SALT MINERALS AND WATERS FROM SOILS
IN KONYA AND KENYA
CENTRALE LANDBOUWCATALOGUS
0000 0086 7032
£«A-/3<U?c/,
o%
Promotor:dr.L.vander Plas,persoonlijk hoogleraar
Mn G201
âtjr
Lideke Vergouwen
SALT MINERALS AND WATERS FROM SOILS
IN KONYA AND KENYA
Proefschrift
terverkrijgingvandegraadvan
doctorindelandbouwwetenschappen,
opgezagvanderectormagnificus,
dr.H.C.vanderPlas,
hoogleraar indeorganische scheikunde,
inhetopenbaarteverdedigen
opvrijdag 22mei 1981
desnamiddagstevieruur
indeaula
vandeLandbouwhogeschoolteWageningen
&rV\2,tO^
^
BIBLIOTHEEK LH.
0
6 MEI J98I
°NTV„TÏJDSCHR.
»
uo?20tjKK>
STELLINGEN
1.Tenonrechte concluderenHallamenEugsterdatindeaardkorstdefugaciteit
vanNH,enkeleordesgroterisdandefugaciteitvanstikstof;ditgeldt alleen
bijonwaarschijnlijk lage totaalspanningenvanstikstof+ammoniakinmetamorfe
enmagmatische vloeistoffen.
Hallam,M.,andEugster,H.P., 1976.Contrib.Mineral.Petrol.,57,p.227-244
2.Skarnsincontact-halo'svanintrusies zijnniet altijdmetasomatisch veranderde carbonaatgesteentes.Skarnvormingkaneveneensplaatsvindentenkostevan
hoornrotsen.
3.ZoutuitbloeiingenopNederlandse bakstenenbestaangrotendeelsuitdesulfaten
eugsteriet,syngenietengips.
4.KittrickbestrijdtdeconclusievanGiltrapenChurchmandatbijdeoplosexperimentenmetdeColonymontmorillonietdeoplossingnaverloopvantijdinevenwichtismeteenvaste fase,opgrondvanhetonjuist citerenvanhunresultaten.
Kittrick,J.A., 1978.Soil Sei.Soc.Am.J., 42,p.524-528.
Giltrap,D.J., and Churchman,G.J., 1977.Geochim.Cosmochim.Acta,41,p.387-392.
5.Lateritisatieenferralitisatie zijngeochemischtweegeheel verschillende
processendievaakverwardworden.
Maignien,R., 1966.Review ofresearch on laterites.UNESCO.
Jenny,H., 1980.TheSoilResource.Springer Verlag.
6.Hetheeft slechtsdanzinomfijnkorreligemineraal associatiesmeteenscanning electronenmicroscooptebekijken,wanneer dezeuitgerustismeteensysteem
waarmeedeindividuelemineralen chemisch onderzocht kunnenworden.Vergissingen
alsgemaaktdoorDriessenenSchoorlenEswaranenCarrerakunnenzovermeden
worden.
Driessen,P.m., and Schoorl,R., 1973.J. Soil Sei.,24,p.436-442.
Eswaran,H.,andCarrera,M., 1980.International symposium onsalt affected
soils,Karnal.p.237-293.
7.DemethodewaarbijdoorvereenvoudigingvandechemischesamenstellingGibbs
energieënvanchemischgecompliceerdenatuurlijkemineralenwordengeschatof
vergeleken,leidttotongeoorloofdeconclusies.
Huang,W.H.,andKeller,W.D., 1972.Am.Min.,57,p.1152-1162.
Reesman,A.L.,1978.ClaysClayMin.,p.217-225.
3.Bijtheoretischevoorspellingenoverdezoutsequentiediebijverdampingvan
eenwaterigeoplossing zalneerslaan,moetterdegerekeninggehoudenwordenmet
devormingvanmetastabielezoutmineralen.
Ditproefschrift.
9.Wanneerbijverdampingvaneenwaterigeoplossingmagnesiumneerslaatniet
alseencarbonaatmaaralseensilicaat,impliceerttoenamevandeCa
+Mg -
concentratiebijverdereverdampingnietnoodzakelijkerwijs afnamevandealkaliniteit.
Ditproefschrift.
10.Vrouwenindezogenaamde "hogerefuncties"vormennogsteedseentegrote
uitzondering.
11.Erlijkteeninverserelatietebestaantussenwetenschapsbeleid enwetenschapsbeoefening inNederland.
zieookSpernaWeiland,J., 1981.UniversiteitenHogeschool,p.311-322.
12.Delaatstestellingvandemens isinmoederaarde.
LidekeVergouwen
Saltmineralsandwaters fromsoils inKonyaandKenya.
Wageningen,22mei1981.
CONTENTS
7. Introduction
1
2. Generalinformation about salt-affeoted soils
3
2.1Introduction
3
2.2 Someconcepts
4
2.3 Soilclassification
2.4Problems arisingwith irrigationandcroppinginaridregions
7
9
2.5 Reclamationofsalt-affected soils
10
2.6 Reclamationofsodicsoils
10
2.7 Preventionofresalinisation
12
2.8 Cropping
13
3. Salt-affected soils in the GreatKonya Basin in Turkey
14
3.1 Physiography
14
3.2 Geology
15
3.3 SoilTypes
3.4 Correlationofoccurrenceofsaltswith soil type
17
20
3.5 Samplingandanalyticalprocedure
21
3.6 Correlationbetweenthefabricandmineralogyofthesalt
associations
21
3.7Methodsofwater analysis
23
4. Salt-affected soils in Kenya
27
4.1Physiographyandclimate
27
4.2Geology
27
4.3AmboseliBasin
30
4.4LakeVictoria
4.5EastofLakeTurkanaandtheChalbiDesert
4.6Samplingandanalyticalprocedure
32
33
35
4.7Correlationbetweenthefabricandmineralogyofthesalt
associations
35
5. Brine evolution
in closed basins
91
5.1 Introduction
41
5.2 Chemicalevolution
42
5.3Acquisitionofsolutes
43
5.4Evaporation
43
5.5Applicationtoirrigationofsalt-affected soils
6. Application
of the model of brine evolution
55
to Konya and Kenya
56
6.1Salts
56
6.2Watercompositioninthesystem (Na+K)-Ca-Mg-Cl-S04-(HC03+C03)
57
6.3WatercompositioninthesystemMgO-SiCL-CaO
58
6.4Concentrationofindividual ionsversuschlorineconcentration
61
6.5Conclusions
66
7. Mineralogy
67
7.1Identificationofminerals
67
7.2Crystallographicproperties
67
7.3Eugsterite,anewsaltmineral
69
7.4Konyaite,anewsaltmineral
75
7.5TwonewoccurrencesandtheGibbsenergyofburkeite
79
7.6Scanningelectronmicroscopy appliedtosalinesoilsfromthe
KonyaBasininTurkeyandfromKenya
8. Salt assemblages and evaporation
84
experiments
91
8.1 Introduction
91
8.2Evaporationexperiments
91
8.3PhasediagramsinthesystemNa-Ca-SO.-C02-H20
97
109
8.4SolubilitycalculationsinthesystemNa-K-Mg-Ca-Cl-S04-H20
9. Stable isotopes
in waters from the Konya- and Amboseli
a reconnaissance
Basins;
study
119
9.1 Introduction
9.2Stable isotopesofhydrogenandoxygen
9.3Factorsaffectingtheisotopiccompositionofnaturalwaters
9.4Analyticalproceduresanddata
9.5Discussions
119
119
120
123
125
10. Summary
129
77. Samenvatting
131
72.
References
VOORWOORD
Hethangtvaneenveelheid van invloeden afwelk zoutuitkristalliseert en
welkevormhetaanneemt.Zowerdenookvormeninhoudvanditproefschriftexternbeïnvloed.Vanhendietotdeuitkristallisatie hiervanbijdroegenwilik
noemen:
Prof.Dr.L.vanderPlas,mijnpromotor,wiensdeur inruime zinaltijdvoormij
openstond,
Drs.E.L.Meijer,mijnkamergenoot,die zijnthermodynamischinzicht belangeloos
metmij deeldeenaltijdbereidwasmee tedenken,
Prof.Dr. Ir.P.Buringh,Prof.Dr.R.D.Schuiling,Dr. Ir.N.vanBreemen,
Drs.N.M.deRooij,Drs.B.W. Zuurdeeg,metwie ikverhelderende discussies
voerde,
Prof.H.P.Eugster,Dr.Y.Gat,Dr.C E . Harvie,Mrs.NancyMö11er-Weare,
Dr.Y.Tardy,Prof.J.H.Weare,whokindly suppliednecessary information,
Ing.H.W. Boxern,Dr.Ir.P.M. Driessen,Dr.A.Mermut,Drs.L.Touber,
Dr. Ir.W.G.Sombroek,Drs.R.F.vandenWegen Ir.W.G.Wielemaker,die
mij begeleidden opmijnverzamelreizen inKenyaenTurkije,
Dr.R.KreulenenJ.Meesterburrie,diede isotoop-analyses uitvoerden,
Drs.J.M.A.R.Wevers,diedeprecessieopnames verzorgde,
Mevr.A.M.A.Baas,L.Th.Begheijn,J.D.J.vanDoesburg,A.J.Kuijper en
E.J.Velthorstvanwegehunanalytische inspanningen,
Dr.M.Barton,diedeskundig ensnelmijnEngelse tekst zuiverde,
G. BuurmanenO.D.Jeronimusvoorhet tekenwerk,
Z.vanDruutenvoordefotografie,
Mevr.A.BouterenMevr.C.B.Beemsterdiehettype-werk verzorgden
enOlafenElleke.
0
ED
^ %
ovtf
^
^#
<>
PUDERT
J(*
<?
PIETER
\&
-f
^
«i
BRAM
-^
EEF
^^
*
^
Soi
%ru
vcy
M#
?
^
T
9
S
^
amboseli
*
c
\ ^
^
ïft
SALTMINERALS AND WATERS FROM SOILSINKONYA AND KENYA
o0
c/i,aie
9N
NOt\-
1. INTRODUCTION
Salt-affected soilsarecharacterizedbythepresenceofhighly soluble
salts,mainly chlorides,sulphatesandcarbonates,which influenceplantgrowth.
Suchsalts increasetheosmoticpressureinthesoilmoistureandconsequently
plants cannot takeupenoughwater.Thisresultsinphysiologicaldrought,a
typicalphenomenonforplants growingonsaline land.Howmuchplants suffer
dependsonthequantityofsaltsandonthetypeofsaltspresentinthesoil
solution.
Salt-affectedsoilsoccurinmanypartsoftheworld,mainlyinwarmarid
andsemi-aridregions,whereprecipitationistoolowtoleachthesalts.Soil
salinisationisanormalprocessinthoseregionsweregroundwaterissaline
andatashallowdepth.Irrigatedlandinratneraridregionsoftenbecomes
saline,because solublesaltsaccumulateinthesoils,particularlyifinsufficientprovisionsaremadeforleachinganddrainage.
Thereclamationofsalt-affected soilsinordertomake themsuitablefor
agriculturalproductionisasubject thathasbeenstudiedformanyyears.There
areseveralbooksandarticlesdealingwith salt-affectedsoilsandinparticular
withsaltsatthesoilsurface,inthesoilprofileandinthegroundwater.In
studiesofsalt-affected soilsthesolublesaltsareusuallydissolvedandthe
totalamountofsaltsand/ortheconcentrationofvarious ionsinthesolution
aredetermined.Only littleattentionispaidtothemineralogicalcomposition
ofthesesalts.
Most saltsinsalt-affectedsoilsoriginatebyevaporationofgroundwateror
irrigationwater.Slightdifferencesinthecompositionofthedilutewaters give
risetodifferenttypesofsaltassemblagesandaccordinglytodifferent typesof
saltprovinces.Conversely,theevolutionofthegroundwater compositionduring
furtherevaporationislinkedtothemineralswhichformfromthesewaters.
Thisstudytreatstherelationbetweenthemineralogical compositionofthe
saltassemblagesandthegroundwatercomposition.
Sampleswere takenfromthreedifferentregionsinKenyaduringtheautumnof
1977andfromtheGreatKonyaBasininTurkeyduringthesummerof1978.Inadditiontosamplesofsaltefflorescences,samplesofthegroundwaterswere taken
atthesamelocalities.
These twocountrieswere chosenfortworeasons:
1.Thesalts inKenyaconsistmainlyofsodiumcarbonateswhereas thoseinthe
GreatKonyaBasinconsistmainlyofsodium-magnesiumsulphates.
2.Regions inbothcountrieshavepreviouslybeenstudied by soil scientists
fromtheDepartment ofSoilScienceoftheAgriculturalUniversity inWageningen.
Specialreference ismade tothestudiesofdeMeester (1970)andDriessen (1970).
The latteralsostarted thestudyofsaltminerals intheGreatKonyaBasinin
Turkey.Soilscientistsworkingatatrainingproject inKisii,Kenya,fromthe
samedepartment,andDutchsoilscientistsworkingattheKenyaSoilSurveyin
Nairobiprovided thenecessarypedological informationfor thesamplingprogram
inKenya.
Thecrystallographic andmorphologicalproperties ofmanysaltmineralsare
notwelldescribed.Someoftheseminerals,occurring inoneormoreofthesamplingareas,wereselected toinvestigate theseproperties indetail.
Moreover,twonewminerals'havebeendiscovered anddescribed.
2. GENERAL INFORMATION ABOUT SALT-AFFECTED SOILS
2.1 INTRODUCTION
Salt-affected soils (Halomorphicsoils)areindicatedonthenewSoilMap
oftheWorld (1:5.000.000)byFAO/Unesco (1974-1980)bytwomajor soilunits:
SolonchaksandSolonetz.
Solonchaks
*
aresoilswithahighsalinity (EC >15ms/cm)within 125cmof
thesoilsurface.
TheSolonchaksaresubdivided intofourmappingunits:
OrthicSolonchaks,themostcommonSolonchaks,
Gleyic
"
,Solonchakswith ground water influenceintheupper50cm,
Takyric
"
,Solonchaks incracking claysoils,
Mollic
"
,Solonchakswithadark-coloured surface layer,oftenhigh
inorganicmatter.
Soilswith lowersalinity thanSolonchaksaremappedasa"Salinephase"of
othersoilunits.
Solonetz
aresodium-rich soilswith accumulationofclayinthesubsurface
*
soil.This clayhasanESP> 15%andthestructureiscolumnarorprismatic.
TheSolonetzaresubdivided intothreemappingunits:
OrthicSolonetz, themostcommonSolonetz,
Gleyic
" , thosewithgroundwaterinfluenceintheupper50cm,
Mollic
" , thosewithadark-coloured surface layer,oftenhigh
inorganicmatter.
Soilswithalowerpercentageofexchangeable sodium thanSolonetzaremapped
as"Sodicphase"ofothersoilunits.
ReferenceismadetoFAO/Unesco (1974)forprecisedefinitions.Itisevident from theseshortdescriptions thatsalt-affected soilswhicharecharacterizedbythepresenceofhighlysolublechloridesandsulphates (Solonchaks)are
distinguished fromthosewithhighlysolublecarbonatesandbicarbonates (Solonetz)inwhichthestructureofthesoilsisdeterioratedandconsequently there
isapoormediumforrootgrowth.
* Theseabbreviationswillbeexplained inSection2.2.
Besides therealSolonchaksandSolonetz therearealsosoilshavingintermediatecharacteristics,which areoftencalledSolonchak-Solonetz orsalinesodicsoils (Richards,ed.,1954;theso-called "Handbook 60").Theterm"sodic"
israthernew;inolderpublications thename "alkali soils"wasoftenused,
butthisnameisnow abandoned.
There isnotmuchconfusion aboutthecharacteristics,properties and
agricultural evaluationofthevarious salt-affected soils.However,thereis
muchconfusiononhowsuchsoilsshouldbeclassified orpresented asmapping
unitsonsoilmaps. Insection2.3 theclassificationofsalt-affected soils
willbedealtwith inageneralmanner;specificandoftencomplicateddefinitionsanddetailswillbeavoided.
2.2 SOME CONCEPTS
Insection2.1 someabbreviationshavebeenintroduced.Theseterms,commonly
used insoilscience,andsomeotherrelevant facts,whicharenotfamiliarto
non-pedologists,willbeexplainedbeforedealingwiththeclassificationof
salt-affectedsoils.
Insomeclayminerals cations arereplaced byothercationswithalower
valency,forexampleSi
byAl
orAl
byMg .Thisproduces anegative
chargeatthemineralsurfaceswhichenablesthemtoadsorbpositively charged
ionsfromthesoilsolution.
Organicmatter canalsoadsorb cations atthesurfaceandexchange these ions
withothers.
Theadsorbed ionsdistribute themselvesover thenegatively charged surfaceof
theclayminerals according toaBoltzmann-typedistribution (fig.2.1a and
2.1b).This istheresult ofanequilibrium'betweentheattractive forcesof
thesurfaceandthetendencyofthe ionstodiffuse away.The layerofliquid
whichextends fromthemineral surfacetothepointwhere theconcentrationof
theionshasdropped totheconcentrationofthesoilsolution iscalled the
diffusedoublelayer (DDL).
Thethicknessofthediffusedouble layerisdependentonthechargeofthe
adsorbed ions;divalent ionsareattractedmorestronglytothesurfaceand
theirDDL isthinnerthaninthecaseofmonovalent ions.Thethicknessofthe
DDLisalsodependentontheconcentrationofthesoilsolution;thethickness
decreaseswhentheconcentrationoftheions inthesoilsolutionincreases.
Whenwater isintroduced intoadrysoil,theionsexertaswellingpressure,
_
E
E
+
+
+
- +
+
- +
+
—
_ ++
+
- +
z
++
+
+
+
+
—+
—+ +
+
z +
—+ + +
-
+
+
+
+
+
+
Fig. 2.1a and 2.1b Distribution of cations against the surfaceof aclay
mineral:c=concentration;x=distance from the surface;DDL=diffuse
double layer, (modified afterBolt &Bruggenwert, 1976)
theclayparticlesaredrivenaway fromeachother,thesoilswellsandthe
permeability decreases.Thisswellingismostpronouncedwhenmuchsodium,a
monovalent ion,isadsorbedandwhen little saltisdissolved inthewater,as
isthecasewithrainwater,becauseunder suchconditionstheDDLisatits
maximum.Whenabundantwaterisavailable,thesoilpeptises,thewaterandthe
clayparticles formakindofconcentrated suspensionandtherearenoaggregates
anymoreandthesoilstructuredisappears completely.
Thephysicalpropertiesofasoilarethusdependentupontheadsorptionof
sodiumontheadsorbingcomplexandupontheconcentrationofthesolutesin
thesoil solution.Thedependenceonthesetwoispresentedinfig.2.2.
Theamountofcationswhichasoilcanadsorbbycationexchangeiscalled
1z+
-1
thecation-exchange-capacityor CEC,expressedasmmol-M .kg ofsoilin
whichMdenotesthecationandzisitscharge.Theseunits conformtoSI,the
Système International d'Unités (Brinkman, 1979).BeforetheintroductionofSI,
theCECofasoilwasexpressedasmilliequivalent .kg ofsoil.ThemilliequivalentisnotrecognizedinSI.
Theexchangeable sodiumpercentageor ESPisthepercentageoftheexchange-
Fig. 2.2 Dependence of thesoil
structure on the electroconductivity
(Ec)of the soil solution and the
exchangeable sodiumpercentage (ESP)
inthe soil.ForexplanationofEc
andESP,see text, (afterBolt &
Bruggenwert, 1976)
ablecationsontheadsorptioncomplexwhichconsistofsodium.
ESP=
exchangeable sodiummmolNa .kg ofsoil
r-^
, —
x100
mmol-M . k g ofsoil
CEC
Thesodiumadsorptionratioor SARisanothertermwhichiscommonlyused.
Thisratioexpressestherelative activityofsodiuminsoil-andirrigation
watersandisexpressedas
Na
SAR
v/(Ca
2+ +
Mg 2 + )/2
1 z+
_3
withallconcentrations expressedasmol-44 .m
A saturated
soil
paste isasoilsampletowhichsomuchwaterhasbeen
addedthatallofthevoidsarefilledwithwater.Thepastestartstoglisten
andisabletoflow.Themoistureextracted fromthispasteiscalledthe
saturationextract.
Theelectricalconductivityor Ea isthereciprocalofelectricalresistivityexpressedasmillisiemenspercm (ms/cm).Before introductionoftheSI
system,theEcwasexpressedasmmho/cm.
2.3 SOIL CLASSIFICATION
All soil classification systemshavehierarchies ofgroupsofsoils (taxa).
Thismeans thatthere aredifferent levelsofclassification.At thehighest
levelssoils areclassified according tocertaincharacteristics whichareconsidered tobemore important thanothercharacteristicswhichareused todistinguish soilsatthe lower levels.Insomesoil classification systemssaltaffected soilsareseparated fromothersoils onthehighest levelsofclassification.Inother systemssoils are firstclassified according toother criteria
and at lower levelsofclassification thepresence ofsoluble salts is indicated.
The lastprocedure is,forexample,followed inthenewAmerican systemofsoil
classification "SoilTaxonomy" (SoilSurvey Staff, 1975).Heretherealsaltaffected soilsbelong tothesoilorderoftheAridisols.Therearetwosubordersviz. the Argids (soilswithclay accumulation inthesubsurface horizon)
and the Orthids
(soilswithout suchaclayaccumulation).At thethird levelof
classification there isagreatgroupnamed "Natrargids",
whichareArgidswith
ahighsodiumpercentage attheadsorption complex,andare indicated asthe
SolonetzontheSoilMap oftheWorld (FAO/Unesco, 1974-1980).At this third
levelofclassification (greatgroups)thereare the "Salorthids"
inthesub-
orderofOrthids,whicharetheSolonchaks ontheSoilMap oftheWorld. Inthe
oldAmerican systemofsoilclassification (1938)thenames Solonchak soilsand
Solonetz soilswereused.
IntheFrenchsystemofsoil classification (CPSS, 1967)the salt-affected
soils areseparated fromallother soilsonthehighest levelof classification.
There isa"Classedessols sodiques"witha"Sous-classe des solssodiques à
structurenondegrade"towhichbelong the"Groupedessolssalins" (Solonchak).
Totheother "Sous-classe dessols sodiques àstructure dégradé"belongs a
"Groupedessolssalins àalcalins" (Solonchak-Solonetz) and a"Groupedessols
sodiques àhorizonB" (Solonetz). IntheprovisionalnewFrenchsystem (Fauck
etal, 1979)thereareatthehighest level "LesSelsols".One ofthegreat subclasses iscalled "Halisols"or "Selsols haliques"andthis is subdivided
according tothedominant saltspresent inthe soil,forexample "halisols
à chloruredesodium".Another subclassofthe"Halisols"referstothe
"halisols carboxiques"or "carboxihalisols",the formerSolonetz.
IntheRussian classification of soils,Solonchaks andSolonetz aredistinguished atahigh levelofclassification. Theoriginofbothnames is
Russian.
TheSolonchaks canbesubdivided into external Solonchaks with soluble
salts throughout thewhole soil,and internal Solonchaks with soluble salts
inthe subsoilorsubstratum.The Solonchaks canbesubdivided accordingto
thecompositionofthe salts.The following typesexist:nitrate-,nitratechloride-,chloride-,sulphate-chloride-,chloride-sulphate-,sulphate-,
sulphate-soda-,soda-,borate-solonchaks.
External Solonchaks canalsobesubdivided accordingtotheir morphological
features.Thefollowing typesofexternal Solonchaksmayoccur:
1. Flooded
Solonchaks
These soilsareperiodically underwater.Evaporationofthesurfacewater
producesacoherentwhite crust,sometimesofseveral centimeters thickness.
This crust seals thesubsurface therebypreventing further evaporation.Therefore,the soilunder thecrustisratherwet.
2. Puffed Solonchaks (puffyorfluffy Solonchaks)
Theupperpartofthesoilisvery looseandpuffybecauseofthepresenceof
sodium sulphateneedleswhichpushthesoilparticles apart.Theuppermost
layerissometimesalittlebit crusty (sodiumchloride)whichholds thesoil
togetherand counteracts evaporation.This typeofsoilformsbyevaporation
ofgroundwater.
3. Sabbakh soils
(or wet mineral
soils)
These soilscontainmagnesium and calcium chlorideswhich areextremelyhygroscopic.Theyattractmoisture fromtheairandare,therefore,darkerincolour
thanthesurroundings,especiallyinthemorning (SabbakhisArabic formorning).
There aremanyotherways inwhich salt-affected soilshavebeen subdivided,
althoughmostoftheseclassifications arenotofficial systemsofsoil classification.Sometimesasubdivisionismade accordingtotheoriginofthesalt.
Thefollowing typesexist:
1. Closed basin saline
soils
Weathering products fromthehinterland aretransportedbysurface andsubsurfacewaters intothebasin,theyaccumulate andcannotberemoved fromthe
basin.Evaporationofthewater causessaltstobedepositedinandontopof
thesoilprofile.These soilswillbetreatedindetailinchapter5.
2. Marine saline
soils
Thesesoilsoccur along coasts eitherbyflooding,seepageofseawaterorby
seaspray.
3. Alloahthonous
airblown saline
soils
Salts fromother localitiesaretransportedbythewindandformnewsaline
soilsattheplacesofdeposition.
4. Anthropogenic
saline
soils
Salinesoils causedbytheinfluenceofmanbymeansofirrigationandflooding
without concomittant drainagetoremovethedissolved salt.Theirrigation
water evaporatesandsaltsareformed.
This short introduction indicates thattherearemany criteriabywhich
salt-affected soilsmaybeclassified.Itdependsonthepurposeofinvestigationwhichclassification systemwill'bechosen.Theclassification fromthe
pointofviewofsoilsystematicsisquitedifferent fromonethatissetup
tocharacterizetheproductive capacityforagriculturalpurposes.Forthe
latteritisnecessarynotonlytoknowtheEC andESPvalues,butalsowhich
saltsarepresentinthesoil,thesoilsolutionandinthegroundwater.A
simpleECmeasurement alone givesonlyavery rough ideaofthetotal,amount
ofsaltsbecausedivalent ionshaveadifferent contributiontotheelectroconductivity thanmonovalent ions.MoreovertheECmeasurements arecarried
outonasaturationextractandthewater contentofdifferent soilsunder
fieldconditions isextremelyvariable.Therefore,theosmoticpressureofthe
soilsolution,whichthisECmeasurementismeanttoindicate,maybeincorrect.
Inthisstudytheterms salt-affected
soils and Solonchaks
willbeused.
2.4 PROBLEMSARISING WITH IRRIGATIONANDCROPPING INARID REGIONS
Inaridregions irrigationisappliedfortwodifferentpurposes:
1.water supplyforplants
2.improvementofchemicalandphysicalpropertiesofthesoil.
Before soilswhichshowapoor naturalvegetationareusedforagriculture,it
mustbeascertainedwhether theyarechemicallyandphysically suitablefor
growingcrops.
Saltsmustberemoved fromsoilswhicharetoosaline,andiftoomuch sodium
isadsorbedontheadsorption complexinthesoil,thesodiummustbeexchanged
withother cations.Inthefollowing sectionsthereclamationofsuchsoils
willbedealtwith.
10
2.5 RECLAMATION OFSALTAFFECTED SOILS
Irrigationisthemethodwherebywater issupplied forcropgrowth in
aridregions.Ifthesoilsaresalt-affected,thesaltsmust firstberemoved.
Usually,salts areremovedby leaching;somuchwater isaddedtothesoil
thatallsaltsaredissolved and transported downwards intheprofile.
Thedependence ofthesolubilities ofmostnaturallyoccurring binary
saltsontemperature isshowninfig.2.3 andfig.2.4.The solubilities are
expressed ing/kgwater and inmole/kgwater respectively.
Thegyosumand anhydritedataarethemeancurvesoftherangesgivenby
Hardie (1967).TheothersolubilitydataarefromLinke (1965).Sodium salts
arepresent inmostsaltefflorescences,atleast inthecaseofsalinesoils
whichoriginatedbyevaporationofgroundwater.Whenthesodiumsaltsare
dissolved intheirrigationwater,caremustbe takenthattheSARvalueof
theirrigationwaterremains lowinordertoprevent thesoilfrombecoming
sodic.
Leaching insummerhas theadvantagethatthesolubilityofmostsaltsincreaseswithtemperature (fig.2.3 and fig.2.4),but ithas thedisadvantage
thatevaporation ishigh.
Ifthere isahard layerofcalciteorgypsumpresent inthe soil,whichis
impermeable and impenetrable toroots,this layermustbe brokenmechanically,
forexampleby subsoiling.Suchanatural gypsumsupplyisvery favourable
whenthesoil issodicaswell (seesection2.6).
2.6 RECLAMATION OFSODIC SOILS
Iftoomuchsodium isadsorbed attheadsorptioncomplex,whichmakes the
soilunsuitable forgrowingcrops,thefirst irrigationwater isused forimprovement ofthephysicalproperties ofthesoil.Thesodiummustbereplaced
bydouble-charged cationssuchascalciumandmagnesiumbecausesodiumcauses
thesoiltoswellandtobecome imoermeablewhenitbecomeswetand further
irrigationisuseless.
Theconductivity ofthefirst irrigationwatermustbehigh,because thedispersionoftheclayparticlesdependsnotonlyontheESPofthesoil,but
alsoonthesaltcontentofthewater (seefig.2.2).Usually gypsum
(CaS0..2H20)isaddedtotheirrigationwater.Theelectroconductivity increases andcalcium ionsbecome availabletoreplace thesodium ions.
11
XI
cu
CU
4-J
CO
•-H
rH
•H
J3
CO
CU
4-1
>1
^^
u
a.
3
CU
C
X
i-I
cu
O
co
I-H
CU
•O
u
3
cu
J2
4-)
•4-1
S-l
CU
4-J
O
CU
CU
O
CN
raÏC
0
D. 0 0
B .ü
-o
CU
o
4J
T)
4->
•rH
e
a)
.c
C
4J
t—1
CO
co
S co
<U
P.
CU
<r
CO
4-t
rH
CO
CO
CN
>>
a
3
O
U
"0
co
•
oo C
•H
-O
[K
|
CU
.c
co
-o
o"_
co
>%
CO
U
•H
4J
4J
CU
t—1
O
S
0
tu
o#
4J
>^
.a
-a
CU
co
3
CU
e
o
e
CO
co
C
MH • H
e
o
O
•r4
4-1
cfl
»H
4-1 " Ö
13
CU
4J CO
•H
CO
t—1 CU
•rH u
X I p.
>*
3
X
^O
CU
co
e
CU
CU
u
CU 3
J = 4J
4J CO
S-l
4-1 CU
O
o.
CU
o
e
CU
•a
s
CU
4-1
-C
i-t
m
CN
CO
CO
CU
O
CN
SC
00
M
CO
3
O
>H
•a
r^
U
CO
• •Ha
u
>>
x;
xi
4J
4-1 4-1
rH
CO
CO
c • SH
a.
C
U CO
o 4-1
CU
cu
o *a
C 0
0 3
o .c
>>
.c
c
00
•rH
CO
Pu -a
00
T3
^-N O
CU T 3
C -H
•H
>
i-H « H
-0
T3
CU
CU
-C -Q
co
CO 4J
T-) CO
v_- g
e
CN
r H CO
CJ (3
CO O
O -H
4-1
u ra
o 4.1
u
[U
12
Ca
Na
Na
+CaS0 4 .2H 2 0
Na
Ca
+Na 2 S0 4+H 2 0
-Ca
Iftheclaycontentofthesoilisveryhighandmuch sodiumisadsorbedat
thisclay,themaximum electroconductivity reachedwiththeadditionofgypsum
maynotbehighenoughtoavoidpeptisationoftheclayparticles.Thisis
becausegypsumhasalowsolubility (2.1g/kgwaterat25 C ) .Insuchcases
amore soluble saltsuchascalciumchloride,whichhasasolubilityof828g/kg
waterat25 C,isadded.However,thissaltismuchmore expensive thangypsum.
Sometimes sodium saltsareaddedinadditiontogypsuminordertoincreasethe
electroconductivityoftheirrigationwater.Inthisparticular casetheprocess
ofreplacing sodiumbycalciumisobviouslyslower.
Sodiummustberemoved fromtheadsorptioncomplex,especiallyinregionswhere
thecompositionoftheirrigationwater cannotbekeptundercontrol,suchas
inregionswithintermittentrainfall.Rainwater containsfewdissolved salts
andarainshower causestheclayparticlestodisperse immediately,iftheESP
istoohigh.
2.7 PREVENTION OFRtSALINISATION
Inaridregionsreclamationandirrigationofsoilsisuselesswithout
drainage. Irrigationcausesthewater leveltoriseandbothcapillary rise
andevaporationwill causetheharmfulconstituentstoreachtheroot zone
andthesurfaceagain.
Ifnaturaldrainageisabsent,anartificialdrainage systemmustbebuiltto
keepthewater levelatacriticaldepth.Evaporationmustbereducedasmuch
aspossible.Thiscanbeachievedbyadensevegetationcover,e.g.lucernein
timesoffalloworafterharvesting.Theplantsprovide shade,therootsincreasethepermeabilityandtheyaddorganicmattertothesoil.This organic
matter increasesthedissolutionofcalcium-carbonatebecauseofthereleased
carbonic acidandinadditionnitrogenisaddedtothesoil.
13
Ingeneral,enoughwatermustbeaddedtothesoiltopreventtheupward
movement ofthesalts.This ismorewaterthanisrequired tobalance the
consumptionbythecropsplus thewaterwhichevaporates.
2.8 CROPPING
Oncethesoilhasbeenreclaimed,cropscanbegrown.There isanintimate
relationbetweenthesoilcondition,thequalityofthe irrigationwaterwhich
isavailable,and thetypeofcropswhichcanbegrown. Ifthe irrigationwater
israthermineralised,thesoilcannotbereclaimed completelyandonlysalt
tolerantcropscanbegrown.Ontheotherhand,thechoiceofcropsdetermines
thedegree towhich thesoilmustbereclaimed.
Saline solutionsareharmful toplantsbecauseoftheosmoticpressurethe
solutionexertsontheplant,which suppresses theabsorptionofwaterbythe
plant,andbecauseofthetoxiceffectsoftheindividual ionspresent inthe
solutions.Thesetwoeffectswillact simultaneously.
Theconcentrationoftheconstituents inthesoil solutionwill increaseby
evaporation.This leadstoanincreaseoftheosmoticpressure,but alsothe
soilmoisture tension increases.Bothlimittheavailability ofwater tothe
plants.
Inaridregions,manyfactors,whicharesometimesdifficult toconsider
simultaneously,mustbeaccounted beforeachoicecanbemade aboutthetype
ofcrop togrowunder thosespecificcircumstances andaboutthedegreeof
soilreclamationwhich isrequired.
14
SALT-AFFECTED SOILS IN THE GREAT KONYA BASIN IN
TURKEY
3.1 PHYSIOGRAPHY
TheGreat KonyaBasinissituated about300kmsouthofAnkaraontheCentral
AnatolianPlateauonwhichseveralbasinsofvaryingsizeoccur.Theinsetin
fig.3.2showsthelocationoftheKonyaBasininTurkey.
Thealtitudeisabout1000mabovesealevel.ThehighestpeakintheBasin
istheKaradag,avolcanicandlimestonemassif,2390mhigh.Theareaof
2
theBasinisabout 10,000km.Therearesomeperennial lakessuchasHotamis.
GölandAkgöl.TheBasinhasnonaturalhydrographieoutlet.
The climateintheKonyaBasinissemi-aridwith cold,moistwintersand
hot,drysummers.Temperaturesof-25Carecommoninwinterwhereasinthe
summertemperatures reach35C.Theaverage temperature inwinterisabout
0Candinsummer about20C.Thereisafrost-freeperiodofonly165days.
Precipitationisrestrictedtoautumn,winterandspring.Inwinterprecipitationisintheformofsnow;foratleastthreemonthsperannumthebasin
iscoveredwithsnow.Thesummersarenormally dry.Theaverage precipitation
is about300mm/yearalthoughitcanvarybetween250mmand350mmdepending
upongeographiclocation.Evaporationexceedsprecipitation;theaverage
evaporationisabout930mm/year.
Thereismuchevidence thatthebasinwasoncealake.Abandoned shore
linescanbefoundattherimsofthebasin andsandbarriersrunparallelto
the formershoreline.Furthermore therearedepositsoflimestone containing
freshwater fauna.Terrarossalayers,which occur100-200mbelowthesoil
surface indicate that this lakemusthaveperiodically driedup.
Thereisarcheological evidence thatthebasinhasbeendryfromatleast
6500B.C.because fromthat timethebasinhasbeeninhabited (Mellaart,
1964).Ataboutthattimetheclimatemusthave changed;inthePleistocene
thetemperaturewasalittlelowersothatprecipitationexceeded evaporation
whereasevaporationnowexceedsprecipitationandthebasinisdrymostof
thetime.
15
3.2 GEOLOGY
TheBasinoriginatedwhen,during theTertiaryupfolding oftheAnatolian
belt,theinnerAnatoliandomecollapsed andsubsidence tookplace.Throughout
mostoftheTertiary andthePleistocene thelargedepressions soformedwere
filledwithwater,forming inland seas.
TheBasinisborderedbytwoalpineorogens.TheTaurusorogenforms theborder
inthesouthandwest,whereas theAnatolidesborder theBasintothenorthand
east.These twomountainchains consistofPaleozoic sediments andupperMiocene
limestones.UltrabasicrocksoccuronalargescaleintheTaurusmountains.
From theEocene-Oligocene onwardsclasticmaterial (clay,marl,sand,gravel
andconglomerate)wasdeposited intheBasin.During theMio-Pliocenenonclasticfreshwater limestonewasdeposited and inmanyplaces this formsintercalationsbetweenmarlandclay. Insomeplacesratherthick freshwaterlimestonebeds canbefound.Theformationoflimestone terminated abruptly andthe
upperpartsofthe sediments are formedofclasticmaterial.
TheKonyabasinshowsmany featuresofvolcanicorigin.Thevolcanicactivitystarted intheUpper-MiocenewhentheCentralAnatolian Plateauwas still
rising andthebasinsoriginated,whichbecame filledwithsediments.Volcanic
activityreached amaximum inPliocenetime.During theQuaternary,volcanic
activity started again.Thevolcanicproducts ofthesetwoperiodsdifferas
follows (Jung&Keller, 1972):
1.Themainpartofthevolcanicmaterial from the Upper Mioaene periodconsistsofintermediate tobasicvolcanics suchasandesites,latites,and
andesiticbasalts.Thesevolcanicsbelong totheso-called andesiticbelt
whichranges fromthesouthof ItalytoPakistan.Examples ofthistypeof
volcanismaretheErenlerDag-AlaçaDag-complexwestofthetownofKonya,
theKaraDag inthecentreoftheBasin,theKaraçaDagNEofthevillage of
Karapinar andtheHasanDag-MelendizDag-massifnear thetownofNigdeNEof
theBasin.Someacid lavassuchasdacites andrhyolites alsooccur.
2.The Quaternary volcanicactivityproduced ontheonehandextremely acid
obsidians andvitrophyres andontheotherhand isolated cindercones,made
upofbasaltic scoriae,andmaars.Thecindercones formachainfromthe
KaraDagMassiftowards thetownofBor intheNEoftheBasinwhich suggests
a correlationbetweenthistypeofvolcanism andyoung faulttectonics.
Nowadaysvolcanicactivity ismanifestby saline springsnearAkhüyük and
theancient cityofTiana (localitiesT16andT25respectively,seefig.3.2).
16
e
o
17
Legend tofig. 3.1
Q
Np
Nm
0
E
Mj
C
P
PC
D
S
DI
V
Quarternary (mainly alluvial and lacustrine calcareous clays)
Neogene (mainly Pliocenehorizontally stratified freshwater limestone)
Neogene (mainlyMiocenemarine limestone)
Oligocène (mainly gypsum or soluble salts)
Eocene (mainly calcareous Flysch deposits)
Mesozoic (mainly Jurassic and Triassic limestone)
Cretaceous (mainlymarlswith Serpentine and radiolaries)
Palaeozoic (mainly marmor)
Permocarboniferic (mainly limestone and marmor)
Devonic (mainly stratified limestone)
Chrystalline Schist
Diorites
Volcanic deposits
a mainlyAndesit andDacit
b mainly Basalt
t mainly Tuff
11mainly Quarternary lava
ThespringsnearTianaarestillvisitedbypatients suffering from internal
diseases.
Fig. 3.1showsthegeologyoftheGreatKonyaBasin'scatchmentarea.
3.3 SOIL TYPES
Fig. 3.2presentsasimplified soilmapoftheKonyaBasin (deMeester, 1971).
Foradetailedmapreferenceismadeto"SoilmapoftheGreatKonyaBasin"
(deMeester, 1971).Fig.3.3showsthecross sections.Thesoilassociationson
themapoffig.3.2weredefinedbytheirphysiographic features.Thefollowing
associationsaredistinguishedonthemap(deMeester, 1970):
TheUplands (U).Theserepresentthesurroundingmountainswhichconsistof
limestoneandvolcanicsandthevolcanoesinthebasinitself.
TheColluvialSlopes (C). These formsteep talusesatthebaseoftheuplandsandconsistofrockdebris fromthemountains.
TheBajadas (B). Thesearegently slopingpiedmontplains consistingof
finegrainedmaterial fromtheuplands.
TheTerraces (T).TheseareremnantsofNeogeneterraces;thehigherterracesarecutbyerosiongullies,thelowerterracesareflat.
TheAlluvial Plains (A).These consistofsediments transportedbythe
rivers intothebasin.Theyrange fromcoarsesandtoheavyclay.
TheLacustrine Plains (L).Theseareveryflatmarlsoilswhichcoverthe
flooroftheAncientLakeinthecentreofthebasin.They includetheoldsandy
beachridges.
o
u
CU
4-1
U-i
CO
"^s
•r4
4J
O
01
CO
1
CO
CO
O
1-1
> 1
o
<u
.ü n
u o
s lxH
ca
ca
ca
60
e
•H
i-H
>.Pe.
ü
o c
a
CO
« CU
4->
ca J 3
CU 4J
u
o
C4-4
O
<u
.C co
4-1
0)
• H
CH 4-1
O -r4
1-1
p. ca
ca CJ
e o
j-t
e
o
01
• H J3
4-1 4J
ca
•H
TS
o c
o ca
m
CO ao
0) • H
a j CH
*a
00
•H
Ol
[ u T3
CU
CU
co
19
Legend tofig. 3.2
Ur
Uv
Cr
Cv
Te
Th
Tc
Br
Bv
Aa
Ab
Ac
Ad
Ae
LimestoneUpland Soils
Volcanic Upland Soils
Limestone ColluvialSoils
Volcanic Colluvial Soils
Flat Terrace Soils
Undulating Terrace Soils
SoftLimeSoils
LimestoneBajadaSoils
Volcanic Bajada Soils
ÇamurlukFanSoils
Former Backswamp Soils
Car^ambaFanSoils
DeliFanSoils
SelerekiFanSoils
I Te
Uv I Au I Cr l u
I Cr I Br
I Lr I
Tc
Af
Ag
Ah
Ak
Am
An
As
Au
Lm
Lr
Lp
Lo
Mf
Dd
CakrnakFan Soils
ZanopaFanSoils
BorFanSoils
Meram and SilleFanSoils
May FanSoils
BayatFanSoils
AyranFanSoils
SoilsofMedium Sized Fans
Marl Soils
Sandridge Soils
Sandplain and Beach Soils
Old Sandplain Soils
Marsh Soils
Sand Dunes
Lr I Cr I
I
I Lr I
Ur I Cr I
I Lp I
Te
lake
I
I
Lo
Lm
I Te I Cr I
iLrl Lm I
Fig. 3.3 Three schematic cross-sections through theGreatKonyaBasin.
Forposition,see fig.3.2 (fromdeMeester, 1970)
Bv
I Uv
Ur
(m)above
sea level
20
"
approximate extend of
the Ancient Konya Lake
extend of the
present Lacustrine Plain
secondary depressions
constituent plains
uplands
Fig. 3.4 Extent of theLacustrine Plain in theBasin,of the constituent
plains and of thedepressions inthem, (fromdeMeester, 1970)
TheAeolianPlain (D).Thisplainconsistsofsandduneswhichhave formed
recentlyasaresultofovergrazing.
TheMarshes (M).Theyoccuratthefootofthealluvial fansandnear lakes.
3.4 CORRELATION OFOCCURRENCEOFSALTSWITH SOIL TYPE
Salt-affected soilscanbefoundinthelowerpartsofthebasin.Themost
extensive saltefflorescencesoccuronthelacustrinemarl soils (Lm).Continuous
saltcrustscanbefoundinthedepressionsinthebasinfloor (seefig. 3.4),
whichisfloodedforatleastafewmonthsperyear.Herethesaltoriginatesby
evaporationofsurfacewater (flooded solonchaks).Somealluvialplainsoilsare
very saline;thesalinityofthesesoils increases towardsthecentreofthe
basin.Onthelowest terracesoils,thesoft limesoils (Tc),salt efflorescences
canbefound.
TheMarshsoils,whichcoverthelowestpartsofthebasin,aregenerallyvery
saline,buttheyaresituated alongperennial lakesandaremoistduringthe
wholeyearsonoefflorescencescanform.
21
3.5 SAMPLINGANDANALYTICAL PROCEDURE
Asmanydifferent types ofsaltefflorescences aspossiblewere sampled
duringthesummerof 1978.Thesamplingsitesareshownonfig.3.2.
Whenthesaltconsisted ofaloosepowder,smallpolyethylene tubeswere filled
withthesalt.Coherentcrustswerepacked intoiletpaperandtransported in
plasticboxes.Samplesofgroundwaterwerealso takenwhenpossible,but in
manycases itwas impossible topenetrate thesoilwith ahandauger,and in
somecases thegroundwater levelwas toodeeptosample.Water samples from
someinflowriverswerealsotaken.
Twodifferent typesofbottleswereused forwater samples:a250ml glass
bottlewhich isimpermeable toCCuwasused forsamples tobeanalysed for
carbonates anda250mlpolyethylenebottleforsamplestobeanalysed for
Na,K,Ca,Mg,Cl,SO.and silica.Theglassbottleswerebrowntominimize
asfaraspossiblealgalgrowth.Allbottleswerepacked inaflanel cloth
covertoavoidpenetrationoflightandriseoftemperature.A fewdropsof
chloroformwereaddedtothecontentsofpolyethylenebottles.As little airspaceaspossiblewas left inthebottlesafter filling.
Temperature,pH,Ehandelectroconductivityweremeasured atthe sampling
site.Immediatelyafterarrival inWageningenthecarbonatesystemwas analysed.
Na,K,Ca,Mg,Cl,SO.andsilicawere analysedwithinfourmonthsaftersampling.Allanalyseswereperformed induplicate.Theanalysesofthewater
samples arereported inTable 3.1.
The saltmineralswereidentifiedbymeansofX-raydiffraction analysis
withaNoniusGuinier cameraFR 552withahigh-resolving JohanssonKcxjmonochromator,usingCoKcxjradiation (X=0.17889nm). Table 3.2 liststhesalt
mineralswhichhavebeenfound intheBasin.Table 3.3 represents thesaltassociations atthedifferentlocalities.
3.6 CORRELATION BETWEEN THE FABRIC ANDMINERALOGY OFTHESALT ASSOCIATIONS
Bloedite,thenardite andhalitearethemostabundantsaltminerals inthe
KonyaBasinand,therefore,theydominate themorphology ofthesaltassemblages.
Whenbloedite isthedominantspecies,thesaltassociation isvery fluffy
andairywith largeopenholes and channelsthroughthecrustyparticles.When
itoccurs incoherentcrusts,thecrust isveryfragileand friableandon
touching itdésintégrâtesintosmallparticles.Whenitoccurs intheformof
powder,thepowder isvery loose,againwithlargeholesandpores.These
22
O u~i Oi r -
— i£> -T — s r
co CM en — r - r - -
f ) mCT\vß f
O c n o c N r - o c M c N c n O O m e M — -3- m O c
1 u-1 f ) O O C
O O O O O O O O O O O -
'
O Oi r-~ CM L
ON • * m o •
O O O
O
• • J ^ m C i ' B C O I s i O ^ f O ' Û I N M
o \ o ON co co f
l O c o c o t
CM CM <J\ •
CM —
>o ^o ^o o o
I N n - IN (N
J o o~> <r -^- co c
- CM co m o CM •
)CN W - J O (
- O 00
• (N kA - J C l »O I
• CM CO — CO I
- O — o N in N M O N (
CM O O O 0> i
• O O O •£> CO O r-~
- r» o r- o
o en co o - * r
(•-»• — co
—
—
CM r-- en — r-% \©
CM — CM
• M »
M i/i N
-* —
Vf
m
LnencMOCTivOO-T
- • ï t o n t N c o o
r - — m co
-r^\Dr-~r-i-^r»-cooot
(O VO t
co co r - r-. r - r
ooooooooooooooooooooooooooooooo
O 0 0 O 0 0 C 0 C T » O O ~ O m O O O v 0 O O O O O O O \ 0 C M O O O O O O O
m - s j - r - o r ^ v £ ) O O r - o c M c o r - O v £ > o o c M O r - c M i o > o m — <r CM o r - o o
-^
cncMcM — r - r - c N —
co O co — 0> O < • O^ O t r i vu
- I N - J ^ I A C O C M ^
\0 \0
CM C M C M C M
-J —
~* —
O — CO \ 0
u u «
ai -u -u -u x:
fflc5(i]4Jfflci!iDü
3 3 3 ca 3 3 3
C(] T3 "O M T3 T3 -Ö
ö0 G C <U C d d
O) 3 3 3 .,
• 4J C > 0 ) - . S0 0 « 0 0 0 > 0 0 0 !
3
3 - -
3
3
3
3
3
3
3
o o o o v o > o o o o o o o o
•
1 \D CO 0~i O -
3
-o-D'O'O'O'OTj'a
^ ^ G d d d d d d d
3 4 1 3 3 3 3 3 3 3 P
'
'
•
- - i - H t i V j ^ i l - i M t - i M H
l û S O O O O O O O O
23
observations contrastwith thoseofDriessenandSchoorl (1973)whostatethat
bloediteformsglass-likecrusts.
Thenarditeoccurs assmallneedleswhichpushthesoilparticles inthe
upper layerofthesoilapart.This layerbecomesverypuffed andwhenwalking
onthesoilonesinksafewcentrimetres init.
Halite formsacoherentsmoothsaltlayerwhichsealsthesoilandprevents
furtherevaporationofthegroundwater.This layer iseasytoremove fromthe
soil.
Themorphology ofmixed saltassociations dependsupontherelativeproportionsofthesethreemostabundant saltminerals:whenhalitepredominates the
association ismorecoherent,sealinganddenser;whenthenarditepredominates
theassociationislooseandpowdery;whenbloeditepredominates,the association
ismorefluffyandairy.
Thenardite inassociationwithhaliteadapts itselftohalite,butthenarditewithouthalite forms loose,powderyassociations.
Bloeditemakes theassociationmore fluffyandairy.
3.7 METHODS OFWATERANALYSIS
Themethodsofwateranalysiswhichwereusedaredescribed indetail in
Begheyn CI980). Inbriefthefollowingmethodswereused:
pH. Immediately afteropeningthebrownglassbottle,thepHwasmeasuredwith
anOriondigital ionanalyzermodel 801A.
Carbonate.
Thewaterwas filtered througha0.2 \m micropore filter.CO,-and
2-
HCO,-analyseswereperformedbyPotentiometrietitrationwithlUSO. 0.1 N
or 0.05 N andC02(aq)withNaOH 0.1 or 0.05 N.Thesteepestslopesofthe
titrationcurveswere considered tobetheendpoints.
CI .ChlorinewasanalysedbyPotentiometrietitrationwithAgNO,0.02 Nusing
anAg~Selectrode.
2-
SO.. Sulphatewasdetermined turbidimetricallyasBaSO.obtainedwith
BaCl ? .2H 7 0inamediumofnitricacid,aceticacid,orthophosphoric acid,
bariumsulphatesuspensionandamixtureofacaciagumsolutionand acetic
acid.
Na and K . Thealkali ionswereanalysedbymeans ofaPerkinElmerAtomic
AbsorptionSpectrofotometer afteradditionofanaluminiumnitratesolution.
2+
Ca
2+
and Mg . TheearthalkaliionswereanalysedbymeansofaPerkinElmer
AtomicAbsorptionSpectrofotometer afteradditionofhydrochloricacidand
a lanthaniumchloridesolution.
24
HSiO
. Silicawasanalysedcolorimetricallyasblue3-silico-molybdenicacid
afteradditionofanammoniummolybdatesolution,sulfuricacid,tartaric
acidandareductionmixtureofNa o S0,,metolandK-,S„Or-solution.
2 3'
225
Table 3.2 Saltminerals found intheKonyaBasin
Aphtithalite (aph)
K 3 Na(S0 4 ) 2
Arcanite (arc)
K 2 S0 4
Bloedite (bl)
Na 2 Mg(S0 4 ) 2 .4H 2 0
Burkeite (brk)
Na 6 C0 3 (S0 4 ) 2
Calcite (cal)
CaCO
Dolomite (dol)
CaMg(C0 3 ) 2
Epsomite (eps)
MgS0 4 .7H 2 0
Eugsterite (eug)
Na 4 Ca(S0 4 ) 3 .2H 2 0
Glauberite (gl)
Na 2 Ca(S0 4 ) 2
Gypsum (gps)
CaSO,.2H.0
4 2
Halite (hal)
NaCl
Hexahydrite (hxh)
MgS0 4 .6H 2 0
Konyaite (kon)
Na 2 Mg(S0 4 ) 2 .5H 2 0
Loeweite (lw)
Na ]2 Mg 7 (S0 4 ) )3 .15H 2 0
Mirabilite (mb)'^
Na„SO..10H„0
2 4
2
Nesquehonite (nqh)
MgC0 3 .3H 2 0
Schoenite (sch)(picromerite)
K 2 Mg(S0 4 ) 2 .6H 2 0
Sodaniter (sdn)
NaNO
Starkeyite (st) (leonhardtite)
MgS0 4 .4H 2 0
Thenardite (tn)
Na„S0.
2 4
Trona (tr)
NaHCO .Na^CO .2H0
1)Mirabilite couldnotbe identified bymeans ofX-raydiffraction analysis
because transport ofmirabilite proved tobe impossible.Mirabilite isassumed tobepresent fromobservations of thecrystals insitu.
25
Table 3.3 Salt associations in theKonyaBasin
loc.no.
salt association
fabric
along a temporary
pond
bl,hxh,gps
very fluffy,very friablesalt
crust;breaks into small
particles on touching
onthe soil surface
bl,hxh,gps
like 1but alittlemore
coherent
on the soil surface
tn.gl
small salt particlesmixed with
loose soil particles
tn,aph
like4a
bl.hal.gl
sealing,bumpy,coherent salt
crustonhard soil
typeof
sampling site
d:bl,gps
soft,fluffy powder sticking
tosmall,round aggregates
bl,hal,hxh,
sch
hard,coherent,very bumpy
salt crust
on the soil surface
eps,bl,gps
like1
on the soil surface
bl,hxh,gps
like 1,but alittlemore
coherent
small flinters
alongacraterwall
hal,gps,sdn
along acrater lake
hal,tn,nsqh, hard pieces of salt
eug
a dry,temporary lake
hal,gl,gps
sealing,flinterlikecrust
hal,tn,eug
like9a
tn
like4a
10
on the soil surface
tn,aph
loose,powdery,fluffy salt
11
on the soil surface
bl.eug
loose,very soft powder
13
on the soil surface
hal,gps,eug
like9a
tn.eug
like1
14
on the soil surface
hal,gps
like4a
15
on the soil surface
hal,gps,brk
coherent,flinterly,sealing
crust,easily removable from
the surface
hal,tn
like 15a,but alittle thicker
16
spring deposit
hal,tn,eug
hard pieces
gps.cal
ooliticmaterial
17
on thesoil surface
bl.eug
coherent,airy,hard salt
crust
26
Table 3.3 continued
loc.no.
19
typeof
sampling site
fabric
salt association
on thesoil surface a:hal,gl,eug
b: bl,hal,gps
sealing,coherent crust,
difficult toremove from
thesoil
like4d
21
on the soil surface
bl,hxh,gps
like1
22
on thesoil surface
bl,hxh,gps,kon
like1
23
on thesoil surface a:bl,hal,gps
b:bl.hal.gl
24
25
on thesoil surface a:hal,gl
around thespring
ofTiana
26
on thesoil surface
28
on thesoil surface
29
like 1,mixedwith soil
particles
like 19
coherent,sealing,bumpy
salt crust
b:hal,gl,tn,gps,
eug
like 24abutmorebumpy and
airy
hal,tn,nqh,gl,
coherent,sealing,thinsalt
crust
eug
hal,tr,brk
like 15a
openpatches
hal,tn,aph
like4c
between grass
tn.aph
like4a
on thesoil surface a:hal,tn
b:hal,eug
c: tn,aph
soft,coherent saltbetween
straws
like9a
like4a
31
on thesoil surface
hal,tr,brk
bumpy,hard salt crustwith
airunder thebubbles
32
on thesoil surface
hal,tr,brk
like31
33
dry,temporary lake a:tn,hal,eug
like9a
b: tn,eug
coherent,bumpy crustwith
airunder thebubbles
c: tn
like4a
35
on thesoilsurface a:bl,hxh,gps,kon
36
on thesoil surface a:hal,bl
b: bl,gps,lw
b: hal,bl,stk,kon
37
on thesoil surface
hal,bl,gps,kon
loose,very fluffy powder
like35a
sugarlike,coherent crust
like35a
like4d
27
4. SALT-AFFECTED SOILS IN KENYA
4.1PHYSIOGRAPHYANDCLIMATE
Samplesofsaltaffected soilshavebeencollectedfromthreedifferent
regionsinKenya.Theseregionsare(seefig. 4.1):
-TheAmboselibasinalongtheKenya-TanzaniaborderatthefootofKilimanjaro
-AfewlocalitiesalongLakeVictoria
-EastofLakeTurkana (theformerLakeRudolf)andtheChalbiDesert.
Theclimatediffersinthesethreeregions.
TheAmboseliareahasanaridclimatewithanannualrainfallof255-510mm.
Therearetworainyseasons,oneinMarch-April,theotherinNovember-December.
ThedriestmonthsarebetweenJuneandOctober.Themeanannualtemperatureis
22°C.
ThedistricteastofLakeTurkanahasadesertclimate.Rainfallisnormally lessthan255mm/yrapartfromMt.Kulalwhererainfallis255-510mm/yr.
Inthisregionthereisonlyonerainyseason,betweenMarchandMay,andthe
meanannualtemperatureis27 C.
AlongLakeVictoriathereismorerainfallthanintheothertwodistricts.
AtHoma Islandtheannualrainfallis1015-1270mmwhileatSindoandLuandait
is760-1015mm.Therainfallisnotrestrictedtoadefiniteseason;thereare
maximainAprilandDecember.Themeanannualtemperatureis24 C (KenyaAtlas).
4.2GEOLOGY
AllthreeareasaredominatedbythevolcanicsassociatedwiththeEast
African (orGregory)RiftValley.
InKenya,twotypesofvolcanismoccur (Williams,1970):
-fissureandmulti-centreeruptions (plateauvolcanics)
-centralvolcanoes
Thevolcanicsofthe fissure
and multi-centre
eruptions
consistmainlyof
basalts,phonolitesandtrachytes.
Basalticeruptionsoccurred fromtheMioceneonwards.Miocenebasaltscanbe
28
Fig.4.1 Map showing thedistribution of themainvolcanic associations
inKenya (afterWilliams, 1970). The areasof sampling are indicated,
a.AmboseliBasin,b.area alongLakeVictoria,c.areaeast ofLake
Turkana.
29
FISSURE ERUPTIONS OR MULTI-CENTRE ERUPTIONS
Basalt-
Bosolt - phonolite -
bosanite
melonephelinite
Phonolite-
Trachyte -phonoNtic
phonolitic t r a c h y t e
trochyte
Comendltepontellerite
Mugearltetrachyte-rhyolite
QUATERNARY
•
•
•
•
+ + + *
+ + *
•
MAJOR CENTRAL VOLCANOES
NepheKnite - phonolite
Bosolt-trachyte-phonolite
Phonolite-trochyte
Basalt- trochyte
P L I O C E N E TO RECENT
A
A
^
foundinthecentralandnorthernpartsoftherift.ThePliocenebasalts form
mostofthefloorandshouldersoftherift,whichdevelopedatthattime.
DuringtheQuaternary,basaltic eruptionsoccurredtotheeastoftherift
valley.
ThephonolitesareofupperMioceneageandcover largeareasbotheastand
westoftheriftinthesouthernhalfofKenya.
Thetrachytesrange fromupperPliocenetoPleistoceneageandthey form large
sectionsoftheriftfloorinsouthernKenya.
Thevolcanicsofthe central
volcanoes
consistmainlyofnephelinite-
phonolites,phonolites,basaltsandtrachytes.
DuringtheMiocene,largecentralvolcanoes originated inwesternKenya.
Thevolcanicproductsbelongtothenephelinite-phonolite series.Duringthe
PlioceneandQuaternary,majorvolcanic activityofthistypeoccurred along
thefloorofthenewly formedriftandalsobuilt some largevolcanoes (e.g.,
Mt. KenyaandKilimanjaro)whichoccurtotheeastoftherift.
Thepost-Miocenevolcanoes consist alsoofbasalts,trachytesandphonolites.
Fig.4.1showsthedistributionofthemainvolcanic associations inKenyaand
thelocationofthethreeareasvisited.Thegeologyoftheseareaswillbe
dealtwithindetailinthenextthreesections.
30
4.3 AMBOSELI BASIN
ThereaderisreferredtoWilliams (1972)foradetaileddescriptionofthe
regional geologyoftheAmboseliBasin.
ThebasinisborderedtothenorthandwestbymetamorphicPrecambrianrocks
suchasmarbles,schists,gneissesandgranulites.Towardsthesouthandeast
thebasinisborderedbytheKilimanjarovolcanicrocks,whichconsistmainlyof
olivinebasaltsbutnephelitesandphonolites alsooccur.Thesmall eruption
centresinthebasin itselfaresurroundedbyPleistocene sedimentswhichinclude lacustrine clays,marlsandsilts.ThePleistocene sedimentswhichare
knownastheAmboseli LakeBedshavebeensubdivided stratigraphically into
three formationsbyWilliams (1972).Thestratigraphy frombottomtotopis
asfollows (Stoessell, 1978):
- Sinya Beds. TheseareexposedatSinyatothesouthofLakeAmboseli.Three
metresofhard,massivegreensepioliteoccurontopoffivemetresofmassive,
whitedolomite.Pocketsandveinsofmassive,whitesepioliteandpastel
keroliteoccur throughoutthedolomiteandthegreensepiolite.
- Amboseli
clays.
These consistofmassive,green lacustrine clayswitha
thicknessofupto60munderthelakewhichrestuponabasal conglomerate
consistingofdolomiteandsepiolite clast.Theclays consistofcalcite,dolomite,sepiolite,1runclayandK-feldsparwhichisprobably authigenic.Inthe
southernhalfofLakeAmboseli largegaylussite crystalsoccurintheAmboseli
clays,directlyunderthefewcentimetres alluviumwhich formthefloorofthe
lake.
- Old Tukai Beds. These consistofsilts,clays,clay-clastaggregates (metamorphicandvolcanicdetritus)andcaliches,upto25mthick.Themineralsin
theclay-clastareauthigenic.
InmanyplacesQuaternarydepositscanbefound,suchasalluvialsoils,
sandysoils,dustyvolcanic soilsandwindblown clayeysiltsandsands.Asoil
mapofthebasinisavailable,madebyL.Touber,UNDPFAOWildlifeManagement
Project (inpress).
Thefollowing soiltypesoccurintheAmboseliBasin:
Plain soils
Soilsdevelopedonalluvium,predominantly derived fromundifferentiated rocks
ofthebasement system.
Soilsdevelopedonwindblowndeposits,predominantly derived frompyroclastic
rock.
Soilsdevelopedonalluvium,predominantly derived frompyroclasticrock.
Soilsdevelopedonrocksofthebasement system,undifferentiated.
31
Lavaflows
Soilsdevelopedonbasicigneousrocks,richinferromagnesianminerals.
Bottomlands
Soilsdevelopedonalluvium,predominantlyderived fromundifferentiated volcanicrock.
Piedmont
plains
Soilsdevelopedoncolluviumandalluvium,predominantlyderived fromcrystallinelimestone.
ExternalSolonchakscanbefoundontheplainsoils.Salt efflorescences
occurespeciallyonthefloorofLakeAmboseli,atemporary lake.Forthe
Fig. 4.2 Map showing theAmboseli Basin (after Stoessell, 1978)and the
locationof the sampling sites.Theregional surface-and groundwaterflowpattern isindicated byarrows.
locationofthesampling sites,seefig.4.2.
From theKilimanjaro,moltensnowandrainwater flowassurface streams
andsubsurfacewater inthedirectionofthebasin.Most streams aredryat
thefoothills.WateralsoflowsfromthePrecambrianrocks inthenorthtothe
basinbut theregionswhere thesaline soilsoccurarefedbywater fromthe
Kilimanjaro.
4.4 LAKE VICTORIA
Sampleshavebeentakenfromdifferent localities onHomaPeninsulaand
fromthesurroundings ofthevillages SindoandLuanda inthecoastalareaof
LakeVictoria (seefig.4.3).Thegeologyofthisareaisdescribed inSaggerson
(1952).
HomaPeninsula isdominatedbyHomaMountain,which isacentralvolcano.
ItsactivitybeganintheMiocene and lasteduntil latePleistocene times.Homa
Mountainisatypicalcarbonatitic complexwithcarbonatite,nepheliniteand
phonolite.On thelowerslopesandonHomaPeninsula thevolcanics arecovered
withPleistocene andrecentsediments.Thesedimentarydeposits thuspartly
formedwhenthevolcanowasstillactive.ThePleistocenedeposits are lacustrine
andconsist ofanalternationoftuffs,clays,sands,gravelsand thinlimestones.
Sampleshavebeentakenfromthreedifferent localities onHomaPeninsula:Lake
Simbi,MitiBiliandKanam.LakeSimbi,avolcanicmaar,originated inhistorical
times.Theexplosionthatproduced theventprobablybrokethroughtheoldcourse
oftheRiverAwach,whichnowflowseastofthelake.LakeSimbi isfedby
groundwater,springs andrains.Saltshaveformedalong theshoreofthelake.
AtMitiBili,ahotspringoccursonlyafewhundredmetres fromLakeVictoria.
Surface andsubsurfacewaterhasformedexternal solonchaks betweenthespring
andthelake.
AtKanamsaltefflorescenceshave formedalongadryriverbed.
ThehinterlandofSindoandLuandaisformedbytheKisingiriVolcanic
complex,oneofthebestexamplesofaMiocenenephelinite-phonolite central
volcano inwesternKenya.Theplainalongtheshoreofthelakeconsistsof
Pleistocene clays,sandsandgravels.Onthisplain,salt efflorescences
developed inthevicinityofthesmallvillages SindoandLuanda.
33
Kendu
Miti Bill'
Kanam
Rusinga
? :....."•'Mt.Homa \
Island
..-'""'
1
Lake
\
Victoria
/5
J Mi t i
i Bili.
4/
|
1
1
HomaBay
\3
'l
1
,V3 -
.
2J
-.>...\ .....
10km
Fig. 4.3 Map showing the localities of the sampling sites along LakeVictoria.
4.5 EASTOFLAKE TURKANA AND THECHALBI DESERT
Little isknown about the geologyofthe area eastofLake Turkana(the
former Lake Rudolf)ando fthe geology ofthe Chalbi Desert.A soilmap
1:1.000.000ofthis whole district isavailable (Sombroek,inpress).
The Chalbi Desert consists ofCenozoic lacustrine sediments. This desert
is barrenofvegetation. Extensive External Solonchaks cover the Chalbi Desert.
Normally, this salt plain isaccessibleb ymotor vehicle,but during the sampling trip (November 1977), the plain was flooded duetounusually heavy rainfall. Rainwater together with water draining from Ethiopia caused the desert
to becomeatemporary lakeandsampleso fthe salts could onlyb etakenata
few places where the rimsofthe desert couldbereached (at Kalacha, Maikona
and North Horr) (see fig. 4 . 4 ) .Thesoilsofthe Chalbi Desert are classified
as Orthic Solonchaks (very poorly drained, deep, firm, excessively saline clay
1
1
34
|
20 km
North Horr
/
I
\
\
\
\
/
/\s'
/
J
S
/
^*
,.—*
f/
El Molo
\KMount
l
. !Kul al
Loien- \
galani /
-~^V,Kargi
i
i
\ /
\ ^
/
/
/
—i
Fig. A.4 Map showing the localities of thesampling sites intheregioneast
ofLakeTurkana and theChalbiDesert.Thedotted line indicates atrackfor
vehicles.
withpuffed saltycrust)andTakyric Solonchaks (poorlydrained,deep,very
firm,excessively salinecracking clay).
Saltefflorescences occuratsomesitesalongtheshoreofLakeTurkana.
Thisdistrict consistsmainlyofQuaternaryolivinebasaltsofthefissureand
multicentre type.Theregionwherethesaltefflorescences occur,nearElMolo
andLoiengalani (seefig.4.4),isdominatedbyMt.Kulal,aPliocene basaltic
centralvolcano.Thesaltoccurrencesarerelatedtothedrainageofthewater
flowingfromMt.KulaltowardsLakeTurkana.ThesoilsalongtheshoreofLake
TurkanahavebeenclassifiedasSolonchaks gravellyphase (imperfectly drained,
deep ,moderately calcareous,stronglysaline,gravelly clay loam,with gravelly
desertpavement).
35
4.6 SAMPLINGANDANALYTICAL PROCEDURE
Thesamplingandtheanalyticalprocedurewerethesameasdescribed insection3.5.
Unfortunately thesamplingprogramwaspartiallydisturbedbyunusually
heavyrainfall.TheAmboseliBasinand theregionalongLakeVictoriawere
visitedbefore therainsstarted.Allofthewater samples fromAmboseliand
somewater samplesfromlocalitiesnearLakeVictoriaweretakenbeforethe
rainsbegan.Bothdistrictswerevisited againafter therains.LakeAmboseli
wasno longeraccessiblebymotorvehicleandmostofthesalthaddisappeared.
AlongLakeVictoriathesalthadalsodisappeared andthereforeonlysome
additionalwatersampleswere takenaftertherains.Thefirstsampling trip
toLakeTurkanawasunsuccessfulbecausetheexceptionallyheavyrainfall
caused thetrackstobecome impassable.Thisregionwasonlyvisited after
therains,whichhadmadethedesertgreenandfullofflowers.
Table 4.1 showsthewateranalyses.Table 4.2 showsthesaltminerals
whichhavebeenfound. Intable4.3 thesaltassociations atthedifferent
localitiesarepresented.
4.7 CORRELATION BETWEEN THEFABRIC ANDMINERALOGY OFTHESALTASSOCIATIONS
Themostabundantsalts intheareasvisited inKenyaarehalite,thenardite,tronaandthermonatrite.Thesemineralsdominate themorphology ofthe
saltefflorescences onthesoilsurface (seetable4.3).
Thenarditeandhaliteshowthesamemorphological featuresasinKonya.
Thethenardite needlespushthesoilparticles apartandcausetheupper layer
ofthesoiltobecomepuffed.Halitehas asealingeffect.Itmakes thecrust
moredenseandcoherent.
Allassociationswiththermonatrite areveryfluffy.When itoccurs in
crusts,thesecrustsareveryfriableandbreakeasily intofluffypieces.
Enoughporesandchannelsarepresenttoavoid sealingoftheunderlying soil.
Thecrusts inwhichtronaoccursareusuallymorerigid andcoherent than
thoseinwhichthermonatriteoccurs.Thesurfacesofthetronacontaining crusts
areveryirregularandusuallytheyareveryairywithmanyvoids and channels
asisthecasewiththermonatrite.
36
- <j\ m ci
— — m (M oo \D
• er. o
m m m o (N <r
o
co co r » co c
•<T
r - \ o i—. co c — CM
moinoocr.r^-a' —
j o - J r-. - * co
o d
o cr> o
1 c \ ~* o CM r-
•— —
; —o o O O
•—o — —o -
r»
o
<£> r~~. -3 m f
1 — O i N CN M O " 1
—o o o
—
• \DO O O
OOCNC
CM CM —
) CO CM UO -3" Ol O
o o o
•o o o o o o o
O co
I CM ~ co vO CO u*ï —
M M N N tM -
00 CM CO — C
uO O CM O •
O O O O O O O O O O -
o o o o o o o o o o o o o -
m oio
(nOM
00o —
odd
u"ï u i
-T M3
m o u-i <r CM —
r - — co co i£> r— <r
— rM o v c j \ CMr - o i
CM\O — c M O O O r -
a i - M 3 n o O
O O CM CM <r v r v t
M O O f l M ^ N 4
cor— CM
O >ûM N Ci- CM-J-— — O COOiror— o-ir--\D~j-incNcoco
ror--o>— O
N M <tm m fi
co CM coro m
CMco— (
cocooNr-r-r-*r-.r-
œ N h. isoi (
r-cocoo^coco
> r—o co
. o i o i o i o i o i o i t o c o o i r
.OlOlco
O O O O O O O O
o c N O c o o m o c M
o-j-cou-icocMOm
o o o o o o
o CM o o co —
O O O O M3 r-~
O O O O O O O O O O O O O O O
O O O O O O O O C 0 C 0 O O O O O
oiOOr--r--Or--or-« — cocococoo
n t O O U C N M / l O N
— \ß COI— O
M l-i l-i
tl U tl U M
(0 tO tO nj 4-J
VJ
JJ J-J 4J 3 r~( 4-1 4-1 4-1 -U 4-1
3 3 3 3 < u o a 3 3 3 3 3
eo-o O O C J O - O - Ö - O T 3
c d c c c d ö d
c a c c a
3 3 3 3 3
- ^ 3 . H .H 3 3 3 3 <•
O O O O O
p - h ft ftIJ I o o o o
00 00 00 00 W
60 00 M OD M
333 32
3 3 T 3 3 3 U 3 0
o o c o o > o >
M M a W) M IJ ÖO IJ
O O O O > IJ
M UI l-i M •I-I o .
00 00 00 Û0 U tn
lj
• CM CO -T - j in o
1 vD CO •
•-
I IIIII4II
to
C
O
^
O
—l
n) (0 X i—i n)
J - -C
3 M
U u J- m ö
<g u u u v
2 t 4 Ü Z fe J
pq w co pa PQ c
.O £ -H -H .H -H -H -H Ëg4-l4-14J4-14-l4J-
• H - H PO
.J
e e a xi o
41 n] tl) « d TJ
^ e c G to ö
W C O S S S S S S S J ^ W :
37
Table 4.2 Saltminerals observed in theKenya sample sites
Amboseli
aphtithalite (aph)
K 3 Na(S0 4 ) 2
aragonite (ar)
CaCO
burkeite (brk)
Na 6 C0 3 (S0 4 ) 2
calcite (cal)
CaCCL
dolomite (dol)
CaMg(C0 3 ) 2
gaylussite (gay)
NaCO .CaCO .5H„0
pirssonite (prs)
NaCO .CaCO .2H20
thermonatrite (tm)
Na 2 C0 3 .H 2 0
thenardite (tn)
Na 2 S0 4
trona (tr)
NaHCO .Na2C0 .2H20
Turkana
burkeite (brk)
Na 6 C0 3 (S0 4 ) 2
eugsterite (eug)
Na 4 Ca(S0 4 ) 3 .2H 2 0
gypsum (gps)
CaSO,.2H„0
4 2
halite (hal)
NaCl
thenardite (tn)
Na„S0,
I 4
trona (tr)
NaHCO .Na 2 C0.2H 2 0
Lake
Victoria
burkeite (brk)
Na 6 C0 3 (S0 4 ) 2
calcite (cal)
CaCO^
eugsterite (eug)
Na.Ca(S0.)Q.2H.0
4
4J 2
gaylussite (gay)
Na„CO .CaCO .5H0
gypsum (gps)
CaSO..2H.0
4 2
halite (hal)
NaCl
nahcolite (nah)
NaHCO,
thermonatrite (tm)
Na 2 C0 3 .H 2 0
thenardite (tn)
Na 2 S0 4
trona (tr)
NaHCO .Na 2 C0.2H 2 0
38
Table4.3 Salt associations observed inKenya
loc.no.
salt association
typeof
sampling site
fabric
Amboseli
1
againstwallof
profile pit,20cm
a: tn,tr,aph
thin,slightlybumpy,rather
flat and friable coherent
salt crust
ditto, 70cm
b: tr,cal,gay
yellowish powder
tr,cal,gay
on thesoil surface
like la
on the soil surface
a:tr
soft,fluffy,coherent salt
crust
onthe soil surface
b: tm,tr,cal
like4a
onthe soil surface
c:tr,cal,prs,
tm
onthe soil surface
d: tr,tm,gay
flat,thin,coherent salt
crust
fluffy,small,loosesalt
particlesmixed withother
soilparticles
alongpond
a: tr,cal,gay,
tm
like4a
alongpond
b: cal,ar
salt peels
onthe soil
surface
tr,cal,gay,
tm,aph
very fluffy,coherent salt,
breaks intopieces on touching
on the soil
surface
tr,brk
coherent,puffed crust,
irregular surface
onthesoil
surface
a: tr,tm,gay
very loose,fluffy,irregular
shaped salt particles
on thesoil
surface
b: tr,aph
like 8a,butmore coherent and
littlemore yellowish
on thesoil
surface
c: tr,aph,gay,
prs
coherent,thin,irregular,
dense salt crust
9
on thesoil
surface
12
onthesoil
surface
on the soil
surface
tr,prs,cal
a: tr,brk
slightly fluffy,coherent,
soft crust
very irregular,airy,coherent
salt crust
like 12abutmore friable
b: tr,tm,brk
c:tr
tr,aph,gay,
cal
like4a
fluffy,friable,coherent
salt crust
13
onthe soil surface
near swamp
15
onthe soil surface
16
on the soil surface
a:tn
thin,dense,bulged salt crust
on the soil surface
b: tn
like 15
17
on the soil surface
tn
tn,aph
white salt fluffs between
clay particles
like 12a
39
Table 4.3 continued
loc.no.
typeof
sampling site
salt association
fabric
Turkana
Maikona
on thesoil surface
a:hal.tn
thin salt layer deposited
betweenbasaltic gravel onthe
soil,not removable asacrust
from the soil
b:hal,gps
Kalacha 1
onthe soil surface
hal.tn.eug
Kalacha 2
onthe soil surface
hal,tn
powdery salt onquartz pieces
glassy salt layeron the soil
surface
NorthHorr onthe soil surface
Furaful
ElMolo
Oasis
hal,tr,brk
on the soil surface
Simbi
thin,flinterly,dense,smooth
salt crust
thin,coherent,easyremovable,
rough salt crust
hal,tn
on thesoil surface
tn
on the soil surface
very irregular,thin,coherent,
puffed salt crust
a: tn,hal
on the soil surface
Lake
likeKalacha1
flinterly,smooth,white salt
crust
b:tn,hal,brk
likea,but alittlemore glassy
and yellowish
Victoria
on thesoil surface
near lake
a: tr,tm
ditto
very fluffy saltparticlesmixed
withclay
b:tm.tr,gay
Miti Bili 1around spring
likea
tr,cal,gay, thin,bumpy,hard salt layeron
nah
MitiBili 2 on thesoil surface
a: tr
onthe soil surface
surface
dark coloured,hard salt layer
b:tr,tm
airy,white,irregular salt
crust
Miti Bili 3on thesoil surface
tr,gay
MitiBili 4 onthe soil surface
a:tr,tm,hal,
thin,white,smooth salt film
very fluffy salt crust
brk
b: tr,hal
Kanam
on the soil surface
likeMiti Bili2b
c:tr,nah,gay
thin,flinterly film
d: tr,gay
bumpy, irregular saltcrust
e:tr,hal,brk
airy,irregular salt crust
a:tr,nah
thin,flintery crustwitha
rough surface
b: tr,tm,gay
powdery salt oncoarse gravel
c: tr,hal,gay
d: tr,hal,tm,
gay
e: tr,gay
puffed,rough crust
likec
likeMiti Bili4d
40
Table 4.3 Continued
loc.no.
Sindo
typeof
sampling site
onthe soil surface
salt association
a:hal,tn
b: gps
c:hal.eug
Luanda
on thesoil surface
eug,gps
fabric
salt filmonclay
white spots
likea
white spots indark soil
41
5. BRINE EVOLUTION IN CLOSED BASINS
5.1 INTRODUCTION
Theoccurrenceofsalt-affected soilsisdependentbothongeologicaland
climaticfactors.Salt-affected soilsformeasilyinsemi-aridregionswhere
evaporationishighandusuallyexceedsrainfall.
Theoriginofsalt-affected soilsmaybeanthropogenicornatural.Anthropogenic
salt-affected soilsoriginatewhenduring irrigationnoconcomitantdrainageis
appliedtoremovethedissolved salts,aswasdiscussedinchapter2.Natural
salt-affected soilspreferentiallyoccurinhydrographically closedbasins.
Geologicallyspeaking,twodifferent typesofclosedbasinscanbedistinguished (EugsterandSurdam 1973,SurdamandSheppard 1978)whichshowacompletelydifferenthydrology.Bothtypeshaveatectonicorigin:
a.Playa-lakecomplextype (seefig.5.1)
A Playa-lakecomplexisabroad flatbasinsurroundedbymountains.Rivers descending intothebasinfromthemountainshaveusuallydisappearedbefore
reachingthecenterofthebasin,whichmayconsistofaperennialorannual
lake.Thegroundwater flowisshallow.Most clasticdebrisisdepositedin
alluvial fansatthefootofthemountains.
b.Rift-systemtype (seefig.5.1)
Inarift-system closedbasinanannualorperennial lakeoccupiesthefloor
ofanarrowflatvalleywhichisborderedbysteepmountains.Thegroundwater
flowisdeepandusuallydischarge takesplaceattheedgeofthelakeby
springs.Asspringsaredevoidofclasticmaterialnoalluvial fansoriginate.
TheKonyaBasininTurkeyisatypicalexampleofaplaya-lakecomplextype
(comparefig.5.1 andfig.3.3).LakeMagadiwhichislocatedintheEastAfrican
riftvalleyinKenyaisatypicalexampleofarift-systemtype.
Thegenesisofsaltaffected soilsinclosedbasinsisasfollows:
Dissolvedmaterialisbroughtinbyrivers,groundwatersandsprings fromthe
surroundings intothebasin.Duringevaporation,theremaining solutionbecomes
moreandmore concentrateduntilitbecomes saturatedwithrespecttosome
minerals,whichthenprecipitate.Itdependsonthepositionofthewaterwhere
42
A
PLAYA-LAKE COMPLEX
SANDSTONE
ALLUVIAL FANS
IGNEOUS AND
METAMORPHIC
ROCKS
(ALKALINE 1EARTH CARBONATES)
Fig. 5.1 Hydrologyandbrine evolutionin
aplaya-lake complex (A)anda
rift system (B).Solid arrows,rainwater;GW,groundwater circulation path,
(fromSurdam&Sheppard, 1978)
thesaltwillaccumulateifthewaterevaporatesassurface-water,forexample
thewaterofatemporaryorpermanentlakeinthecenterofthebasin,whitesalt
rimsformaroundthelake.Mostofthetime,however,thewater evaporates
as
groundwaterwithinthesoil.Mineralswillprecipitateatdifferent levelsinthe
soilprofiledependingontheirsolubility.Thelesssolublemineralswillprecipitateinahorizonsomewhereinthesoilprofile (internalSolonchak)andthe
more solublemineralswillaccumulateatthesoilsurface (externalSolonchak).
5.2 CHEMICAL EVOLUTION
Thechemicalevolutionofwatersinclosedbasinscanbeseparated intothree
mainphases:
-acquisitionofsolutes
Dilutewatersacquiresolutesbyweatheringreactionswith soilsandbedrock
(GarrelsandMcKenzie 1967).
-evaporation
Thedilutewatersevaporate,whichcausesconcentrationofthedissolved
material leadingtoprecipitationofminerals (EugsterandHardie1978,
HardieandEugster 1970).
43
- re-solution
Partial re-solutionofsaltsdeposited onthesoilsurface takesplace.This
influences thecomposition ofthegroundwater (Drever andSmith 1977;Jones,
Eugster andRettig, 1977).
5.3 ACQUISITION OF SOLUTES
Chemicalweathering ofprimary silicatesbyCO^-chargedwaters causes the
solution tobecomeenrichedwithcations,silicaandbicarbonate.Theweathering
reactioncanberepresentedby (GarrelsandMcKenzie, 1967):
primary silicate+ILO+CO,-*clayminerals +cations +HCO~+H.SiO° (aq)
(1)
During weathering of igneousandmetamorphicrocks,HCO-,willbethepredominant
anioninthesolution. Weathering ofpreexisting evaporites canproduce solutions
2*
inwhichCI andSO. arethepredominant anions.Theybothmayalsobe essential
constituents ofrainwater. Inaddition,chlorineoftenoriginates fromfluidinclusions inminerals,andsulphatemay enter solutionbyoxidationofsulfides,
forexamplebyweathering ofpyrite.
5.4 EVAPORATION
Dilutewatersofdifferent compositionbehave indifferentwaysduringevaporation.Thedirection inwhichthesolutionwill evolve iscompletely determined
bytheratiooftheionspresent inthedilutewaters after thestageofacquisition.Thisprinciple canbedemonstrated withasimple example:the precipitation
ofNaCl (Al-Droubi, 1976).A fictivesolutioncontainingonlyNa andCI ions
willbeconsidered.Undernatural circumstances suchafictive solutionwillnever
occur,because carbonate-ions arealwayspresent inwaters inequilibriumwith an
atmosphere containing C0 ? . Withcarbonate-ionswehave thecomplication thatthey
areinvolved inthefollowing equilibria (Stumm&Morgan, 1970):
(thevaluesoftheK's areat25°C,H 2 C0 3 =H 2 C0 3 +C02(aq))
C0, r -. +H,0ÎH,C0,
2(g)
2
H 2 C0 3 JH +HC0 3
K= -„
2 3
*
(H?C0,)
= 10
-1,47
(2)
PCQ2
K1 =
(H+)(HC0")
3H±= 10
(H2C03)
-6.35
(3)
44
(H+)(CO^)
Hco" t H ++co::"
-10.33
(4)
10
(HCO~)
K
H20 XH +OH
_(H)(OH)
w
H20
•14
10
InafictivesolutioncontainingonlyNa andCl ions,theseionsmustbe
presentinequalamountstomaintainelectroneutrality.Duringevaporationthe
concentrationsofNa andCI willincreasebythesamefactoruntilthesolution
becomessaturatedwithrespecttoNaCl.Thesaltwillprecipitate,Na andCI
willbetakenfromthesolutioninequalamountsandduringfurtherevaporation
thecompositionofthesolutioninequilibriumwithhaliteremainsconstant
(point1infig.5.2).IfinadditiontoNa,K isalsopresentasacation,
theelectroneutralityrestrictionimplies[Na]+[K]=[CI]and[CI]/[Na]>1
(squarebracketsdenotingmolarconcentrationsandroundbracketsdenotingactivities).Duringevaporation[CI]/[Na ]remainsconstant,theconcentrations
following,forexample,line1 infig.5.2untilthesolutionbecomessaturated
withrespecttoNaCl.AtthisstageanequalamountofNa andCI willbetaken
fromthesolutionbutduringfurtherevaporationtheconcentrationofchlorine
willincreaseandas(Na).(CI)=constant,theconcentrationofNa will
log(Na+)
4-
(Na>(CO
2
0
-2
~l
-4
1~"
-2
~!
1
r-
0
2
4
log(Cl)
Fig.5.2 EvolutionduringevaporationofasolutioncontainingNa-andCI-ions
incaseintheinitialsolution:1.[Na+]=[CI];la.[CI]>[Na+]and
lb.[Na+]>[Cl"].
(5)
45
decrease.Theconcentrations followthelinedescribingthesolubilityproduct
ofNaCl (direction2infig. 5.2).Ontheotherhand,whenforexampleSO. is
+
+
presentinadditiontoNa andCI,[CI]/[Na ]<1andthesolutionwillevolve
intheoppositedirection (direction3infig.5.2).
?-
Duringtheprecipitationofcarbonates,analogousrestrictionstothosegoverningtheprecipitationofNaClmustbevalid.
If,duringevaporation,calciteprecipitates (HardieandEugster,1970):
2+
21.Ca andCO, mustbelostfromthesolutioninequalmolarproportions;
2+ 22.Theactivityproductofcalcite((Ca)(CO,))mustremainconstantatconstantPandT.
InasolutioninequilibriumwithatmosphericCO-containingonlyCa ,H,CO,,
HCO,,andOH thefollowingelectroneutrality equationisvalid:
2[Ca 2+ ]=2[C02~]+[HC0~]+[0H~]-[H+]
(6)
[H],[HC0~]and[0H~]canbeexpressedintermsof[CO,-]asfollows:
V
[H+] -K l K 2 P <
T C 0 2- [Cof]
%:o-
(7)
[HC0 ]
3 - r2 •Pco2 • K- T a f rœfh
(8 )
T 0 H - [OH"]= Ü*
(9)
T H +[H]
Intheseequationsy.meanstheactivitycoefficientofioni.Thefollowing
relationexistsbetweenactivityandconcentration:
y- .concentration.=activity, ory-[i]=(i)
(10)
Substituting intoequations7to9thevaluesfortheK'sat25Candp m =10 ,
+
:itutingtheobtainedresults intoequation6,
neglecting[H]andsubstituting theobtainedresults intoequation6,thefollowingequationresults:
y2 2Y 2 2u
J- -31. 1 7 5
U
J4 9 5 ^ 3
2+,
™
2
,
.
,
,
,
„
3
2-A
^
,
0
.
r
rnr
n
[Ca' ] = [CO'~]+ 1(10
. — - ^ - [CO' V + 10 * ^ . — ^ - [CO'rnn] 2Z)
«-*
TTA T T
OH"
3
-J
YITtr<n,
HCO,
C0 9
LU
2
(11)
^
46
or, m activities:
(Ça2^) < ^ P
T
Ca2+
T
io-°- 495 (co 2 -)^
103 J 7 S (C02~)a
+
1(-
CQ2-
'OH
+
3
T,HCOv
}
(12)
Inthedilutestage,activitycoefficientsmaybeneglectedwhichyields:
[Ca2+]=[CO2-]+10°- 7 9 5 [C0 2 "F
(13)
ThisequationrepresentstherelationbetweentheCa -andtheCO, -concentrationsinthesystemCaO-CO^-H-,0duringevaporationunderatmosphericconditionsat25°C (line1onfig.5.3).Whenthesystembecomessaturatedwith
respecttocalcite,calciteprecipitatesandthecompositionofthesolution
remainsconstantatacompositionindicatedbypoint1'infig.5.3.IfNa,K,
Mg, CI,andSO.,whicharenormalconstituentsofnaturalwaters,arepresentin
log(Ca2+)
A<0
-4
-5
log(C0|")
Fig. 5.3 Evolution of the systemCaO-CO-HOduring evaporation incase
0, Ie1.A> 0and 1 D .A<0.
47
additiontotheionsmentioned inequation6,theelectroneutrality equation
becomes:
[Ca 2+ ]= [CO^"]+i([HCO:]+[OH"]-[H+])+i([Cl"]+2[S02"])-i([Na+]+[K+]+2[Mg2+]) (14)
v _ ^
^
/ v
*
^
y
A
[C02~]+ 10~ 0,795 [CO 2 ~:P
atsmall I (=ionic strength)
2+
2IfA iszerotherelationbetweentheCa -andCO,-concentrations during
evaporation canagainbedescribed by line 1.Whentheconcentrations ofthe
other ionsbecome toolarge,theactivity coefficientsmaybecome important
which influences thepositionofcurve 1infig. 5.3,ascanbeseenfrom
equation(11).
IfA ispositive, [Ca ]isinexcess overcarbonate andduringevaporation the
2+
2Ca -andCO,-concentrations aredescribed byacurveontheCa-richsideof
curve 1 (forexamplecurve 1) .
a
b
Fig. 5.3 suggests thatcurves 1,1 and 1 areparallel,but asthe absolute
valueofA increases duringevaporation,these curves arenotparallelbut their
verticalseparation increases.
Whenthesolutionbecomes saturatedwithrespect tocalcite,Ca
2+
andCO, will
beremovedfromthesolutioninequalmolar amounts andthecompositionofthe
solutionwill changeaccording tothelinerepresenting thesolubility product
ofcalcite indirection 2 (fig.5.3).
WhenA isnegative,thecompositionofthesolutionwill changeinthe opposite
direction(3).
Theprecipitation of aaleite
is the first
decisive
step intheevolutionof
abrine (HardieandEugster, 1970;Eugster andHardie, 1978).Thesolutionbecomeseither carbonate-rich orcarbonate-poor,depending ontheratioofcalciumtocarbonate intheinitial solution,which iscorrelated tothesignof
A of(14).
Thenext cationwhichusuallywillberemoved fromthesolutionisMg
Nesbitt (1974)found that inwaterswith intermediate (Ca 2+ +Mg2+)/(HCO~)
ratiosmore andmoreMgwillbe incorporated inthecalcite asMg isenriched
inthesolutionrelative toCaasaresultofcalciteprecipitation until
finallyprotodolomite andmagnesitemaybe formed.Gacetal (1977)have shown
experimentally that intheLakeChadenvironmentMgandSicoprecipitate which
finally leads totheformationofMg-richmontmorillonite.Hardie (1968)found
2-
48
coprecipitationofMgandSiintheformofsepioliteatSalineValley,California.
Letusconsiderwhathappens theoretically toasolution if,duringevaporation,calciteandsepiolite coprecipitate.Mgprecipitatesassepiolite accordingto(GarrelsandMcKenzie, 1967):
2Mg 2+ + 3Si0 2 (aq) + (n+2)H20 * Mg^ijOg.nh^O + 4H
(15)
log(Si02(aq)
i
SiO
log(Ca
sepiolite
CaO
MgO
calci te
Fig. 5.4a MgO-SiO„-CaO-diagram.During coprecipitation of sepiolite and calcite fromasolutionwith compositionP, thecomposition of the resulting
solutionmoves somewhere intheshadedarea.
Fig. 5.4b Activities of components pivot around join 1-2 during coprecipitationof calcite and sepiolite.
Thecompositionofthesolutioncanberepresented inaMgO-Si02-CaO-diagram
(fig. 5.4 a ).Theactivitiesofthecomponentscanbeplotted alongaxesperpendiculartotheplaneofcomposition (fig.5.4) .Fromtherelation
Ca 2 + +H 7 0 $ CaO+2H +itfollows that log(CaO)isanalogoustolog(Ca+ )+2pH
9-*.
+
and fromMg +H 2 0$MgO+2H itfollows that log(MgO)isanalogousto
CaO
calcite
+
)
49
log(Mg2+)+2pH.
During calciteprecipitation atconstantPp n :
log(Ca2+)+ 2pH=constant
(point 1infig.5.4 )
(16)
which follows fromthereactionequationCa +C0 ? +ILOÎCaCO„+ 2H.
During sepioliteprecipitation:
31og(Si0 2(aq) )+2(log(Mg2+)+2pH)=constant
(17)
because thesolubility product ofsepiolite (18)isconstant.
(HV
K =•
=constant
^2+)2(S02(aa/
Therefore,point 2infig.5.4 isfixed.
Ifcalcite andsepiolite coprecipitate fromasolutionwith compositionP,the
composition oftheremaining solutionplots somewhere inthesegment formedby
thejoins ofcalciteandtheoriginal solution andofsepiolite andthat solution
(shadedarea infig. 5.4 a ). Theactivities ofthecomponents insolutionare
represented bythedissection from theaxesbyaplanewhichpivots around the
linebetweenpoint 1and 2infig.5.4 .During sepioliteprecipitation,Mgand
Si areremoved fromthesolution inamolar ratioof2/3.Ifatthemomentsepiolitestarts toprecipitateMg/Si> 2/3,whichmeans thatthecomposition ofthe
solutionplots intheCaO-MgO -sepiolite triangleoffig.5.4 ,thisratioincreaseswithfurtherprecipitation. Conversely,whenthecompositionplots in
theCaO-SiO?-sepiolitetriangle,theMg/Siratiodecreases during precipitation
ofsepiolite.
Inthefollowingdiscussionwhatmayhappen toasolution ifsepiolite starts
tocoprecipitatewithcalcitewillbedealtwithsystematically. Itisassumed
thatatthemoment sepiolite andcalcitestart tocoprecipitate:
[Mg]/ [Si]> 2/3.During furtherevaporation andcoprecipitation thisratioincreases.
- If (SiO?(. O isconstant,itfollows that
log(Mg )+2pH=constant
61og(Mg2+)=-26pHI
6
o.
I_^
61og(Ca )=-26pHf
,, rr 2+. ,. n, 2+,
61og(Ca )=61og(Mg ) .
(18)
50
As (Si02(. J isconstant,andMg/Simust increase,itfollows thatMgmust
increaseandthereforeCaincreases alsoandpHdecreases.
-If(SiO,, ,)increases,itfollowsthat
log(Mg )+2pHdecreases.Thisispossible onlyif
a.2pHdecreasesmorethanlog(Mg )increases•>6log(Ca )>Slog(Mg ) both+
b.2pHincreases lessthanlog(Mg )decreases-+[<5log(Mg )|>6log(Ca ) Mg+Ca
c.2pHdecreasesandlog(Mg )decreases•+log(Mg )-I- log(Ca )t
d.2pHdecreasesandlog(Mg )remainsconstant•+log(Mg )constant log(Ca )+
e.2pHremainsconstantandlog(Mg )increases->-log(Mg )+
log(Ca )const.
BecauseMg/Simust increaseand(Si02(- •>)increases,Mgmustincreasetoo,so
onlycasea.fulfillstherequirements.
-If(SiCu, -.)decreases,itfollows that
log(Mg )+2pHincreases.Thisispossibleif
a.2pHdecreases lessthanlog(Mg )increases-+61og(Mg )>61og(Ca ) botht
b.2pHincreasesmorethanlog(Mg )decreases-+|Slog(Ca )|>|<51og(Mg ) | both+
c.2pHincreasesandlog(Mg )increases •*• log(Ca )+
log(Mg )t
d.2pHremains constantandlog(Mg )increases•>log(Ca )=const.
log(Mg )+
e.2pHincreasesandlog(Mg )remainsconstant -*log(Ca )+ log(Mg )=const.
BecauseMg/Simust increaseand(Si0 9r ,)decreases,
casesa-eareallvalid.
L aa
l
J 7+
However, case b. is only valid if |log(Mg )|<2/31og(Si02(-
,).
Thesametypeofreasoningashasbeenappliedfor[Mg]/[Si]>2/3,canbeapplied
forthecases [Mg]/[Si]=2/3andfor[Mg]/[Si]<2/3.
Theresultsaresummarizedintable5.1.Asonlydirectionsinwhichthesolution
evolvesareindicatedandabsoluteconcentrationsarenotgiven,activitycoefficientscanbedisregarded.
Thefollowing symbolsareusedinthetable:
Ca=log(Ca 2+ )
Mg=log(Mg 2+ )
Si=log(Si0 2(aq) )
J =concentration increases
1 =concentration decreases
- =concentrationremains constant
Thecompositionofthewateratthemoment sepiolite startstocoprecipitate
with calciteplots:
^ intheMgO-CaO-sepiolitetriangle
A intheCaO-SiCu-sepiolite triangle
/^onthelinejoiningcalciteandsepiolite
51
Table5.1
A
t
t
1
t
t
1
óMg=óCa
6Mg<6Ca
6Mg>6Ca
16MgI>|6Ca1
—
Mg
Ca
pH
1 1
t
l
1
1
Si
\
\
t
1
t
!
\
—
t
t
—
i
—
1
t
|6Mg|<2/3|6Si|
A
t
t
t
1
1
I
I
t
óMg<óCa
ÓMg<1ÓCa1
óMg=2/3óSi
|óMgl=2/3lóSil
Si
Mg
Ca
pH
t
Si
Mg
Ca
pH
t
\
\
\
\
t
t
t
t
\
t
t
t
\
—
t
1
t
\
—
1
1
1
t
\
lóMgl<|6Cal
ôMg=6Ca |óMgl>óCa6Mg<6Ca
Mg<2/36Si
Theproblemcanalsobeapproached fromthesideoftheelectroneutralityequation,rewritten from(14):
2
2[Ca2+]+2[Mg2+]=[HCOl]+2[C0
M O H _ ] - [ H + ]+[Cl'>2[S02L"]-[Na+]-[K+] (19)
ä
\
£,
alkalinity
/ v__
/
.it
A'
52
Fromthecarbonate equilibria (equation 2to 5)anintimate relationbetween
pH,Pco ? andalkalinity canbederived (vanBeekandvanBreemen, 1973).At constantPCO?a n increaseofalkalinity impliesanincreaseofpHandviceversa.
When,during evaporation,calcite andsepiolite coprecipitate,A'remains constantly zeroortheabsolutevalueofA' increases.Forthe caseA'remains
zerowecanderivethe followingrelations:
a.alkalinity remains constant
-*•pHconstant andbecause log(Ca )+ 2pH=constant-*• (Ca )isconstant.
From (19)itfollowsthat theMg-concentrationmust alsoremainconstant.
Thiscombination isonlypossiblewhenthecompositionofthesolutionatthe
moment sepiolite startstocoprecipitatewithcalcite (further called
"comp.
")plotsonthejoinbetweensepiolite andcalcite intheCaO-SiCL-
MgOdiagram (seetable5.1).
b.alkalinity increases
•*•pHincreases-»•(Ca )decreases.Ifalkalinity increases,thesumoftheCa
andMg concentrationsmust also increasewhichfollowsfromequation (19)and
therefore theMg-concentrationmust increasemore thanthe Ca-concentration
decreases.Fromtable 5.1 itfollows that "comp.
"plots intheMgO-CaO-
sepiolite triangle.
c.alkalinity decreases
•*-pHdecreases->(Ca )increases and therefore theMg-concentrationmust
decrease faster thanCaincreases."Comp.
"plots intheCaO-SiCu-
sepiolitetriangle.
Thesametypeofreasoning canbe followed forA' isnegative andbecomesmore
andmorenegativeduringevaporation andforA' ispositive andbecomesmoreand
morepositiveduringevaporation.Theresults aresummarized intable5.2.
Iftable 5.1 and 5.2.arecombined,therelationsbetween 2[Ca ]+2[Mg ] ,
thealkalinity,theMg/Si-ratio and thesignofA'(see formula (19))during
coprecipitation ofcalciteandsepiolite canbederived.Theserelations are
represented infig.5.5.Thefigureshows that ifMgprecipitates asasilicate
instead ofacarbonate,increaseof [Ca ]+ [Mg ]concentrationdoesnotnecessarily implyadecrease inalkalinity andviceversa.This isincontrast tothe
conclusionofSurdamandSheppard (1978).Theprecipitation of sepiolite
the second decisive
maybe
step intheevolutionofbrines.
Ifthealkalinitydecreases aftertheprecipitationofcalciteand sepiolite
aNa-Ca-Cl-SO.oraNa-Mg-Ca-Cl-SO.-brineresults.Insuchbrinesthe
53
Table 5.2
la
A' = O
alk.
—
pH
—
2+
„ 2 +„
Ca
+Mg
Ca
—
2+
—
2+
—
Mg
A
"comp.
,b
ic
t
t
t
l
t
1
1
1
1
1
A
A
sep.
2a
A' > 0
1
t
t
t
—
!
alk.
—
PH
—
2+ 2+
Ca +Mg
„ 2+
Ca
2+
Mg
"comp.
"
sep.
t
!
A
A
3a
A' < O
alk.
—
pH
—
2+ 2+
Ca +Mg
Ca
Mg
!
2+
2+
!
"comp.
sep.
2b
A
3bl
t
t
2cl
1
2 c2
!
\
1
1
t
1
1
t
t
1
H
A
A
•
3 b2
3 b3
!
—
1
t
t
t
t
1
t
H
A
A
•
—
2 c3
3C
t
t
!
\
A
54
Mgt Cas
MgtCatSitA
MgtCal S l i A
MgtCatSit A
"
A
MgtCal
sepiolite
^Sil
HgO
MgICal
<45
MgtCalSII
CaO
cal ci t e
The composition of the water plots:
A on the l i n e joining calcite and s e p i o l i t e
MgtCatSitA
HglCalSiI
/^K
HglCalSil
»
A
MglCaSil
AMglCalS1l
AMgICalSli
AMgtCalSil
A in the s e p i o l i t e - calcite - MgO t r i a n g l e
A in the s e p i o l i t e - c a l c i t e - Si0 2 t r i a n g l e
A
random
t concentration increases
I concentration decreases
— concentration remains constant
a l k a l i n i t y ([HCOj)+ 2[C0j ])
2+-,
2+-,
Fig. 5.5 Thedirection inwhich 2([Ca ]+[Mg ])andthealkalinity evolve,
when during evaporation calcite andsepiolite coprecipitate, inrelation with
the initial Mg/Si ratio ofthesolution andthesign of A ' (see(19))atthe
moment thetwominerals start to coprecipitate.
2+
2precipitationof gypsum isthe third deoisive
step. TheCa -andSO.-concentrationswillchangeinoppositedirectionswhenduringevaporationgypsum
precipitates.
Whenthealkalinity increasesaftercalciteandsepiolitecoprecipitation,it
dependsontheinitialMg/SiratioofthesolutionwhethertheNa-CO-.-SO.-Clbrinewill containMgornot.
Inthiswaysixessentiallydifferentbrinescanoriginateafterprecipitation
ofcalcite,precipitationofsepioliteandprecipitationofgypsumascanbeseen
infig.5.6(modifiedandsimplifiedafterÊugsterandHardie,1978).Fromthese
brinessixessentiallydifferentsaltassemblagesanddifferenttypesofsolonchaksoriginatethroughevaporation.
Sodium,unfortunatelythemostharmfulionforvegetationandforthephysical
propertiesofthesoil,andchlorinearealwayspresentinthesaltassemblages
ofexternalsolonchaks,ascanbeseenfromtheflow-diagramoffig.5.6.It
dependsontheinitialcompositionoftheinflow-waterswhetherthesaltswill
containMg,Ca,SO.andcarbonateinconsiderableamounts.Asmalldifferenceof
theMg/Siratiointhedilute stagemaybeenoughtodeterminewhetherthe
groundwaterwillshowMg-enrichmentornot,whichinturngivesrisetocompletelydifferenttypesofsolonchaks.
55
Icalcite
Fig. 5.6 A flow sheet of theevolutionofwaters during evaporation calcite-,
sepiolite-and gypsum-precipitation. For explanation of the triangles,see
fig. 5.5. (modified and simplified afterEugster &Hardie, 1978)
5.5 APPLICATIONTOIRRIGATIONOFSALTAFFECTED SOILS
Theflowdiagramoffig.5.6canbeusedintheoppositedirectionwhenirrigationisappliedonsalt-affectedsoils.
Forexample,ifirrigationisplannedforasalt-affected soilwithsaltsof
theNa-Cl-SO.-typeonitsoilsurface.Thefirst irrigationwaterisusedfor
leachingofthesesalts.Ifthecropwhichwillbegrownissulphate-sensitive
butchlorine tolerant,suchasPartheniumargentatumGrayorLinumL.,sufficient
Camustbeaddedtothefirstirrigationwater,causingthemolarratioofCa/SO.
tobecomelargerthan 1.ThisCacanbeaddedforexampleascalciumchloride.
Subsequentevaporationandprecipitationofgypsumwill lowerthesulphateconcentrationandtheconstituentsofthesoilsolutionwillbeNa,CaandCI.
56
APPLICATION OF THE MODEL OF BRINE EVOLUTION TO
KONYA AND KENYA
6.1SALTS
Ifweconsiderthesaltswhichhavebeenfoundonthesoilsurface (seetable
3.2and4.2)thereappearstobeastrikingdifferenceinthetypeofsalt
minerals foundinKenyaandthosefoundintheKonyaBasin.
Thesaltswhichoccuratthesoilsurfacearethemostsolubleones.Theless
solublesaltminerals,whichdeterminethecompositionoftheresidualbrine,
haveprecipitatedatanearlier stagesomewherewithinthesoilprofile.
Konya
IntheKonyaBasinthesaltsaregenerallyoftheNa-Mg-SO.-Cl-type.Halite
andsulphatesofsodiumandmagnesiumarepredominantbutpotassiumandcalcium
bearing sulphates alsooccur.Nochloridesofmagnesiumandcalciumoccur.From
theflowdiagramofFig.5.6itisclearthatprecipitationofcalcite,sepiolite
andgypsummusthaveprecededtheprecipitationofthemore solublemineralsat
thesoilsurfaceinordertogenerateabrineoftheNa-Mg-SO.-Cl-type.
IntheAkgölplain (seefig.3.4)different typesofsaltsoccur.Sodium
bearingcarbonatessuchastrona,thermonatrite,andburkeitehavebeenfoundin
additiontohaliteandthenardite.
Onlytheprecipitationofcalciteisrequiredforthegenerationofthebrine
oftheNa-CO_-SO.-Cl-typefromwhichthesaltsfoundintheAkgölplainhave
originated.
AlongthecraterwallofKrater Gölinthecentreofthebasin,sodaniter,a
sodiumnitrate,occursinassociationwithhaliteandgypsum.
Thebasiniscoveredwithmarl soils;therefore,itisdifficulttosee
whetherthecalciteisarecentprecipitateorisofdetritalorigin.Gypsumis
alsopresentinconsiderable amounts,bothinthesoil itselfandatthesoil
surfacewhereitisassociatedwiththemore solubleminerals.
SepiolitewasnotfoundintheKonyaBasin,butatthetimeofsampling there
wasnoreasontosearchforitcarefully.
Kenya
Thesaltassociations observedinKenyaaredifferent fromthoseobservedin
57
theKonyaBasin.Themoststrikingdifference istheabsenceofMg-saltsfrom
thesaltassemblagesonthesoilsurface inanyoftheKenyan localitiesvisited.
InAmboselisodiumcarbonates suchastronaandthermonatritearepredominant.
Calcium ispresent incalciteand inthe double-salts gaylussiteandpirssonite.
Inaddition,somesulphatesoccur,butnochlorideshavebeenfound.Thetypeof
minerals intheAmboseliBasincanbestbedescribedbytheNa-CO,-SO.-Cl-typeof
fig. 5.6.Despite thefactthat chlorideshavenotbeenfound,themineralassociationcanbedescribedbythistypebecausechlorinedoesnottakepart ina
decisiveprecipitationreactionand,therefore,when itispresent intheinitial
dilutewater,occursinallsixtypesofbrinesoffig.5.6.Thistypeofassociationoriginates ifcalciteand/orsepioliteprecipitation areassumed.In
Amboselibothcalciteandsepiolitearepresent inconsiderableamounts
(Stoessell, 1978).
The saltassociations atMiti Bili,KanamandLakeSimbialongLakeVictoria
showthe sametypeofminerals asthosedescribed fromtheAmboseliBasin.However,halite ispresent inadditiontoabundanttronaandthermonatriteand,
furthermore,sepiolitehasnotbeenfound.At SindoandLuandaonlysulphateand
chloridewereobserved.Theseassociations canbecharacterisedbytheNa-Cl-SO.type.Calciteandgypsumprecipitationaretheoreticallyrequired togenerate
thistypeofbrineandabundantcalciteandgypsumare indeedpresent.
IntheNorthofKenya,eastofLakeTurkanaand intheChalbiDesert,the
saltmineralsaremainlyoftheNa-Cl-SO.-typewithhaliteandthenarditeasthe
mainrepresentatives.
Thistyperequirestheprecipitationofcalciteandgypsumand theseminerals
havebeenfound.Sodiumcarbonatesoccur locally atNorthHorr inthenorthof
theChalbiDesert.Themainsaltrepresentatives aretronaaridthenarditeand
thebrineatthis localityisoftheNa-CO,-SO.-Cl-type.Suchabrinecanonly
originate ifthecompositionofthediluteinflowwater issuchthatcalcite
precipitationcauses thefinalbrinetobecomealkaline.
6.2 WATERCOMPOSITION INTHESYSTEM (Na+K)-Ca-Mg-Cl-S04-(HC03+C03)
Aswith thesaltassociations,therearepronounceddifferences inthecompositionofgroundwateratthedifferent localitieswhere thesaltefflorescences
havebeensampled.
Thewateranalyses (seetable 3.1 and 4.1) canbeplotted inthesystem
(Na+K)-Ca-Mg-Cl-S04-(HC03+C03).Themolarratiosoftheanionsareplotted ina
C1-S0.-(HC0,+C0,)-composition triangleandthemolarratiosofthecationsare
58
plotted ina (Na+K)-Ca-Mg-triangle (fig.6.1).Suchdiagrams give information
about theratioofthe ionspresent inthesolution;theydonotgive informationabout totalconcentrations. Itgoeswithout sayingthat thesumofthe
anion-concentrations ofaspecificwater isthesameasthesumofthecationconcentrations.Thecompositionofthewaterswillbedealtwithonthebasisof
thesediagrams.
Konya (Fig. 6.1a)
The threesolid symbols inthediagramrepresent theinflowrivers -the
Çarsambafromthesouthwest (locality T2), the Zanopafromthesoutheast (localityT12)and theBorfromthenortheast (localityT27,fig.3.2).Theratio
carbonate/Cl+SO.forthese inflowwaters ismuch largerthanforthemoreconcentrated groundwaters.Asaresultofevaporation andprecipitationofminerals,
thecompositionofthesolutionchanges insuchaway thattheratioofthe
anionsmovesawayfromthecarbonate corner and theratioofthecationsmoves
towards thesodiumcorner.
Thewaters fromAkgölareindicatedwith adifferent symbolbecause the
mineralogicalcompositionofthesaltassemblages intheAkgölplaindiffers
fromthatatother sites.TheAkgölwaters areratherpoor insulphateandmagnesium.
Kenya (Fig. 6.1b)
Fromfig.6.1b itisobviousthatthecompositionofthewatersfromKenyais
different fromthatoftheKonyawaters (fig.6.1a).Thedilute inflowwaters
arepoorer inCaandMgthantheKonya-inflowwaters.The compositionofthemore
concentratedwaters showspronounceddevelopment towards thesodium-rich sideof
thediagram.Withregard totheanioniccomposition,theratio carbonate/Cl+SO.
inboththedilutewatersandtheconcentratedwaters isveryhigh,especially
inthewatersfromAmboseliand thosefromLakeVictoria. Itisevidentthatthe
waters fromSindoandLuandaareexceptional,justasthemineralogy ofthesalt
associations atthese localities areexceptional.SomeTurkanawatersalsoshow
arelativelyhighsulphateandchlorinecontent.
6.3 WATERCOMPOSITIONS INTHE SYSTEM Mg0-Si02-Ca0
Aplotofthewater compositions inthesystemMgO-SiO?-CaOcanrevealwhether
sepiolitewasaprecipitate (seefig. 5.4a); ifcalciteand sepiolitecoprecipitateduringevaporation,thecompositionoftheconsequent concentratedwaters
59
Na + + K +
a.KONYA:
• A Akgöl waters
CO^~ + HCOJ
"9
CTORIA
CO3
+ HCO3
"g
Fig. 6.1 Composition ofwater samples plotted in the system
Ca-(Na+K)-Mg-Cl-SO-(HCO+CO )from theKonya Basin inTurkey (a)
and fromdifferent regions inKenya (b)expressed inmole percent.
Solid symbols:surface inflowwaters;open symbols:groundwaters.
Luanda
60
plotinthesegment formedbythejoinsofcalciteandthedilute inflowwaters
andofsepioliteandthesewaters (segmentP,Q,MgO,Rinfig. 5.4 a ).
Konya
Fig. 6.2ashowsaplotofthewaters fromKonyaintheMgO-SiCU-CaO-system.
MostwatersplotintheMgO-calcite-sepiolite triangleandthemore concentrated
watersplotinthesegmentQ-MgO-R-dilutewaters.
AlthoughnosepiolitehasbeenfoundintheKonyaBasin,thepatternasshownin
fig. 6.2aisanindicationthatsepioliteoranother correlatedMg-silicatehas
precipitated.
Somewaters fromtheAkgölplainandthreeotherwaters formanexceptionto
thispattern.However,thisisnotsurprising becauseithasalreadybeennoted
thatthesaltassociationsarealsodifferentandthatforthegenerationofthe
Akgölwatersnosepioliteprecipitationisrequired theoretically. Finally,the
compositionofthedilute inflowwatersintheAkgölplainisnotknown.
Kenya
Fig. 6.2bshowsthecompositionofthewaters fromKenyaintermsofthe
systemMgO-SiCL-CaO.AgainastrikingdifferencewiththeKonyawatersis
apparent fromthisdiagram.
a. KONYA:
b KENYA:
SiO,
SiO,
2
o
i A AMBOSELI
A A A k g ö l waters
/ \
•0 T U R K A N A
• O Other waters
• O LAKE V I C T O R I A
S
L
sepiolite
R
CaO
(calcite)
Sin d o
a
•
Luanda
MgO
Fig. 6.2 Composition of water samples p l o t t e d in the system Ca0-Mg0-Si02
from the Konya Basin in Turkey (a) and from d i f f e r e n t regions in Kenya (b)
expressed in mole per cent. Solid symbols: surface inflow w a t e r s ; open
symbols: ground w a t e r s .
CaO
(calcite)
61
Thedilutewater fromtheregionalongLakeVictoriaplotsonthejoin
betweencalciteandsepiolite andthemoreconcentratedwatersplotabove this
join. Thispattern isconsistentwiththecoprecipitationofsepiolite and calcite.TheexceptionalpositionofthewatersfromSindoandLuandaisclear.The
generationofthetypeofbrineatthese localitiesdoesnotrequireprecipitationofsepiolite.
Thewatersboth fromtheAmboseli BasinandfromtheregioneastofLake
TurkanaandtheChalbiDesert showaratherscatteredpatternaroundthejoin
betweencalciteandsepiolite.No conclusions canbedrawnfromthesepatterns.
6.4 CONCENTRATION OF INDIVIDUAL IONSVERSUS CHLORINE CONCENTRATION
Theevolutionoftheconcentrationofthe individual ionscanbeevaluated
byplotting theconcentrations oftheseionsagainstchlorine concentrationof
thesamesample.Thechlorine concentration canbeconsidered ameasureofthe
concentration factorofthebrinesbecausethechloridesofthemain compounds
occurring innaturalgroundwaters and surfacewaters areverysoluble (Eugster
andJones,1977,1979;Jacks, 1973).
Chloridesprecipitate onlyatthefinal stageofevaporationandhence ifan
ionisnot lostfromasolutionbyprecipitationoranotherremovalmechanism,
itsconcentration factor isthesameasthatofchlorine.Therefore,ifthelog
oftheconcentrationofthisspecific ionisplottedversus the logofthe
chlorine concentration,thepoints formastraight linewith slope 1.Ifanion
isremoved fromthesolutionandthe logofitsconcentration isplottedversus
itschlorineconcentration,wewill findapointwhichfallsbelowthislineof
slope 1andviceversawhenanionisenrichedwewillfindapointwhichfalls
above thisline.
Konya
Infig.6.3 the logoftheconcentrations ofNa,K,Ca,Mg,sulphate andthe
alkalinity areplottedversus the logofthechlorineconcentration inthesamples fromtheKonyaBasin.Itmustbeemphasized thatthemostsolublesalts
whichcanbe foundonthesoilsurfacehavenotyetprecipitated fromthese
solutions.Againintheseplots theAkgölwatershavebeendistinguished from
theotherwaters.
TheplotofNaversus CIshowsthatduring evaporationsodium is,like
chlorine,averyconservative element.Thesodiumconcentrations oftheAkgöl
waters andtheotherwatersplot onastraight linewithslope1.
62
a. KONYA
•
1-
"3"
O
c/1
•
• •• ••
••
±
A
A
0•
A
A A
.
-1
I
.
I
••
•
• ••
A
. A• A A
• A A
•
A
A
•
-1
-I
l
•
•
o
' **.. A
A
•
•
•
o> 1
A
0
•
2-
2
.
.
2
0
\
I
I
i
.
i
i
.
3
l o g Cl
••
-
Ol
o
•
• ••
A
1 -—
o-
••
•
A
-
°1
-
in
D) 0
o
A
4 . I
-1
^
'A' .
•
.
I
I
2
3
l o g CI
.
I
-1
A
4
-1
A
,
•
. A A&A^.
• *
. A
.
.m
•
.
I
,
••
•• •
A
0
^AA.
.
A
•
-1
•
•
'I 4
-1
.
I
,
I
,
2
t
I
3
l o g CI'
Fig. 6.3 Plots of the logarithm of ionic concentrations versus the logarithm
of chlorine concentrations ofwater samples from theKonyaBasin inTurkey.
All concentrations expressed inmolm ,apart from alkalinity (eqm ) .
Solid triangles:surface inflowwaters;open triangles:waters from the
regionnearAkgöl; solid dots: otherwaters.
i
,
i *
63
Potassiumshowsfor themostpart thesame linearrelationship as sodium
apartfromasmallgroupofsampleswitha logCl ofabout 1which showsalargerincrease inpotassium concentration thancanbeexplainedby simpleevaporationalone.TheAkgölwatersaredepleted inpotassium;theyshowonlya
slight increase inpotassium-concentration.
Theabsolute Ca-concentration gradually increasesduringevaporationwitha
slope lessthan 1.Therefore,Caisalreadybeingremoved fromthesolutionat
thedilutestage.TheremovalofCacaneasilybeexplainedbytheprecipitation
ofcalcite,which isconsistentwiththegradualdecreaseofthealkalinity.The
Akgölwaters showjusttheoppositebehaviour.Thealkalinity,whichishigher
atthesamechlorine concentration thanthealkalinity oftheotherwaters,increases.TheCa-patternoftheAkgölwaters showsnoconsistentbehaviourbut
theCa-concentration isinanycase lowerthantheCa-concentrations ofthe
otherwaters.
Thepatternofsulphate israther scatteredbut ingeneralsulphate isconserved till logCl isabout 2.Fromthatstageonasmallamount ofsulphate is
removed fromthesolution,but theconcentration ofsulphate inthe solution
still increases.
Atthisstageofevaporationsulphate ispossiblyremovedby theprecipitation
ofgypsum,whichhasbeenfound ininternalsolonchaks (Driessen, 1970).
Silica islostfromthesolutionfromthedilutestage.Thesilicaconcentrationfirstslowly increases and thereafterdecreases.Thedecreaseofthe
silicaconcentrationmaybean indication thatsilicaprecipitates incombinationwithothercomponents andnotas,forexample,silicageloropalinesilica.
Precipitationofsuchspecieswould causethesilicaconcentration toremain
constantatconstantpH.As thepHintheKonyaBasinshows littlevariation
(onlyatAkgölisthepHsomewhathigher thanintherestofthebasin),pH
variation isnotthereasonforthe inflectionofthesilicaconcentration.The
silicaconcentration oftheAkgölwatersremains atarather constant levelso
intheAkgölplainsilicamayhaveprecipitated asalowtemperaturepolymorph
of Si0 2 .
Themagnesiumconcentration initially increasesand showsapositivecorrelationwithchlorine,althoughthere isa largespread inthedata.From
logCl ofabout 2the increaseofthemagnesium concentration levelsoffand
finallymagnesiumconcentrationdecreases.Thedecrease inmagnesiumconcentrationmaybecorrelatedwiththedecrease inthesilicaconcentration,both
being causedby sepioliteprecipitation.Themagnesium concentrations inthe
Akgölwatersareclearly lower thanthose intheotherwaters atthesamecon-
64
centrationofchlorine.ThepatternofMg isirregularbut correspondswith
theirregular Ca-patternoftheAkgölwaters.Thismaybe anindicationthat
atAkgölmagnesium isnotremovedbysepioliteprecipitationbutbyacoprecipitate ofMgandCasuchasmagnesium calciteordolomite.
Kenya
Conclusions drawnfromplotsoftheconcentrationofthedifferent ionsversuschlorinearevalidonlyfortheAmboseliBasinbecausethehomogeneity of
theinflowwatershasnotbeenestablished fortheotherregionswhichwere
sampled inKenya.Analyses fromthesamples ofthisstudyhavebeen supplemented
by theanalyses givenbyStoessellandHay (1978).StoessellandHay'swatersampleswere collected inAugust,1975,fromstreams,wells,swampsandpits.
Thewaters fromAmboseli show lessclearpatterns thanthewaters fromKonya,
butgeneral trendscanbededuced (fig.6.4).
Sodiumagainappearstobeaveryconservative element.Noremoval takes
placeuntiltheveryconcentratedstage.
Potassium isalsoconservative inthedilute stagebut fromalogCl ofabout
0.5 asuddenremovalofpotassium takesplace andthepotassium concentration
evendecreasesduring furtherevaporation.Potassiumremoval canbe explained
bytheprecipitationofaphtithalite (K,Na(SO.)?)which isfound atdifferent
localitiesnotonlyonthesoilsurfacebut evenatadepthofaboutonemetre
inthesoilprofile itself.Sulphate is,likepotassium,conservative inthe
dilute stagebutagainat logCl ofabout 0.5 andhigher sulphate islostfrom
thesolution,althoughitsconcentration stillincreases.Whereas,during
aphtithaliteprecipitation,boththesulphate andthesodium concentrations
increase,thepotassium concentrationmustnecessarilydecreaseduring further
evaporation.This iscompatiblewith thesodium,potassium andsulphatepatterns.
Boththealkalinity andCainthesolutionincreaseuptilla logCl ofabout
-0.5isreached.Fromthatpointonwardsbothareremoved fromthesolution.
Alkalinity still increasesbut theCa-plotshows aninflectionpointandthe
Caconcentrationdecreases.Thecontrastingbehaviour ofCaand alkalinity
during evaporationcanbeascribed totheprecipitation ofcalcite.
SiandMg showthesamebehaviour.Theybothremaincompletely insolution
till logCl isabout -1, thenboth areremoved fromthesolutionandtheirconcentrations remainatarather constant level and,finally,their concentrations
decrease.Decreaseofbothmagnesium andsilicacanbeascribed tothecoprecipitationofsepioliteandcalciteonlywhentheCa-concentrationdecreases
andthealkalinity increases (seefig. 5.5).This isthecaseatAmboseli so
65
AMBOSELI, KENYA
o
o
CT>0
•
O
•
o
-
•
• ": A
-•
•
1
••
• a*
i
•
-1
•
_
•
o„ o
o
•
•
•
2
A
i
-2
.
i
,
i
-1
*
1
2
log
.
Cl"
1
i
-1
1
2
log Cl"
O
Ol
o 0
*o o- o. .
,/fr. Al
•
-1
°
o
•
o
I
i
° o•
I
I
0
- 2 - 1
1
2
log C l "
o
•:. * . •
V. A
'
0
-1
-2
-1
1
2
log Cl"
'•
•
9
o
. o
••
. • AO
. :• A
•
.
o
o
•
•
•
•
i
I
1
I
Fig. 6.4 Plots of the logarithm of ionicconcentrationsversus the logarithm
of chlorine concentrations ofwater samples from theAmboseli Basin inKenya.
All concentrations expressed inmolm--*,apart fromalkalinity (eqm~3).
Solid dots:data from Stoessell &Hay (1978); solid triangles:surface inflow
waters; open circles:groundwaters.
1
66
thebehaviourofMgandSicanbeexplainedbysepioliteprecipitation.
6.5CONCLUSIONS
The saltsobservedintheKonyaBasinwerecomparedwith thoseobservedin
Kenya.Variousdiagramswereusedtoinvestigatetheevolutionofthemain
speciesinsolutionduringevaporationandtotestwhethertheprincipleof
chemicaldividesisapplicabletotheareasstudied.Thesediagramsarealso
ofgreathelpinestablishingthehomogeneityoftheinflowwaters.
Konya Basin
IntheKonyaBasinthebrinesandsaltmineralsaremainlyoftheNa-Mg-SO.-C
type.ThesaltmineralsandbrinesattheAkgölplainareoftheNa-CO,-SO.-Cltype.
Withinthebasin,nolossofsodiumandpotassium takesplaceuntilthevery
concentrated stage.Thebehaviouroftheother solutescanbeexplainedbythe
successiveprecipitationofcalcite,sepiolite,orarelatedMg-silicate,and
gypsum.
ThewatersfromtheAkgölplaincanbedistinguished fromtheotherwaters
anditisconcluded thatthesewatershaveadifferent origin.InAkgöl,sodium
andpotassiumarealsoconservative elements.ThecongruencyoftheCa-andMgpatternsfavoursdolomiteormagnesium calciteprecipitationoversepioliteprecipitation.Theconstancyofthesilicaconcentrationisprobably controlledby
theprecipitationofapolymorphofSiCu suchasopalinesilicaorsilicagel.
Kenya
ThewatersandsaltmineralsinKenyaareingeneralverycarbonate-rich.The
waters fromtheAmboseliBasinandthose foundalongLakeVictoriaareofthe
Na-CO,-SO.-Cl-typeapartfromthelocalitiesofSindoandLuandawhichareof
theNa-SO.-Cl-type.Thewaters eastofLakeTurkanaarealsomainlyofthe
Na-S04-Cl-type.
ThedescriptionoftheevolutionofthebrinesisrestrictedtotheAmboseli
Basinbecausethehomogeneityoftheinflowwatersintheotherdistrictswas
notestablished.InAmboseliNaisavery conservative elementasintheKonya
Basin.MgandSiarelostfromthesolution fromalogCl valueofabout -1, Ca
andalkalinityareremoved fromalogCl valueofabout-0.Sand,finally,K
andSO.areremoved fromalogCl valueofabout0.5.Thebehaviourofthese
solutescanbedescribedbycalcite,sepiolite andaphtithalite precipitation
respectively.
67
7. MINERALOGY
7.1 IDENTIFICATION OFMINERALS
Saltmineralsofboth thenaturallyoccurring saltefflorescences andthe
saltsprecipitated intheevaporationexperiments inthe laboratory (seesection8.2)were identifiedbymeans ofX-raydiffraction analysiswith aNonius
GuiniercameraFR552withahighresolvingJohanssonKcxjmonochromator,using
CoKairadiation (X=0.17889nm). Table 7.1 showsasurveyofall identified
saltmineralspresent inthesamplesofthisstudyandthecharacteristic dvaluesoftheirX-raypatterns.Thesecharacteristic linepositions arenot
necessarily thestrongest ones.
Ithappenednot infrequently thatasaltassemblageshowedanumber ofunidentiableX-rayreflections.TwounknownX-raydiffractionpatterns belonging
tounknownmineralsoccurredregularly.Specialattentionwaspaidtothesetwo
unknownminerals.Bothcompounds couldbesynthesized,analysed andtheircompleteX-ray diffractionpatternwasdetermined.Oneofthose,themineralnamed
"eugsterite",isaccepted asanewmineralby theCommission onnewmineralsand
mineralnames ofthe IMA (seesection 7.3).Theotherone,named "konyaite",is
submitted totheCommissiononnewmineralsandmineralnamesoftheIMA.Fora
descriptionof"konyaite",seesection7.4.
7.2 CRYSTALLOGRAPHIC PROPERTIES
Thecrystallographicproperties,suchascrystalsystem,crystalclass,
spacegroup,unitcelldimensions and theopticalpropertiesofallminerals
listed intable 7.1 aredescribedextensively inDana,Volume II (1957)apart
fromthoseofthemineralsburkeite,darapskite,gaylussite,loeweiteand
starkeite andofcoursethenewmineral eugsterite.Thedataofdarapskite
canbe foundinEricksenandMrose (1970),thoseofstarkeite inBaur (1962),
thoseof loeweite in Fang (1970) andthoseofgaylussiteweredescribedby
Maglione (1968).Crystallographicdataofeugsterite andkonyaitearegivenin
section 7.3 andsection 7.4respectively.Burkeite isamineralwhich isvery
rarelydescribed inliterature.A specialpaperwaswrittenaboutthismineral
asitoccurs frequentlybothinKenyaandintheKonyaBasin,cf.section7.5.
Table 7.1 Characteristic d-valuesof saltminerals.Intensities between
brackets
mineral
d-values (Ä)
1. aphtithalite
K 3 Na(S0 4 ) 2
4.09(30)
2. arcanite
K-SO.
1 4
3.
aragonite
CaCO
4. bloedite
Na 2 Mg(S0 4 ) 2 .4H 2 0
4.162(41)
2.940(75)
2.997(92)
3.396(100)
4.56(95)
3.795(75)
6. calcite
CaCO
3.86(12)
7. darapskite
Na 3 (N0 3 )(S0 4 ).H 2 0
10.3(100)
8. dolomite
CaMg(C0 3 ) 2
9. epsomite
MgS0 4 .7H 2 0
2.902(100) / 2.886(76)
3.273(52)
4.28(30)
5. burkeite
Na 6 (C0 3 )(S0 4 ) 2
10. eugsterite
Na.Ca(S0,)_.2Ho0
4
43 2
11. gaylussite
Na 2 C0 .CaCO .5H0
2.839(100)
1.977(65)
3.29(95)
3.526(80)
3.25(100)
2.801(100) / 2.777(55)
3.035(100)
2.495(14)
3.46(35)
2.871(30)
3.69(5)
2.886(100)
2.192(30)
5.99(22)
5.35(26)
4.21(100)
9.20(39)
5.50(64)
4.50(33)
6.41(95)
4.50(45)/4.43(40)
12. glauberite
Na 2 Ca(S0 4 ) 2
6.22(8)
13. gypsum
CaSO..2H.0
7.61(45)
14. halite
NaCl
2.821(100)
15. hexahydrite
MgS0 4 .6H 2 0
5.45(50)
16. kainite
KMgClSO .3H0
8.115(69)
7.771(83)
17. konyaite
Na 2 Mg(S0 4 ) 2 ,5H 2 0
12.01(57)
4.541(100)
18. loeweite
10.3(80)
Na ]2 Mg 7 (S0 4 ) 13 .15H 2 0
3.18(75)
2.677(24) / 2.659(22)
3.428(100)
3.21(100)
3.13(100)
3.11(80)
3.06(55)
2.87(25)
4.28(90)
1.994(55)
5.10(45)
9.41(25)
2.015(15)
1.628(15)
4.39(100)
7.372(100)
3.08(86)
4.002(88)
3.960(70)
4.29(95)
4.04(95)
19.nahcolite
NaHC0 3
5.91(16) 4.84(25) 3.482(30)
2.956(70) / 2.936(100)
20.niter
KNO.
4.66(23) /4.58(11) 3.78(100) /3.73(56)
2.647(55) / 2.632(30)
3.17(100)
3.082(25) / 3.059(35)
2.662(41)/
69
Table 7.1 Continued
mineral
d. values
21. northupite
Na 6 Mg 2 Cl 2 (C0 3 ) 4
8.08(50) 4.96(30) 2.697(75)
22. nesquehonite
MgC0 3 .3H 2 0
6.48(100) 3.04(30)
23. pirssonite
Na 2 Ca(C0 3 ) 2 .2H 2 0
5.13(70)/4.93(60) 2.654(90)/2.565(80)/2.506(100)
24. sodaniter
NaNO
3.89(6) 3.03(100)
25. schoenite/
picromerite
K 2 Mg(S0 4 ) 2 .6H 2 0
7.14(25) 4.16(85)/4.06(95) 3.71(100)
26. starkeite
(leonhardtite)
MgS0 4 .4H 2 0
6.83(45) 5.43(75) 3.951(65)
27. thenardite
Na 2 S0 4
4.66(73) 3.178(51) 3.075(47)
2.473(100)
2.617(55)
2.311(25)
2.646(48)
ScanningElectronMicroscopywasappliedtothesaltmineralswhicharethe
mainconstituentsofthesaltefflorescencesinordertorevealthecrystalhabit
andthemorphologyofthesesalts (seesection7.6).
7.3 EUGSTERITE,ANEWSALTMINERAL
Ooouvvenae
EugsteritehasbeenfoundinKenya,intheGreatKonyaBasinandonbricks
ofDutchriverclays.
- Kenya. InKenyaeugsteritehasbeenfoundalongtheshoreofLakeVictoria,
atSindoandLuanda (seefig.4.3).Thereitoccursasasurfacemineralin
associationwith thenarditeandhalite.ItisalsofoundatKalacha,Turkana
district,eastoftheChalbiDesertinthenorthofKenya (seefig.4.4).In
thisareavast saltefflorescences occurandeugsteriteisfoundinassociation
withhaliteandthenardite.Thegroundwatersatthelocalitieswhere eugsterite
hasbeenfound,areoftheNa-SO.-Cl-type.
- Konya Basin. IntheKonyaBasineugsteritehasbeenfoundinassociationwith
sulphatesandchloridesofsodiumandmagnesium.Itoccursinthefollowing
associations:eugsterite,halite,thenardite;eugsterite,bloedite;
70
eugsterite,halite,gypsum;eugsterite,halite,glauberite;eugsterite,halite,
glauberite,thenardite,nesquehonite.This lastassociationwasfoundaroundthe
historicalmedicinal springofTiana.Formulaeofthemineralsaregivenin
table3.2.
Theanalysesofthewatersatthesiteswhereeugsteritehasbeenfound
bothinKenyaandintheKonyaBasin,havebeenplottedinfig.7.1.
Fig. 7.1 Composition of groundwaters at the localitieswhere eugsterite has
been found in theKonyaBasin inTurkey and indifferent regions inKenya
plotted inthe systemCa-(Na+K)-Mg-Cl-SO,-(HCO+CO )expressed inmolepercent.
Type localityistheplayanorth-eastofthevillageofKarapinarwhichis
situated alongthemainroad fromKonyatoEregli.Insummer thisplayais
coveredwithaflintywhite saltcrustwhichconsistsofhalite,thenarditeand
eugsterite.
- Bricks.
Eugsteritehasbeenfoundinsaltefflorescencesonfreshly laidbricks
ofDutchriverclaysinassociationwithgypsum.
Chemioal
analysis
Inordertodeterminethecompositionofeugsterite,sincetheminuteamounts
ofmaterialavailableprecluded conventional chemicalanalysis,apreliminary
investigationwasmadebyDTA.
A naturalsampleofeugsteritewith thenarditeandhalite showsasharpendo-
71
thermiereactionbetween165°and185°.AMettlerTA2000BDTAapparatuswas
used.Thereactionproduct contained glauberitesoitseemed likelythat
eugsterite couldbeasodium-calcium-sulphate-hydrate. Inaddition,qualitative
analysiswithanelectronmicroscopeequippedwithanenergy-dispersive system
established thatNa,CaandSarethemajorcomponents.
Themineralwassynthesized,becausethenaturallyoccurring eugsteritealways occurswithothermineralsandbecausethegrainsizewasvery small(see
fig. 7.2).Solutionsofsodiumsulphateandgypsumwereevaporatedonawater
bathof60 Candeugsterite formedinconsiderable amountsinsolutionswith
amolarratioNa/Ca>4togetherwith thenarditeandgypsum.Eugsteritedidnot
formfromthesamesolutionsatroomtemperature.Thenarditeandgypsumcould
beseparatedwithaheavy liquidwithd=2.50.Eugsterite occurredbothinthe
heavier fractiontogetherwiththenarditeandinthelighterfractiontogether
withgypsumbecauseofitsstickycharacter.Thetinyeugsterite fluffswere
separated fromthefractionwithgypsümbyhandpickingunderabinocular.The
gypsumcontaminationofthesamples (samplesize approximately 1mg)wasdeterminedwiththeaidofaDTAapparatus.These samples showedinadditiontothe
largeendothermicpeakofeugsteriteonlyverysmallnegligible gypsumpeaks.
Na-andCa-contentofthesampleswereanalysedbyatomicabsorption.Sulphate
couldnotbeanalysedaswellbecauseofthesmallsamplesize.Fromthefraction
ofeugsteritewiththenardite,weight lossafterheatingwasdeterminedwitha
TGAapparatus (DupontTA990). InthisfractionallCaandKLObelongsto
eugsteritesotheELO/CaOratioofeugsterite couldbedetermined.These sampleswere largeenoughtoanalysesulphateaswell.Resultsofanalysesare
givenintable 7.2.Combinationofthesetwosetsofanalyses leadstothe
idealformulaNa 4 Ca(S0 4 ) 3 .2H 2 0.
X ray
crystallography
TheX-raydiffractionpattern (table7.3)wasdeterminedforsynthetic
materialasthesefilmsweremuch lesscomplicated thanthefilmsofthenatural
occurrences.Onlythepatternofthenarditeorgypsumhadtobesubtracted.The
linepositionsofnaturaleugsterite agreeexactlywiththoseofsynthetic
eugsteriteandcomparisonofmany filmsshowednoshiftoflinesatall.
Ithasnotbeenpossibletocalculatetheunit cellbymeansofthepowder
data.Withacomputerprogram (Visser,1969)many solutionswereobtainedbut
nonewasverysatisfactory.Effortstogrowlargercrystalsforasingle-crystal
analysisdidnotsucceed.
72
Table7.2 Chemical analysesofsynthetic samples
D.T.A.s amples
2
1
Nammol
9.15 x ÎO -3
Cammol
3
2.33 x 10~
Na/Ca
9 0
2 25
3.93
T.G.A.s amples
X
X
3
10
10
-3
9 15x10" 3
-3
2 33x 10~ 3
4
3.93
1
2
Nammol
0 150
0 019
Cammol
0 0226
0 0045
SO, mmol
4
H O mmol
0 10
0 0138
0 049
0 0087
H20/Ca
2 16
1 93
Morphology,crystallograpy
and physical
and optical
properties
Themineral formsclustersofthinfibres.Thenaturally-occurring fibres
haveathicknessof0.5-1.5urnandtheyareupto40ymlong.Thesynthetic
fibresare2-6urnthickandupto200ymlong (measuredonSEMpictures,see
fig. 7.2).Themineraliscolourlessandtransparent;itshardnessisverylow
anditissolubleinwater.Thedensity couldnotbemeasuredduetoitssticky
nature.Intheheavy liquidwithd=2.50,theupper fractionconsistsofmixturesofeugsteriteandgypsum,whereasthelower fractioniscomposedof
eugsteriteandthenardite.
TheSEMpictureofthesyntheticmaterialshowsthesymmetrymostprobably
tobemonoclinicwitha~c=116.
Opticallyitisbiaxial,buttheaxialanglecouldnotbedeterminedon
accountofthefibroushabit.Refractive indicesare1.492< a , P , Ï < 1.496;
birefringence=0.004,n //b,n A c= 27°.
Discussion
Eugsteriteisaverycommonsaltmineralwhichformsduringevaporationof
nonalkalinewaters.Themineralissynthesized easilyat60°whileatroomtemperatureitdidnotform.Thismaybeanindication thatitisametastable
73
Table7.3 X-raydiffractionpowderpatternofsyntheticeugsterite
LinepositionsmeasuredbyDr.J.W.VisserattheTechnischFysische
Dienst atDelftusingaGuiniercamerawithCuKalradiation,
A=0.15406nm,andaspecialdensitometerforGuinierfilms.
d(R)
I
d(£)
I
d(£)
12.62
9.20
6.32
5.96
5.50
1
39
4
1
64
2.545
2.458
2.291
2.284
2.231
5
5
5
5.35
4.64
4.60
4.58
4.50
1
5
5
8
33
2.214
2.178
2.170
2.158
2.148
4.20
3.860
3.819
3.622
3.590
4
7
5
3
2
3.454
3.428
3.233
3.211
3.150
I
2
1.7126
1.6999
1.6809
1.6632
1.6530
61
4
5
6
4
2
2
<1
<1
2
1.6459
1.6198
1.6053
1.5901
1.5829
<1
3
2
1
2
2.132
2.112
2.043
2.013
1.9837
2
2
1
7
2
1.5523
1.5418
1.5318
1.5234
1.5148
3
<1
<1
<1
2
32
100
4
10
4
1.9546
1.9431
1.9291
1.9199
1.9081
3
4
<1
<1
1.5046
1.4963
1.4645
1.4599
1.4536
<1
3
1
<1
1
3.118
3.065
3.054
2.973
2.936
<1
15
6
2
16
1.8935
1.8835
1.8494
1.8406
1.8313
<1
3
5
1
5
1.4293
1.4178
1.4021
1.3869
1.3820
2
3
2
1
<1
2.893
2.797
2.763
2.746
2.727
8
19
25
46
8
1.8125
1.7998
1.7922
1.7648
1.7619
6
4
2
1
2
1.3654
1.3555
1.3470
1.3353
1.3299
1
2
<1
3
<1
2.671
13
1.7273
T
0
3
1.3234
74
mineralundernaturalcircumstances.
Hydroglauberite istheonlyotherknown sodium-calcium-sulphate-hydrous
mineral.However,theX-raypatternofhydroglauberite differs essentially
fromthatofeugsterite (Slyusareva,M.N., 1969).Anunknown sodium-calciumsulphate-hydrate,calledphaseX,hasbeenfoundbyL.A.Hardie,JohnsHopkins
University,Baltimore,inSalineValleyandbyR.C.ErdoftheU.S.Geological
Survey intheDeathValley saltpanassociatedwithglauberite,thenarditeand
gypsum (bothfromL.A.Hardie,unpublished Ph.D.thesis,JohnsHopkinsUniversity).Thismineralwasnotchemically analysedbutHardieassumed ittobethe
sameaswhat isknownas"labile salt",asyntheticdoublesalt
(2Na2SO..CaS0..2H20) (HillandWills, 1938).NoX-raydataareavailablefor
"labilesalt".TheX-raypowderdiffractionpattern givento"labilesalt"by
Conley andBundy (1959)isforthenardite,which isalreadynotedby Braitsch
(1971,p. 76).ThelinepositionsoftheX-raypowderpatternofHardie's
phaseXagreefairlywellwith thatofeugsterite.Thereare,however,some
differences inintensities.Itisveryprobable thateugsterite isthesameas
Hardie'sphaseXandas"labilesalt".
Themineral iscalled "eugsterite"afterHansP.Eugster,TheJohnsHopkins
University,Baltimore,Maryland,whohasextensively studiedtheoriginand
mineralogy ofsalinelakes.
Themineraland thenamewere approvedbytheCommissiononNewMinerals
andMineralNames,IMA.A revisedversionofthissectionwillbepublished in
AmericanMineralogist (inpress).
J^"
^
-
f
r
'
/
Fig. 7.2 S(canning)E(lectron)M(icroscope)pictureofnatural eugsterite in
associationwithhalite.
75
7.4 KONYAITE,ANEWSALTMINERAL
(co-authors J.D.J,vanDoesburg andL.vanderPlas)
Occuvrenoe
Konyaitehasbeenfoundinsaltefflorescencesonsaltaffected soilsin
theGreatKonyaBasininTurkey.Itoccursinthefollowingassemblages:
konyaite,calcite,gypsum,hexahydrite,bloedite;konyaite,bloedite,hexahydrite,gypsum;konyaite,bloedite,halite,starkeite;konyaite,bloedite,
gypsum,halite,calcite;konyaite,calcite,gypsum,hexahydrite;konyaite,
loeweite,hexahydrite; (forformulaeoftheseminerals,seetable3.2).
The type localityisintheGreatKonya BasininTurkey alongtheroad
fromEreglitoNigdenearthevillageofCakrnak.
Synthesis
and chemical
analysis
Becausemostmineralswhich occurinassociationwithkonyaite contain
Na-SO.and/orMgSO.,itwasconsidered likely thatkonyaite alsocontains these
components.Themineralhasbeensynthesizedbyevaporationofasolutionof
Na~SO.andMgSO.inamolar ratio 1:1inawatch-glassintheovenatatemperatureof35 C.Inthiswayawhitepowder consistingofpurekonyaite formed.
Larger crystals formedbyevaporationofasaturated solutionusingasecond
largerwatch-glassasacover glass;whitepowderish salt consistingofpure
konyaite crept overtherimofthewatch-glassandinthecentre transparent
single crystals formed.
Table7.4showsthechemical analysisofsynthetickonyaite.Sodiumand
magnesiumwere analysedbymeansofAtomicAbsorptionSpectrometry.Thesulphate contentwasdeterminedbymeansofgravimetric analysis.ThewatercontentwasdeterminedwithaTGAapparatus (DupontTA990).Theanalysis leadsto
theideal formula Na 2 Mg(S0 4 ) 2 -5H 2 0.
Table7.4 Chemical analysisofsynthetic konyaite
Na
Mg
SO/,
H20
%
%
%
%
1
2
mean value
13.06
6.99
53.93
25.25
13.14
6.89
54.14
25.63
13.10
6.94
54.04
25.44
13.05
6.90
54.51
25.54
99.52%
100.00%
theoretical
76
Table 7.5 X-ray diffraction powder pattern of Konyaite
d(Ä).
obs
12.01
7.62
6.00
5.76
5.67
5.19
4.806
4.677
4.541
4.409
4.202
4.180
4.017
4.002
3.960
3.916
3.797
3.592
3.448
3.434
3.413
3.337
3.315
3.278
3.156
3.130
3.101
3.080
2.988
2.937
2.881
2.855
2.834
2.812
2.801
2.778
2.724
2.710
2.659
2.639
2.611
2.597
2.572
2.558
2.532
2.504
2.479
2.469
2.443
2.432
d(&)calc
.
12.010
7.616
6.005
5.757
5.670
5.192
4.803
4.674
4.539
4.408
4.201
4.181
4.015
4.003
3.960
3.912
3.797
3.589
3.449
3.431
3.414
3.338
3.315
3.278
3.157
3.131
3.101
3.081
2.991
2.937
2.879
2.858
2.835
2.811
2.799
2.776
2.72.2
2.709
2.660
2.638
2.613
2.596
2.572
2.560
2.533
2.506
2.479
2.466
2.445
2.433
I
hkl
57
12
2
19
9
9
27
2
100
59
65
17
9
88
70
3
9
36
10
12
33
19
12
21
22
21
8
15
31
15
1
1
13
16
35
22
35
12
51
50
10
55
9
2
2
1
6
1
5
13
020
011
040
101
031
121
110
131
120
112
122
130
002
060
012
132
140
032
101
150
111
042
121
152
103
113
160
052
141
133
202
212
062
170
222
211
221
232
013
231
023
242
223
180
091
112
163
122
043
171
d(&).
obs
2.415
2.402
2.337
2.323
2.309
2.300
2.270
2.250
2.232
2.222
2.204
2.183
2.158
2.132
2.117
2.101
2.090
2.078
2.060
2.052
2.038
2.030
2.007
2.000
1.994
1.978
1.962
1.957
1.954
1.946
1.935
1.928
1.920
1.903
1.897
1.868
1.852
1.843
1.825
1.819
1.808
1.800
1.790
1.782
1.755
1.738
1.732
1.723
1.720
1.715
d(£)calc
.
2.416
2.402
2.337
2.323
2.309
2.301
2.269
2.250
2.231
2.223
2.204
2.183
2.157
2.132
2.117
2.100
2.090
2.078
2.061
2.052
2.038
2.029
2.008
2.001
1.994
1.980
1.963
1.957
1.954
1.947
1.936
1.930
1.919
1.904
1.898
1.869
1.853
1.843
1.823
1.818
1.807
1.801
1.791
1.782
1.756
1.739
1.731
1.722
1.720
1.716
I
hkl
11
25Ï
220
262
142
253
39
36
14
4
5
5
2
7
8
8
14
8
7
3
1
1
11
4
10
5
4
4
23
12
3
4
6
13
10
10
4
2
7
12
25
4
7
10
10
10
2
3
11
9
2
3
14
12
7
0101
240
134
152
092
224
250
1100
162
1102
247
260
282
0102
211
193
221
004
014
270
024
1_L_L2
292
103
113
174
182
303
044
133
215
1120
105
314
342
324
0131
0122
335
163
1102
IUI
362
212
0140
77
Table 7.5 continued
obs
1.707
1.691
1.679
1.668
1.663
1.655
1.643
1.636
1.630
1.623
1.610
I.602
X ray
d(Ä)calc
.
1.707
1.692
1.679
1.669
1.663
1.657
1.644
1.636
1.630
1.623
1.610
1.601
I
10
3
3
3
3
4
4
4
4
9
3
4
hkl
222
0113
0132
084
364
242
1141
335
291
252
345
330
d(X)obs
.
1.578
1.564
1.555
1.539
1.533
1.522
1.515
1.507
1.498
1.475
1.468
I
hkl
9
6
1
1
1
1
6
3
9
2
10
0142
d(X) ,
calc
1.578
1.565
1.556
1.540
1.534
1.522
1.516
1.508
1.498
1.476
1.469
225
134
0104
144
1150
195
116
2111
0J_6_1
265
Crystallography
About 25X-ray reflectionsofnaturally occurring konyaite havebeen obtained
bycomparisonoftheX-ray diffractionpatternsoftheassemblages inwhich
konyaite occurred after subtractionofthereflectionsoftheother known
minerals inthese assemblages.Itwasimpossibletoseparate konyaite from
itsassociatingminerals.Therefore,thecompleteX-ray diffraction pattern
wasdetermined forsyntheticmaterialwithaNonius Guinier cameraFR552with
ahighresolvingJohansson Kajmonochromator,using CoKajradiation (A=0.17889
run).Table7.5showstheX-raydiffractionpattern.AnX-raypowder photograph
ofthelarger single crystalsproved that thisX-ray patternbelongstoonepure
mineralandnottoamixtureofmoreminerals.Thepurityofthesingle crystals
wasestablishedbymeansofsingle crystalphotographs.Theintensities were
measuredwithadensitometer.
Unit celldatawere obtained fromthepowder datausingacomputer program
(Visser, 1969).Themineralismonoclinic,a=0.8066nm,b=2.4025nm,
c=05783nm,beta=95.375 .From single crystal rotation-,Weissenberg-and
precession-photographsthefollowingunit celldatawere obtained:monoclinic,
a=0.5783nm,b=2.402nm,c=0.9474nm,beta=122.04°.
Theseunit cellsobtained fromthepowder dataandthesingle crystalphotographsrespectivelyareidentical.Inaplaneperpendicular totheb-axis,the
c-axisobtained fromthesingle crystalphotographs,whichisparalleltothe
morphological c-axis,formsthediagonaloftheunit cellobtained fromthe
powderdata.Theunit cellobtained fromthesingle crystalphotographsis
proposed because orientationofthecrystallographic c-axisparalleltothe
morphological c-axisisfavourable.Thespace groupisP2 1 /n,Z=4 (number
78
ofmoleculesperunit cell).
Morphology, crystallography,
physical
and optical
properties
Thewhitepowderishkonyaite consistsofclustersofflakeswithathicknessofabout3urnandadiameterofabout50ym (measuredonSEMpictures).If
konyaite growsinlarger singlecrystals thesecrystalsareprismatic,transparent,2mmlongand0.4mmthick.Thehardnessofkonyaiteisverylowandit
issolubleinwater.Thedensitymeasuredbymeansofheavy liquidsis2.088.
Thedensity calculatedonthebasisof4moleculesperunit cellis2.097.
Opticallythemineralisbiaxialpositive,n =1.464,n =1.468,n =1.474.
Birefringence=0.10.2V=74°( 2 V c a l c u l a t e d a = 79°).n a//b;n"c= 30°.
Chemical
stability
Konyaiteisavery instablemineral.Itwasidentifiedinthesamplestaken
fromtheKonyaBasinwithinthreemonthsafter sampling.Twoyears laterin
somesampleskonyaitewasstillpresentbutinothersitcouldnolongerbe
identified.Synthetickonyaite couldneitherbeconservedintheovenatthe
temperatureofsynthesis35 Cnoratroomconditions.Inbothcasesittransformedbetweenoneandfivedaystobloedite.
Grinded samplesofsynthetickonyaitewere conservedindifferent exsiccators
atroomtemperaturewith 01,201,50$,80°srelativehumidityrespectively,afew
samplesperexsiccator.Aftersixweeksallexsiccators contained sampleswhich
were completelyorpartly transformedtobloedite.Inadditionat20%and80$
relativehumidity sampleswerepresentwhichconsisted stillofpurekonyaite.
Ungrinded samples transformedmoreslowly.Thelargesinglecrystalsdidnot
transformaftersixweeks.
Itcanbeconcludedthatkonyaiteismetastablewithrespecttobloediteat
roomtemperature.Minor factors suchasgrainsizedeterminewhetherittransformstobloedite.
Themineralisnamedforthelocality.
ItissubmittedtotheCommissiononNewMineralsandMineralNamesofthe
IMA.
79
7.5
MINERALOGICAL MAGAZINE, SEPTEMBER 1 9 7 9 , VOL. 4 3 , PP. 3 4 1 - 5
Two new occurrences and the
Gibbs energy of burkeite
LIDEKE
VERGOUWEN
Department of Soil Science and Geology, Agricultural University,
PO Box 37,Wageningen, The Netherlands
SUMMARY. Burkeite (Na 6 C0 3 (S0 4 ) 2 ) is found as a
surface mineral in the north of Kenya associated with
thenardite and halite and in the Konya basin in Turkey
associated with tronaandhalite.Thecelldimensionsofa
natural burkeite are a= 7.070, b=9.220, c= 5.173Â.
The Gibbs energy of burkeite isestimated in relation to
other sodium carbonates and sulphates. A Gj 7J9S-15
(burkeite) = 3594.2 + 3 kj.
T w o new occurrences of the salt mineral burkeite
(Na 6 C0 3 (S0 4 ) 2 ) were found in Kenya and in
Turkey as a surfacemineral on saline soils. Up till
now burkeite has only been found in Searles
Lake California (Haines, 1957, 1959).There burkeite is found in a drillhole from a depth of 28 ft
to 465 ft always associated with trona
( N a 2 C 0 3 •N a H C 0 3 • 2 H 2 0 ) and in association
with one or more of the following minerals: halite
(NaCl), hanksite ( 9 N a 2 S 0 4 •2 N a 2 C 0 3 •KCl),
borax ( N a 2 B 4 0 7 • i o H 2 0 ) , thenardite (Na 2 S0 4 ),
pirssonite ( N a 2 C 0 3 •C a C 0 3 •2H 2 0), gaylussite
(Na2C03-CaCCv5H20),
and
northupite
( N a 2 C 0 3 • M g C 0 3 •NaCl).
Geological setting and pangenesis
a.Kenya. Burkeite was found during a sampling
trip in summer 1977 in the north of Kenya along
the eastern shore of Lake Turkana (the former
Lake Rudolf) near the small village of Loiengalani.
Most of the area east of the lake is covered with
Miocene and Pliocene basalts. Some spectacular
eruption centres can be seen.
Along theshore alacustrine laminated clay loam
occurs, which was deposited when the level of the
lake was a few metres higher. In some places this
impermeable clay is covered by a very permeable
gravelly basaltic material. Water which infiltrates
the gravelly basaltic material cannot penetrate the
underlying clay and moves on the interface of the
gravel and the clay. This water flowing down
the surface of the clay evaporates and forms a salt
crust. The salt crust consists of halite and thenardite. Around some cracks in the clay the salt looks
a bit more glassy. This salt contains burkeite.
S(canning) E(lectron) M(icroscope) pictures show
rosettes of thenardite needles on burkeite. Halite
occurs both in cubes and in patches with no
crystalline form on burkeite and between the
needles ofthethenardite. Sometimes the thenardite
needles are overgrown with a layer of halite. Fig.1
is a SEM of pure burkeite from Kenya.
b.Turkey. InTurkey burkeite wasfound during a
sampling trip in summer 1978in the Konya basin,
which is a lacustrine plain in Central Anatolia. The
Uplands of the basin are mostly formed of limestone formations but in some places also of
andesitic volcanics. In the basin itself some stratovolcanoes occur. The plain consists of salt-affected
clayey marl soils.From theedgessomealluvial fans
spread into the basin. Huge areas are covered with
powdery and crusty salt, especially near the lower
edges of the alluvial fans. These salts are mainly
chlorides and sulphates of sodium and magnesium,
but in the centre of the basin near Ak Gol sodium
carbonates can be found. This is the only place
where the pH of the ground water was found to be
higher than 8. Here burkeite was found associated
with halite and trona. They form a white salt crust,
which formed asa resultofévapotranspiration and
seals the surface.
Chemical analysis
Due to the small grain-size of the burkeite (a few
micrometres measured on a SEM picture)and to its
intimate intergrowth with other salt minerals, it is
impossible to separate burkeite from its associated
salt minerals. Therefore, no quantitative analysis
could be performed and only some qualitative
analyses have been made by means of an electron
microscope equipped with an energy-dispersive
system. In this way it was established that Na, S,
and C are the major components.
Crystallographic properties
X-ray powder diffraction data of the natural
burkeites were obtained with a Nonius Guinier
342
L. VERGOUWEN
FIG. I. Scanning Electron Micrograph of burkeite from Kenya (made by TFDL, Wageningen).
camera using Co-Ka, radiation, X= 1.7889Â. The
data are listed in Table I. The calculated d-values
show reasonable agreement with the measured
ones. The cell dimensions were calculated using a
program written by Visser (1969). From powder
diffraction data the space group isdetermined to be
either Pmmi or P222.
Stability and thermodynamic properties
of burkeite
Eugster and Smith (1965) constructed an a CO]
<*HjO diagram for mineral reactions involving
sodium carbonates and sodium sulphates. They
—
labelled the fields of stable associations under the
assumption that burkeite (a mixed carbonatesulphate) occurs with carbonates. This restriction is
valid in the Searles Lake environment, as well as
for the locality in Turkey, where burkeite has been
found with trona and halite(seefig.2a).The burkeiteinKenya isfound withthenardite and haliteonly,
so for the Kenya environment the labels of the
stability fields have to be changed accordingly (see
fig. 2b). Comparison of the two diagrams shows
that the association burkeite+ trona occurs under
more restricted conditions than the association
burkeite + thenardite.
There are no thermodynamic data available for
GIBBS ENERGY OF BURKEITE
343
j
\
Tn+Nh
•/
Nh
1
M
^v
\.2/
B+Nh
T n
y / \
/
Tr
\+
3/^:::::::^
\ :•:::::•::::T1
M+Na
V:B +Tr:\
O
t'
Y
B +Tm
-.detail
t"\ '
\
CN
V::::;:
u
Vwi*.'
M-l
O
\ . • . ( . • .
JLB+Na
Ü
a c t i v i t y of H_0
a c t i v i t y of H-0
Tn
(b)
(a)
Na
.^
0.69 0.88
0.81 0.88
FIG. 2a and b. a H l o-aco, diagrams for the system Na 2 COj-Na 2 S0 4 -H 2 0. For abbreviations see Table II.
B= burkeite.Infig.2afieldsarelabelled assuming that burkeiteoccurswith carbonates.Infig.2bfieldsarelabelled
assuming that burkeite occurswith sulphates.Thecoordinates ofpoints i,2,and3 infig.2aare: iPco, = !0~ 4atm.,
fl
H,o = 0-81;2PCOi= to" 26 atm.,<jHl0 = 0.29;3PCo, = 10"> atm-> "Hfi = 0.01.Intheenlargements ofthedetailsthe
slopes of thefieldboundaries arearbitrary but their relative steepness hasbeen maintained.
82
L. VERGOUWEN
344
T A B L E I. X-ray data for burkeite
Burkeite from Kenya
NBS synthetic burkeite
hkl
9.199
4.609
4.510
9.220
4.610
4.512
4175
3.862
3.804
4-175
3.862
3.803
3-537
3.440
3.300
3-535
3-442
3.301
3-075
2.806
2.778
2.643
2.588
2.490
3073
2.805
2.783
2.642
2.587
2.490
030
2-347
2.305
2.283
2.191
2.150
2-145
2.106
1.980
1.925
1.906
1.899
2-349
2.305
2.283
2.192
2.149
2.145
2.105
1.979
1.931
1.906
1.902
112
1.768
1.768
400
1-737
1-737
051
1.680
1.647
1.635
1.675
1.648
1.635
1.616
1.615
1-559
1.550
1-559
1-549
1.546
1-537
1.529
'•547
1-537
1.528
a 7.070 b 9.22c
Pmmi 3r F 2 2 2
010
020
on
101
120
i n
200
021
210
220
211
031
002
012
040
310
140
122
301
041
032
240
132
222
103
113, 340
250
023
251
203
242
060
213
c 5.173 Â
hkl
llh
9.215
4.607
010
8
020
5
4-507
4.172
OH
17
101
3-854
3-795
3.526
120
4
40
3-439
3307
3-072
2.801
2-777
2.640
2.583
2.488
2-345
2.305
2.279
2.191
2.147
2.142
2.105
1.978
1.929
1.904
1.898
1.784
1.764
1-735
1.722
1.673
1645
1.635
1.627
1.614
1-557
1.548
1-545
1.536
1.526
III
200
75
80
021
19
210
3
030
220
•7
100
211
55
03I
75
002
75
012
112
5
6
040
11
3IO
4
I40
3
122
14
11
3OI
04I
6
O32
12
240
30
I32
222
25
I50
4OO
051
OO3
103
II3.34O
250
322
O23
25I
203
242
060
213
a 7.055 b 9.215
c 5.167 Â
30
2
17
8
3
2
3
4
4
4
5
6
5
6
5
burkeite, but an approximate Gibbs energy of
burkeite can be calculated from the Gibbs energies
of the other minerals plotted in fig. 2(seeTable II).
The equilibrium fCOi between thermonatrite and
nahcolite can be calculated as follows:
Tm+ C 0 2 ± ? 2 N h
(1)
AGR = -RT\nK
= +RT\nfCO];
at 298.15 K
/co, = io" 4 bar.
In fig. la it can be seen that at thisfCOl and at an
aHl0 where thenardite and mirabilite are in equilibrium with each other (point 1 in fig. 2a) burkeite
is unstable with respect to the assemblage
thenardite + trona, which means that at these conditions the AGR of the following reaction must be
positive:
6Tn + 2 T r t ; 3 B + C 0 2 + 5 H 2 0 ;
(2)
/ c o 2 = icT 4 bar AGR > 0
3GB > 6GTn + 2GTr - Geo, «rin/co2-5GH2o-5«7-lnaHlo(3)
aUl0 can be calculated from the equilibrium:
Tn+ioH20±îM
(4)
AGR = + i o R T l n a H 2 o ;
at
at 298.15 K
ÖH2O = O.8I.
Substituting this value as well as/ C o 2 = 1 0 * m t n e
inequality (3), the following result is obtained:
GB > - 3 5 9 6 0 2 kJ.
(5)
In fig. 26 it can be seen that the equilibrium
between burkeite and natron + mirabilite is situated at a higher a H2 o than the equilibrium between
thenardite and mirabilite (where a H î o =
0.81).
at
"H 2 O =
0-81
B+ 3 0 H 2 O ± ; N a + 2M
AGR > o
(6)
GB < GNa+ 2GM - 3OGH2O- 3°KTIna H2 o
GB < - 3592-40 kJFrom (5) and (7) it follows:
-3596.02 kJ < GB < -3592.40 kJ.
(7)
The accuracy of these values depends on the
accuracy of the Gibbs energies of the individual
minerals which take part in the reaction equations
(1, 2, 4, and 6).As there are no uncertainties given
TABLE II. Gibbs energies of the minerals
Name
Formula
A
Nahcolite (Nh)
Natron (Na)
Thermonatrite (Tm)
Trona (Tr)
Thenardite (Tn)
Mirabilite (M)
Carbon dioxide (g)
Water (liq.)
NaHC0 3
Na 2 C0 3 ioH 2 0
Na2C03H20
NaHC0 3•Na 2 C0 3-2H 2 0
Na2SO„
Na2S04ioH20
-851.86
- 3428.98
-1286.55
-2386.55
-1269.985
-3646.540
co
2
H 0
2
References: 1.Garrels (R. M.) and Christ (C. L.), 1965.(1cal. = 4.184J).
2. Robie (R. A.),Hemingway (B.S.), and Fisher (J. R.), 1978.
G/298.,5kJ/mol
-394-375
-237.141
Reference
83
GIBBS ENERGY OF BURKEITE
for theGibbsenergiesofthesodium carbonates,we
cannot give an exact uncertainty for the Gibbs
energy of burkeite. Therefore, we estimated:
AG,/: 298.15 (burkeite) = 3594.2+ 3 kj.
The Gibbs energy of burkeite is calculated from
equilibrium relationships with respect to other
minerals. As soon as the values of the Gibbs
energies of the other minerals change, the Gibbs
energy for burkeite has to be changed accordingly.
Once the Gibbs energy of burkeite is known, the
positions ofall other equilibria in the diagrams can
be calculated. It can be seen that burkeite can form
under the atmospheric condition fCOl = io~ 3 ' 5
atm and that during formation thesolution must be
rather concentrated (activity ofH 2 0 lessthan 0.88),
a stage which iseasily reached during evaporation.
Acknowledgements.IthankDrE.L.Meijer forhelpingme
with the unit-cell computer program and for his helpful
discussions.
REFERENCES
Eugster (H. P.)and Smith (G. I.), 1965.J. Petrol. 6,473.
Garrels (R. M.) and Christ (C. L.), 1965. Solutions,
MineralsandEquilibria. Harper and Row edn., New
York, 450 pp.
Haines (D. V.), 1957. Geol. Surv., open-file report.
1959. US Geol. Surv. Bull.1045-E, 139.
Robie (R. A.), Hemingway (B. S.), and Fisher (J. R.),
1978. Ibid. 1452, US Government Printing Office,
Washington.
Visser (J. W.), 1969.J. Appl.Cryst.2,89.
[Manuscript received 30 October 1978]
345
7.6 SCANNING ELECTRON MICROSCOPY APPLIED TOSALINE SOILS FROMTHEKONYA BASIN
INTURKEYANDFROM KENYA
This sectionispartof
Vergouwen,L.,1981,Scanning ElectronMicroscopy applied tosaline soils from
theKonyaBasininTurkeyandfromKenya.In:E.B.A. Bisdom (ed.):Submicroscopyofsoilsandweathered rocks. 1stworkshopofUndisturbed Soil
Materials (IWGSUSM),Wageningen, 1980.Pudoc,Wageningen,theNetherlands,
inpress.
Summary
SaltsoftheKonyaBasininTurkeyandfromdifferent regionsinKenyawere
investigatedbyScanning ElectronMicroscopy. This techniquewasusedtoexamine
the crystallographic propertiesoftheindividual salt crystals,toinvestigate
theirmorphologyandtheir influenceonthemorphologyofthesaltassemblagesin
which theyarepresentandtostudytherelations betweenthedifferent salt
crystals.
Thedifficultiesaredescribedwhich arisewhenthisvery friablematerialis
handled.Ananalysing equipment isessentialforidentificationofthedifferent
minerals. Backscattered electron scanning imagesandX-ray imagesarealsoof
greathelp.
Halite formsasmooth crust sealingthesoil; trona,bloediteandhexahydritemakethesalt crustvery fluffy.Thenardite occurs intwocrystal forms.
Procedure
Sampleswere selectedwith different saltassociations. Fromeachsamplea
piecewasusedforX-ray diffraction analysis,todeterminetheexactmineralogical contentofthesaltpiece,usingaGuinier camerawithCoK1radiation;
a corresponding piecewasusedforSEManalysis.
The typeofmaterialwasvery complicated tohandle.Only coherent material
whichdidnotbreak intopiecesontouching couldbeused.Thisisinfactalreadyapreselection.Forexampleitwasnotpossible tohandlethethenardite
crystals fromthepuffed solonchaks becauseoftheir fragility.Many thermonatrite samples disintegrated aswellonhandling.
Thesampleswere gluedwith silverpaintwith tolueneassolventonasmall
aluminium stub.Alcoholassolventisnotsuitablebecausethesalts dissolve
85
init.Thissilverpainthastheadvantage overnormalglue thatitisconductive
atthesametime.Thesampleswereprepared atthesamedayofusingthem.
Preparationafewdaysbeforeuseappeared tobeuselessbecause thesamples
disintegrateeasily.Thematerialwas coated firstwith acarbon layerand then
with agold layer.A carboncoating isrequiredbecauseotherwise thegolddoes
notadheresufficiently tothematerial.Thegoldwasevaporated thermally.With
thismethod thedistancebetween thehotspotand thesample islarger thanwith
thesputteringmethod.The temperature ofthesamplemustbekeptas lowaspossiblebecause somemineralscontaincrystalwater.Thermal evaporationhasthe
additionaladvantage thatthegold layerbecomes thickerwhich isfavourable
becauseoftheirregular surfaceofthesamples.A disadvantageofthiscoating
method isthatthevacuum inthechambermustbehigher (10 torr)thaninthe
caseofsputtering (0.1 torr).
Thesamplesveryoftenshowcharging.This iscausedbythe factthatthe
sidewhich isfixedonthealuminiumstubisusuallyvery irregular.Therefore
noteverypartofthespecimen isincontactwith thealuminiumstubbutonly
somelocalspotsarefixed.Thisproblemisusually less severeontherimof
thespecimen.Therethenicest imagescouldbeseen.
Toreduce theeffectofchargingwe tried tocoatthespecimenwithosmium
whichhasahigherconductivity thangold.Thiswasnotverysuitablebecause
osmium isfirstdissolved indistilledwater andthenthesolutionisevaporated
inacoveredpetribowl inwhichthespecimen isalsoplaced.Thedistilledwater
withthedissolved osmiumprecipitated onthesampleanddissolved thesalt
material.Smallholescouldclearlybeobservedwith theSEM.
Somespecimenswere investigatedwith theSEMbeforecoating themtosee
whether thecoatingprocedurewould causemineral transformations.This seemed
nottobe thecase.
TheSEMpicturesweremadewithaJEOLJSM-35cequippedwithEDAXPV 9100
withanECONwindowlessdetector.Polaroid filmwasused toobtaindirectresults
because ofthedifficultmaterial.Themineral-content ofthesample isknown
fromtheX-raydiffractionanalysis.Allmineralswere chemically identified
withEDAX.Althoughthisanalysing system isnotquantitative,atleastnotwith
samples coatedwithgold,itappeared tobe indispensable.Wrong identifications
couldbeprevented inthisway formineralswhichoftenshoweddifferentcrystal
habits thanexpected. Inothercasesmineralswere identifiedwhichwerepresent
insuchsmallamounts thattheycouldnotbedetectedbyX-raydiffraction.
86
Results
Themineralogy andmorphology ofhalite,thenardite,bloedite,hexahydrite,
eugsterite,tronaandburkeitewere investigated.Thecrustswere 1/2 -3mm
thick.No zoningcouldbedetected,neithermacroscopicallynormicroscopically
with alightmicroscope.
Eugsterite isanewsaltmineralwhichoccursbothinKenyaandinKonya.The
constituting elementsofeugsterite couldonlybedeterminedbymeansofEDAX
becausethemineralcouldnotbeseparated fromitsassociatedminerals.This
qualitative indicationmade itpossible tosynthesize themineraland toanalyse
thissynthesizedmaterial quantitatively (themineralhasbeenapprovedbythe
CommissiononNewMineralsandMineralNamesofthe IMA).FromSEMpictures the
monoclinicnature ofthemineral couldbededuced and theanglebetweentheaandc-axiscouldbemeasured.
Fig. 7.3a showsthenardite crystalswithbloediteandeugsterite.Thenardite
wasexpected togrowinneedle-type crystalsascanbeseenmacroscopically in
puffed solonchaks.Inassociationwithother saltminerals,thenarditehasthe
tendencytogrowinthecrystalline formasshowninfig.7.3.Inevaporation
experiments inbeakersthenardite growsasneedles.Infig.7.3b itcanbe seen
thatthesmallbloedite crystalshaveprecipitated onthethenardite inalater
stage.Theeugsterite (fig.7.3c)hascrystallized attheexpenseofthethenardite.Fig.7.3 isanexample ofthenecessityofananalysingequipment.Without
suchanequipmenttheeugsteriteneedleswouldhavebeenattributed toeither
gypsumor thenardite andthethenarditewouldnothavebeenrecognized.Figs.
7.4a and 7.4b showthedifference inanalysisbetweentheeugsteriteneedles
andthethenarditecrystals.Fig.7.4c isanexampleofthenarditegrowing in
itsneedle-type form.
Figs. 7.5a and 7.5b areanexampleof-halite.Thesmoothsurface iscontinuouswithonlyfewpores.This isthetypeofcrustwhichsealsthesurface,
preventing furtherevaporation.Thesmallparticles onthesmoothsurfaceare
calciteandquartz.
Fig. 7.6a showswellcrystallizedhalite cubes inequilibriumwithbloedite.
Bothminerals seemtohavecrystallized atthesametime.Theyhave completely
intermingledwitheachother.Thebloeditecrystalshaveasomewhatsmallersize
andarepacked somewhat looserthanthehalite.Fig.7.6b isanenlargement of
thebloedite.
Figs. 7.7a and 7.7b showtronaminerals.Tronacrystallizes asbars,both
criss-cross inalldirections and inradiating fans.These tronabars causethe
87
saltassociationtobecomevery fluffy.
Figs.7.8and7.9showtheapplicationofdifferenttechniques.
Infig.7.8athetopologyofasaltcrustmainlyconsistingoftronabarswith
different irregularitiescanbeobserved.
Fig.7.8bisabackscattered electronscanning image (BESI)ofthesamespot.
WithBESI,thedifferenceincontrastiscausedbythedifferenceinchemical
compositionandnotbythetopologyasinthecaseofsecondaryelectronimages.
A comparisonoffigs.7.8aand7.8bindicates thattheirregularitiesonthe
topsofthetronacrystals correspondwiththelighterpartsofBESI.Spot
analyses showthatthesmallgrains thatoccur throughoutthesamplearequartz
grainsandthatburkeitereplacesthetronacrystals.Thepresenceofthese
mineralsisconfirmedbyX-raydiffraction analysis.SiandSX-ray imagesof
thesamespotshowthedistributionofthesilicaandtheburkeite (figs.7.9a
and 7.9b).
Figs. 7.10aand7.10bisanexampleoftheassociationhexahydriteand
bloedite.Hexahydriteconstituesthemainpartandthesmallbloeditecrystals
aredepositedonthehexahydrite.Thiscrusthasavery largeporosity.IdentificationwithEDAXwasagainindispensableinthiscase.Themiddleoffig.
7.10ashowsclearlythephenomenonofcharging.
Discussion
and
conclusions
Scanningelectronmicroscopyincombinationwithananalysingequipmentis
averyuseful techniquetoevaluatethesubmicroscopicalpropertiesofsalt
minerals.Themorphologicalpropertiesofthesalts,whichweredetermined
macroscopically,couldbeestablishedwiththeSEM.
Thesurfaceofhalitecrustsisindeedaverysealing smoothcrustwithmany
pores.
Bloediteispresentincrustsassmallrounded crystalswhicharelooselypacked.
Especiallyinassociationwithhexahydrite thesecrustsarevery fluffy.
Driessen&Schoorl (1973)assumedbloedite crystalstoberesponsibleforthe
sealingofthesoilsurfacebutdetailed investigationofmany crusts containing
bloedite contradict theseassumptions.
Tronaminerals seemtoincreasetheporosityofthesaltcrustaswell.
Whensalinesoilsareirrigated,thesoilscoveredwithhalitecrustsneed
ploughingtoavoidrunoffoftheirrigationwater.Theporosityoftheother
typesofcrustsisusually suchthattheirrigationwatercanpenetratethecrust
easily.
ssiwa
o.
J3
(3
0)
T3
e
0!
0)
60
U
e
.a
c
w
E-l
3
3
u
o
T3
U
c •
01 <u
.C -u
H -H
•
S-4
n) <u
•u
en w
.
ÛO
r- 3
• <u
M
•H m
t* o
o)
89
Fig.7.5a. Smoothhalite crustwith quartz and calcite grains,b.Enlargement.
w%
'y>
w.
v' •''**.
'SS? 1 ^' 11 ".'*
Fig.7.6a.Halite cubes intermingled withbloedite.b.Enlargement ofbloedite
ft *#.*# r*>;'
X P '
OT*
. rn/. *$>£
.,• *r
Î-if^i t k / ! r . •10/im
Fig.7.7a.Trona crystalsbetween aphtithalite.b.Fanof tronacrystals.
90
.Jfs^.»y»r~ J,
,<•*
f->*i
"S
Fig.7.8a. SE-image of trona,locally transformed intoburkeitewith scattered
quartz grains,b.BESI-imageof the samespot.
Fig.7.9a. SiX-ray image of the same spot as fig.7.8.b.SX-ray imageofthis
spot.
Fig.7.10a.Hexahydritewith smallbloedite crystals,b.Enlargement of
hexahydrite.
All SEMpicturesweremade atTFDL,Wageningen.
91
8. SALT ASSEMBLAGES AND EVAPORATION EXPERIMENTS
8.1 INTRODUCTION
Inthis chapteradescriptionisgivenofthedifferent salt associations,
inordertoexaminetheconditions underwhichthesaltsprecipitated,totest
whethertheassociations represent equilibrium assemblagesandtoconsiderthe
processes thatoperated during evaporationandprecipitation.Inordertosolve
someoftheseproblems,thewaterswhichwere sampled inthefieldwereevaporatedinthelaboratory.Acomparisonwasthenmadebetweenthesalts precipitated
fromthesewatersinthelaboratoryandtheassemblages foundinthefield.
The chapter isdivided intothreeparts.Inthefirstparttheresultsof
theevaporation experimentswillbegiven.Inthesecondparttheassemblagesin
thesystemNa-Ca-SO.-CCL-TLOwillbeexamined. Saltassemblages belongingtothis
system occurmainlyinthevisited localities inKenyaandonlyinafewlocalitiesintheKonyaBasin.
Itisimpossibletocalculatewhichmineralswill theoreticallyprecipitate from
specificwatersinthis systemduring evaporationbecauseuntilnownoactivitycoefficientsforcarbonate-ions inconcentrated solutionshavebeenavailable.
Therefore,theassemblages inthis systemhavebeen investigated withtheaidof
log P Q Q -logaj-[Qdiagrams.Inthethirdpartthesalt associations inthesystemNa-K-Ca-Mg-SO.-Cl^-TLOwillbeconsidered. Salt assemblages belongingtothis
systemoccurmainly intheKonyaBasin.
8.2 EVAPORATION EXPERIMENTS
The ground-andsurfacewaterswhich havebeensampledintheGreat Konya
BasininTurkeyandindifferent localities inKenyawere evaporated inevaporatingdishesinthelaboratory.Thewater sampleswere first filtered througha
0.2 ymmicropore filter.Theamountofwaterwhichevaporated fromeach sample
varied between 150-300mldependingontheamountofwater leftafter analysis
ofeachsample.Therelativehumiditywasapproximately 401andtheroom temperature 20 C+2.Thesamples evaporated inabouttwoweeks.Fig.8.1shows
aphotographofsuchanexperiment.
During suchexperiments fractional crystallization occursbecausethesalts
precipitate againstthesidesofthedishesandduring further evaporationof
92
Table 8.1 SaltAssemblagesprecipitated fromthewater samples fromKenya
inthelaboratory compared withnatural assemblages
Typeofwater
Saltassemblages found inthe
log
field 2 )
»
Amboseli
laboratory
3)
waters
Am 3
groundwater
-1.30
tr,tm,hal,tn
tm,hal,KNaSO,
tm,aph,hal
groundwater
Am4
-2.72
tr
tr,tm,hal
tm,tr,cal
hal,cal
tr,cal,prs,tm
tr,tm,gay
pond
Am5
-3.05
tr,cal,gay,tm
cal,ar
tm
tr,hal,tm
tr,tm,hal
Am6
groundwater
-1.67
Am8
groundwater
-0.98
tr,cal,gay,tm,aph
tr,tm,gay
tr,aph
tr,hal,syl,KNaSO^
syl,hal,tm
hal,syl
tr,aph,gay,prs
Am11
Am 14
Turkana
riverwater
-2.60
hal,syl,cal
riverwater
-1.82
cal,hal
waters
Maikona
groundwater
-2.79
hal,tn
pth,kai,hal
hal,kai,gps
hal,gps
Kalacha I
Kalacha II
NorthHorr
kai,hal,gps
hal, tn,ntp
groundwater
-2.44
hal,tn,eug
-2.35
hal,tn
-2.89
hal,tr,brk
hal,cal
dpsk,ntr,tn,hal
groundwater
tn,hal,ntr,dpsk
tr,hal,brk
groundwater
tr,hal,brk
brk,hal
Furaful
riverwater
Loyengalani spring
-3.59
tn,hal
tn,hal
-2.78
tn,hal
ntp,hal,cal
tn,hal,brk
1)logp r n calculated fromHCO~-concentrationandpHmeasured inthefield.
Equilibrium constants fromStummandMorgan (1970).
2)Thesearethesaltassemblages formed intheimmediatevicinity ofthesite
ofsamplingof thegroundwater.
3)Thesaltsdeposited intheevaporating dishes.The sequenceofthedifferent
assemblages isthesameasinthedishes,fromtheupperrim till thebottom.
4)b.r. beforerains
5)a.r. afterrains
93
Table8.1cont'd
Loc.
Saltassemblages found inthe
2)
laboratory 3)
log
Typeofwater
P
C0„
field
Lake Victoria Waters
Simbi 1
spring
-2.17
tm,tr
sdn
Simbi2
groundwater
-1.51
tr,tm,gay
sdn, gps
MitiBili 1 spring
b.r. 4>
-0.10
tr,cal,gay,nah
sdn, dpsk,tm,tr,hal
MitiBili 1 spring
a.r.5)
-0.16
tr,cal,gay,nah
tm,tr,hal,brk
sdn, dpsk,tm,tr,hal
tm,tr,hal,brk
tm,tr,hal,brk
MitiBili2 groundwater
a.r.
MitiBili 3 groundwater
a.r.
-1.95
tr,tm,hal
tr
.
-2.08
tr,tm
tr,gay
tr, tm,hal
tr, tm,hal,brk
tr, tm,hal,brk
tm,hal,brk
tr, tm,hal,brk
MitiBili4 groundwater
a
-2.05
tr,tm,hal,brk
'r*
tr,hal
tr,nah,gay
tr, tm,hal,sdn,dpsk
tr, tm,brk,hal,sdn
brk, tr,hal,tm,sdn
tr,gay
tr,hal,brk
MitiBili 5 surfacewater
a.r.
tm,hal,brk
-3.71
tr, tm,hal
cal, hal,syl,tm
LakeVictorialakewater
-3.32
Kanam 1b.r.groundwater
-1.30^ tr,nah
sdn, dpsk,hal
tr, tm,gay
sdn, dpsk,hal
tr,hal,gay
Kanam 1a.r.groundwater
-1.38[ tr,hal,tm,gay
tr,gay
Kanam2a.r. groundwater
tm,tr,hal,brk
tm,tr,hal,brk
tm,hal,brk
-2.45'
tr, tm,hal,brk
Luanda
groundwater
-1.0
eug,gps
bl,tn,kon
tn, hal,bl,kon
tn,hal
Sindo
groundwater
-0.83
hal,tn
gps
kai, hal
tn, lw,hal,kon
lw,hal,kon
Forabbreviationsoftheminerals,seetable 2.2and 3.2.Inadditiondarapskite
(dpsk),Na (NO)(S0,).H0;pentahydrite (pth),MgS04.5H20;kainite (kai),
KMgClSO,.3H20;northupite (ntp),Na 6 Mg 2 Cl 2 (C0 3 ) 4 .
94
Table 8.2 Salt assemblages precipitated fromthewater samples from the
Konya Basincompared with natural assemblages
Loc.
Typeofwater
log
Salt assemblages found inthe
3)
2)
laboratory
field
1)
P
C0,
Tl
pond
-2.53
bl, hxh, gps
eps,hal,hxh
T4
groundwater
-1.28
tn, gl
hal,bl,hxh,kon,lw
tn, aph
hal,bl,hxh,lw,kon
bl, hal, gl
bl, gps
bl, hal, hxh, sch
T5
T5
irrigation
channel
groundwater
T4 hal,hxh,gps,lw
-0.97
hal,hxh,gps
-2.27
eps,bl,gps
T5 hal,hxh,gps
hal,hxh,gps
T6
groundwater
•1.1!
bl,hxh,gps
hal,hxh
hal,hxh,ar
T8
crater lake
-2.64
hal, tn,nsqh,eug
hal,eps
hal,eps
T9
T10
groundwater
groundwater
-1.75
-1.46
hal,gl,gps
hal,hxh
hal, tn,eug
hal
tn
hal
tn,aph
lw,hal
hal
Til
groundwater
-1.79
bl,eug
pth,hal
pth,gps,hal
T13
groundwater
-1.57
hal,gps,eug
tn,hal,lw,hxh
tn,eug
gps,hal,hxh,tn
gps,hal,hxh,tn,kon
T14
groundwater
-1.08
hal,gps
gps,lw,hal,kon
gps,hxh,lw,hal,kon
gps,hal,hxh,bloed,kon
T15
groundwater
-1.76
hal, gps,brk
hal,tn
hal,tn
T15
T16
irrigation
channel
-3.19
spring
-0.23
hal, cal,lw
hal,cal
hal, tn,eug
gps,cal
hal,tn
95
Table8.2 cont'd
Loc.
Typeofwater
log,.
P
T17
groundwater
1)
Salt assemblages found inthe
field Z'
C0„
J
2
-1.86
bl,eug
laboratory
gps,hal,hxh, kon,lw
gps,hal,hxh,kon,lw
gps,hal,lw,hxh,kon
T22
groundwater
-1.71
bl,hxh,gps,kon
gps,hxh,lw, hal
eps, gps,lw,hxh,hal
eps, gps,hal
T23
groundwater
-1.87
bl,hal,gps
T24
groundwater
-2.22
hal,gl
T25
spring
+0.03
hal,tn,nsqh,gl,eug hal,tn,Na & Mg 2 Cl 2 (C0 3 ) 4
bl,hal,gl
pth,hal,gps
gps,hxh,pth,hal
hal, gps
hal,gl,tn,gps,eug
hal,MgCO .3H20
hal,gps
brk,tn, hal
T26
groundwater
-0 53
T27
river
-1 97
T28
groundwater
-1 28
hal
tr, brk
brk, Na 6 Mg 2 Cl 2 (C0 3
hal
hal
tn, aph
hal, gps, cal, hxh
tn, aph
T29
T31
groundwater
groundwater
-1 74
-2 51
tn
tn, hal
hal
eug
cal, hal
hal
tr, brk
hal
tr, tm,hal,brk
hal, brk, tr
groundwater
-2 .84
hal
T33
groundwater
-1 .99
tn, hal, eug
hal, hxh, kon
tn, eug
hal, hxh, kon
T35
groundwater
-2 .05
tr, brk
hal, brk, tr
T32
tn
hal
bl, hxh, gps, kon
hal, hxh, kon
bl, gps, lw
hxh
hal, lw
hxh, hal
lw, hxh, kon
T36
groundwater
-1 .30
hal
bl
hal, hxh, pth
hal
bl, stk, kon
hal, pth, hxh, gps
T37
groundwater
-1 .76
hal
bl, gps, kon
hal
hxh
gps
hal, hxh
1), 2 ) ,3) seetable8.1
For abbreviationsoftheminerals seetable3.2and 4.2.
96
Fig.8.1 An example of anevaporation experiment.
thesolution,thesaltmineralsarenolongerincontactwiththesolution.
Thesaltsattheupperrimhaveprecipitated first.Thesaltatthebottom
isamixtureofthelast-formed crystalsandofcrystalswhichsinktothebottomduringthewholeprocessofevaporation.Usuallyonlywell crystallized
halite cubessanktothebottom.Evaporationinevaporatingdishesinthelaboratorycloselyreproducestheprocesswhichtakesplaceduringevaporationof
groundwaterinnaturebecauseinthelatter casefractional crystallizationalsooccurs.
ThesaltmineralsproducedinthedisheswereanalysedbyX-raydiffraction.
Enoughsaltprecipitated fromthemore concentratedwaterstoenabledifferent
analysesofthesaltassociation,beginningwiththeprecipitateoftheupper
rimandgoingdownwards towardstheprecipitateinthecentre.Thelessconcentratedwaters,suchasriverwater,formedaverythinsaltfilmagainstthe
sideofthedishespermitting onlyoneanalysistobemade.
Table8.1and8.2showthemineralswhich formedinthedishes fromthe
97
watersfromKenyaandKonyarespectively.For comparison,thenaturalsalt
assemblages foundintheimmediatevicinityofthesamplingsitearealsogiven
togetherwiththeequilibrium log ~PQQ °fthewaters (calculatedusingthe
analysedHCO,-concentrations),thepHmeasuredinthefield (seetable3.1 and
4.1)andtheequilibriumconstantsat25 Cand1bar,takenfromStummand
2Morgan, 1970.Theequilibrium logpco 0 °ft n ewaterscontainingCO, calculated
9
-
2
-J
usingtheanalysedCO,-concentrationandthepHmeasuredinthefieldappears
tobeingeneralafactor0.5higher thanthatcalculatedbymeansoftheHCO,concentration.Thismaybeduetothefactthatactivitycoefficientshavebeen
neglected,whichhasagreatereffectinthecaseofbivalent thanmonovalent
ions.
8.3 PHASEDIAGRAMSINTHESYSTEM Na-Ca-S04-C02-H20
The system
m-COg-üg)
Thestability fieldsofthemineralstrona,thermonatrite,nahcoliteand
natroninthesystemNa-CO?-H?0canbestudiedwithalogpçg -loga^ Q
diagramatconstantPandTinwhichp „ 0 isthepartialC02-pressureand
a H Qisthewater-activity. Itwouldbemorerigouroustouse £QQ7, t n e ^US~
acityofCO2,insteadofPQV,> but,becauseP c o isameasurablequantityand
becausethedifferencebetweenp ™ andfCg isnegligiblysmall,itisconvenienttouseppQ.
Theslopeofthefieldboundariesbetweenthestability fieldsofminerals
insuchaplotisdeterminedbytheratioofILOtoC0 2 moleculesinthereaction
equation (Garrels&Christ,1965).Thereactionequationsandtheir corresponding
slopesare(forabbreviationsandformulaeoftheminerals,seetable8.3):
2tr
t 3tm+CO,+2H 2 0
tr+C0 2
t 3nah+H 2 0
2tr+25H20 $ 3nat+C0 2
tm+9H 2 0 $ nat
tm+C0 2 *2nah
2nah+9H 2 0Înat+C0 2
slope -2
(1)
1
25
(2)
(3)
0 0
(4)
(5)
(6)
0
9
IftheGibbsenergyofformationofthesemineralsisknown,thepositionof
theselinesinalogp r „ -log SL,0 diagramcanbecalculated.
Forexampleforreaction(1):
AG°=-RTlnK=-2RTIna R0-RT Inp œ
inwhich
AG R = 2AGfQ^0)
2AGf(tr)
+A G £ ( C o 2 ) + 3 A G f ( t m )
I£AGScanbecalculated,thepositionofthefieldboundarybetween thestabilityfieldsoftronaandthermonatriteisfixed.Thesameprocedure canbeappliec
totheotherminerals inthesystemNa 2 0-CCL -H 2 0. Itisclear that theresultantdiagram isdependentuponthevaluesoftheGibbs energies oftheindividualminerals,andmanydifferentvalues arereported inthe literature.Fig.
8.2 and fig.8.3 showthediagramcalculatedwithtwodifferent setsofGibbs
energies offormation -thoseselectedbyAl-Droubi (1976)andthose selected
byGarrelsandChrist (1965) (seetable8.3).
Table 8.3 Gibbs energies of formation ofminerals andother species
in thesystem Na.O-CaO-SO,•C0r - H O at25°Cand 1bar (AG°)
AG"f
Species
Chemical formula
nahcolite (nah)
NaHCO,
natron(nat)
Na 2 CO 3 .10H 2 O
Gibbs energyo
f formation
- 851.68
(1)
- 851.86
(2)
-3429.58
(1)
-3428.98
(2)
trona (tr)
NaHCO .Na 2 C0 3 .2H 2 0
-2381.7
(0
-2386.55
(2)
thermonatrite (tm)
Na 2 CO .H 2 0
-1289.59
(1)
-1286.55
(2)
gaylussite(gay)
Na CO .CaCO . 5 ^ 0
-3384.04
(1)
-3372.64
(3)
pirssonite (prs)
Na 2 C0 3 .CaC0 3 .3H 2 0
-2661.79
(1)
-2661.79
(3)
thenardite (tn)
Na2S04
-1269.99
(4)
mirabilite (mb)
Na„S0..10H.0
2 4
z
Na,CO,(SO.).
6 3 42
CaCO
-3646.54
(4)
-3594.2
(5)
-1129.40
(6)
- 394.37
(4)
- 237.14
(4)
burkeite(brk)
calcite(cal)
2(g)
H 2 0 (1)
(1)Al-Droubi (1976).Thesevalues havebeencalculated frompKvaluesand
fromvalues for the species in solution fromRobie et al. (1978).
(2)Garrels &Christ (1965). (3)Hatch (1972). (4)Robie,Hemingway and
Fisher (1978). (5)Vergouwen (1979). (6)Helgeson (1969).
kJ/mol
99
e
4-1
O
nt
01
H
H
u
ffl
. >-,
o
ni)
Ctl
-rH
Tl
T)
CU
•u
1
I—t
O
*—
•H
-O
r)
o
VJ
oi
ï~i
ra
<
3
^
o
XI
CN.-1
en
o
T)
CO
ta
OU
o
oCN
U
0)
i—1
,—<
1
i
CU
CN
<)
CNO
1
01
<">
u
o. a
t>0
R
o
•J
(11
4J
ca
!>J
30
CU
4J
>^
co
tu
.c
en
CIt
4-1
ai
•o
o
•r - l
H
e
ctl
T>
o
ti
100
InthediagramcalculatedwiththevaluesofGarrels&Christ,the stability
fieldoftronaisconsiderably enlargedattheexpenseoftheother threecarbonatesincomparisonwith thediagramcalculatedwiththevaluesofAl-Droubi.
Thisdemonstrates thedependenceofthecalculatedmineral stability fieldson
thechoiceofthermodynamicdata.
Inthe following considerations thevaluesofGarrels&Christhavebeenused
because,contrarytoAl-Droubi'svalues,thesearemore consistentwiththeobservationsmadeinthefieldandintheexperiments.
The system Na-Ca-COn-H„0
Thestability fieldsofthemineralsgaylussite andpirssonitecanbesuperposedonthesodium carbonatediagram,assuming calciteoraCaCO..-containing
mineraltobestableover thewhole logp m -loga„»range.Gaylussiteand
pirssonite arerelated accordingtothereaction:
prs+3H 2 04 gay
slope oo
(7)
Oncemore,theGibbsenergiesofformationdetermine thea„„atwhichthese
twominerals areinequilibriumwitheachother.TheGibbsvalues selectedby
Al-Droubi (1976)yieldanequilibriumwater activitybetweenpirssoniteand
gaylussite thatiscompletelydifferent fromthatcalculatedwiththevaluesof
Hatch (1972) (seetable 8.3).Theequilibriumwater activity calculatedwith
Hatch'svaluesisgreaterthan 1,whichmeansthatgaylussitewouldalwaysbe
metastablewithrespecttopirssonite.Asgaylussite occursmuchmore frequently
thanpirssonite,Hatch'svalues areprobablynotrealistic.
Thestability fieldsofgaylussite andpirssonite aredelineatedbythefollowingreactions:
prs
*tm+cal+H 2 0
slope o o
(8)
prs+C0 2 %cal+2nah+H 2 0
3prs+C0 2 X 2tr+3cal+H 2 0
3 gay+C0 2 %2tr+3cal+10H20
gay+C0 2 $ cal+2nah+4H 2 0
gay+5H 2 0 Î cal+nat
'1
1
10
4
oo
Infig.8.4thesecalcium-containingminerals aresuperposedonthesodiumcarbonatediagramcalculatedwithGarrels&Christvalues fortheGibbsenergiesofthesodiumcarbonates.Thegaylussite andpirssonite stability-fields
areboundedbyreactions (8),(10),(11),(12)and(13) (seefig.8.4).From
thisdiagramitisconcluded thatbothpirssonite andgaylussite canformin
equilibriumwithtrona.Thesamediagramcalculatedwith theGibbs energy
(9)
(10)
(11)
(12)
(13)
101
01
u
<u
e
H
H
•i"in
^__^
•§-
|J
!_
•i•e-
»
=,
_£
c
>
5
\ "
/
«
**' jS"^
/ \
is.
/
I
\
\
»,
H
al
4J
00
M
ai
H
tui
ot
S
/
I
*
i
IN
-
öl)
o
r-i
/
1
1
/
o
Z
H
ci)
•J
4-1
un
CO
m
a)
-C
>
1
1
01
co
1
1
1
1
CI)
>
O
1
—|—
VJ
P.
CM 0)
-Q
C)
C-)
f>
Cslc/3
01)
|-
\
1
(U
1
O
<r 4J
1
iJ
1
O
W
IfiS?-'
i
«
Sln ^
S 1
c
• T4
H
•a
I )
O.
|^
O
.C
|ï
a)
g
a)
1o
1
c/i
<N
T—1
U-l
Ü
1
i
1
i
*"
<l)
^r
S
J
S
\
I
<•>
où
•H
ta
ai
^m
^
m
ö
•H
H
011
>. a
ou
a
m
u
i
o
61)
o
w
i
c*
O
C
•r-1
m
4-1
G
cO
CJ
r
a
m
•H
4-1
•rH
(1)
cj
a.
011
o
i-t
al
>-i
m
c
•r-t
H
U
o
m
J2
4J
CU
•U
•H
C)
,—1
m
o
fi
al
u
on
ni
•H
XI
al
CJ
01)
>
Cl
4-»
C
C
Cl)
CO
cil
fi
u
eu
••H
71
C0
CO
1
M
m
oCN al
Cl)
.a
S3 r>
n
CM 4-1
ai
33
011
1CNl T - l
o
al
i-i
O
u
l
C.J
cil
1
al
•H
M U
n
t)
Cl-
OU
o
r-J
1
01
S5
fi
Cl)
•U
C0
01)
•H
ta
fi
011
C
01)
VJ
1
n
Cs
X
al
01)
o
t—i
1
•H
CN
C
f )
al
Cu
•H
1)
4J
.e
u
C
ou
o
o
c) r - 1
1
en Cl)
r-l
o
i)
O
c
al
c_)
.e
r->
-tf
00
C
•
eu
cal
CO
0)
• H
•a
102
valuesofthesodium carbonates selectedbyAl-Droubishows thatpirssonite
cannot forminequilibriumwithtrona.Becausepirssoniteandtronaoften
occurinthesamemineral assemblages innature,preference isgiventothe
values selectedbyGarrels&Christ.
Al-Droubi (1976)statesthatathighCO^-pressurestheassociation calcite+
nahcoliteisunstablewithrespecttogaylussiteandthatthelatterisfavored
byhighp ™ .Itisclear fromboth fig.8.4andfromreaction (12)thatthis
statement isincorrect.
The system Na-SO -CO -H 0
Themineralsburkeite,thenarditeandmirabilitecanberepresented inthe
systemNa-SO.-CCu-lLOinadditiontothesodiumcarbonatesmentioned insection8.3.Thenarditeandmirabilitearerelatedviathereaction:
Tn+10H20 X Mb
slopeo o
(14)
Thestability fieldofburkeiteinrelationtothesodiumcarbonatescanbe
determined assuming either thenarditeormirabilite (dependingonthewater
activity)orasodiumsulphatecontainingmineraltobestableoverthewhole
logCCL-loglLO-range.TheGibbsenergyofburkeiteasdeterminedinsection
7.5isusedtocalculateitsstability field (fig.8.5).TheGibbsenergyof
burkeitewasfoundusingGarrels'valuesoftheGibbs energiesofthesodiumcarbonatesanditis,therefore,consistenttousethisvaluetodeterminea
stability fieldofburkeiteinadiagram constructedwiththeGarrels&Christ
Gibbsenergyvaluesofthesodiumcarbonates.
The following reactionsdelineatethestability fieldofburkeite:
brk+
CO ? +
3brk
+ C0 ? +
3brk
+
brk+
C0 ? +
30H20
H?0
2nah+2tn slope -1
5H 2 0
6tn+2tr
"
-5
65H 2 0
6mb+2tr
"'-65
2mb+nat
"
c
^3
The system Na-Ca-SO -HO-CO
Fig.8.6showsalogpco 7 "l°ga H?0-diagramofthesystemNa-Ca-SO.H-O-CCL.This figureisacombinationoffig.8.4andfig.8.5andisconstructedundertheassumption thatbothaCaC0_-containingandaNa?SC\-containingmineralispresent overthewholelogPQ-, -logau Q -range.With
thisassumption,minerals suchaseugsterite,glauberiteandgypsum,whichalsobelongtothesystemNa-Ca-SO.-FLO-CO,arenotstable,whichisconsistent
(15)
(16)
(17)
(18)
103
logp
Fig. 8.6 Logpeg - logajjg-diagramofminerals inthe system
Na-Ca-SO,-CO„-H„0assuming aCaCO-and aNa„SO,-containingmineral
tobepresent over thewhole logp
CO„
loga
H20
range.
104
withtheabsenceoftheseminerals fromnatural assemblages containing sodium
carbonates.
Fig.8.6shows thattheburkeiteandgaylussitestability field contractin
comparisonwith thoseintheNa-Ca-ŒL-FLO (fig.8.4)andtheNa-SO.-CCL-FLO
(fig.8.5)systems.Theirbounding reactionis:
gay+2tn X brk+cal+5FLO
slope c^=>
(19)
Wheneverthesulphate concentrationpermitstheformationofburkeite,pirssonite
ismetastablewith respecttoburkeite+calcite.
Ingeneral,iftronaoccursinassociationwith thermonatrite,theconditionsof
formationmusthavebeensomewhere along line 1-2(fig.8.6),whichmeansap r n
-3S
belowatmospheric conditions (10 ' atm.)ifequilibriumhasbeenmaintained
duringtheprocessofprecipitation.Theassemblage trona-nahcolitemust have
formed under equilibrium conditions describedbyline 1-3,atap r n aroundor
higher thanatmospheric CO,-pressure.
The largerthenumberofminerals formedinequilibriumwith eachother,at
constantPandTandwithadefinitenumberofcomponents,themore restricted
arethepossible conditionsofformation.Forexampletheassemblage trona,
thermonatrite,burkeite,canonly formatconditions describedbyline1-7;
trona,thermonatrite,gaylussitebyline7-2;trona,nahcolite,gaylussiteby
line5-3,etc.(seefig. 8.6).
Explanation
of symbols used in fig.
8.4, fig.
8.5 and fig.
8.6
Theminerals occurringinfig.8.4containthecomponents Na~CO,andCaCO,
inadditiontoCCLandFLO.Ifthecompositionofthemineralsisrepresentedby
a linewithNa_CO,ononesideandCaCO,ontheother side,allsodium carbonates
plotontheNa7C0,-sideofthis lineandcalciteplotsontheCaCO,-sideofthe
line.Gaylussiteandpirssoniteplotinthemiddleofthelineasonemoleof
gaylussiteorpirssonite iscomposedofonemoleNa^CO,andonemoleCaCO,in
additiontoILO(seeinsetinfig. 8.4).Wheremoreminerals coincideatthe
samepoint alongtheline,theonethatappearsinacertainstability field
isnamed.Forexample
i
i means:inthis field thermonatrite isstable
inassociationwithcalcite.
tmDrs
i *i i means:inthis fieldpirssoniteisstableeitherwith thermonatrite
orwith calcite.Thermonatriteandcalcitecannot coexistheresinceatthe
boundarybetweenthesetwofieldsthethermonatriteplus calcitereacttogive
pirssonite.
Theminerals occurring infig.8.5containthecomponents Na^CO,andNa^SO.
inadditiontoC0 ? andFLO.Ifthecompositionoftheminerals isrepresented
2
105
onalinewithNa.CCLononesideandNa^SO.ontheother,burkeiteplotsata
distance2/3,becauseonemoleofburkeiteiscomposedoftwomolesofNa~S0.
andonemoleNa?CO_ (seeinsetinfig.8.5).Themineralsoffig.8.6canbe
representedinthecompositional triangleNa ? CO,,Na~SO.andCaCO,(seeinset
fig.8.6).
Forexample:
means:inthisfield,theassociationnahcolite,thenarditeandcalciteisastableassociation.
nah
tn
means:inthisfieldburkeiteisstableeitherwith tronaandcalcite
t r / '\tn orwith thenarditeandcalcite;tronaandthenardite cannot coexist,
cavA
means:inthis fieldgaylussiteisstableeitherwith thenarditeand
tr
tn tronaorwith thenarditeandcalcite;tronaandcalcitecannotcoexist.
Attheboundarybetweenthestability fieldsofthelasttwoexamples,thereactionburkeiteplus calcitegivesgaylussiteplus thenardite (crossingofthetwo
joinsinthetriangle)takesplace.
Thismethodofrepresentingstableassociationsofmineralsandtheircompositions along linesorintrianglesisextensively describedinKorzhinskii
(1959).
Salt assemblages
in the system Na-Ca-SO .-CO„-H„0
Theaimofthissectionistoinvestigatewhethertheassemblages belonging
tothesystemNa-Ca-SO.-CCu-h^Owhichhavebeenfoundinthefieldandinthe
evaporatingdishesareequilibrium assemblagesandwhether theseassemblagescan
beexplainedbymeansofthelogp„„ -loga„„-diagrams previouslydiscussed.
Intable8.4theagreementsanddifferencesbetweenthetwosetsofsaltassemblageshavebeengiven.
Lake Victoria.
Thesamples alongLakeVictoriaatSimbi,MitiBiliandKanamall
containtrona.
AtMitiBilithecalculatedequilibriumCCL-pressureofthegroundwater
originating fromthespringdecreasesdownstreambutisalwayshigher thanatmosphericCCL-pressure (seetable 8.1).ThesaltssampledatlocalityMitiBili1
(seefig.4.3)precipitated fromsurfacewater.Water coming fromaspringflows
overthesoilsurfaceandevaporates.TheCCu-pressure liessomewherebetween
thecalculatedvalueofthespringwater (10~ ' bar)andatmosphericCCL-pres-35
sure (10 ' bar).Ifthesaltassociation foundaroundthespring (trona,calcite,gaylussite,nahcolite)hasformedinequilibriumatp
m
of10 ' bar
-257
106
Table 8.4 Differences and similarities of salt assemblages belonging tothe
systemNa-Ca-SO,-CCL-H.O found inKenya and those inevaporating
dishes precipitated fromwaters sampled inKenya.
Kenya (see table 4.3)
Lake Victoria
Evaporation experiments (see table 8.1)
Lake
Victoria
- trona ispresent inallassemblages -tronadoesnot occur inall assemblages
- thermonatrite isnot always present - thermonatriteoccurs inall assemblages
- nahcolite occurs
-nahcolite doesnot occur
- burkeiteoccurs
-burkeite occurs
- gaylussite occurs,not in
-gaylussitedoesnot occur
associationwith burkeite
- pirssonite doesnotoccur
-pirssonite doesnot occur
- nitrates donot occur
-nitrates occur
Amboseli
Amboseli
- tronaoccurs inall assemblages
-trona doesnot occur inall assemblages
- nahcolite doesnot occur
-nahcolite doesnot occur
-,aphtithaliteoccurs,no sylvite
-both aphtithalite and sylvite occur
- burkeite doesnot occur
-burkeite doesnotoccur
- bothpirssonite and gaylussite occur -pirssonite and gaylussite donot occur
(point5,fig.8.6),degassingmusthavetakenplacebeforeprecipitationaccordingto:
2HC0~*COj"+C0 2 +H 2 0
(20)
Attheother localitiesatMitiBili,thesaltefflorescencesprecipitated from
waterwhichwasbroughttothesoilsurfacebycapillaryrise.AgaintheCCLpressureofthesewatersmustbesomewherebetweenatmosphericCCL-pressureand
thecalculatedCCL-pressureofthegroundwaterwhichishigher thanatmospheric
CCU-pressure.Thesaltassociations formedinthiswayshowassemblagescharacteristicforlowCCLpressures,iftheyrepresent equilibriumassemblagesof
minerals.Forexample,theassemblagetrona,thermonatrite,gaylussitewould
have formedunder conditionsdescribedbyline7-2infig.8.6whichmeansat
CCL-pressuresbetween 10~ " and10" " bar.Suchadropinp m undernatural
circumstancesishighly improbable.Therefore,theseassociations (formedunder
conditionsofcapillaryrise)mustbeassumedtobenon-equilibriumassemblages.
Theformationofnahcoliteisapparently inhibitedalthoughthisisastable
sodiumcarbonateunderthecalculatedCCL-pressuresinthewaters.
ThepresenceofburkeiteatMitiBiliexplainstheabsenceofpirssonitein
theassemblages.Burkeitedoesnotoccurinsamplesinwhichgaylussiteispresentandthesetwomineralscanonlycoexistatonespecificwateractivityat
107
whichreaction (19)takesplace (seefig.8.6).
Amboseli.
Thesaltssampled atAmboseli (seetable 4.3) originated by evaporation
ofgroundwater.ThepartialCCu-pressuresoftheAmboseli groundwatersvaryfrom
-074
-305
10 ' to 10 ' bar (seetable 8.1), well aboveatmosphericCCu-pressure.
JustasalongLakeVictoria,tronaispresent inallsamples containing sodium
carbonates.Nahcolitehasnotbeenfoundandthermonatriteiswidespread.Once
more,manyassemblages occurwhichcouldonlyhave formedundermuch lowerCCupressures -iftheyprecipitated underequilibrium conditions.
Because someofthewatersatAmboseli contain significant amountsofpotassium,aphtithalite isformedboth inassociationwithpirssonite andwithgaylussite.Theaphtithalite controls theactivityofsulphate insuchaway that
burkeitedoesnotform.Theabsence ofburkeite inthesesamples explains the
presence ofpirssonite.
Konya Basin. IntheKonyaBasininTurkeysodium carbonates occuronly ina
restricted areaaroundAkgöl,atlocalities T15,T26,T31 andT32.The carbonate
associations areverysimple,mainly trona,burkeite.Thisassociation isstable
overavery large logpçg -logau Q rangebelow log PQ-^ =-2.61 bar (fig.8.5).
TheCCU-pressuresofthewaters atthese localitiesvarybetween 10 ' and
10"3.19bar.SomeoftheseCO,pressures fallwithinthetrona,burkeitestabilityrange.Incaseofthehighervaluesdegassingwillhave takenplacebefore
precipitation ofthesaltassemblage.
Evaporating
dishes. A fewwater samples taken intheLakeVictoria areaproduced
nitratesuponevaporation inthe laboratory.Thenitrate concentrationofthe
waterswasnotdetermined atthe timeofanalysisbut thehighnitrate content
explains insomecasesthegreatdifferencebetweenanion-andcationconcentrationoftheanalysed ions (seetable4.1).For example thewaters fromSimbi
appeared tobeverynitrate-rich afterevaporation inthe laboratory. Inthe
field,however,onlycarbonates -nonitrates -were found.
Thewaters fromthespringatMitiBilishowanotherremarkablefeature.
Fromawater sample takenbefore therainy season,nitratesprecipitated inthe
evaporationdishes inassociationwithcarbonates.Nonitrates formed atall
fromawater samplefromthesamespring collected onemonth laterafterabundantrainfall.Thesamephenomenonwasobserved ingroundwatersamples takenin
adryriverbed atKanam (fig.4.3).
Intheevaporationexperiments itwasobserved thatthermonatrite isthe
108
firstsodiumcarbonatetoprecipitate,sometimes inassociationwithtrona,in
mostdishes inwhich anysodiumcarbonatesprecipitated.Fromtherimdownwards
towardsthecentreofthedishestheratiooftronatothermonatrite increases.
Theinitial(XL-pressureofthewaters liessomewherebetweenthecalculated CO ? pressure (seetable 8.1) andatmospheric(XL-pressure.Apparently,thesame
process takesplace inthedishes asundernatural circumstances.Thermonatrite
forms intheexperimentsalthoughnahcoliteortronaarethestablesodiumcarbonates at(XL-pressures that arehigherthantheatmosphericvalue.Thesameis
observed inthefield-Thiscanbeexplained asfollows:Once thermonatrite
precipitates,twoprocessesoccur simultaneously:
1.Evaporationcausesallconcentrations insolutiontoriseand thereforealso
the(XL-content.Itiswellknownthatthediffusionof(XL throughthesolutionair interfase isaslowprocess andthe(XL-pressureofthesolution,therefore,
willincrease.
2.Thesodiumconcentration ishigher thanthealkalinity inmostwater samples
fromwhichthesodiumcarbonatesprecipitated. Inthecaseof thermonatrite
precipitationthesameprocesswill takeplace asinthecaseofcalciteprecipitationwhichwasexplained insection 5.4.Duringprecipitationofthermonatriteand furtherevaporation,thesodium concentrationwill increaseandthe
alkalinitywilldecrease causing the(XL-pressure ofthewatertodecrease.
Thesetwoprocesses counteract eachother and,therefore,itis impossible
topredictwhetherthe(XL-pressure ofthe solutionwill increaseordecrease
during thermonatriteprecipitation.An increasing(XL-pressure favours theformationoftronaoverthemonatrite.
Gaylussiteandpirssonitedidnot form intheevaporationexperiments.In
thosecaseswhere gaylussite formedundernatural conditions from carbonate-rich
samples takenfromtheLakeVictoriaarea,burkeite formed intheevaporating
dishes.Thismightbeduetothefollowing tworeasons:
1.DegassingofCO,causestheCa-concentrationtobecometoolowfortheformationofgaylussite.Duringfractionalcrystallisationcalcite isno longer in
contactwith thesolutionandcannotreacttoformgaylussite asitwould dounderequilibriumconditions.
2.TheactivityofH~0reachesalowervalue intheevaporatingdishes thanundernaturalcircumstances.At lowwateractivities gaylussite ismetastablewith
respecttoburkeite (seefig.8.6),provided thatsufficientsulphate ispresent.
109
Itmustbeemphasized that theconclusions inthis section,whichhave been
drawnon thebasisof thermodynamic data,arevalid only ifthelatter arevalid.
Thus,thermonatriteprecipitates either asametastable phasewithrespect to
tronaandnahcolite.or thethermodynamic data onwhich this conclusion isbased
arenotvalid.
8.4 SOLUBILITY CALCULATIONS INTHE SYSTEM Na-K-Mg-Ca-Cl-S04-H20
Most saltminerals intheKonyaBasin inTurkeybelong tothe system
Na-K-Mg-Ca-Cl-SO.-FLO,theso-called "seawater system".Much experimental work
hasbeendone inthis system inorder toexplain salt sequences inmarine evaporitedeposits.Especiallyvan 'tHoff andcoworkers (1905,1909,1912} determined
manyphasediagrams inaneffort toelucidate theseawaterproblem.Phase diagrams
areuseful as longas the system isnottoocomplicated. Formulticomponent systemshowever,phasediagrams have their limitations andmineral solubility predictions insuchsystems require atheorywhich describes thebehaviour ofelectrolyte solutions inanaccurateway.First ashortreviewof thedevelopment of
the theory describing suchsolutionswillbegiven.
The activity ofaspecies iisdefinedas:
a. =y.m.
l
(221
v
'l l
J
where a. =activity ofspecies i
y•=activity coefficientof speciesi
m. =molality ofspecies i.
ForanelectrolyteA ,B which dissociates inv+kationsA and v-anions B
(v++v-=v ) ,themean ionic activity coefficient y isused,which isdefined
as:
iV/
ïf-ïr
(23)
Indilute solutions themean ionicactivity coefficient canbe calculated using
theExtended Debye-Hückelequation:
-A|z z _ | V f
l o g Y+=
-
^=
1+BaVl
whereA and Bareconstantswhich involve the temperature and thedielectric
constant inadifferentway
(24)
110
a =distanceofclosestapproachoftheions
2
I=ionicstrength,definedas I= \ Em.z.wherem. isthemolality of
everydifferent ionicspeciesiand z. isthechargeofioni
z+=electricchargeofthekation
z_= electric chargeoftheanion.
TheExtendedDebye-Hückelequation isvalid forsolutionswithanionicstrength
lessthan0.1.Scatchard (1936)describes thedeviationathigherionicstrengths
betweentheexperimentalmean ionicactivitycoefficient andtheactivitycoefficientcalculated fromtheExtendedDebye-Hückel equation.Thenewequation
becomes:
-A|Z+Z ivr
logy + =
=—-+B!I
(2S)
1+BaVT
Thefirstterm isthenormalExtendedDebye-Hückelequationandthesecond term
iscalled the"Scatcharddeviationfunction".Scatchard foundthatthefunction
B.appearstobeeitherconstantoverawiderangeofionicstrengthsor isa
slowlyvarying functionofI.
Harned (1935)didmanyexperimentswithmixedelectrolytesandhe foundthe
followingrelation,knownas"Harned'srule",betweentheactivity coefficient
ofanelectrolyte Bandthemolalityofanother electrolyte C.
log Y B =i°gy B ( 0 ) - < v c
(26)
wherey R ,,istheactivitycoefficient ofBinasolutionwhichcontainsonly
saltBanda R isamixingparameter,called"Harned'.scoefficient" (Harned,1935;
Robinson&Stokes,1970).
Wood (1972,1975)combined theScatcharddeviationfunctionwithHarned's
ruleresultinginanequationwhichappearedtobeapplicabletoconcentrated
brinesandwhichagreedratherwellwithexperimentaldata.Hisequationisas
follows:
logY + =
-A|zz IVF 2vv_
n
'-=+
(B.+S^ y J I
1
1+BaVT
v
j=11;>1
(j*i)
(27)
where a- • isaconstantmixingparametercharacterizingtheinteractionofthei
andjcomponents
y- istheionicstrengthfractionofthej component.
111
Pitzerandcoworkers (1973,1974,1975)describethethermodynamicpropertiesofbrinestohigh ionicstrengthsbyavirialexpansionoftheactivity
coefficientusing second,thirdandhighervirial coefficients.Referenceis
giventoPitzeretal (1973,1974,1975)andHarvie&Weare (1980)astheequationsarerathercomplicated.Scatchard (1968)andPitzer (1973)indicated that
theuseofonlyasecondvirial coefficientisnotsufficienttopredictthermodynamicpropertiesofsolutions aboveaionicstrengthof4.Thirdorhigher
virialcoefficientsarerequired.RecentlyHarvie&Weare (1980)developedthe
methodofPitzerandcoworkers (1973,1974,1975)formineral solubility calculations.Harvie&Weare (1980)objected againsttheWoodmodelasthisisonly
a secondvirialcoefficientmodel.Another objectionagainsttheWoodmodelis
thattheHarned's coefficientswhichareusedinthemodelarenotinternally
consistent (deRooij,personalcommunication).Harvie&Weare (1980)demonstrated
thatsecondandthirdvirial coefficientsaresufficienttoperformmineral solubilitycalculations.Theyderived chemicalpotentialsandthirdvirialcoefficients fromternaryexperimentalphasediagrams.Unfortunately,theircalculationsdidnotyetincludeparametersforsystemswithcarbonates.
ThecomputerprogramofHarvie&Wearepermitsthepredictionofprecipitation
ofmineralsequences from solutions,forexamplethesequencesinmarineevaporites (Eugsteretal, 1980;Harvieetal,1980).Withtheprogrambothsequences
formedunderequilibrium conditionsandsequences formedduring fractional crystallisationcanbetraced.Precipitationunder equilibrium conditions allows
backreactionofpreviously formedmineralswiththesolution.During fractional
crystallisationamineralwhichonceprecipitatedisnolongerincontactwith
thesolution.
The system Na-Mg-SO -H 0
Mineralsprecipitating fromasolutionintheternarysystemNa-Mg-SO.-H^O
canbedepictedinam^ a gg -n w g o -diagram,whichisshowninfig.8.7.This
diagramhasbeencalculatedbyHarvieandWeare (1980)fromexperimentaldataat
25°C.'
Suchdiagrams indicatewhichmineral(s)areinequilibriumwiththesaturated
solutionaswellasthecompositionofthesaturated solution.If,forexample,
a solutionwithaninitialcompositionPevaporates,mirabilitewillbethe
firstprecipitateinequilibriumwithasolutionwithcompositionQ.Duringfurtherevaporationandprecipitationofmirabilite,thecompositionofthesolution
will changealongthesaturation lineofmirabiliteuntilthesolutionbecomes
112
2.0
mirabi1ite
y
bloediteN
P./
1.0
/
/
SOLUTION
1 .0
2
0
„„
.. ,
epsomiteV
-°
"'MMgSO.
en
3
'°
Fig. 8.7 Calculated mineral solubilitiesinthe systemNa-Mg-SO-H„0at25C
(fromHarvie&Weare, 1980)
alsosaturatedwithrespecttobloediteatcompositionR.Thephaseruleshows
thatRisaninvariantpointatconstant temperatureandpressure.Therefore,
thecompositionofthesaturated solutionnolongerchanges onceithasreached
compositionR.Atthispointbothmirabiliteandbloeditewillprecipitate.It
isevident thattheNa?S0./MgS0.ratiointheinitial solutiondetermineswhether
thefinalprecipitatewillconsistofmirabilite+bloediteorofepsomite+
bloedite.
The system
Na-Mg-SO
-Cl-HO
Additionofchlorinetotheternary systemoftheprecedingparagraph results
inthequaternaryreciprocalsystemNa-Mg-Cl-SO.-FUO.Fig.8.8isaJänecke
diagramofthissystemat25 C,ascalculated recentlybyHarvie&Weare (1980).
Theseauthors claimabetteragreementwithexperimentaldatathananyearlier
effort.Onthechlorine-freerighthand sideofthesquare,thesame sequence
mirabilite-bloedite-epsomiteappearsasintheternarydiagramoffig.8.7.
InthisJänecke typeofdiagramonlytheratioofthecomponentsinasaturated solutioncanbefoundandnottheirabsolute concentrationsasinfig.8.7.
The labelsinthefields indicatethefirstmineral specieswithwhichasolution,
plottinginthatfield,becomes saturated.Forexample,thenarditewillbethe
firstmineraltoprecipitatewhenasolutionwith initialcompositionPevaporates.During furtherevaporationandthenarditeprecipitationthecompositionof
thesolutionwillmoveawayfromthesaltpointofthenardite (thepointatwhich
solidthenarditeplots)alongaline throughtheinitial compositionP(see
113
bischofite
kieserite
MgS04
~Q-eps.
hxh.ks
hexahydrite
halite
epsomite
bloedite
hal.
Fig. 8.8 Jänecke diagram of thereciprocal systemNa-Mg-SO,-Cl-HOat 25 C
fromHarvie &Weare (1980).For abbreviations of saltpoints,see table3.2.
Inaddition,bi:bischofite (MgCl .6H0 ) ;ks: kieserite (MgSO .H0)
arrowinfig.8.8)untilthecompositionofthesolutionreachesthefieldlabelledbloedite.Atthis compositionQboththenarditeandbloeditewillprecipitate
andthecompositionofthesolutionwillchangealongtheboundarybetweenthe
thenarditeandbloeditefieldsuntiltheinvariantpointRisreached,where
thenardite,bloediteandhaliteareinequilibriumwithasaturated solution
withcompositionR.Thefinalprecipitatewill consistofthesethreeminerals.
Thesquarecanbedividedinthreesubtriangles (seefig.8.8).Allinitial
solutionsplottinginsubtriangleNa2Cl2-Na2SO.-bloeditewillterminateatinvariantpointR;allinitialsolutionsplottinginsubtriangleNa2Cl2-bloediteMgSO.will terminateatinvariantpointSandallinitial solutionsplottingin
subtriangleNa ? Cl ? -MgCl 2 -MgS0.willterminateatinvariantpointT.Ateachinvariantpointthreemineralsareinequilibriumwithasaturatedsolution.
The system Na-K-Mg-Cl-SO -H 0
ThequinarysystemNa-K-Mg-Cl-SO.-rLOcannotberepresentedinatwodimensionaldiagramunlessanadditionalrestrictionismade.Fig.8.9isaJänecke
projectionofthisquinarysystemat25°Cwiththerestrictionthatallmineral
114
Fig. 8.9 Jäneckeprojection at 25 Cof the systemNa-K-Mg-Cl-SO,-H„0calculated byHarvie &Weare (1980).Allmineral zonesdenoted are saturated with
respect tohalite.Abbreviations of saltpoints,see fig.8.7.Inaddition,
leo: leonite (KMg(S0 ) .4H„0);kai:kainite (KMgClSO,.3H 0 ) ;car: carnallite
(KMgCl3.6H20).
zonesarealsosaturatedwithrespecttohalite.Alongthepotassium-freeMg-S0.
sideofthetrianglethesamemineral sequencecanbeseen (thenardite,bloedite
epsomite,hexahydrite,kieserite,bischofite)asthesequence alongtheboundary
ofthehalitefieldinfig.8.8.
The labelsofthefieldshavethesamemeaningasthelabelsinfig.8.8and
theevolutionofasolutionwhichissaturatedwithrespecttocertainminerals
isalsoderivedinthesameway.ForadetaileddescriptionoftheuseofJänecb
diagrams,thereaderisreferredtoBraitsch (1971).Boththeratioofthecompo
nentsinthesaturated solutionandtheratioofthequantityofmineralswhich
haveprecipitated fromthissolutioncanbederived fromthisdiagram.
Thediagramcanbedividedinsubtrianglesbyjoiningthesaltpointsofthe
differentminerals.Thesubtriangleinwhichaninitial solutionplotsdetermine:
theinvariantpointatwhichtheevaporationwillend.Forexample,initialcompositionsplottinginsubtrianglethenardite-aphtithalite-bloedite willterminât'
atinvariantpointSatwhichpoint thenardite,aphtithaliteandbloeditecoprecipitatewithhaliteinequilibriumwithasaturated solutionwithcompositionS
Threemineralsareinequilibriumwithasaturated solutioninadditionto
115
haliteataninvariantpointinthissystem.
The system Na-K-Ca-Mg-Cl-SO -H20
ThephaserelationsinthesenarysystemNa-K-Ca-Mg-Cl-SO.-1-LOcanberepresented inatetrahedronwithMg,K,CaandSO.atthecornersandallmineral
zonessaturatedwithrespecttohalite.Thistetrahedroncanbesubdivided into
subtetrahedra formedbyfoursaltpointsofminerals.Evaporationofaninitial
solutionplottinginacertainsubtetrahedronmust terminateattheinvariant
pointatwhichthefourminerals formingthecornersofthesubtetrahedroncoprecipitate.Fourmineralsareinequilibriumwithasaturated solutioninadditiontohaliteataninvariantpointinthissystem.
Itisevidentthatthephaserelationsaretoocomplicatedtovisualizein
twodimensions.TheprogramofHarvie&Weare (1980),whichhasbeendeveloped
forthissystem,permitsthecalculationofthemineral sequences thatprecipitate fromasolutioninthissenary system.
Salt assemblages
in the system Na-K-Mg-Ca-Cl-SO -H„0
Somestrikingdifferencesbecomeapparentbycomparingthesaltassemblages
foundintheKonyaBasinandtheassemblagesintheevaporatingdishesprecipitated fromthewaters sampledintheKonyaBasin.Table8.5showsthesimilarities
anddifferencesbetweenthetwosetsofsaltassemblages.Theassemblagescontaining sodiumcarbonateshavenotbeentakenintoaccountinthistable(see
table8.3).
Theanalysesoffivecarbonate-poor samples, 19, T8,T5,T11andT13from
Turkey (forchemicalanalyses,seetable3.1)andthecarbonate-poor Luandasample fromKenya (seetable4.1)were selectedandrunwiththecomputerprogramof
Harvie&WeareinSanDiego,CaliforniabyNancyMöller-Weare.Thecalculation
wasdoneassumingthatequilibriumprecipitation takesplace.Beforerunningthe
samples,thewateranalyseswereallchargebalancedwithinthestatedpercent
deviation.Thenecessaryconcentrationchangesweremadeonthemajorion(ifobvious)oronallthemajor ionsiftheir concentrationsweresimilar.Thecarbonate-ionswere subtracted fromthesolutionsbyassuming calciteprecipitation.
Inallselectedsamplesgypsumwasthefirstmineraltoprecipitate.Glauberiteformedinthenext stagebybackreactionofgypsum.Halitewasthethird
mineralwithwhichthesolutionsbecamesaturated,followedbythenarditeand/or
bloedite.Thepotassium-bearingminerals formedinalaterstage.Accordingto
116
Table 8.5 Differences and similarities of salt assemblages belonging tothe
systemNa-K-Mg-Ca-Cl-SO^-H^O found intheKonya Basinand those in
evaporating dishesprecipitated fromwaters samples intheKonya
Basin.For abbreviations ofminerals,seetable 3.2.
Konya Basin (see table 3.3)
Evaporation experiments (seetable 8.2)
bloedite occurs inmany assemblages
inthe followingassociations:
bl-hxh-gps
-hal-gl
-eps-gps
-hxh-sch-hal
-eug
-hal-gps
-gps-lw
-hal-stk-kon
-hal-gps-kon
-hxh-gps-kon
bloedite occurs inonly twodishes in
thefollowingassociations:
bl-hal-hxh-kon-gps
-hal-hxh-kon-gps
thenarditeoccurs inthe following
associations:
tn-gl
-aph
-hal-eug
-gps-hal-eug-gl
-hal-aph
thenardite doesnot occurwith
Mg-bearingminerals
thenardite occurs inthe following
associations:
tn-gps-hal-hxh
-gps-ha1-hxh-kon
-hal
-hal-lw-hxh
thenardite occurswithMg-bearing
minerals,notwith bloedite
loeweite occurs only once inthe
following association:
lw-gps-bl
loeweite occursvery often inthe
following associations:
lw-hal-b1-hxh-kon
-gps-hal-hxh
-hal-hxh-tn
-gps-hal-kon
-gps-hxh-hal-kon
-hal-cal
-gps-eps-hxh-hal
-hxh-hal
aphtithalite istheonly potassium
bearingmineral,nopolyhalite
nopotassiumbearing minerals
glauberite occurs
no glauberite
eugsterite occurs
no eugsterite
notalwayshalite
alwayshalite
konyaite occurs rarely
konyaite occurs often
117
thecomputer calculations thefollowingmineralsprecipitate fromthe selected
samplesunder theassumptionofequilibrium precipitation:
gypsum
[-backreaction
glauberite
halite
thenardite
bloedite
bloedite
polyhalite
aphtithalite
The sampleswererununtil SL, „reached avalue ofabout 0.70.At thiswateractivity twosamples (T9and theLuanda sample)ended atan invariant point
(bloedite,thenardite,aphtithalite,glauberite,halite).
Itmustbe concluded from acomparisonofboththenatural salt assemblages
and thesaltassemblages formed inthe laboratorywith theequilibrium sequences
calculatedwith theprogram ofHarvie &Weare,that theassemblages formed inthe
field,mainly due toévapotranspirationofgroundwater,approachmuch better
precipitation under equilibrium conditions than theassemblages formed inthe
evaporatingdishes.
There issomeevidence forthenon-equilibrium conditions inthedishes.
Eugsterite can,asglauberite,beregarded asamineralwhich formsattheexpenseofgypsumduring equilibriumprecipitation,although it isametastable
mineral contrary toglauberite.Eugsterite andglauberite arebothabsent inthe
evaporating dishes.Theminerals loeweite and konyaite,bothbeingmetastable
minerals atambient conditions,formed often inthedishes.Furthermore,thenarditeandeitherhexahydrite orepsomitecannot form inequilibriumwith each
other inthesameassemblage (seefig.8.9).Inthedishes thenardite occurs
often inassociationwithhexahydrite.This association canonlyhave formed
during aprocess offractional crystallisation.
Inthefield equilibrium conditions predominate.Botheugsterite and glauberite,whichprecipitate bybackreaction ofgypsumwith thesolution,have formed
widespread undernatural circumstances.Apparently all thenardite,ifithas formed inMg-rich solutions,hasbackreactedwith thesolution toMg-bearingmineralsbecause thenardite doesnotoccur inassociationwithaMg-bearing mineral
inthefield,notevenwithbloedite.Notallassemblages found inthe field
118
have formedunder equilibrium conditions.Thecoexistence ofthenardite and
gypsumand thepresence ofeugsterite andkonyaiteprove thatalsoduring évapotranspirationboth fractional crystallization and formationofmetastable
minerals takesplace.Konyaite evenhaddisappeared insome samples twoyears
after sampling,whichproves themetastability ofthismineral.Suchnonequilibriumprocesses occuronalessextensive scale than inthe evaporating
dishes.
Prediction ofmineral assemblages insalt efflorescences and in evaporating
dishes isdifficult,evenifaccurate activity coefficients forbrines areavailable. Inbothcases theprocesses areamixture ofequilibrium conditions,fractionalcrystallizationand formationofmetastableminerals.The salt assemblages
innatural saltefflorescences arebestpredicted by assumingprecipitationunder
equilibrium conditions,whereas themineral assemblages inevaporating dishes in
the laboratory arebestpredicted byassuming theformationofmetastablemineralsand fractional crystallization.
119
STABLE ISOTOPES IN WATERS FROM THE KONYA- AND
AMBOSELI BASINS; A RECONNAISSANCE STUDY
9 . 1 INTRODUCTION
Inrecentyearsthestableisotopecompositionofnaturalwatershas increasinglybeenusedtotracetheoriginofthesewaters,andtheirsubsequent
historyduringevaporationand/or interactionwithsoilsandrocks.Thisstudy
isprimarily concernedwiththeevolutionofevaporating solutionsandtheprecipitationofsaltminerals fromthesesolutions.
Itwasfeltthatthestableisotopiccompositionofthewater samplesfrom
theKonyaBasininTurkeyandfromtheAmboseliBasininKenyacouldgiveadditionalinformationontheprocesses involved suchastheevolutionofthe
evaporating solutionsandtheprecipitationofsaltminerals fromthesesolutions.
Theopportunitytodeterminetheisotopiccompositionofthewatersamples arose
duringthepresentstudybutbecausethesampleswerenottakenwiththisaimin
mind,onlythosesamples couldbeusedwhich fulfilledtherequirementsfor
isotopicanalysis.
9.2THESTABLE ISOTOPESOFHYDROGENANDOXYGEN
Hydrogenandoxygen,bothconstituentsofwater,occurasmixturesofdifferent isotopes,namelyhydrogenwithmass 1(99.9841),deuteriumwithmass2
(0.016V)andthree isotopesofoxygenwithmasses 16(99.761),17(0.041)and
18 (0.201).Thedifferentmassesoftheseisotopes causethemtobehavein
slightlydifferentwaysinchemicalreactions,althoughtheirchemicalpropertiesareidentical.Consequently,theisotopiccompositionsoftwooxygenand/
orhydrogenbearing compoundswhichareinequilibriumwill,ingeneral,be
slightlydifferent.Thisdifference,definedastheequilibrium isotopefractionation,increasesastheequilibriumtemperatureislowered.Ingeneralthe
heaviest isotopeisconcentratedinthephaseinwhichitismost strongly
bound.
Theisotopiccompositionofasampleismeasuredwithamassspectrometer.
Unknownsamplesaremeasuredrelativetoastandardofknownisotopiccomposition.Thestandardusedforwater analysesistheso-called SMOW (Standard
120
MeanOceanWater). Isotopicvalues arecommonlyreported inthe6-notation, in
which the isotopic compositionofanunknownsample (x)isrelated tothe isotopiccompositionofastandard sample (st)inthefollowingway:
R -R
6X = ~ K. 51.1000
st
1ft16
1ft
In this formula Rdenotes theisotope ratios D/Hor 0 / 0 .Consequently 6 0
and 6D ofSMOW are0.The6-value isexpressed in /oo(permil) andhasan
analytical precision which isingeneral oftheorder of1 /ooforÖDand
o
1ft
0.1 /oof o r ö 0.Positive values of6 indicate that thesample is enriched
1ft
inDor 0relative tothestandard andi s ,therefore,heavier; negative values
indicate that thesample hasless oftheheavy isotope than thestandard and is,
therefore, lighter.
9.3 FACTORS AFFECTING THEISOTOPIC COMPOSITION OFNATURAL WATERS
The oceans arethelargest reservoir ofwater ontheearth'ssurface.All
meteoric waters andthelakes, rivers andgroundwaters originating from themare
directly orindirectly derived from theocean reservoir.
The isotopic compositions ofmany meteoric waters ranging from tropical rainwater toAntarctic iceandsnow have been measured (Craig, 1961; Epstein et al,
1965, 1970). Forthese waters there isa linear relationship between 6Dand6 0
givenby:
6D =86 1 8 0+5
This line, called theMeteoric Water Line (MWL), determines theisotopic composition ofalmost allprecipitations asa function ofgeographic latitude,
topography, andlocal climatic conditions.
1ft
In older papers this equation isalso given as6D=86 0+ 10because at that
time theAntarctic precipitations hadnotyetbeen analysed.
The linear relationship between thehydrogen andoxygen isotopic compositions
is caused bythefact that thecondensation ofwaters from theearth's atmosphere
in theform ofrain orsnow isessentially anequilibrium fractionation process,
and that allofthese waters derive ultimately from a single source ofwater,
namely ocean water. Thefractionation ofD/Hisproportional tothe 0/ 0
fractionation intheratio of8:1.Astemperature decreases, both fractionations
increase proportionally.
Some deviations from thelinear relationship given above dooccur. Inthe
Eastern Mediterranean,which isrelevant tothediscussion oftheisotopic
1ft
121
results oftheKonyaBasin,theprecipitations followa linegivenby (Gat,
1966):
6D=86 18 0+22
Otherexceptions tothenormal linearrelationship are the isotopiccompositions ofwaters fromclosedbasins.Waters inclosedbasinsareusuallyconcentrated salinewaterswhich formbyevaporationofprecipitationwaters.The
isotopiccompositions ofsuchwaters alsoshowa linearrelationship,buthave
a slopeof 1-6, instead of8as fornormalwaters.Suchlines,whicharecharacteristicforevaporationprocesses,intersect theMWL at thepointofthemean
oftheprecipitations inthatparticulararea.
Another typeofwaterswhichdonot followtheMWLarethe (usuallyhot)
springwaters,thathavehadameasurabledegree of interactionand exchange
with solidrocks.Rocks constitue alargereservoir ofoxygen,becausealmost
allminerals canbeconsidered ascompoundsofoxides,butonlyasmallreservoirofhydrogen,asmostminerals containno,oratmostonlyasmall amount of
hydrogen.Water,onthecontrary,isa largereservoir bothofoxygenandhydrogen. Ifacertainamount ofwater interactswith arockbody,withwhich itis
not inisotopicequilibriumunderthegivenconditions,the isotopeexchangewith
therockwill causeasignificant shiftoftheoxygenisotopiccompositionof
thewater.Thehydrogen isotopiccompositionremainsvirtuallyunchangedduring
thisprocess,because thereisrelatively littlehydrogentoexchangewith.
Theseeffects areessentially causedbytherelative sizesofthe exchanging
reservoirs. If,forexample,hydrogenandoxygen isotopes areexchanged between
1kgofwaterand 1.6 kgofkaolinite,amineral certainly richer inhydrogen
thanmostminerals,the finaloxygenisotopecompositionwould be for 501determined by thecontributionofthekaolinitewhereas thehydrogen isotopiccompositionwould onlybedetermined for less than Z% bythekaolinite contribution.
Thismeans thattheoriginalD/H ratioofthewaterwould hardly changeby the
water-rock interaction.Fig.9.1 showstheMeteoricWaterLine (MWL),theEast
MediterraneanWaterLine (EMWL),andtheevolutionofthe isotopic composition
ofaspecificprecipitateduringevaporationandwater-rock interactionrespec18
tivelyina6D-6 0diagram.
Notonly the isotopiccomposition ofanaturalwater changesduringevaporation,butalsoits chemicalcomposition changesasalinearfunctionofits
degreeofevaporationas longas itisnot saturatedwithrespect tooneormore
saltminerals.Ontheotherhand,awater canalsochange itschemical com-
122
S,80 or 8°
&D(°/oo)
60-
^„pnTition Une
40
dissolution of salts
without
I
-10
-
8
-
I
6
I
-
4
I
-
2
I
0
I
2
1
4
evaporation
1
s a l ini ty
6
,8
S 0(% o )
F i g . 9 . 1 E a s t M e d i t e r r a n e a n Water L i n e
(EMWL) D = 8 6 1 8 0 + 2 2 . M e t e o r i c Water
L i n e (MWL) D = 86 0 + 5 . E v a p o r a t i o n
t r e n d , slope 5. Water-rock i n t e r a c t i o n
trend, slope + 0.
Fig. 9.2 The isotopic compositionof anevaporatingwaterbody
undernatural circumstanceswith
a given relativehumidity reaches
a steady state.Dissolutionof
salts increases the salinity
without changing the isotopic
composition.
positionbydissolutionofsaltswithout changingitsisotopic composition.
The simplecaseofevaporationofasinglebatchofwaterinacompletely
dryatmosphere (relativehumidity0)is,ofcourse,never realized innature.
ExampleswhichcloselyapproachthissituationhavebeendescribedbyFontes&
Gonfiantini (1967).Inmost casesofevaporationundernatural conditions,with
a givenhumidity,inwhichthevaporhasadefined isotopic composition,the
isotopiccompositionoftheevaporatingbodyofwater reachesasteadystate.
Theprocessiscomplicatedbythefactthatthereisusuallyacontinuousinflowofnewwater intotheevaporatingbody.Themathematicsofsuch situations
becomerather complicatedandarebeyondthescopeofthis study.
Theevolutionoftheisotopic compositionofanevaporatingbodyinwhichno
evaporation,butonlydissolutionofsaltstakesplace,isshown schematically
18
ina6Dor6 0versus salinitydiagraminfig.9.2.
Itcanbeconcluded thatthehistoryofanaturalwatercanbetracedby
measuringitschemicalanditsisotopiccomposition.Itbecomespossibleto
determinetheisotopiccompositionoftheprecipitationinaspecific areaand
123
todistinguishbetweeneffects ofevaporation andofinteractionsbetweenwater
androcksorspecificminerals.
9.4 ANALYTICAL PROCEDURESANDDATA
For the6Ddetermination,thewaterswere reduced tohydrogengaspriorto
o
18
analysisbyreacting thewaterwithuraniummetal at 800 C.For the6 0
measurement,thewaterswere equilibratedwith small amounts ofCO,ofknown
isotopiccompositionat 25 C;thefractionationbetweenwater andCCLis
accuratelyknown.Analyseswereperformed byDr.R.KreulenandJ.Meesterburrie
ofthedepartment ofGeochemistry,VeningMeineszLaboratory,RijksUniversiteit,
Utrecht,onamass spectrometerMicro-mass602.
From theanalyticalmethodsused,itisclearthat inthecaseoftheoxygen
isotopecompositions activityratios areinfactmeasured insteadofconcentrationratios (thesampleandthestandard areequilibratedwith smallamounts
ofCCU),whereas inthecaseofD/H ratios concentrationratios aremeasured
(allthehydrogenofthewater sampleandstandard isconverted tohydrogengas).
Indilutesolutions thedifferencebetweenconcentrations andactivitiesis
negligible.Forsomeofthehighly concentrated brines thissalteffectbecomes
measurable.
Sofer&Gat (1975)determined thattheMg-concentrationhas the largest
effectontheactivityoftheisotopes.Therefore,theÓDofthesamplewith
thehighestMg-concentrationhasbeenrecalculated,using theformulagivenby
Sofer &Gat (1975),toshowthedifferencebetweenthe6Dexpressed inconcentrationratiosandthe6Dexpressed inactivityratios.Thisdifference,calculated forsample 37fromtheKonyaBasin,is-1.5/oo,i.e. around theanalytical
precision.Becausethecorrections aresmaller fortheothersamples,thecorrectionshavenotbeenapplied inthediscussionoftheresults.
Table 9.1 summarizes thedataobtained. Inaddition,theCl-concentrations
havebeengivenasanindicationoftherelative salinities.Thedatahavebeen
plotted onfig.9.3 and 9.4 (KonyasamplesandAmboselisamples respectively).
Thelocalitiesofsampling canbe seenonfig.3.2 and fig.4.2.Tables 3.1 and
4.1 showthechemicalanalysesofthewatersamples.Thedataonprecipitation
intheKonyaBasin,aswellastwosamples fromtheCar§ambariverasreported
bySentürketal (1970)havebeenincluded infig. 9.3.
124
18
Table9.1 6 0andÓDvaluesofwatersamplesfromtheKonyaBasininTurkey
andfromtheAmboseliBasininKenya
locality
ó'80°/oo (SMOW)
ÓD°/oo (SMOW) logCl(mol/m)
Konya
1
+4.7
+12
1.1
-37
0.3
2
-5.8
5a
-11.0
5b
-11.1
6
-4.2
-24
0.5
8
+2.3
-3
2.9
9
-1.7
12
-11.4
-65
-65
-35
-71
-70
0.6
0.6
3.2
-1.1
13
-10.5
14
-9.9
2.3
16
+10.0
17
-9.6
22
-9.8
-64
1.4
26
-9.3
-65
1.5
27
-7.8
-45
-0.1
28
-9.6
-62
0.8
-67
2.1
-9
2.7
-64
1.5
29
-10.3
-66
0.9
31
-11.6
-80
1.4
32
-11.0
33
-2.3
-35
3.2
37
-6.5
-44
3.0
3
-0.4
-9
4
-4.8
-24
5
+14.1
6
-5.3
-25
-17
-74
0.9
Amboseli
+63
0.6
0.5
1.8
-0.2
8
-4.0
11
-5.0
13
-1.9
-8
0.2
14
-4.8
-23
-0.5
-18
1.1
-0.6
125
6D(SM0W)(%o)
40
4
6
8 10
18
6 0(SM0W)(%o)
Fig. 9.3 Isotopic composition ofwaters from theKonyaBasin.Symbols:
+,precipitationwaters from§enturk et al. (1970).A andA;Riverwaters from
this study and §enturk et al. respectively. C:Çarsambariver;Z:Zanopariver;
B: Bor river.• andO otherwaters from this study and §enturk et al.respectively. S:SuglaLake.Figures aredescribed in thetext.
9.5DISCUSSION
Konya
Ascanbeseenfromfig.9.3,mostprecipitationsamplesaswellastheriver
18
samplesintheKonyaBasinplotwellabovetheMWL(6D= 8 6 0+ 5),andare
18
closetotheEastMediterraneanWaterLine (6D= 8 6 0+22).Thespreadof
theindividualvaluesisconsiderable.Therefore,nosinglecombinationof6D
18
and6 0foraparticularwatercanbeconsideredasthestartingcomposition
forgroundwatersintheKonyaBasin.Arangeofpossibleprecipitationsisindicatedinfig.9.3.AlthoughtheÇarçambariveristhemostimportantsingle
sourceofwaterfortheBasin,othersmallriversliketheZanopaandtheBor
riverscontributealsotoitswaterbudget,asdoeslocalrainfallinthebasin.
CraterLake(no.8),withinthemouthofanextinctvolcano,derivesallofits
waterfromlocalrainfall.
126
6D(SM0W)(%o)
100
16 18
6180(SM0W)(%o)
Fig. 9.4 Isotopic composition ofwaters from theAmboseli Basin inKenya.Symbols:•,riverwaters;• otherwaters;O inferred compositionof localprecipitation.
Fromthespreadofthepossibleprecipitationsitseemsevidentthataconsiderablescatteroftheisotopiccompositionoftheevaporatedwatersistobe
expected.Therefore,wehavetolimitthediscussiontosomegeneralremarks,
whichcanbemadeonthisbasis.
Gat(1979)discussesdifferenthydrologiemodelsinwhichevaporationtakes
place.InhisterminologymostofthehydrologiesituationsintheKonyaBasin
whichgaverisetosaltformationcanbeconsideredas"terminalgroundwater
lakes"(typeA.,)(seefig.9.5),or"groundwaterlakeswithlowwaterstand"(type
B ? ).CraterLake(no.8)comesclosestotypeC,a"perchedlake"withoutconnectionwiththeregionalgroundwatersystem.TheKa§inhanisample(no.1)isa
"terminalsurfaceflowlake",typeA..LakessuchastypeA-,ortypeC,characterizedbyhighevaporation,showhighlyenrichedisotopiccompositions,afact
bornoutbythedataobtainedinthisstudy.TheKa§inhanisampleshowsinfact
thelargestenrichmentinheavyisotopesofanyofthenormalground-andsurface
waterslistedintable9.1.
ThelakesamplesKa^inhani(no.1),CraterLake(no.8)andPlaya(no.9)showa
largeshiftinisotopiccompositionfollowinganevaporationtrendwithslope5
(seefig.9.3).InthegroundwatersbelongingtotypeB?-waters(groundwater
127
lakeswithlowwaterstand),theevolutionofthewatersthroughcapillary
evaporationresultsinanincreasingsalinity,withoutlargechangesinthe
6-values.Fig.9.3showsindeedthattheisotopiccompositionsofmostwaters,
evenofthosethatshowaconsiderablesalinity (seetable9.1),havenotshifted
farfromtheisotopiccompositionoftheprecipitations.
Asinglesampleisunique,asitcannotconceivablybederivedexclusively
byevaporationprocessesfromanyoftheavailablesources.Thisisthesample
18
ofthespringsatAkhüyük (no.16)whichhasanextremelyhigh6 0-valueof
+10 /oo.Mostprobablythiswateroriginatedbyevaporationfromlocalrainwaters,bringingitfirsttoahypotheticalcompositionofroughly6D=-10 /oo,
-IQ
.-.
6 0=+2 /oo,therebygreatlyincreasingitssalinity.Afterthisevaporation
step,thewaterinfiltratedintothesubstratumtoconsiderabledepthandunderwentrock-waterinteractionatelevatedtemperatures,therebyshiftingtheoxygen
isotopecompositionfromthehypotheticalvalueof+2toitspresentlevelof
6 0=+10 /oo,afterwhichthewatersfroundtheirwayupwardsalongafault
zoneonwhichthehotspringsaresituated.Thegeologicalsettingofthe
springsatAkhüyükmakesthishypothesisacceptable.
Amhoseli Basin (Kenya)
TheAmboselibasininKenyamostcloselycorrespondstotheterminalsurface
flowlakes(typeA)inGat'sclassification.Thegroundwatersinsuchasituation
showmoderateenrichmentintheheavyisotopes (seefig.9.4),whereasthesurface
samplescanbehighlyenriched.Thisisexactlyashasbeenfound,themostenrichedsampleat6D=+63°/oo,and6 0=+14.1°/oobeingthesurfacesample
ofthelakeinanadvancedstageofevaporation,whereastheothersamplesrepresentthelocalgroundwaters.
1
terminal
groundwater lake
low water stand
/ \
terminal surface
flow lake
perched lake
Fig. 9.5 Different hydrologicalmodelswhich giverise to salt formation.
Typenumbers and figures fromGat (1979).
128
Themost consistentexplanationofthe isotopedataoftheAmboselibasin
seems tobe thatthe localprecipitation isnot lyingontheMWL,but isinfact
lyingonalinesimilartotheEastMediterraneanWater Line;thiscouldwellbe
thecaseastheAmboselibasin is,liketheEasternMediterranean,characterized
byasemi-arid climate.Theestimated compositionofthe localprecipitationhas
been indicated infig.9.4.
Theuseofthe isotopiccompositionofwaters inorder toobtainquantitative
informationonthesourceanddevelopmentofthegroundwaters inclosedbasins
would require amoredetailed samplingprogram,preferablywithsampling carried
outover aperiodoftime.Inadditionhumiditymeasurements shouldbemade,
both inandabovethesoilprofile.
129
10.SUMMARY
This studydealswith therelationbetween themineralogical composition of
saltassemblages andthecomposition ofgroundwaters fromwhich these salts
precipitated.A comparisonwasmadebetween salts andwaters sampled inthe
KonyaBasin inTurkey andwaters sampled inthreedifferent regions inKenya.
The chemical composition ofwaters fromrivers entering theKonyaBasinis
different from thecomposition of those fromrivers inKenya.The initial composition oftheseriversdetermines thetypeofminerals thatwill precipitate
during evaporation oftheseriverwaters.Theratio ofcalcium tocarbonate at
themoment thesolutionbecomes saturatedwith respect tocalcite,usually the
firstmineral toprecipitate,determineswhether thefinal solutionwill become
carbonate-rich orcarbonate-poor.Theratio ofmagnesium tosilicaand theratio
ofcalcium tosulphate atthemoment thesolutionbecomes saturatedwith respect
tosepiolite andgypsumrespectively,determine themagnesium and sulphate contentofthefinal solution. Inthisway sixdifferent types ofconcentrated brines
originate and sixdifferent typesofsaltassemblages precipitate from these
brines during evaporation.
The concentrated waters and thesaltassemblages intheKonyaBasinbelong
mainly totheNa-Mg-SO.-Cl-type.The evolutionofthegroundwater composition
and the typeofmineralswhichprecipitated from these groundwaters canbeexplainedby assuming successive precipitation ofcalcite,sepiolite and gypsum.
The concentratedwaters and thesaltassemblages fromKenyabelongmainly to
theNa-CCL-SO.-Cl-type.Thebehaviour of thedissolved species and thetypeof
saltminerals canbe explained by assuming calcite andsepiolite precipitation
only.
Thecrystallographic properties ofsomesaltmineralsweredetermined by
meansofX-ray diffraction analysis and themorphologicalproperties by Scanning
ElectronMicroscopy. Thepresence ofhalite causes asalt crust tobecome dense
and sealing.Theporosity ofasalt crust increaseswhenbloediteor trona is
present.
Prediction ofthesequence ofsaltmineralswhichwillprecipitate froma
concentrated solutionrequires anaccurateknowledge ofthethermodynamicpropertiesofconcentrated electrolyte solutions.Unfortunately, theseproperties are
stillunknown forcarbonate-rich solutions.Only recently atheorywas developed
130
for carbonate-poor solutions.
Thewatersampleswereevaporated inthelaboratory inevaporating dishes
under ambient conditions.Themineralogicalcomposition oftheprecipitates was
comparedwith theassemblages thatoccur inthe field.
The carbonate-containing saltassemblages bothfromthefield and fromthe
laboratory experimentswere investigated bymeans of log PQQ -loga^ Q diagrams. Itappeared thatneither thefieldnor thelaboratory saltassemblageswere inequilibriumwith theCO^-pressuresofthesolutions fromwhich
theyprecipitated.TheseCO^-pressureswere calculated fromtheanalytical data.
Thecarbonate-free saltassemblages from theKonyaBasinwere investigated
bymeans ofJänecke-diagrams.For afewselected samples thetheoretical mineral
sequenceswere calculated under theassumption ofequilibrium precipitation
withthehelpofthecomputerprogram ofHarvie &Weare (1980). Itappeared that
prediction ofmineral assemblages insaltefflorescences isdifficult evenwith
a sound thermodynamic theory.Saltassemblages innatural salt efflorescences
couldbest bepredicted assumingprecipitation under equilibrium conditions,
whereasmineral assemblages inevaporating dishes inthe laboratory could best
bepredicted assuming bothmetastablemineral formation and fractional crystallization.This conclusion issupportedby thepresence of thetwonewminerals,
konyaite andeugsterite,whichhavebeendiscovered inthis study.Botharemetastablemineralsunder ambient conditions.
Theoxygenandhydrogen isotopic composition ofsomewaters from theKonya
BasininTurkey andofthewatersamples from theAmboseliBasin inKenyawere
determined. The isotope fractionation causedby evaporation isdifferent in
groundwater andsurfacewater.Processes ofevaporation andwater-rock interactioncouldbe distinguished.
131
11. SAMENVATTING
Zoutegrondenzijnnietgeschiktvoorlandbouw.Daaromishetvanhetgrootstebelangtevoorkomendatgeschikte landbouwgrondenverzouten.Bovendienmoet
veelaandachtgeschonkenwordenaandeverbeteringvanreedsverzoutegronden.
Zoutegrondenkomenvoorinsemi-arideenaridegebieden.Doordehogeverdampinginzulkegebiedennemendeconcentratiesvandeopgeloste stoffeninhet
oppervlaktewatereninhetgrondwater toe.Wanneerdezestoffennietvoldoende
afgevoerdworden,zullenzealszouteninofopdebodemneerslaan.Natuurlijk
gevormde zoutegrondenontstaanveelaldoorverdampingvangrondwaterinhydrografischafgeslotenbekkens.Ookdoormenselijke invloedkunnengrondenverzouten
wanneerinaridegebiedenirrigatiewordttoegepastzonder zorgtedragenvoor
voldoendedrainage.
Bodemkundigendiezichmetzoutegrondenbezighouden,zijnoverhetalgemeen
nietgeïnteresseerdindemineralogische samenstellingvandezouten.Gewoonlijk
lossenzijdezoutenopenbepalendeconcentratiesvandeverschillende ionenin
oplossing.Vanuit landbouwkundigoogpuntisditbegrijpelijk omdatplanten lijden
vantehogeconcentraties zouteninopgelostevorm. Inditproefschriftis
eenverband gelegdtussendemineralogischesamenstellingvandezoutenende
chemischesamenstellingvanhetgrond-enoppervlaktewaterwaaruitzeneerslaan.
MonstersuitdrieverschillendestrekeninKenyaenuithetKonyabekkenin
Turkijewerdenmetelkaarvergeleken.Detweeverschillende gebiedenwerdenom
devolgenderedenenuitgezocht:
1.DezoutenuitKenya zijnvaneengeheelander typedandieuitKonya.
2.BeidegebiedenwarenreedseerderbestudeerddoorNederlandse bodemkundigen
vandeLandbouwhogeschoolinWageningenenvandeKenyaSoilSurveyinNairobi.
Hierdoorwasvoldoendebodemkundigekennisaanwezigenbestonddemogelijkheid
voordeskundigebegeleiding.
Naastmonstersvandezoutuitbloeiingenwerdenookmonstersvanhetgrondenoppervlaktewatergenomen.
DesamenstellingvanderivierwatersdiehetKonyabekkenbinnenstromenverschiltvandeKenyaanserivierwatersamenstelling.Deverhoudingvandeopgeloste
stoffeninhetrivierwaterbepaaltuiteindelijkhettypemineralendatbijverdampingzalneerslaan.
132
Calciet isgewoonlijkheteerstemineraaldattijdensverdampingneerslaat.Deverhoudingtussencalciumencarbonaatindeverdundeoplossingbepaaltofdegeconcentreerdeoplossing carbonaat-armof-rijkzalworden.Wanneer
deoplossingverzadigdraaktaansepioliet isdemagnesium-silicaverhoudingbepalendofdeuiteindelijkeoplossingmagnesium-rijkof-armzal zijn;decalciumsulfaatverhoudingophetmomentdatdeoplossingverzadigdraaktaangipsbepaalt
hetsulfaat-gehaltevandegeconcentreerdeoplossing.Zoontstaanzesverschillendehoofdtypesgeconcentreerdeoplossingenwaaruit zesverschillende types
zoutenneerslaan.Dezetypesbevattenallechloride.
InhetKonya-bekkenbehorendegeconcentreerdewatersendezoutmineralen
grotendeels tothetNa-Mg-SO.-Cl-type.Deontwikkelingvandegrondwatersamenstellingenhetbijbehorende typemineralenkanverklaardwordenwanneeruitgegaanwordtvandeopeenvolgendeprecipitatievancalciet,sepiolietengips.De
geconcentreerdewatersende zoutmineralenuitKenyabehorenhoofdzakelijk tot
hetNa-CO,-SO.-Cl-type.Hetgedragvandeioneninoplossingenhetvoorkomen
vandittype zoutmineralenkunnenverklaardwordendoordeopeenvolgendeprecipitatievancalcietensepioliet.
Dekristallografische enmorfologische eigenschappenvaneenaantalzoutmineralenwerdenmetbehulpvanröntgendiffractie analyseenelektronenmicroscopiebepaald.Wanneerhaliet involdoendemate aanwezig isineenzoutassociatie,vormtdezeeendichte,afsluitendekorst.Bloedietentronavergrotendeporositeitvaneenzoutkorst.
Wanneerwewillenvoorspellenwelke zoutmineralenuitgeconcentreerde oplossingenzullenneerslaan,moetenwedethermodynamischeeigenschappenvanelectrolytoplossingenkennen.Totophedenzijndezeeigenschappenvoorcarbonaatrijkeoplossingennietbekend.Zeerrecentiseentheorievoorcarbonaat-arme
oplossingenontwikkeld.
Allewatermonsterswerdeninindampschalenonder laboratoriumomstandigheden
verdampt.Demineralogische samenstellingvanhetneerslag isvergelekenmetde
mineraalgezelschappendieinhetveldondernatuurlijke omstandigheden zijnneergeslagen.
Zoweldenatuurlijkecarbonaathoudendemineraalgezelschappenalsdegezelschappendie zichinhet laboratoriumgevormdhebben,werdenonderzochtmetbehulpvanlogpço -loga^g-diagrammen.Hetbleekdatdezebeidegroepen
mineraalgezelschappenniet inevenwichtwarenmetdeCO?-drukvandeoplossingenwaar zeuitneersloegen.DezeŒL-drukwerdberekenduitdeanalytischegegevens.
DecarbonaatvrijemineraalgezelschappenuithetKonya-bekkenwerdenonder-
133
zochtmetbehulpvanJäneckediagrammen.Vaneenaantalmonsters ismetbehulp
vanhetcomputerprogrammavanHarvie &Weare (1980)berekend inwelkevolgordede
zoutenonderevenwichtsomstandighedenzoudenneerslaan.Hetbleek,dat zelfsmet
eengedegenthermodynamischetheoriehetmoeilijk iseengoedevoorspellingte
doenomtrentde zoutendiegevormd zullenworden.Degezelschappendieonder
natuurlijkeomstandighedendoorevapotranspiratievangrondwaterontstaanzijn,
blijkenbeterverklaard tekunnenwordenwanneerprecipitatie onderevenwichtsomstandighedenwordtaangenomen.Degezelschappendieindeverdampingsexperimentenonder laboratoriumomstandighedenontstaanzijn,kunnenbeterverklaardwordenonderdeaannamedatmetastabielevormingvanmineralenenfractionelekristallisatieplaatsheeftgevonden.Dezeconclusieismedegebaseerdophetvoorkomenvandemineraleneugsterietenkonyaiet.Dit zijntweenieuwemineralendie
tijdensdezestudieontdektzijn.Beidemineralen zijnmetastabielbij 1atmosfeer
drukenkamertemperatuur.
Dezuurstof-enwaterstofisotoop-samenstellingvaneenaantalwatermonsters
uithetKonya-bekkenenvandewatermonstersuithetAmboseli-bekken inKenya
werdenbepaald.Deisotoop-fractioneringingrondwatereninoppervlaktewater
onder invloedvanverdamping isverschillend.Erkoneenonderscheid gemaaktwordentussenwaterdatverdampinghadondergaanenwaterwaarbij intensieveinteractiemetgesteentenhadplaatsgevonden.
134
12. REFERENCES
Al-Droubi,A., 1976.Géochimie des sels etdes solutions concentrées par evaporation.Modèle thermodynamique desimulation.Application aux sols salésdu
Tchad.Thèse,Université Louis-Pasteur,Strasbourg,France, 177p.
Aubert,G., 1967.Classificationdes Sols.Edition 1967.Trav.CPSS,Grignon.
Beek,C.G.E.M. van,Breemen,N.van, 1973.Thealkalinity of alkali soils.J.
Soil Sei.,24,p.129-136.
Begheyn,L.Th., 1980.Methods of chemical analyses for soils andwaters.AgriculturalUniversity,Wageningen,theNetherlands.
Braitsch,0., 1971.Saltdeposits,their originand composition. Springer-Verlag,
NewYork, 297p.
Brinkman,R., 1979.Ferrolysis,asoil-forming process inhydromorphic conditions.
Pudoc,Wageningen,theNetherlands.
Brinkmann,R., 1976.Geology ofTurkey.Elsevier,Amsterdam,Oxford,NewYork,
158p.
Conley,R., andBundy,W.M., 1959.Mechanism ofgypsification.Geochim.Cosmochim.
Acta, 15,p. 57-72.
Craig,H., 1961.Isotopicvariations inmeteoricwaters.Science, 133,p.17021703.
Dana,J.D., 1951.System ofMineralogy,vol.II.JohnWiley &Sons,NewYork.
Drever,J.I.and Smith,C.L., 1978.Cyclicwetting and drying of thesoil zone
asan influence on thechemistry of groundwater inarid terrains.Am.Jour.
Sei.,278,p.1448-1454.
Driessen,P.M., 1970.Soil salinity and alkalinity in theGreatKonyaBasin,
Turkey.Pudoc,Wageningen,theNetherlands,99p.
Driessen,P.M.,and Schoorl,R., 1973.Mineralogy andmorphology of saltefflorescences onsaline soils in theGreatKonya Basin,Turkey.J. Soil Sei.24,
no. 4,p.436-442.
Epstein,S., 1970.Antarctic icesheet:stable isotope analyses of Byrd station
cores and interhemispheric climatic implications.Science, 168,p.1570-1572.
Epstein,S.,Sharp,R.P.,andGow,A.J., 1965.Six-year record of oxygenand
hydrogen isotopevariations inSouthPolefirn.Journ.Geophys.Research,
70,p.1809.
135
Ericksen,G.E.,andMrose,M.E., 1970.Mineralogical studies of thenitrate
deposits of Chile.II.Darapskite,Na (NO )(SO).H0.Am.Min.,55,p.
1500-1517.
Eswaran,H.,and Carrera,M., 1980.Mineralogical zonation in saltcrust.
International Symposium onsalt-affected soils,Karnal. p.237-293.
Eugster,H.P., 1970.Chemistry and originof thebrines ofLakeMagadi,Kenya.
Mineral. Soc.Amer.Spec.Paper 3,p.215-235.
Eugster,H.P.,andHardie,L.A., 1978.SalineLakes.In:Lerman,A.,ed.,Lakes:
Chemistry,Geology and Physics.NewYork,Springer-Verlag,p.237-293.
Eugster,H.P.,Harvie,C.E.,andWeare,J.H., 1980.Mineral equilibria inasixcomponent seawater system,Na-K-Mg-Ca-SO,-Cl-H0,at 25 C.Geochim.Cosmochim.Acta,44,p.1335-1347.
Eugster,H.P.,andJones,B.F., 1977.Thebehaviour ofpotassium and silica
during closed basin evaporation. 2nd IAGC SymposiumWater-Rock Interaction,
Strasbourg,II,p. 1-12.
Eugster,H.P.,andJones,B.F., 1979.Behaviour ofmajor solutes during closed
basinbrine evolution.Am.J. Sei.,279,p.609-631.
Eugster,H.P.,Smith,G.I.,1965.Mineral Equilibria inSearles LakeEvaporites,
California.J. Petr.,6,p.473-522.
Eugster,H.P.,and Surdam,R.C., 1973.Depositional Environment of theGreen
River Formation ofWyoming:APreliminary Report.Geol. Soc.Amer.Bull.,
84,p. 1115-1120.
Fang,J.H., andRobinson,P.D., 1970.Crystal Structures andmineral chemistry
of double-salt hydrates:II thecrystal structure of loeweite.Am.Min.,55,
p.378-386.
FAO/Unesco, 1973.Irrigation,drainage and salinity:an international source
book.Unesco,Paris.510p.
FAO/Unesco, 1971-1978.SoilMap of theWorld, 1:5.000.000.Unesco,Paris.In
twoof thefour following languages:English,French,Russian and Spanish.
Vol. 1-10.
Fauck,R.,Lamouroux,M.,Perraud,A.,Quantin,P.,Roederer,P.,Vieillefon, J.,
and Ségalen,P., 1979.Projet declassificationdes sols.O.R.S.T.O.M.,
Paris.
Fontes,J.C., andGonfLantini,R., 1967.Comportement isotopique au coursde
1'evaporationd'eaux bassin Sahariens.Earth Plan. Sei. lett.,3,p.258-266.
Gac,J.Y.,Al-Droubi,A.,Fritz,B.,and Tardy,Y., 1977.Geochemical behavior
of silica andmagnesium during theevaporationofwater inChad.Chem.Geol.,
19,p.215-228.
136
Garrels,R.M., and Christ,C.L., 1965.Solutions,Minerals and Equilibria.
Harper and Row,NewYork,450p.
Garrels,R.M., andMackenzie,F.T., 1967.Originof thechemical compositionof
some springs and lakes.In:Equilibrium concepts innaturalwater systems.
American Chemical Society,Washington,D.C.,p.222-242.
Gat,J.R., 1966.The isotopic composition of atmosphericwater intheMediterraneanSeaarea and their interpretation interms of air-sea interactions,
presented at the 20th Congress of theComité International de l'Exploration
Scientifique de laMerMéditerranée,Bucharest.
Hardie,L.A., 1965.Phase equilibria involvingminerals of the system CaSO,Na.S0,-H0.Ph.D.Thesis,TheJohnsHopkins University,Baltimore.
Hardie,L.A., 1967.Thegypsum-anhydrite equilibrium at 1atm.pressure.Am.
Min.,52,p.121-200.
Hardie,L.A., 1968.Theoriginof therecentnon-marine evaporitedepositof
SalineValley,Inyo County,California.Geochim.Cosmochim.Acta,32,p.
1279-1301.
Hardie,L.A., and Eugster,H.P., 1970.The evolution of closed-basinbrines.
Mineral. Soc.Am.Spec.Paper 3,p.273-290.
Harned,H.S., 1935.Some thermodynamic properties ofunivalent halidemixtures
inaqueous solution.J.Amer.Chem.S o c , 57,p.1865.
Harvie,C.E.,andWeare,J.M., 1980.Theprediction ofmineral solubilities in
naturalwaters:theNa-K-Mg-Ca-Cl-SO,-H_0 system from zero tohighconcentration at 25 C.Geochim.Cosmochim.Acta,44,p.981-997.
Harvie,C E . ,Weare,J.H.,Hardie,L.A., and Eugster,H.P., 1980.Evaporationof
seawater:calculated mineral sequences.Science,208,p.498-500.
Hatch,J.R., 1972.Phase relationships inpart of the system sodiumcarbonatecalcium carbonate-CO„-H„0at 1atm.pressure.Ph.D.,Univ.ofIllinois,
Urbana-Champaign.
Helgeson,H.C., 1969.Thermodynamics ofhydrothermal systems atelevated temperatures and pressures.Am.J. Sei., 267,p.729-804.
Hill,A.E.,andWills,J.H., 1938.Ternary systemsXXIV.Calcium sulfate,sodium
sulfate andwater.J.Am.Chem.S o c , 60,p. 1647-1655.
Jacks,G., 1973.Chemistry of groundwater inadistrict inSouthern India.J.
Hydrol., 18,p.185-200.
Jones,B.F.,Eugster,H.P.,and Rettig,S.L., 1977.Hydrochemistry of theLake
Magadi basin,Kenya.Geochim.Cosmochim.Acta,41,p. 53-72.
Jung,D., andKeller,J., 1972.DiejungenVulkanite imRaum zwischenKonyaund
Kayseri (Zentral-Anatolien).Z.Deutsch.Geol.Ges.,123,p.503-512.
137
Korzhinskii,D.S., 1959.Physicochemical basis of theanalysis of the paragenesis
ofminerals.Consultants Bureau,Inc.,NewYork, 142p.
Linke,W.F., 1965.Solubilities of organic andmetal organic compounds.Amer.
Chem.S o c ,VanNostrand.
Maglione,G., 1968.Présencede gaylussite etde trônadans les "natroniëres"du
Kanem (pourtournord-est duLac Tchad). Bull.Soc.Fr.Minéral.Crystallogr.,
91,p.388-395.
Meester,T.de (ed.), 1970.Soilmap of theGreatKonya Basin.Agricultural University,Wageningen,theNetherlands.
Meester,T.de (ed.), 1970.Soils of theGreatKonya Basin,Turkey.Pudoc,Wageningen, theNetherlands,290p.
Meester,T.de, 1971.Highly calcareous lacustrine soils in theGreatKonyaBasin,
Turkey.Pudoc,Wageningen,theNetherlands, 169p.
Mellaart,J., 1964.Early cultures of theSouthAnatolian Plateau.Anatolian
Stud.XIII.
Nesbitt,W.H., 1974.The study of somemineral-aqueous solution interactions:
Ph.D. thesis,JohnsHopkinsUniv.,Baltimore, 173p.
Pitzer,K.S., 1973.Thermodynamics of electrolytes.I.Theoretical basis andgeneral equations.J.Phys.Chem., 77,p.268-277.
Pitzer,K.S., 1975.Thermodynamics of electrolytes.V.Effects ofhigher-order
electrostatic terms.J. Solution Chem., 4,p.249-265.
Pitzer,K.S.,andKim,J.J., 1974.Thermodynamics of electrolytes.IV.Activity
and osmotic coefficients formixed electrolytes.Am.Chem. S o c , 96,p.57015707.
Pitzer,K.S.,andMayorga,G., 1973.Thermodynamics of electrolytes.II.Activity
and osmotic coefficients for strong electrolytes with one orboth ions univalent.J.Phys.Chem.,77,p.2300-2308.
Pitzer,K.S.,andMayorga,G., 1974.Thermodynamics of electrolytes.III.Activity
and osmotic coefficients for 2-2 electrolytes.J. Solution Chem., 3,p.539546.
Richards,L.A., 1954.Diagnosis and improvement of saline and alkalisoils.
Agriculture handbook no.60.U.S.Department ofAgriculture, 160p.
Robie,R.A.,Hemingway,B.S.,and Fisher,J.R., 1978.Thermodynamic properties
ofminerals andrelated substances at 298.15K and 1bar pressure and at
higher temperatures.U.S.Geol. Surv.Bull. 1452,456p.
Robinson,R.A., and Stokes,R.H., 1970.Electrolyte solutions.Butterworths,
570p.
Saggerson,E.P., 1952.Geology of theKisumuDistrict.Report no.21.Geological
138
SurveyofKenya,Nairobi.
Scatchard,G., 1936.Concentrated solutions of strong electrolytes.Chem.Rev.,
19,p.309-327.
Scatchard,G., 1968.The excess free energy and related properties of solutions
containing electrolytes.J. Am.Chem.S o c , 90,p.3124-3127.
Sentürk,F., Bursali,S.,Omay,Y.,Ertan,I.,Güler,S.,Yalçin,H.,andOnhan,
E., 1970.Isotope techniques applied togroundwatermovement in theKonya
plain.From:IsotopeHydrology 1970.Proceedings of anIAEA syposium, 9-13
march 1970.International Atomic Energy Agency,Vienna,1970.
Slyusareva,M.N., 1969.ZapiskiVses.Mineralog.Obshch.,98,p. 59-62. (in
Russian).
Sofer,Z.,and Gat,J.R., 1975.The isotopic composition of evaporatingbrines;
effect of the isotopic activity ratio insaline solutions.Earth Planet.Sei.
Lett., 26,p. 179-186.
Soil Survey Staff, 1975.Soil taxonomy: abasic system of soilclassification
formaking and interpretating soil surveys.Soil Conservation Service,U.S.
Dept.ofA g r i c ,Washington,D.C.,AgricultureHandbook No.436.
Sombroek,W.G., 1980.Legend of theexploratory soilmap ofKenya,scale
1 :1000000.Internal communication no.22,Kenya Soil Survey,Nairobi.
Stoessell,R.K., andHay,R.L., 1978.The geochemical originof sepioliteand
kerolite atAmboseli,Kenya.Contr.Miner.Petr.,65,p.255-267.
Stoops,G.,Eswaran,H.,and Abtahi,A.,1978.Scanning electronmicroscopy of
authigenic sulphateminerals insoils.In:M.Delgado (ed.),Proceedings
fifth internationalworkmeeting on soilmicromorphology,Granada,Spain,
p. 1093-1113.
Stumm,W.,andMorgan,J.J., 1970.Aquatic chemistry.An introductionemphasizing chemical equilibria innaturalwaters.Wiley.Interscience Ed.,New
York,583p.
Surdam,R.C.,and Sheppard,R.A., 1978.Zeolites insaline alkaline lake deposits
In:Sand,L.B.,andMumpton,F.A., eds.Natural zeolites:Occurrence,properties,use.Elmsford,NewYork,PergamonPress,p.145-174.
Touber,L.Reconnaissance soilandvegetation survey of the Amboseli-Kibwezi
area. Inpreparation.
United StatesDepartment ofAgriculture, 1938.Soils andMen,Yearbook of
Agriculture,1938.
Van 'tHoff,J.H., 1905.ZurBildung ozeanischer Salzablagerungen.ErstesHeft.
Vieweg. 85p.
Van 'tHoff,J.H., 1909.ZurBildung ozeanischer Salzablagerungen. ZweitesHeft
139
Vieweg,90p.
Van 'tHoff,J.H., 1912.Untersuchungen über dieBildungsverhältnisse derozeanischen Salzablagerungen.Precht and Cohen,eds.Akad.Verlags ges.,Leipzig,
374p.
Vergouwen,L., 1979.Twonew occurrences and theGibbs energy ofburkeite.Min.
Mag., 43,p.341-345.
Vergouwen,L., 1981.Mineralogical composition and origin of saltefflorescences
intheKonyabasin,Turkey and inKenya.N.Jb.Miner.Mh.,p. 23-34.
Vergouwen,L., 1981.Scanning electronmicroscopy applied to saline soils from
theKonyaBasin inTurkey and fromKenya.In:E.B.A.Bisdom (ed.):Submicroscopy of soils andweathered rocks.1stworkshop of theInternationalWorkingGroup onSubmicroscopyofUndisturbed SoilMaterials (IWGSUSM),Wageningen,
1980.Pudoc,Wageningen,theNetherlands,inpress.
Vergouwen,L., 1981.Eugsterite,anew saltmineral.Am.Min. ,inpress.
Visser,J.W., 1969.A fully automatic program for finding theunit cell from
powderdata.J. Appl. Cryst.,2,p.89.
Williams,L.A.J., 1970.Thevolcanics of theGregory RiftValley,EastAfrica.
Bull.Volcanol.,34,p.439-465.
Williams,L.A.J., 1972.Geology of theAmboseli Area.Geological Survey ofKenya,
Report no.90,Nairobi,86p.
Wood,J.R., 1972.Prediction ofmineral solubilities inconcentrated brines.A
thermodynamic approach.Ph.D.Thesis,Baltimore,Maryland, 121p.
Wood,J.R., 1975.Thermodynamics ofbrine-salt equilibria. I-The systems NaClKCl-MgCl-CaCl-H0and NaCl-MgSO-H„0at 25°C.Geochim.Cosmochim.Acta,39,
p. 1147-1163.
140
CURRICULUM VITAE
NahetbehalenvanheteindexamengymnasiumbètaaanhetStedelijkGymnasium
teUtrecht,begondeschrijfsterin1969haar studiegeologieaandeRijksUniversiteitteUtrecht.Inmei1972deedzijhaarkandidaatsexamenmethoofdvakken
geologieenscheikundeenbijvakkenwiskundeennatuurkunde.Indecember1976
werdhetdoctoraalexamengeologieafgelegdmetalshoofdvakgeochemieendebijvakkenalgemeneeneconomischegeologieenchemische thermodynamica.Injanuari
1977tradzijindienstalspromotie-assistentebijdeafdelingBodemkundeen
GeologievandeLandbouwhogeschoolinWageningen.GedurendehetpromotieonderzoekwerdenreizengemaaktnaarKenyaenTurkijewaar zoutmineralenverzameldwerden.