Jona A„ Harrisca

(JU^S^*j^^^

STATE OF ILLINOISHENRY HORNER, Governor

DEPARTMENT OF REGISTRATION AND EDUCATIONJOHN J. HALLIHAN. Director

DIVISION OF THE

S T A r E G E (3 L O G I C A E S U R \^ E YM. M. LEIGHTON, Cliicf

URBAN

A

REPORT OF IXVESTIGATIOXS — NO. 66

INVESTIGATION OF THE EFFECT OF HEAT ON THECEAY MINERAES ILIJTE AND MONTMORIEEONITE

R. E. GRIM AND W. F. BRADLEY

Reprinted from the Journal of the American Ceramic Society,

Vol. 23, No. 8, pp 242-248, 1940

I'RINIED BV AUTHORITY OF THE STATE OF ILLINOIS

ERBAXA, ILLLXOIS

1 940

STATK OF ILLINOISHON. HENRY HORNER, Governor

Dl.PAK'rMKNT OF REGISTRATION AND FDUCA'IIONHON. JOHN J. HALLIHAN, Director

BOARD OFNATURAL RESOURCES AND CONSERVATION

HON. JOHN J. HALLIHAN, ChairmanEDSON S. BASTIN, Ph.D., Geology HENRY C. COWLES, Ph.D., D.Sc, Forestry.WILLIAM A. NOYES. Ph.D.. LL.D., Chem.D., D.Sc.,

Chemistry

LOUIS R. HOWSON, C.E., EngineeringWILLIAM TRELEASE, D.Sc., LL.D , Biology

(Deceased)

ARTHUR CUTTS WILLARD, D.Engr., LL D.,President of the University of Illinois

S T A T E GEOLOGICAL SLR V E Y D I V I S I O NU7-bana

M. M. LEIGHTON. Ph.D., Chief

ENID TOWNLEY, M.S., Assistant to the ChiefJANE TITCOMB, M.A., Geological Assistant

GEOLOGICAL RESOURCESCoal

G H. CADY. Pli.D., Senior Geologist and HeadL C. McCABE, Ph.D., Assoc. GeologistJAMES M. SCHOPF, Ph.D., Asst. GeologistJ NORMAN PAYNE, Ph.D.. Asst. GeologistCHARLES C. BOLEY, M.S., Asst. Mining Eng.

hidnslrial Minerals

J E LAMAR, B.S., Geologist and HeadH B WILLMAN, Ph.D.. Assoc. GeologistDOUGLAS F. STEVENS, M.E., Research AssociateROBERT M. GROGAN, Ph.D., Asst. GeologistROBERT R. REYNOLDS, B.S.. Research Assistant

Oil and Gas

A H. BELL, Ph.D., Geologist and HeadG V COHEE, Ph.D., Asst. GeologistFREDERICK SQUIRES, B.S., Assoc. Petr. Eng.CHARLES W. CARTER, Ph.D.. Asst. GeologistF. C. MacKNIGHT. Ph.D., Asst. GeologistROY B. RALSTON, B.A., Research AssistantWAYNE F. MP:ENTS, Research Assistant

Areal and En\^ineerin^ Geology

GEORGE E. EKBLAW, Ph.D., Geologist and HeadHARRY McDERMITH, B.S., Asst. Topographic Eng.RICHARD F. FISHER, B.A., Research Assistant

Subsurface Geology

L. E. WORKMAN, M.S., (k-ologist and HeadELWOOD ATHERTON, Ph.D., Asst. GeologistMERLYN B. BUHLE, M.S., Asst. GeologistI. T. SCHWADE, M.S., Asst. GeologistFRANK E. TIPPIE, B.S., Research Assistant

Stratigraphy and Paleontology

J MARVIN WELLER, Ph.D., Geologist and HeadCHALMER L. COOPER, M.S., Assoc. Geologist

Petrography

RALPH E. GRIM. Ph.D., PetrographcrRICHARDS A. ROWLAND, Ph.D., Asst. Geologist

Physics

R. J. PIERSOL, Ph.D.. PhysicistDONALD O. HOLLAND. M.S.. Asst. PhysicistPAUL F. ELARDE, B.S., Research Assistant

GEOCHEMISTRYFRANK H. REED, Ph.D., Chief ChemistW. F. BRADLEY, Ph.D., Assoc. ChemistG. C. FINGER, Ph.D., Assoc. ChemistHELEN F. AUSTIN, B.S.. Research Assistant

Fuels

G. R. YOHE, Ph.D., Assoc. Chemist in ChargeCARL HARMAN, B.S., Research Assistant

Industrial Minerals

J. S. MACHIN. Ph.D., Chemist and HeadJAMES F. VANECEK. M.S.. Research Assistant

Analytical

O. \V. REES, Ph.D., Chemist and HeadL. D. McVICKER, B.S., Asst. ChemistGE{)R(;E W. LAND, B.Ed., Research AssistantP. W. HENLINE, M.S., Research AssistantMATHEW KALINOWSKI, M.S.. Research AssistantARNOLD J. VERAGUTH, M.S., Research Assistant

MINERAL ECONOMICSW. H. VOSKUIL, Ph.D.. Mineral EconomistGRACE N. OLIVER, A.B., Assistant in Mineral Eco-

nomics

EDUCATIONAL EXTENSION

DON L. CARROLL, B.S., Assoc. Geologist

PUBLICATIONS AND RECORDS(iEORGE E. EKBLAW, Ph.D., Geological EditorC:HALMER L. cooper, M.S., Geological EditorDOROTHY ROSE, B.S., Technical EditorKATHRYN K. DEDMAN, M.A., Asst. Technical

EditorALMA R. SWEENY, A.B., Technical P^iles ClerkFRANCES HARPER LEHDE, M.S., Asst. Teciinical

Files ClerkMEREDITH M. CALKINS, Geologic DraftsmanLESLIE D. VAUGHAN. Asst. PhotographerDOLORES C. THOMAS, B.A., Geologic Clerk

Consultants: Ceramics, CULLEN WARNER PARMELEE, M.S., D.Sc. and RALPH K. HURSH. B.S.. Universityof Illinois; Pleistocene Invertebrate Paleontology FRANK COLLINS BAKER, B.S., University of Illinois.

Topographic Mapping in Cooperation with the United States Geological Survey.

This Report is a Contribution of the Petrography Division.

ILLINOIS STATE GEOLOGICAL SURVEY

3 3051 00005 6790

(A27774—1500—8-40) July 1, 1940

CONTENTS

I'AGE

Abstract 5

Introduction 5

Procedure 5

Changes in physical characteristics 6

Illite 6

Montmorillonite 8

Changes in mineral composition of ilHte on firing 8

Interpretation of optical data 8

Interpretation of X-ray data 9

Summary of mineral changes in illite 9Changes in mineral composition of montmorillonite on firing 11

Interpretation of optical data 11

Interpretation of X-ray data 12

Summary of mineral changes in montmorillonite 13

ILLUSTRATIONS

FIGURE PAGE

1. Indices of refraction anti representative X-ray diffraction patterns of sample (1) 7

2. Indices of refraction of sample (2) 8

3. Indices of refraction of sample (3) 8

4. Indices of refraction and representative X-ray diffraction patterns of sample (4) 10

5. Indices of refraction and representative X-ray diffraction patterns of sample (5) 11

I3

I

Digitized by the Internet Arciiive

in 2012 with funding from

University of Illinois Urbana-Champaign

http://archive.org/details/investigationofe66grim

IN\ ESTIGATION OF THE EFFECT OF HEAT ON THE CLAYMINERALS ILLITE AND MONTMORILLONTTE*

Bv R. F. Grim W. F. B RAD LEV

ABSTRACT

Samples of three purified illites, one purified montmorillonite, andone natural clay containing montmorillonite, quartz, and limonitewere heated at successive temperatures up to 1400° C. X-ray andoptical determinations were made on all samples. The changes thattake place in the clay minerals, illite and montmorillonite, when theyare heated at various temperatures up to 1400°C. are discussed.

INTRODUCTION

The concept that clays are composedessentially of extremely minute crystal-line particles of one or more of a fewminerals, known as the clay minerals, isgenerally accepted by students of clays.It is further recognized that there are

only three important groups of clayminerals (Table 1), that is, almost allclays are composed of extremely smallparticles (frequently less than 0.001 mm.in diameter) of one or more of these threeclay mineral groups.

Table 1.

—

Lmportant Clay Mineral Groups^

Name Chemical composition Occurrence

Kaolinite (OHjgAl.Si.O.o China clay, fireclay.ball clav, etc.

Illite (OH),Ky(Al,-Fe,-- Shale, fireclay, ballMg,-Mg,i) (Si,_ clav, etc.yAly)0,p

Montmoril-- (OHj^Al^Si^O^„j-- Bentonite, gumho-lonite XH2O til, loess, etc.

apor details concerning other clay minerals and possiblevariations in the composition and properties of clay miner-als generally, see R. E. Grim, "Relation ot Composition toProperties of Clays," Jour. Amer. Ceram. Soc, 22 (5

1

141-51 (1939). Reprinted as Illinois State Geol. SurveyCir. 45.

*Presented at the Forty-Second Annual Meeting.American Ceramic Society, Toronto, Canada, April 9,1940 (Materials and Equipment Division). ReceivedJune 7, 1940.

Ka) T. X. McVay and C. L. Thompson. "X-Ray In-vestigation of Effect of Heat on China Clays," Jour. Amer.Ceram. .Soc, 11 [11! 829-41 (1928).

(b) R. A. Heindl and W. L. Pendcrgast, "Fire Clays.Some F"undamental Properties at Several Temperatures,"ibid., 13 [10] 725-50 (1930).

(c) J. F. Hyslop, "Chemistry of Fire Brick," Jour. WestScot. Iron Steel Inst., 40 (Part 2] 23-28 (1932).

(d) L. Tscheischwili, W. Biissem, and W. Weyl, "Meta-kaolin." Ber. deut. keram, Ges., 20 [61 249-76 (1939); Ceram..\hs.. 18 |10| 279 (1939).

Considerable study has been made ofthe changes that take place in kaolinite^when it is heated at various elevatedtemperatures, but no detailed studyappears to have been made of the effectof heat on illite and montmorillonite.As montmorillonite and illite are thedominant constituents of many clays, andminor but important constituents ofother clays used for ceramic purposes, aknowledge of the effect of heat on theseclay minerals is important for an under-standing of the firing characteristics ofmany clays. Further, the firing behaviorof clays that are mixtures of clay miner-als cannot well be understood until thefiring characteristics of the pure clayminerals have been studied. It is thepurpose of this report to give the resultsof an investigation of the changes thattake place when illite and montmorilloniteare heated at various temperatures up to1400°C.

PROCEDURESamples of illite were purified from two

Pennsylvanian underclays and one argil-laceous Ordovician dolomite found inIllinois. The crude samples were selectedon the basis of previous detailed mineral-ogical analyses of many argillaceous ma-terials, which showed that these materialscontained illite that could be easilyseparated from the other constituents bya simple sedimentation procedure. Theactual purification procedure was as

I 5 1

EFFECT OF HEAT ON CLAY MINERALS

follows: The clays (including the clayfraction of the dolomite after removal ofthe carbonate by dilute HCl) were placedin suspension. Ammonium hydroxide wasused as the dispersing agent because nosalt residue forms when the sample isdried. The suspension was allowed tostand until material coarser than 0.001mm. had settled out. The suspension re-maining was removed and dried on asteam bath. This material, composed ofparticles finer than 0,001 mm., was shownby X-ray diffraction analyses to containonly illite as the clay mineral. .\ smallamount of quartz that could not be re-moved in the sedimentation process wasalso present. Before firing, the dried ma-terial was ground to pass a 100-meshsieve.

It has previously been reported"- thatclay minerals belonging to the illite groupmay vary somewhat in their physicalcharacteristics and their chemical com-position, notably in the K2O content.One illite sample (No. 1) is representativeof illite with relatively high plastic andbonding properties, slightly lower KjOcontent, and smaller ultimate particlesize. The other illite samples (Nos. 2 and3) are representatives of illite with re-latively larger ultimate particle size andlow plastic and bonding properties. Thechemical composition and source of thepurified illite are given in table 2.

One sample of montmorillonite (No. 4)was purified from a sodium bentonite(pH = 9.04) by the same process usedfor the illite samples. The purified samplewas composed of particles less than 0.001mm. in diameter, and X-ray diffractionanalysis showed the presence of mont-morillonite and about lO^/f of quartz.The other montmorillonite sample (No.5) was an impure calcium-hydrogenbentonite (pH = 4.10), containing quartzin grains up to 0.1 mm. and limonite inaddition to the montmorillonite. Thislatter sample was not purified.

Small separate portions (about 0.2gram) of each sample were heated fortwo hours in platinum crucibles in anelectric furnace at 100°, 150°, 200°C., and

'-(a) R. E. Grim, "Relation of Composition to Propertiesof Clays." Jour. Amer. Ceram. Soc, 22 |5| 141-51 (1939):Illinois State Geol. Survey Cir. 45. 1939.

(b) R. E. Grim and W. F. Bradley, "A Unique Clayfrom the Goose Lake, III., Area," Jour. Amer. Ceram. Sor.22 15) 157-64 (1939); Illinois State Geol. Survey ReportInv. 53, 1939.

Table 2.

—

Chemical Composition and Source of

Samples'*

Sample (1), purified illite from Pennsylvanian underclay.Grundy County, 111.; (2), purified illite from Pennsylvanianunderclay, Vermilion County, 111.; (3), purified illite fromOrdovician Maquoketa clayey dolomite, Kane County, 111.;(4), purified montmorillonite from Cretaceous bentonite.VVyo.; and (5) bentonite from Cretaceous, near Pontotoc,Miss.

Sample Nor

(1) (2) (3) (4) (5)

Si02 52.79 51.22 51.10 64.60 57.55AI2O3 24.99 25.91 21.96 20.80 19.93Fe20.s 4.68 4.59 6 38 3.05 6.35FeO 1.10 1.70 1 65 0.46 0.95MgO 2.70 2.84 3.91 2.30 3.92CaO 09 0.16 1.73 0.53 1.94Na20 20 0.17 03 2 60 0.33K2O 5.86 6.09 6 62 0.40 0.59Ti02 0.53 46 0.15 32Ignition

loss 7.14 7.49 6 24 4.74 8.53Total 99.55 100.70 100 08 99.63 100 41H2O+ 6.83 7.14 5.75 4.74 8.51H2O— 5 56 1.45 1.74 8 00 7.43

^Analyses made under the supervision of O. W. Rees,Chemist, Illinois vState Geological Survey.

SO on at 50°C. intervals, up to 1100°C.Additional portions of each sample wereheated in x^lundum crucibles in a gas-fired furnace for two hours at 1200°,1300°, and 1400°C. In order to determinewhether a two-hour firing approachedequilibrium conditions, small portions of

each sample were fired for six hours at450°, 600°, and 800°C. No significantchanges were apparent for these longerheatings, but the gradual appearance ofsome features would suggest that trueequilibrium may not strictly have beenattained. Equilibrium, however, is notcritical in these studies. All temperaturemeasurements were made by thermo-couples. The heated samples werestudied with the petrographic microscope

and by means of X-ray diffractionanalysis.

CHANGES IN PHYSICALCHARACTERISTICS

Illite

The three samples of illite, before firing,were light gray to light gray-green incolor. All of the samples showed the samechanges in characteristics up to about1100°C.; they became very light yellow-red at 250°C., and this color darkenedslightly with increased firing tempera-

ILLITR ./.YD MOXTMORILLOXITE

tures up to about 900°C. At 900°C., thesamples were red-brown and had a loosepowdery character; at 950°C., they re-tained the same color but had formedaggregates that could be broken with

slight pressure. All samples were well

vntrihed bv 1100°C., and thev were com-pletely vitrified at 1200°C. 'Sample (1)at 1300°C., was brown, dense, slightlybloated; at 1400°C., it was a steel-grayglass with many large bloat holes. Sample(2) at 1300°C. was a steel-black glass with

many bloat holes which had disappearedat l400°C. Sample (3) showed a higherdegree ot vitrification at 1100°C. thansamples (1) and (2). At 1200°C., it was adark brown glass with many small bloatholes which had disappeared at 1300°C.and 1400°C.

It is interesting to note that the illitesample (No. 1) with the high plastic andbonding properties and slightly lowerKoO content is the only one that re-mained bloated at 1400°C. Sample (3),

jA* , Jftv sS A A A ;»1 c r\ 1 ^\ ^^\\\ 11 60

^oHO

/N<

liJ

\v^jj

^/^ >^^oXUJQ

y\ >\,\s^ ^/ \

1 t^O

zoo 4 00 600 800TEMPERATURE(DEGREES C)

1000 1200 1400

Fig. 1.—Indices of refraction and representative X-ra>- diffraction patterns of sample (1). The more promi-nent lines ot \-arious phases are designated as follows: I ^ illite, Q= quartz, S = spinel, and M = mullite.

EFFECT OF HEAT ON CLAY MINERALS

which contained the largest amount ofiron, also exhibited relatively greater

vitrification at a lower temperature thanthe other illite samples.

MONTMORII.LONITE

The original samples of montmorillonitewere light yellow-gray (No. 4) and lightyellow (No. 5). At about 250°C., bothsamples became distinctly more yellow-red, and this color became darker onheating to successively higher tempera-tures. Both samples at 900°C. were lightbrown and showed no vitrification. At950°C., they had the same color but hadformed small powdery easily broken ag-gregates, thus showing the beginning ofvitrification. At 1050°C., both sampleswere partially vitrified and slightlybloated; at 1200°C., vitrification wasabout complete. Sample (4) at lvS()0° and140()°C. was brown glass with many largjebloat holes. Sample (5) at 1300° and1400°C. was steel-gray glass, not bloated.

C HANGFS IN MINERAT. COMPOSI-riON OV IIJJTK ON IHRINCt

In'I KRI'KKTA riON OF Op'IICAI. DaTA

The indices of refraction for the threesamples of illite, plotted in figures 1, 2,and 3, show such marked general similar-ity that it may be concluded that aboutthe same changes took place in all samplesof this clay mineral. There was littlechange in the general appearance of all

three samples as observed under themicroscope on heating up to about 800°C.,and it was possible to measure 7 and avalues on all samples heated to thistemperature. The birefringence, shownby the difference between the 7 and avalues, was retained in samples fired at700°C. and then began to decreaseslightly in samples fired at higher tem-

peratures. Samples fired above 800°C.were unlike those fired at lower tempera-

tures. Samples fired between 800°C. and1000°C. had the appearance of a hetero-geneous aggregation of unlike material,

and it was possible only to approximatethe spread of the indices of refraction.

The retention of the birefringence andof the original appearance to a tempera-

ture of 800°C. would indicate that the

NF.iio01 se\\

\^

N,^^^

^.,s

Ny y\ N^^ ^/ \ •-^

Img. 2.—Indices of rcfracrion of sample (2).

Im(;. 3. Indices of refraction of sampk' (3).

illite structure was not completely lostbefore this temperature was reached.The reduction in the birefringence at700°C. suggests that the final destruc-

tion of the illite lattice begins at about

this temperature.

Sample (1), the relatively low K2Oillite with high plastic and bondingproperties, unlike samples (2) and (3)shows an initial increase in indices of re-fraction on firing up to 200°C. The ex-planation for this increase is not entirely

clear, but it probably results from thefiner ultimate particle size of this illite

as compared to the other illite sampleswith the attendant greater difficulty in

removing adsorbed water.

From 350° or 400° to 600°C., there isa decrease in indices of refraction cor-

responding to the loss of (OH) water fromthe lattice.-^ Between 600° and 700°C.,there is little change in the indices of re-

fraction or in the birefringence, suggest-

ing a period of existence of the illite from

which (OH) water has been completelylost prior to its final breakup.

:«U. Hofmann and J. Endell, "Dependence of Base-Exchange and Swelling of Montmorillonite on Preheating,"Z. Ver. dent. Chem., Beihcfte No. 35, 10 pp. (1939).

ILLITE JiYD MONTMORILLOXITK

The optical data gave no indication oithe composition of the material heated

to 800° to 1100°C. other than to indicatea heterogeneous material that is changingcontinually with increasing temperatures.

Samples heated to 1100°C. show thepresence of glass (the indices given onfigures 1, 2, and 3 for samples heated to1100°C. and above are for the glass com-ponent) through which are scatteredmany minute discrete particles that can-not be identified. Samples heated to1300° and 1400°C. sh( needles ofmullite disseminated through the glass.

Interpretation of X-Ray Data"*

As might be inferred from the foregoingoptical examination, the three illitespecimens investigated exhibited parallelsuccessions of phase changes, differingonly in the apparent relative amountsand, to a small degree, in the temperaturesat which new phases developed. Diffrac-tion diagrams illustrating these changesare reproduced only for sample (1). Theclay mineral persists in effect to about850°C. in each sample. In agreement withthe index curves, there are characteristicdifferences in the diffraction diagrambefore and after loss of (OH) water fromthe lattice. The positions of the 001reflections are shifted slightly corre-

sponding to an increase of 1 or 2%, in thecell height, and the relative intensity ofthe 004 and 008 lines are noticeablygreater for the dehydrated state. Above600° to 700°C., the illite diffraction linesgradually disappear, being observed forthe last time at 850°C. Small amounts ofa-FegOg have become apparent at 850°C.,particularly in sample (3), but it seemsprobable that this iron constituted alimonite impurity in the original clays.The basic portion (MgO, FcsOg, AI2O3)of the clay mineral gives rise, in this sametemperature range, to the appearance ofan entirely new phase, a spinel. Thespinels comprise a highly variable group

*X-ray diffraction diagrams were obtained in tlielaboratories of the Chemistry Department of the Uni-versity of Illinois through the courtesy of G. L. Clark.All diflfraction diagrams were registered with FeKa radia-tions in a camera of 6.33 cm. radius (reduced 35%).

•"•T. F. W. Barth and E. Posnjak. "vSpinel Structureswith and without Variate Atom Equipoints," Z. Krist,82, 325-41 (1932).

•'A. Swetsch, "X-Ray Experiments in Ceramics, " Ber.dent, kernm. Ges., 15, 2-14 (1934).

of cubic double (or mixed) oxides whichhave been the subject of extensive re-search.^ The spinel developed in thesethree cases has a unit cube edge of about

8. 15A. and probably represents a com-position (Mg-Fe) (Al2-Fe2)04. Both theamount present and the particle size ofthe phase increase with increased ignitiontemperature through 1100° or 1200°C.By 1300°C., however, the spinel in eachcase has dissolved in the glass. In eachsample, mullite needles appear in vary-ing amounts, first becoming apparent ataround 1100°C. and persisting beyond thesolution of the spinel. The amounts ofspinel and of mullite appear to be com-plementary, and the richer spinel speci-mens are subject to fusion at the lowertemperatures. A minor amount of quartz,from which it was impractical to com-pletely free the illite samples examined,persisted without any inversion to cristo-balite until its gradual solution between850° and 1050°C.

The spinel and mullite phases aboveare of particular interest in comparisonwith the changes on firing of the crystal-line micas. Muscovite^ ignites to leuciteand 7-AI2O3 and finally to a-AX^O.^. Abiotite examined by us yielded leucite,7-Fe203, and a very little of a spinelsimilar to those mentioned before, around1200°C., with Y-FcaOg remaining as acrystalline phase at 1350°C. These 7-ox-ide phases are, of course, themselves

spinel-like phases, but have differentcharacteristic cube-edge lengths.

Summary of Mineral Changes inIllite

Illite loses (OH) water from its latticebetween about 350° and 600°C., retain-ing its essential micaceous character butassuming the form of a new "anhydrousmodification." The final destruction ofthe illite lattice becomes complete around800° to 850°C. The optical data suggestthat the dehydrated lattice persists withlittle change between 600° and 700°C. andthat following 700°C., the final destruc-

tion is gradual. The "anhydrous modi-fication" exhibits a slightly larger unit

cell height and distinctly lower refractiveindices than does the normal mineralwith its (OH) groups intact.

10 EFFECT OF HEAT ON CLAY MINERALS

At 850°C., a spinel, a new phase, ap-pears, which increases in size of particle

and in amount with temperatures up to1100° or 1200°C. The middle sheet ofthe illite lattice, carrying the alumina,

magnesia, and iron, seems to go into the

formation of the spinel, and the alkalis

and the silica from the outer sheets go

into amorphous glass.

By 1300°C., the spinel has dissolved inthe glass. Mullite becomes apparent at1100°C. and persists to the final firingtemperature employed in this work(1400°C.). A vitrified appearance at950°C. suggests that glass formation hasstarted at this temperature. The quartzpresent persists, without any inversion tocristobalite, until it is dissolved in the

glass at about 1050°C.

^1^. - ^i^\. ^l^. Ji^ ^^ ^^

1.50200 4 00 600 800 1000

TEMPERATURE (DEGREES C)I ZOO 1400

Fig. 4.—Indices of refraction and representative X-ray diffraction patterns of sample (4). The more promi-nent lines of various phases are designated as follows: M'^montmorillonite, Q=quart7, C = cristobalite,S=spinel, and M=mullite.

ILLITE JXD MOXTMORILLOXITE 11

CHANGES IN iMINERAL COMPOSI-TION OF MONTMORILLONITF

ON FIRING

Inperpretation of Optical Data

All of the fired samples of purifiedmontmorillonite (No. 4), up to and in-cluding that fired at 800°C., retain thesame appearance under the microscopeand permit the determination of 7 and a

values (figure 4). The birefringence doesnot change at firing temperatures up to600°C., but it gradually decreases as thefiring temperature increases to 800°C.Samples heated above 800°C. show arapid drop in the spread of the indices ofretraction and a changed appearanceunder the microscope. It is suggestedfrom these data that the final destructionof the montmorillonite lattice begins atabout 600°C. and is essentiallv completeat about 800°C.

.^t^i ^^ ^^ MK Jsk iW^ ^^ ^KS^. ^1^ ^^\.

58

1.56

1.54

1.52

1.50

X\ ^^ ~--^ C:::^^ \\ . /

/^^\ /

/

\/ \ / \ \,\./

V

200 400 600 800 1000TEMPERATURE (DEGREES C)

1200 1400

Fig. 5.—Indices of refraction and representative X-ray diffraction patterns of sample (5). The prominentlines of various phases are designated as follows: M'^montmorillonite, Q=quartz, H==hematite, C =cristobalite, S^spinel, and M=mullite.

12 EFFECT OF HEAT ON CLAY MINERALS

The indices of refraction increase asthe temperatures of firing increase up toabout 300°C. As for illite sample (1), it issuggested that this results from the slowloss of adsorbed water because of the verysmall ultimate particle size of the clay

mineral. There is no change in the bire-fringence or decrease in the indices of re-

fraction at firing temperatures up to500°C. Above 500°C., the indices of re-fraction decrease, and above 600°C., thebirefringence also decreases. This sug-

gests that the (OH) water of the latticebegins to come off by 500°C. and is fol-lowed by the final destruction of the lat-tice. Like illite, an "anhydrous modifi-cation" develops with the loss of (OH)water from the lattice, but the tempera-ture interval in which it is stable is muchshorter than that for illite.

Only the range of indices of refractioncan be measured on samples heated totemperatures between 800° and 1050°C.The components existing in this temper-ature range cannot be identified optically,and the optical data merely suggest thata heterogeneous changing mixture ot com-ponents exists.

The sample fired at 1050°C. showedglass (the indices of samples fired to tem-peratures above 1050°C. on figure 4 arethose of the glass) through which manydiscrete unidentified particles are scat-

tered. Samples fired to higher tempera-tures showed successively more glass,and samples fired at 1300° and 1400°C.were largely glass with scattered needlesof mullite.

Only a mean index of refraction couldbe measured for the fired samples of theimpure montmorillonite, sample (5), andthese data are given in figure 5. Theseoptical data are too limited to be very

significant, but the fact that the values

for this sample run about parallel withthose of sample (4) suggest that thechanges in sample (5) were about thesame as those that took place in sample(4).

"(a) G. Nagelschmidt, "Lattice Shrinkage and Structureof Montmorillonite," Z. Krisl., 93 [6| 481-87 (1936)Ceram. Abs., 15 18] 258 (1936).

(b) VV. F. Bradley, R. E. Grim, and G. L. Clark, "Behavior of Montmorillonite on Wetting," Z. Krist., 97216-22 (1937); Ceram. Abs., 17 18] 288 (1938).

(f) E. Maegdefrau and U. Hofmann, "Crystal Structureof Montmorillonite." Z. Krisl., 98, 299-323 (1937); CeraAbs., 18 [2] 58 (1939).

id) See footnote 3.

Interpretation of X-Ray Data

The two montmorillonite specimens in-vestigated differed more from each otherthan did the illite samples among them-selves, but, on the whole, they also ex-hibited similar successions of phasechanges. Numerous investigations^ ofthe behavior of montmorillonite towardlow-temperature water have been re-ported and no attention has been given tothis problem here. Interest is directedonly to the water-free pyrophyllitelikeproduct from which hydration water hasbeen irreversibly lost. This material isfirst observed around 200° to 300°C. asa triple layer structure quite analogousto the micas, and, as in the case of the

illites, an increase in the unit-cell height ofabout 1% is observable by about 600°C.,which it is assumed relates to the loss of(OH) water from the actual lattice. Thisphase, again, might be spoken of as a new"anhydrous modification" of the normalmineral.

The rather pure specimen (No. 4) ex-hibited lines of this "anhydrous modifi-cation", which persisted to about 800°C.By 850°C., the montmorillonite is rathercompletely decomposed, and a spinelphase has appeared. For this specimen,which contains little iron, the phase isapparently spinel itself, the cube length

being 8.05 A., and the estimated amountof the phase developed is the least for thefive samples studied. The crystallizationof the raw unpurified bentonite (No. 5)was distinctly less perfect to start with,and clay lines easily discerned persistedonly to about 600°C. Little more can beobserved until around 850° to 900°C.,where both a-FejOg and a spinel phasebecome apparent. The a-FejOa appearsfirst, suggesting that it develops from thelimonite impurity known to be presentrather than from the clay mineral. Be-

tween 1050° and 1200°C., the amount ofQ;-Fe203 decreases, whereas the cube-edgelength of the spinel grows from about 8.1

to 8.2 A., evidently by dissolution of the

a-FcaOa.

The spinels in each case grow and be-come more abundant and then fuse intothe glass, just as in the case of the illites.

Mullite appears, as before, around 1050°

or 1100°C., and the mullite-spinel propor-

ILLITE AND MONTMORILLONITE 13

tions appear complementary; that is,No. 4 develops very little spinel andabundant mullite, whereas No. 5 de-velops abundant spinel and only a fewmullite needles. Each of the montmoril-lonite samples precipitates considerablecrystalline silica at about 950°C. In thecase of sample No. 4, contaminated withvery hne-grained quartz, it is not certain

whether the newly synthesized sihca isquartz or cristobalite, for inversion is

taking place in the same range. In sampleNo. 5, however, it is distinctly apparentthat quartz precipitates as such and that itlater inverts to cristobalite. In each case,the cristobalite fuses with the glass at

about the same temperature as the spinel.

Summary of Mineral Changes inmontmorillonite

Final destruction of the montmorillo-

nite lattice begins at about 600°C., im-mediately following the loss of (OH) lat-tice water, and is essentially complete at800° to 850°C. The loss of (OH) water isfirst reflected in the optical values at

about 500°C. by a reduction in the in-dices of refraction. A slight increase(about 1%) in the height of the unit cellalso accompanies the loss of the (OH)water.

A new phase, spinel, appears at 850°C.and increases in particle size and abun-dance up to temperatures of aboutnOO°C. By 1300°C., all the spinel is dis-solved in the glass. Mullite appears atabout 1050°C. and increases in abun-dance with higher firing temperatures.As in the case of illite, it appears that themiddle sheet of the montmorillonite lat-tice, carrying alumina, iron, and mag-nesia, goes to form spinel. A considerableportion of the silica of the outer sheets,

however, contributes synthetic crystal-line silica, and the remainder combineswith the alkalis to form glass. The forma-tion of amorphous glass, the developmentof crystalline silica, and the beginning ofvitrification are first evident at about950°C. In sample No. 5, it is clear thatthe crystalline silica is first quartz, whichlater inverts to cristobalite. The cristobal-ite is dissolved in the glass by 1300°C.

Illinois State Geological Survey

Report of Investigations No. 66, 1940