Note on the Structure of Dickite and Other Clay Minerals

NOTE ON THE STRUCTURE OF DICKITE AND OTHER CLAY MINERALS

C. J. KseNrA ANDTou. F. W. Benrrr, GeophysicalLaboratory, Carnegie Insl.itution oJ Washington. Diffraction data from r-ray powder photographsof severalclay minerals taken by Gruner in combination with the results of a structure analysis of mica as published by Pauling have enabled Gruner to propose plausible structures for kaolinite, dickite, nacrite, vermiculite, talc, and pyrophyllite. However, as is well known, diffraction data from powder photographs are, generally speaking,insufficient for establishing uniquely the crystal structure of monoclinic minerals with many parameters. Until verified by further experimental evidence the structures assignedto these minerals should thereforebe regardedas tentative. On account of the physical properties of the material, it is of course very difficult to obtain single crystal measurementsfrom any of the clay minerals,which is the obvious reasonwhy Gruner used only powder photographsin his investigation. Dickite, however, often seemsto grow in somewhatbetter crystals than any of its congeners,and in a sampleof dickite from Pine Knot Colliery, Schuylkill Co., Pa.,1we have been able to find individual crystals large enoughfor a rcintgenographicstudy. A small crystal, 0.4 mm. long, 0.04 mm. broad, and 0.02 mm. thick was mounted on a Weis- senberggoniometer,2 and r-ray photographs were taken with the crystal rotating about the o-axisand also about the 6-axis.A Laue photograph normal to the basewas also taken. The presentpaper is concernedwith the experimentaldata thus obtained, and with the agreementbetween thesedata and the structure proposedfor dickite by Gruner. Oprrcar, Pnopn'nrrBS The plane of the optical axesis normal to the plane of symmetry and inclined 16", rear to the normal to (001); b:l; ai 9:160'

1 Beautiful samples of transparent but minute crystals of dickite (labeled Honess & Williams) were kindly sent to us by Dr. F J. Williams, Pennsylvania State College, to whom we are indebted. 2 We have used much smaller crystals than that of dickite in our r-ray work. They require a somewhat longer exposure, but the resulting photographs are uni- formly more satisfactory than those from large-sizecrystals (seeFig 1). Manipula- tion of such small crystals, however, requires a slightly modified technique compared with that of optical goniometry: the adjustment and orientation can be effected only photographically and with :r-rays

631 632 THE AMERICAN MINERALOGIST optically positive with 2V:70o; no perceptibledispersion. (These measurementswere made with a Fedorov stage and are accurate to 2".) The indicesof refractionareiot:1.562, P:1.565, "y: l.JJ1, all probably within +0.001 (Na-light). These properties identify the mineral as dickite.s

Uwrr Cnr,r, DrlrBNstoNs The geometricalelements of dickite, as derived from our r-ray measurementsare given in Table 1. With them are compared the elements derived from measurements with the reflection goni- ometer by Miers,aand the planar spacingsof Gruner,5who, however,had to use Miers' value of the angleB in order to arrive at the spacings. Tesr-B 1 Gnolro:rnrclr, Er,elreNls or Drcrttr

5.145 8.882 14.3370.5789 .,1 1 .61+2 96" 45' Ksandaand Barth 0.5748 : l: 1.5997 96'49' Miers 5.1+ 8.q4 1442 Gruner

The accuracy of the spacings can be seenfrom Tables 2 to 5 in which the new tr-ray measurements have been listed. The angle B has been determinedwith an accuracyof * 10'. Our values lead to a unit cell containing four moleculesof the formula Al2Si2O5 (OH)*; calculateddensity: p calc.:2.618, density measuredby im- mersion:p:2.62.6 Arourc AnnaNcBunNr The correct spacegroup of dickite, as indicated by the usual er-raycriteria, is either Cl^ or Ci. The differencebetween them is that Cl, has a symmetry center which is lacking in C]. A test for piezo-electricityTgave a negative result, which makes C] improb- 3 Ross, C. S., and Kerr, P. F., The kaolin minerals: (/. S. GeoI Surl., ProI Paper 165 E, pp. 151-180,1931. a \{iers, H. A., Mi.nerol,.Mag., vol.9, p. 4, 1890. 5 Gruner, J. W., The crystal structure of dickite: Z. Krist, vol. 83, pp.394-404, 1932 6 Dick, Allan B., On kaolinite: Mineral, Mag., vol 8, pp. 15-27, 1888. 7 Kindly made by Dr. R. T. Milner of the U. S. Bureau of Chemistry and Soils on two samples of dickite, one from Pine Knot Colliery, Schuylkill Co., Pa., and the other from Red I\4ountain. Colo. JOURNAL MINERALOGICAL SOCIETY OF AMERICA 633 able. In his paper on dickite Gruner8 has considered two different structural arrangements which he has called arrangement A and arrangementB, both isomorphouswith Cl. For both arrangements he has calculatedthe product j F' (the frequencyfactor times the structure factor) for all reflectionsdown to a spacingof 1.300using for them the F-values given in Wyckoff's tables.eBy multiplying

Frc. 1. Equatorial Weissenberg photograph of dickite rotated about the o-axis and taken with copper radiation.

this product by the function of the glancing angle and correcting for the difference in7, we should thus, if either of the two arrangements were correct, obtain figureswhich would be proportional to the observed intensities. An inspectionof the tables 2 to 8 showsthat the agreementbetween calculated and observed intensities is not satisfactorv;10

I Op. ci,t. 'gWyckofi, R. W. G., The Structure of Crystals, 2d ed., 1931, pp.95 and 100. r0The intensity data listed in Tables 2, 3, 4, and 6,7, 8 are all based on Weissen- berg photographs of one and the same crystal of dickite. Except in Table 6, 10 in- dicates the strongest and 1 the weakest intensity of any visible spot. The intensities obtained with Cu-radiation are not directlv comoarable with those obtained with Mo-radiation. 634 THE AMERICAN MINERALOGIST but certain setsof reflectionsare in better agreementwith the cal- culationsthan others.In particular it is noticed that the reflections from the baseand from the front pinacoidare in fair agreementwith the calculations, whereas the reflections from the side pinacoid are very much off (seeTables 4, 5, and 2 respectively). A trial calcula- tion also showed that a slight readjustment of some of the 13 parameter valuesalong the c-directionwould producea still better it between the observed and calculated intensities of the basal reflections. The new r-ray data thus permit of the following conclusions: (1) The shape and dimensionsof the unit cell of dickite have beencorrectly determinedby Gruner. (2) Dickite is composedof discretesheets of cations and anions parallel to the a-b plane stacked on top of each other in the manner describedby Gruner. (3) The two-dimensional arrangement of the several atoms within each sheet is, however, different from any of the arrange- mentsconsidered by Gruner. (4) Sincethe crystal structuresof kaolinite,11nacrite,12 vermicu- Iite,I3and halloysitel4have been calculatedon the basis of the in- correct structure of dickite, the proposed structures of all these minerals cannot be accepted as being correct without further proof. Tasra 2 Rclr,ncnoNs lRoM TH.o Ston Prxacoro (EQueronur- Pnotocn,clE ol a Srwcr-n Cnvsrar- ol Drcrrru, Rorarco Alour rnlr a-Axrs)

Intensity

CUI( Calculatedfor Observed radiation position

020 p 4.43r 8.862 020 ot]-oz 4.430 8 860 0.0 +.7 4 040 qtt@2 2 220 8.880 0 8 0.0 I -l- 060 p |.478 8.868 4+ 060 dt-1- q2 r.477 8 862 6.6 66 b6:loro:3 367[

11Gruner, I.W , Z.Krisl.,voI.83,pp. 75-88,1932' 12Z. Rrist.,vol.85, pp.345-354, 1933. rr Am. Mineral'.,vol. 19, pp. 557-575,1934' 1aMehmel. M., tJber die Struktur von Halloysit und Metahalloysit: Z' Kri'st', vol. 90, pp. 35+3, 1935. TOURNAL MINERALOGICAL SOCIETY OF AMERICA 635

TasLE 3 Rerr,lctroNs lRoM THE Fnoxr PrNecoro (Equeronrar, Pnorocnern ol A STNGLECnvsrar, ol DrcrrrB Roraron ABour rHE 6-Axrs)

MoK Intensity Indices dno radiation Calculated Observed

200 otloz 2 555 5 110 50 J 400 qtlqz 1 278 5.tI2 30 2 600 arta2 0.3 0

d100:J.IIIA ao: 5.1464

Tasro 4 RBuncrroNs lnou rrro Blsn

CuKo and CuKg radiation MoKo radiation

Intensity Indices Radiation d001 0001 Calc.* Obs o02 B 7.t19 14.238 002 ot*az 7.120 t4.240 t7.8 7.118 t4.236 004 p 3.555 14 220 004 dl+q2 3.556 14.22+ 21.8 10 3 551 t4.204 006 {l 2.372 t4.232 006 d7+a2 2 37t t4 226 6.0 2 376 14.256 008 p 1.780 t4.2+0 008 ot1,az | 779 14.232 t.) 1 -781 14.248 00i0 p | 422 14.220 0010 ql |.422 14.220\ t.423 14.230 0010 q2 r.423 14.Xol 0012 p 1.186 t4.232 0012 qI 1.186 t4.232\ .189 14.268 0012 a2 1.186 t4.nzl 0014 p 1017 14.238 0014 q7 1.016 t4.224\ 0.6 .019 r4.266 0014 d2 1.016 14.2241 0016 p .890 14.240 0016 dl .890 14.240\ .890 r4.2m 0016 q2 ,8e0 14.2401 0018 dL .791514.247\ u.l 0018 a2 .7918t4.252) 0020 p .7t26 t4.252 n

doot:14.233it doot:14'243A co : 143324 co : 14.342A' * The calculated intensities of Gruner in this case have been checked by the use of Bragg's F-values and the theoretical chemical formula for dickite. This afiects the intensities in the decimal place only; some intensities of low spacings have not been calculated by Gruner, for them also Bragg's F-values have been used. 636 THE AMERICAN MIN ERALOGIST

Tanrr 5 Mnesurnunxrs lnou RoranoN Prrorocnergs or A SrNcrB Cnvsrer- ol Drcxrrn

The crystal of dickite rotated about The crystal of dickite rotated about a-axis, CuK radiation b-axis, MoK radiation

Layer-line oo/n a0 Layer-Iine bo/n bs 1 5.175 .t.l/.) 1 8.915 8.915 2 2.565 5.130 2 4.M4 8.888 3 1.710 5.130 J 2.972 8.916 - ,,- 2 ao:J.l4JA + 2.222 8.888 q | .782 8.910 6 r.478 8.868 bo:8 ' 898A

Tesln 6 INTENSTTvDere rneu AN EeuAToRTAL Pno:rocna.pu ol a Srucr,n Cnvsret or DrcKrrE, Rourrn ABour rrrE a-Axrs, CuK ReorlrroN

Calculated for position Indices Observed AB

021 4.8 4.8 7 022 4.6 0.4 8 023 2.O 4.6 ?J. 024 0.7 0.7 6 025 0.6 0.5 2 026 0.4 0.3 3 027 0.2 o.7 0 '1- 028 0.4 0.0 ') 029 0.1 n 1 041 o.2 0.4 1 042 o.2 0.4 2 043 0.4 0.8 2- OM 0.8 0.0 5 045 U.5 0.6 3+ 046 0.0 0.4 1 047 0.5 0.4 3+ 048 (r.5 01 4 061 0.0 0.0 0 462 I.J 1.3 t+ 063 0.0 0.0 0 064 0.5 1 JOURNAL MINERALOGICAL SOCIETY OF AMERICA 637

Tanr,t 7 INrnNsrrv Dlr.t rnou ar EQunronr,ql Psorocnc'pu ol e Srxclt Cnvstl.r ol Drcrrrn Rotartn ABour rr{E b-Axis,MoK Rmrenox

Indices Calculated Observed

202 6 202 23.3 10 204 0.7 3 204 7.4 6 206 2.6 A 206 9.6 7 208 4.9 3 208 I.J 2

T.q.sm 8 lNrnNsrrv Dera lnou urn Tnrno Llvnn-LrNr Pnotocnapn on I SrNcr-l Cavsral or Drcrrte Rorarnn ABour rIrE b-Lxts, MoK RmrA:rrox. DrNoranox or SvMsor.s:s:SrnoNc; m:Mnolv; p:Wnnr

fndices Calculated Observed

130 00 0 131 +.J m-

IJI s r32 0.0 0 132 0.0 0

IJD 9.5 s-

I J.) 1.0 n7 r34 00 0 134 0.0 0 135 .) .6 m 135 2.2 ln l.tt) 0.0 0 136 0.0 0 r37 9.8 ,,1 137 nt 138 00 0 138 0.0 0 139 1.5 '4) 139 to.2 nx- 330 0.0 0 331 11.0 s 331 2.5 s- 332 0.0 0 J.) Z 0.0 0

JJJ 2.2 tft

.)J.l 0.9 w

.tJa 0.0 0 11 JJO n,

.t.to 0.0 0