SYNTHETIC .{SBESTOS INVESTIGATIONS, II: X_RAYAND OTHER DATA ON SYNTHETIC FLUOR-RICHTER-

ITE, -EDENITE, AND -BORON EDENITE"

J. A. KonNt aNl J. E. CouBrono,l Electrotechnical Laboralory,(J. S. Bureau of M'ines. Norris. Tennessee.

AlsrnLct

As a portion of a general research program on the synthesis of asbestiforrn minerals,

r-ray and other data have been obtained on the following chemically analyzed synthetic

fluor-amphiboles: richterite, Na(CaNa)Mg5(SinO1)zFr, edenite, NaCa2Mg5(SiB.bAlo.rOr)zF:,

boron edenite, NaCarMgs(Si3 rBo.aOn)zFr. Comparisons are made with the values pre-

viously reported for fluor-tremolite, Ca:Mg5(SirO1)rFz.A detailed indexing of *-ray powder diffraction patterns has been made in the range up

to 76o 20, and accurate unit cell dimensions have been determined The observed cell-

dimension variations are discussed with reference to ionic location and polarization. The

synthesis and analysis of additional specified compositions are needed to elucidate thefactors controlling fibrosity and flexibiiity in layered and allied silicate structures.

fxrnolucuoN

During an extensive study of the synthesis of fluor-amphiboles frommelts, over 100 different batch compositions were investigated; the effectof fluoride concentration and various isomorphic substitutions in thebatch on the development of asbestiform amphibole were evaluated(1). Concurrently, some of the fundamental constants of various "end-members" of the monoclinic fluor-amphiboles were determined.

In the present study, the optical and tr-ray constants of syntheticfluor-richterite, Na(CaNa)l{95(SiaOu)zFr, f luor-edenite, NaCa2l{gg(SL.bAl0.5Ou)2F2, and a boron-containing fluor-edenite, NaCa2NIg5(Srn.uBo.uOtt)2F2, have been evaluated and are compared with the valuespreviously measured for synthetic f luor-tremolite, Ca2l{gb(Si4O11)2F2 (2).

ExpBnriuBNrAL PRocEDURE

The method of synthesis and the techniques employed were the sameas previously reported (2) ; that is, a batch correspondilg to the composi-tion of the desired fluor-amphibole was melted and cooled, during whichtime the amphibole devitrified. Graphite crucibles with screw-on coverswere used to minimize the extent of volatilization and to prevent the in-troduction of impurit ies that might enter the amphibole structure orotherwise cause a deviation from the desired composition.

* A contribution from the Synthetic Minerals Section, Industrial Minerals Branch,

U. S. Bureau of Mines, Norris, Tennessee.

t Present address, Chemical-Physics Branch, Signal Corps Engineering Laboratory,

Fort Monmouth (Hexagon), New Jersey.f Present address, Sylvania Electric Products Inc., Woburn, Mass.

410

SYNTHETIC ASBESTOS I NV ESTIGATIONS

The percentage of each raw material used in compounding the syntheticfluor-amphiboles is given in Table 1. To eliminate the presence of car-

bonates, hydroxides, or other compounds that decompose with liberation

of gases, certain components were initially reacted to form stable, an-

hydrous, nonvolati le materials, such as NazMgSieOa (as glass), CaSiOs,

and X{gB2O+. Others, such as the clay and silicic acid, were dehydrated

before use. fn every instance, the particle size of the ingredients was

minus-100 mesh. After dry mixing in a glass jar with rubber balls, the

batches were packed in the crucibles to hand tightness.

Ta.slr 1, Fruon-AMpgrsolr Bercn CouposrtroNs (Wnronr-Pnn CrNr)

Raw Material Richterite EdeniteBoronedenite

411

Na2MgSi3O3*

CaSiOa+MgF2 (tech. grade)

MeotMgBron*Dehydrated Georgia clay (AtzSi:Oz)Dehydrated silicic acidt (SiOz)

3 4 . 41 A 1

/ . o

1 A I

29.2

1 6 . 92 7 . 7

/ . J

t 6 .9

13.217 .9

t 7 . 228.2

t - o1 4 . 86 . 7

2 5 . 6

100.0 100 .1

* Prepared from reagent-grade chemicals.

t Reagent-grade chemicals.

The batches were fired to 1350o C., maintained at this temperature for5 hours, and then cooled at 36o C. per hour to 1100o C., at which tempera-ture the furnace was turned ofi. The resulting crystalline reaction prod-ucts, consisting primarily of brittle, acicular crystals of fluor-amphibole(at least 80/), werc ground to minus-200 mesh. Beneficiation was car-ried out by heavy-liquid separation using tetrabromoethane and methyl-ene iodide (adjusted to 2.97 and 3.10 gm.f cc., respectively) unti l frac-tions containing at least 95/6 fluor-amphibole (microscopically deter-mined) were obtained. Each of these samples was divided into fouraliquot portions: The first was examined petrographically, and the opti-cal constants determined; the second was chemically analyzed; thethird was used for the *-ray study; and the last portion was filed.

RBsurrs

(l) Chemicol C omposition

The beneficiated samples of the synthetic fluor-amphiboles wereexamined petrographically, and in no case was the estimate of the total

412 J. A. ROHN AND J. E. COMEFORO

Tl.p.nn 2. Crruurcl.r, Axervsrs or Svlttuntrc Fluon--q.lpntnor,rs*

Constituent

Fluor-richterite

Theo-retical

58 .45 58 .88.00.00. 0 0 0 . 1 7

2 4 . 5 1 2 4 . 2 46 . 8 2 7 . 0 07 . 5 1 7 . 2 04 . 6 2 4 . 7 4

- r .94 -2 .00

Theo-retical

Actual

5 0 . 1 7 5 1 . 0 96 . 0 8 6 . 4 7

.00

. 0 0 . 2 124.O5 23 .061 3 . 3 8 1 2 . 3 03 . 7 0 4 . 2 64 . 5 3 4 . 8 9

- r .9 r -2 .06

-lFluor-edenite Fluor-boron edenite

Theo-retical

Actual

5 1 . 1 6 5 2 . 2 1.00

4 . 2 4 3 . 9 1.00 .24

24.52 24 .09t3 .64 12 .693 . 7 7 3 . 9 94 .62 4 . 85

- 1 . 9 5 - 2 . O 4

sio,AhOsBzOaFezo:Mgo

CaONarOF-

o : F

100.00 100.23 100.00 100.22 100 .00 99 .94

* Analysts: H. R Shell, R. L. Craig; samples No. 3848, 3867, and 3850, respectively.Analytical data based on samples dried at 110'C.

impurities greater than 4/6. The fluor-richterite sample contained lessthan t/6 impurity, mostly CaFz. fn the case of fluor-edenite, the majorcontaminant was l-2o/o pyroxene, probably diopside. The remainingimpurity concentration was less than |/6. The extraneous phases influor-boron edenite totaled less than 4/6, oI which approximately 3/ewas forsterite,

The chemical analyses of the beneficiated samples are given in Table

raslr 3' t "'."":::i;iff ,'"";n:i:;T*H;:'" ^" roNrc Rarro s

(WOrz) (xot (YO6) (ZOe)

RichteriteTheoretical

Calculated*

Theoretical

Calculated*

Theoretical

Calculated*

Na (CaNa)

Nao ,r (Car orNao sz)

Na CazNao ss (Car s+Nao m)

Edenite

Boron edenite

Mgr (Si?AD OzzF:(Mgu zsAlo rs) (Si7 rzAl0 88) OzsFz rs

MguMgr sa

MgtMga sz

SitSir o,

OzzFzOzzFz or

Na CazNao ga (Car azNao rs)

(si?B) ozFz(Siz rBo se) Os:F: ro

* Calcuiated from chemical analysis on basis of 22 oxygen atoms. No attempt wasmade to correct for the small amounts of impurities present.

SY NTH ET IC A SBESTOS I N V EST IGAT ION S

2. A comparison between the theoretical and empirical formulas of the

three fluor-amphiboles appears in Table 3.

(2) Opticol Properties

Comparisons of the optical constants reported on various naturally

occurring richterites and. edenites with those measured on the synthetic

fluor-amphiboles are rendered difficult because the natural minerals in-

variably contain significant amounts of cations other than those re-

quired for theoretical compositions (3,4). For this reason, the observed

difierences cannot be attributed entirely to complete replacement of

hydroxyl by fluoride.The refractive indices of the synthetic fluor-amphiboles were measured

at room temperature with a petrographic microscope, using the oil-im-

mersion technique. The results obtained are compared in Table 4 with

those previously measured for synthetic fluor-tremolite.

Tesla 4. SvNrnrrrc Fr-uon-AupnrnolB Oplrcal CoNsraNrs"

413

Tremolite Richteritet Edenitel Boron-edenite

XYZzAc2VSign

1 . 5 8 11 . 5 9 3r .602

2 lo86+'( - )

1.6031 .6141 .622

22"72"

( - )

1.605r - o l /t .624

18'69"

( - )

I .5881 .5981.605

12"750( - )

* Determinations by M. V. Denny; maximum error for X, Y, and Z is +0'00.2;2Y

(3) X-ray Data

No r-ray data are available in the literature on relatively pure syn-

thetic or natural amphiboles of the types herein concerned. In the pres-

ent stud.y, accurate unit cell dimensions have been obtained, and com-

plete diffraction data are given ip the range up to 76o 20.The three syn-

th"ti. fluot-amphiboles studied are monoclinic, with a bimolecular unit

cell.All r-ray diffraction data were recorded using a chart operation in con-

junction with a Philips high-angle goniometer (diffractometer). The

synthetic fluor-amphibole samples were packed in the usual rectangular

aluminu- holders. The instrumental setting used was as follows: scale

factor, 16 (unless rescanning a very strong maximum); multiplier. 1'0

414 J. A. KOHN AND r. E, COMEFORO

(giving a counting rate of 800 counts per second, full scale) I time con-stant, 4 seconds; Geiger overvoltage, 300 volts; divergence slit, 1o1 scan-ning speed, |o per minutel chart scale, ] inch per degree.

Both before and after each chart operation involving a fluor-amphibolepattern, appropriate silicon maxima were recorded using the siliconstandard compact furnished with the instrument. Corrections rangedfrom 0.005o to 0.055o and were read from a curve plotting instrumentalcorrection against 20.

A low-power microscope fitted with a movable-hair ocular, used inconjunction with a photographically processed slide of 200 lines per inch,permitted very accurate 20 readings of the diffraction maxima. For thosepeaks directly concerned with the calculation of the unit-cell dimensions,20 readings to the third decimal place were obtained. For all othermaxima, measurements were taken to 0.005'. Following the procedureestablished by Donnay and Donnay (5), the peaks were bisected at ap-proximately two-thirds of the peak height to obtain the readings.

In the case of fluor-tremolite, the positions of the maxima directly in-volved in the calculation of the cell dimensions were determined by acounting operation in conjunction with the Philips unit. For the presentstudy, a chart operation was found to be of approximately the same ac-curacy (see Table 8) and substantially less time-consuming. After thecomplete pattern of the fluor-amphibole under investigation was ob-tained by scanning down-scale from 76o 20, f.our sharp, unambiguously-indexed maxima in the higher 20 range were selected. Each chosen peakwas then scanned four additional times. 2d values averaged from the 5separate measurements were used for a solution of the quadratic form(6). The unit-cell dimensions thus derived were refined until close agree-ment was obtained between the calculated and observed 2d values ofvarious selected maxima. The final cell dimensions are given in Table 9and compared with those previously obtained for synthetic fluor-tremo-Iite.

Following the last cell-dimension refinement, the positions of all re-flections permissible by the space group symmetry (C 2/m) were calcu-lated in the range up to 76" 28 (approximately 250 potential maxima ineach case). The diffraction data for all resolved maxima (and a few sig-nificant doublets) in this range are tabulated in Tables 5, 6, and 7 forsynthetic fluor-richterite, -edenite, and -boron edenite, respectively. Re-sults bearing on the accuracy of the difiraction data, and ultimately uponthe unit-cell dimensions adopted, are collected in Table 8, and comparedwith the values for synthetic fluor-tremolite. fn the present study, nodiffraction maximum showed a deviation in 20 of. more than 0.02o fromthe calculated value, and the average deviation was in the order of 10-adegrees.

T.lsre 5. X-Ray Drrrn,tcrron Dlrl (Pownrr) rOn SVnrsnTIc Fr-uOt-Rrcntrnl're(S4ace Grouq C 2/m)

2d obs. 20 calc.* Meas. Int. iJ calc.

020110130111200

0402201 1 1137131

330J J I

151061202

350. t J l

421t71171

t32261202351370

9.865"1 0 . 5 117 .4918 .2518 .635

19.762 t . r t s22 .27523.O326.34

3 1 . 9 132 .913 3 . 1 534.73J . ) - . ) . ) . )

37 .80538.6838 .88539 .38541 .515

41.6341 .81M . 2 044.90545 .435

9 .855 '1 0 . 5 1 517 .5018 .25518 .645

19.775z l - l J

22.26523.04526.345

26.50527 .28528 .5530 .3330. 52

31 .9132 .893 3 . 1 5J 4 . / l )

J J . J J . )

37 .80538.66538 .87539.3654 1 . 5 0

41.644 1 . 8 1M.r9544.9145.43

46.4648 .02549.0550.25550.80

52 .58555 .6756. 29557 .20558. 105

J 6 - / l . )

59 .096t .67 56 1 . 9 5 56 3 . 1 0

64.09564.9868. 555

t7o .35s\\70.3ss/70. 5757 2 . 2 373.505

- 0 .01 '+ .005+ .01+ .005+ .01

0- .02

0_ . U I J- .o2

0- . 0 1 5- . 01- .02- . 015

+ .010

- . 0 0 5+ .005- . 0 0 5

- .005- .005_ .01_ .02- . 0 1 5

8.e7e A8.4095.0674.8604 . 7 5 9

4.4894.2053.9933 .8593.380

3 .3603.2653.1242 . 9 M2.926

2 .8032 . 7 2 12.7002.5822.524

2 . 3 7 82 . 3 2 72 . 3 t 52 .2872 . t 7 4

2 . 1 6 72 . 1 5 92.O482.Ot71.9947

1 .9528r.89281 .85561 .8139| .7957

| .7389r .6496r . 6 3 2 71 .6090r .5862

r .57rrt .56201.5026r .49641 . 4 7 2 r

1 .45161.4340l . J o / o

I r .3370\ 1 .3369

1 .33331.30681 .2872

> 10023

1 1

10424

10

150 26.505240 27.285310 28.5422r 30.325151 30.505

+ .015+ .015- .01+ .01s+ .00s

00

+ .01+ .005+ .015

190 46.465510 48 .03191 49.06530 s0 .2750 . 1 0 0 5 0 . 8 1 5

5 1 2 5 2 . 5 7 546t 55.67480 s6.2951 . 1 1 . 0 5 7 . 2 0600 58 .10

552 58.695620 59.1055 5 1 6 1 . 6 80. 12 0 6 r .96M2 63.10

3 . 1 1 . 0 6 4 . 0 8661 64.9755t2 68.545

991\ 7o.3ssJ J Z )

263 70 .5757s1 72 .232 ' 1 2 . 2 7 3 . 4 9 5

I

A

+ .0100

+ .005+ .005

+ .02- . 0 1 5- .005- .005

0

+ .015+ .005+ .01

00

+ .01

60>>100

1 1

/ o9

2076

1 l757

38355

29

82

z

t66

109

32671

51 1

o

1

* Using tr CuKcr (1.54050 A) above 25" 2g and x CuK- (1'5418 A) below 25'20'

416 T. A. KOHN AND J. E. COMEFORO

T.lsrs 6. X-Rey Drrrnac:rroN Derl (Powonn) ron Svxrnruc Fluon-Bnnnrrn(Space Group C 2/m)

20 obs 20 ca.lc.* L20 Meas. Int. d calc.

020 9.825"1 1 0 1 0 . 5 1 51 1 1 1 8 . 1 5200 18.64040 r9.7I

220 21.095131.)O4li zo'rt

240 27.24310 28.54

30 .39531.8732 .74J J . l . 1 . )

34.64535.3736. 14537 .3537 .76

35T 38.5142T 38.7017I 39.24261 41 .815351 44.95

370 45.36190 46.325510 48.0246T 48.47530 50.25

0 . 1 0 . 0 5 0 . 6 6 5550 54.485461 55.735480 56.2051 1 1 . 0 5 7 . 0 3

600 58.095402 59.46

6 I i r . n i 6 r . 7 et . ro . i ' 6J .3os

3 . 1 1 . 0 6 3 . 9 566T 64.76519 68.807r0 69.24751 72.or

2 . t 2 . 2 7 3 . 2 3 5

9.825"1 0 . 5 1 518. r718.64519.725

2r . t t sJ26.37 \\26 .37s127 .24528.54

/30 .40s\\30 .415/

J r . 6 / . ' )32.73533.r4

34.64535 .3636. 145J / . J J

J / . / J J

38.503 8 . 7 139.244 1 . 8 1 544.96

45 .3546.3448.01548.4650.235

50.6654.47 5q \ 7 ?

56.2157 .045

58 .09s9.445

1 6 r . 7 7 \\ 6 1 . 7 8 /63.29

63.94564.7768.78569.2372.0r5

7 3 . 2 4

+ .02

9.002 A8 .4134.8824 .7 594 .501

4.208J3 .377l3 .3763 .2703 .125

(J O17

\t.sie2 .8052 . 7 3 32 . 7 0 1

2.5872 . 5 3 62.4832 . 4 0 62.38r

2 . 3 3 62 . 3 2 +2.2942 . 1 5 82.0 t4

1 .9980r . 9 5 7 7r.8932r .87691.8146

1.80041 .68301 .64811 . 6 3 5 11 .6130

1 .586.5I . . ) . ) .1.)

l r .so06\1.soo3t .4681

1.45461 .43811.36361.3560t . 3 l o 2

r .2913

0"0

+ .02+ .005+ .015

481

6

6

10A A

99

t9

J I

7

71 1128

+ .0050

+ .005- .005+ .015

0- . 01

00

- .005

- . 01+ .01

00

+ .01

221\lsI/330331151

061202r70401350

77598

- .01+ .01s- .005_ . 0 1- .015

- .005- . 0 1- . 0 1 5+ .005+ .015- .005- . 0 1 5

- . 0 1 5

- .005+ . 0 1- . 0 1 5- . 0 1+ .00s

+ .005

zt2I6

12

1 2+3

4

1 2

1

21 1.)23

2

* Using X CuKar (1.54050 A1 aborre 25" 20 and X CuKo (1.5418 A) below 2.5' 2d.

SYNTHETIC ASBESTOS INVESTIGATIONS 417

Tnsrs 7. X-Rly Drrrnecfior Darl (Powoer) r,on SvlttEnrrc Fr.uon-BOnor EonNrtn

(SPace GrouP C 2/m)

20 obs. 20 calc.* Meas. Int. d calc'

+ .0050

- . 0 10

- . 0 1 5

020110130117200

0402201 1 1137131

2+O31022r330J J I

151350\4001. ' J IA a 1

9 .845"10. 541 7 . 5 018 .25518 .675

1 9 . 7 5 52 1 . 1 5 522.32523.0426.39

27 .3 r28.6r30.40531 .95s32.925

3 3 . 1 837 .8538 .67s38 .915

s9 .37541 .8645 .0045.+7

r t I

26r351370\

190 46.48510 48 .1319T 49.04530 50 .370 ' 1 0 . 0 5 0 . 8 0 s

+61 55 .7660T\ (6 16480I1 . 1 1 0 5 7 . 1 8 5600 58.235

620 59 .235402 59.390. 12 .0 61 .963 . 1 1 0 6 4 . 1 1 566T 65.065

< 1 t l

iii" 68.76si\o' 6s.4r730 7r .20575T 72.35s

2 12 .2 73 .50

9 .85 '10 . 541 7 . 5 1 518 .25518 .685

19.7752 t . t 6 522.30523.O452 6 . 3 8

27 .3r528.6130.3953 1 . g s s3 2 . 9 t

3 3 . 1 7 5/37 .8.s\\37 .86t38 .6838.92

39 .36541 .85544.985

14s.47 \\4s .47 |

46.4654 8 . 1 349.0550 .3650.80

.).). / /156.3ss\\s6 .36 i57 .20558.2+

59.22559 .3961 .95564.13t)5 . uo)

168.73s\\68 .77 [

69.4171.20s72.345

/ J . J I

+0.005'0

+ .0150

+ .01

+ .02+ .01- .02+ .005- . 01

s.e7e A8.3955.0644.8604 . 7 4 8

4.4894.1983.9863 .859J . J / O

3 . 2 6 23 . 1 1 82.9392 . 7 9 82 . 7 t 9

2 .698[2 .37s\ 2 . 3 7 42 . 3 2 62.312

2.2872 . 1 5 62.013l1.9e3r\ 1 .9929

1.95261 .88881 .85561 .8104r .7957

r.6469lr.6312i 1 . 6 3 1 1

1 .60891 .5828

1 .55881.55491.49641.45091.3423

Ir.364s\ 1 .363e

1 .3528t.32311 .3050

| . 2 8 7 2

495

z8

65o

7

J J

>>100t6299

20

6

86

47

5

- .005

2221

+ .00s+ .005- . 0 1- .005- . 0 1 5

+

T

+ .o2+ .005- . 0 1

0- . 0 0 5+ .015

0

a

t92oL

8

J

7

J

L

.t

2

00

- . 0 1

.0150

. 0 1

. 0 1

.005

. 0 1

2L

5L

I J

+ .01

* UsingI CuKcr (1.54050 A; uborr" 25" 20 anrl, x CuK. (1'5418 A) below 25'20'

418 J. A. KOHN AND J. E. COMEFORO

Tesre 8. AccuRlcv or. SyNrrrnrrc Fr,uon-AMpnrsolr Drlrn,lcrrox Dern

MaximumDeviation(degrees)

AverageDeviation(degrees)

ContributingPeaks

(number)

TremoliteRichteriteEdeniteBoron edenite

+2.5 X 10-4-4.0x 10-4-2 .5x10 -1*1 .3X to - r

0 .03.02.02.02

4l5 14)38

Frc. 1. Plastic-ball model (8) of a portion of the fluor-termolite structure, .r,r'ith

[d1s0] perpendicular to the plane of the figure.

SYNTHETIC ASBESTOS INVESTIGATIONS 419

DrscussroN

A model of a portion of the fluor-tremolite structure, as derived byWarren (7), is depicted in Fig. 1 to supplement the following discussion.

In richterite, Na(CaNa)Mgb(Si4O1r)2F2, the substitution of 2 Na+ for1 Ca++ presents two geometrical possibilities for the sites occupied bythese cations. Specifically, 1 Na+ can fill the 12-fold vacant sites knownto exist in tremolite (environment similar to the 12-fold positions in mi-ca), while the remaining Na+ proxies for Ca++ in S-fold coordination. Onthe other hand, both Na+ ions may replace Ca++ in the (XO) positions,with Ca++ occupying the (WOu) vacant sites. fn tremolite, where theCa++ ions are more or less free to choose between 12- and 8-fold co-ordination, the latter is preferred, although in calcium phlogopite,CazMgo(SiaAIOro)zF+, where no such option exists, Ca# does occupy theposition of 12-fold coordination. Thus it would seem that in the particu-lar structural environment presented by the tremolite arrangement,Ca++ is more stable in the (XO8) position. Since the structural environ-ment of richterite does not difier radically from that of tremolite, itseems logical to assume that, where possible, Ca+ will again seek posi-tions of 8-fold coordination. Thus it is indicated that Na+ occupies the(WOrz) vacant sites, and the (XO) positions are shared by both Na+ andCa++. This double coordination of Na+ is not unusual, as shown byeckermannite, NaNa2MgaAl(SirOrr)z(OH,F)r, where the (WOrz) and(XO3) positions are filled by this cation.

Tasrn 9. MoxocrrNrc Cnr,r, Drlrsxsroxs on SvNtunrrc Fr-uon-AlrpntBotEs*

Tremolite Richterite Edenite Boron edenite

A O

bo

Co

pCalculated

Density

9 .78 r L 9 .82s18.007 t7.9575.267 5.268

75"29' 75"40'

3.021 g/cm} 3 035

9.84718.0045.282

75010',

3 . 0 7 7

9.80717 .9575 . 2 6 6

/ . ) -JJ

3.042

* Maximum errors are as follows: o6, *0.005: bn, *0.004; c6, *0.006; B, *5'; calcu-lated density (r,), :t0.006.

The change in unit-cell dimensions between tremolite and richterite,as shown in Table 9, can be explained by either of the above assumptions.

Both would result in an increased a dimension because of filling of thevacant sites, since [d1e6] parallels the principal axis of the (WOrz) co-ordination polyhedron. With the limited data available, the contractionalong 6 can be ascribed to substitution of a slightly smaller ion (Na+ for

420 T. A. KOHN AND J. E. COMEFORO

Ca++) in positions between the double chains. Regardless of the actuailocation of the Na+ and Ca++ ions, no noticeable change in the c directionwould result, as the latter parallels the elogation of the SirOrr chains andis essenitally uninfluenced by the occupants of the (WOu) and (XO)positions. Knowledge of the cell dimensions of fluor-eckermannite andvarious richterites, especially those in which Mn++, Bo*r, or Sr++ re-place Ca++, might resolve this question without recourse to a detailedstructural analysis.

Edenite, NaCa2NIg5(Sig b,Alo sO1)zFz, is derived from tremolite bysubstituting one AI+3 for Si+a in tetrahedral coordination and restoringelectrical neutrality by filling the (WOrz) sites with Na+. In boronedenite, NaCarMgr(Sir.s,Bo.sOrr)zFz, B+3 replaces AI+3 in the (ZO) posi-tions. As with richterite, filling of the 12-fold sites in both edenites in-creases the o dimension relative to that observed in tremolite (cf. Table9). This direction is also influenced by the substitution in 4-fold coordina-tion. Thus the o dimension of aluminum edenite is larger than that ofboron edenite, owing to the larger (atOr)-u grouping. It was expected thatthis difference in tetrahedron size between (AIO4)-5 and (BOn)-5 wouldalso be reflected in the 6 direction, with aluminum edenite exhibiting thelarger D dimension. The observed data bear out this expectation. Whencompared with fluor-tremolite, however, the 6 dimension of aluminumedenite remains essentially the same. This may be due either to a cush-ioning effect along b (perpendicular to the direction of the double chains)or to a distortion of the (AlOa)-5 grouping, which would not be unex-pected with a trivalent ion in 4-fold coordination.

The c dimension of aluminum edenite is increased relative to that oftremolite owing to the effect of the larger (A1O4)-5 grouping on the direc-tion of the double chains. fn boron edenite a contraction (relative totremolite) was expected along c, because of the probable smaller size ofthe (BOo)-s grouping in comparison to (SiOa)-a.* This contraction, how-ever, did not materialize, which intimates a distortion on the (BOr)-rtetrahedron. Such a distortion would be in keeping with the relative easeof polarization of this group. The indicated distortions of both the(AIO4)-5 and (BOn)-r groups are such that the tetrahedra seem atten-uated (relative to D) along the c direction.

In the discussion of the fluor-richterite composition, certain isomorphicsubstitutions were suggested as ofiering promise in correlating unit-celldimensions with ionic location. Continuing in this vein, the effect ofpolarization of the tetrahedra could be clarified by the synthesis and

* Analogous to the relative sizes of the (BOo;-s utrd (SlOr;-r tetrahedra in danburite,CaBrSiros(9).

SVNTHETIC,4SBBSTOS INVESTIGATIONS 421

(OH)- by F-, (10) which is basic to the study of f luor-sil icates'

When such data become avaiiable, it may be possible to comprehend

relation to crystal structure.

AcrNowr,ouGMENTS

The authors are indebted to their associates in the Electrotechnical

Laboratory who have contributed to this investigation: to H' R' Shell

and R. L. Craig for the chemical analyses; to l'I. V. Denny for the optical

determinations and the photograph; to \'Irs. G. N{' Huff for many of the

calculations involved in the cell dimension determinationsl and to R. A'

Hatch for his critical review and helpful suggestions'

Rrlnnrncns

1. Counlono, J. E., Errrr,, Wrturlu, amo Hnrcrr, R' A' (1954), Synthetic asbestos

investigations, III: The effect of isomorphic substitutions and fluoride concentration

on the synthesis of fluor-amphiboles: To be published as u. s. Bur. Mines. Rept. Inttest.

2. Couerono, J. E. aNo Konx, J. A., (1954), Synthetic asbestos investigations, I: Study

of synthetic fluor-tremolite: Am. Mineral.,39' 537-548'

3. Lensnn, E. S. (1942), Alkalic rocks of Iron Hill, Gunnison County, Colo': [/' S' Geol'

p . 118 ) .

6.-(1944),Internationale Tabellen zur Bestimmung von Kristallstrukturen: Gebriider

Borntraeger, Berlin, vol. II, revised ed., 454.

7. Wannru, B.E. (lg2g),The structure of tremolite, HrCazMgs(SiO:)s: Z. Krist.,72,44.

8. Hercn, R. A., Courrono, J. E., ann Pacn, N. A. (1952), Transparent, plastic-ball'

crystal structure modeis: Am. Mineral.,37, 58-67.

9. HenuaNn, C., et al (1937), Strukturbericht, vol' II (1928-1932)' 153-155'

10 .Ko r rN ,J .A . , awoHarc r r ,R .A . (1955 ) ,Syn the t i cm ica inves t i ga t i ons ,V l :X - rayandoptical data on synthetic fluor-phlogopite: Am. Mineral",4Orl0-2l'

ManuscriPt receiaed May 20, 1951.