o Nal°'*' K/O'lb Cl.., F%.2 , . , • 5 0 I , , • 0 , , 0 • • la 0 , , 3

SAII9LES , , , , , , , , , ,~

'" 58

?"5865.."62

\" 160

'" 15

}"1312

"10so l

'" 83 t t""

180

""12

"Fig. 3. - Ch~mical variationin series 'rom Santorini. . • surge matrix; • pumice; .. lava.

••3 -2892 - 288, .15l

0-286

,. HO! ~

'91'"'"'"2901290 I

Cl.,*", Na,o% K1O% F.., Cl ..0 0' o , 1 3 4 5 0 ] 4 5 6 , , o , , ,.~

, , I I I ~

rrr \( (~j r "!\ j

~ t\ t )l ~

Fig. 4. - Chemical variations in the series from Vesuvius a~. - • surge matrix or pyrodastic flow(size less than 0.063 mm); 0 pyroclastic flow (size between 050 and 0.71 mm); • pumice.

CHEMICAL VARIATIONS IN PYROCLASTIC SERIES 291

solutions, with a distribution similar to thatof chlorine, but its absolute quantities re­p~nt only a small fraction of the totalsodium content of the samples.

For Vesuvius the situarion is quite dilIerent,because in spite of the significant concen­trations of chlorine in the rocks, very minorconcentrations are observed in leachates(fig. 4).

The extent of leachable chlorine d~s notappear as depending on the pyroclastic natureof the rocks nor on the total content, andmust be accounted for by a different processamong the natural ones which substantiallycan be restricted to:

1) leaching of rocks by percolating waters,with accumulation in levels of slightpenneability;

2) alteration by secondary fumaroles, withenrichment in volatiles;

moreover, taking into account the sodiumchloride character of the solutions obtainedby leaching, we can consider

3) pollution by seawater spray;

4) contribution of seawater to the eruptivemechanism which determined the empla­cement of the studied pyroclastics.

The freshness of the samples, collected onthe: front of quarries under exploitation,~ms to rule our any alteration by fumarolicactivity, while modifications by percolatingwaters cannot be completely excluded.

The observed concentrations in leachates,however, result from a very short and weaktreatmenr of samples with cold water, andwe think that during the time elapsed sincetheir deposition even a reduced ratc ofpercolation through the studied formationswould have ptoduced an almost completcdepletion in their leachable components; anyextent of heavier alteration should haveinrcoduced also modifications in the unifor­mity of the chemical picture, expeciallybetween levels of different granulometry.

Besides this, no significant concentrationsin leachates are observed for samples col­lected at Athinios, which also should havebeen interested by a general process ofsecondary chemical changes.

If we consider the possibility of seawateras providing the excess of chlorine, we thinkthat any pollution by marine spray shouldhave interested to a similar extent all of theexposed formations; moreover, our investi­gation on deposits of the same nature locatedon a cliff near the seaside in the island ofVulcano has demonstrated that a very minorquantity of sodium chloride, if any, can beprovided through a process of that kind.

The contribution of seawater to eruptivemecl1anism has therefore, in our opinion, the:best probability among the processes hereconsidered, and the location of the volcanicsystem of Santorini ~ms to allow such anhypothesis (BoND and SPARKS, 1976).

Significant concentrations of sodium andchlorine in well defined levels in the studiedseries could be thus considered as witnessesof an inflow of seawater into the volcanicconduit, and should pertain to those depositsmore directly associated to phreatomagmaticactivity triggered by such contribution.

As a matter of fact, the highest peaks inchlorine and sodium occur in base surgedeposits, which are thought to be producedby the above mentioned type of eruption.

Nothing similar is observed at Vesuvius,for which an inflow of phreatic waters hasbeen proposed (SHERIDAN et aI., 1981).

At present, no further general explanationcan be provided, keeping in mind the preli­minary character of this research, but wethink that the obtained results stress outthat a deepening of the studies is needed inorder to verify possibilities and limits of achemical approach in the investigation onrecent pyroclastic formations.

AcJmowledgemrtft. - We are gnlleful [0 P. MA­I\"ETTI and R. SAA'TAC.OC£ who provided the samplesof Samorini and Vesuvius, t'CIpectively.

292 N. CORAOOSSI, M. MARTINI

BIBLIOGRAPHY

BAKEI PE. (1979) - Gtolof,iclII IHpt~ts 0/ uol~tlno

prtdictian. )1. Gcol. Soc. Looo., 136. }41-J4'.BAlIEtI F. and GASPAIINI P. (1976) • Vol~tlnk

hl1U1rds. Bull. Int. A". En,!:. GeQ!., 14, 217-232.BA18EIU F., MUIN£ll.l G., ROSI M. and SANTA~

CIOCE R. (1977) - If rischio da Itlhar ntl Sommtl­Vtruvio, in .. I vulcani anivi del]'a~a napolC:la'na,., Napoli.

BoND A. and SPARKS R.S.). (1976) - Tbt Mino/Intruption 0/ Santorini, Grttct. )\. Gee!. Soc.Lond., 1J2, 1-16.

BooTH B. (l9791 . Asstssing 1I01ctlnic risk. JI. Gee\.Soc. Lond.• 136, JJI·340.

BooTH B., CIOASDALE R. and WALKER <".P.L. (1977)- A qU4ntililti~ stud)' 0/ ft~ tbousilnds )'ttlrs 0/IX)/ctlnism on San Migutl, korts. Phi!. Trans.R. Soc., A 288, 271-319.

DELlBlIAS G .• DI PAOU. M., Rosl M. and SAN­TACIOCE R. (1979) - LJ storill trllUiV/l dtl com­pltUO vJdCllnico Sommil Vtsuvio rkOSlfuitll dll1ltsucussioni pirodilstkht' dtl Manu Sammil. Rend.Soc. h. Min. Pets., 3', 411-438.

KOGAIKO LN., KRICMAN LD. and SHAlUDllO NoS.(1%8) . E"ptrimtnlal invtstigatiOnJ 0/ tht tltCl0/ alkalinit)' 0/ silicaft' mtllS on Iht Up"ration0/ jluorint into IM gas phlHt. Geekhimiya, n. 8,948·9'6.

KOCARKO L.N. (1974) - Rofe o/volatiles, in • The

alkaline roch,., «I.. SOren$en, Wilcy aoo Sons.New York.

tau L., PESCATOIE T., BooTH B. and WALKUG.P.L. (1973) • Two pliniiln pumiu./aJl dtpositsfrom Somma·Vtsuvius, Ilal,. Gcol. Soc. Am. Bull.,84,7'9·772-

MARTINI M. (1982) . Tht Pompdan tvtnt 0/ Vt­suvius: chtmical tvidtnct 0/ truptivt' prousus,in .. Vesuvius,. (in press).

SHf.RIDAN M,F. (1979)· Empfoctmtnl 0/ p)'foclasticflows; 0 rtvitw. in '" Ash·llow tuffs,., w. Chapinand ElslOn, Gool. Soc, Am., Spe1:. Paper 180,12'-136.

SHEIIDAN M.F., BARBEll F., ROSI M. and SANTA­CROCE R. (1981) - A motitl /0' PJinian truplion0/ Vtsuvius. Nature, 289, 282-28'.

VITALlANQ C.J., FOIJT ).5. and VITALlASO D.B.(1978) . PttrfXMmicill sllld., 0/ tht' ltphrll St·qutnu t"postd in tM Phi,iI qUII"J, TWa. 1benand the A~n World, London.

W"'LKEI G.PL. (I974) - VoIcilnic haUlrJs {lnd Iht'prtdicl;on 0/ voIcilnic trup/ions. Gcol. Soc. Lond.,Mise. Publ. n, 3.

WAUEI G.P.L. (1977) - Mttodi gtologid ptr IIIvollltilz.iont dtl ,ischio vu{cllnico, in .. I vulcanianivi dell'area napole!ana,., Napoli.

WalCItT ).v., SMITH A.L. and SELf S. (1980) •A Wf),king It,minology 0/ p)'rrxlastic deposits.Jour. Vole. Geolh. Res., 8, 3U-3J6.