THE CRYSTAL STRUCTURE OF THE DIOCTAHEDRAL MICA …clays.org/journal/archive/volume...

6
Clays and Clay Minerals, Vol. 21, pp. 465-470. Pergamon Press 1973. Printed in Great Britain THE CRYSTAL STRUCTURE OF THE DIOCTAHEDRAL MICA 2M2 DETERMINED BY HIGH VOLTAGE ELECTRON DIFFRACTION A, P. ZHOUKHLISTOV, B. B, ZVYAGIN, S. V. SOBOLEVA and A. F. FEDOTOV Academy of Sciences, Moscow, U.S.S.R. (Received 6 April 1973) Abstract The structure of a dioctahedral 2M2 mica was defined by high voltage electron diffraction. The cell parameters are: a = 8-965, b ~ 5-175, c = 20.31 A, fl = 100 ~ 40', Z = 4, space group C2/c. Despite the peculiar character of the layer dispogition(or 4 as ~r~ as), the oxygens of the layers are packed according to the cubic law. Consequently, the interlayer cations K have a trigonal prismatic coordination. The angle of tetrahedral twist is 11~ 20'. The interatomic distances T-O indicate ordered replacements of AI for Si. INTRODUCTION A STRUCTU~Linvestigation of the dioctahedral mica polytype 2M 2 is quite expedient not only because it may reveal some new data concerning a peculiar mica, but also because it is closely connected with problems of crystal chemistry and polytypism of micas in general. Indeed, in principle six possible mica poly- types consisting of centrosymme.trical layers and saris- fying the homogeneity condition are subdivided into two distinct groups: 1--1M, 2M 1, 3T; and 2--2M2, 20, 6H which are distinguished by the relative orientation of adjacent layers, differing by even (group 1) or odd (group 2) numbers of 2n/6. These groups differ diffrac- tionally by positions and intensities of reflections with k = 3n (indices correspond to axes with b = a x/3). If the formation and existence of polytypes of group 1 is quiteclear and natural, the same is not so easy to comprehend for polytypes of group 2. As shown by Bailey (1967), the action of octahedral cations is in favor of a "cubic" packing of oxygens in a layer. Such layers have been designated by Franzini (1969) as layers A in contrast to layers B with "hexagonal" pack- ing of oxygens. Orientations of group 1 give nearly a close packing of oxygens and an octahedral coordina- tion of cations in the interlayers. In contrast to this, in polytypes of group 2 favorable arrangements of oxygens both inside the layer and in the interlayers cannot be simultaneouslyrealized. If, say, all the layers are of type A, the oxygens in the interlayers are directly stacked, one over another, forming prismatic poly- hedra. As a consequence, polytypes of group 2 are less probable and only 2M2 has been found experimentally and then quite seldom. In order to explain the existence of 2M2, Radoslo- rich (1958) supposed a hexagonal geometry of the tetrahedral sheets for such cases. Franzini (1969) has found another possibility for the formation of 2M 2 polytypes by alternation of layers A and B in the struc- ture. In such a condition the interlayers are the same as for group 1 and the ditrigonal geometry of tetra- hedral sheets is not an obstacle. As it is difficult to un- derstand the reason for differences of adjacent layers in one and the same structure, one may suppose that an alternation of mode packing takes place inside each layer in dependence on the relative orientation of the octahedral sheet with the lower and upper tetrahedral sheets. All layers are thus equivalent and the structure is homogeneous, but polar. This model has the sym- metry Cc in accordance with observed reflection absences, in contrast with symmetry C[ proposed by Franzini. Independent of these considerations, the proposals of Franzini were acceptable only for trioctahedral micas, because dioctahedral B-layers would have H: bonds between OH and Ob,s not c0nfirmed by i.r.-data. Meanwhile, dioctahedral 2M 2 micas do exist, as has been shown by Drits et al. (1966) and Sokolova (1966). All this underlines the necessity of a detailed structural analysis of such a mica and that is the pur- pose of this work. STRUCTURE DETERMINATION For a structural study by means of high-voltage electron diffraction, a sample kingly given us by Mkhitarian (1969) was chosen from a number of diocta- hedral 2M 2 micas of different origin and structural order. The sample gave very high quality oblique tex- ture electron diffraction patterns (see Fig. 1). Unfortun- ately, this sample contained some non-separable foreign material. Therefore, the chemical analysis (Table 1) 465

Transcript of THE CRYSTAL STRUCTURE OF THE DIOCTAHEDRAL MICA …clays.org/journal/archive/volume...

Page 1: THE CRYSTAL STRUCTURE OF THE DIOCTAHEDRAL MICA …clays.org/journal/archive/volume 21/21-6-465.pdflayers of the 2M2 mica; thus, the above mentioned 'unfavorable' stacking of layers

Clays and Clay Minerals, Vol. 21, pp. 465-470. Pergamon Press 1973. Printed in Great Britain

THE CRYSTAL STRUCTURE OF THE DIOCTAHEDRAL MICA 2M2 DETERMINED BY HIGH VOLTAGE

ELECTRON DIFFRACTION

A, P. ZHOUKHLISTOV, B. B, ZVYAGIN, S. V. SOBOLEVA and A. F. FEDOTOV

Academy of Sciences, Moscow, U.S.S.R.

(Received 6 April 1973)

Abstract The structure of a dioctahedral 2M2 mica was defined by high voltage electron diffraction. The cell parameters are: a = 8-965, b ~ 5-175, c = 20.31 A, fl = 100 ~ 40', Z = 4, space group C2/c. Despite the peculiar character of the layer dispogition (or 4 as ~r~ as) , the oxygens of the layers are packed according to the cubic law. Consequently, the interlayer cations K have a trigonal prismatic coordination. The angle of tetrahedral twist is 11 ~ 20'. The interatomic distances T-O indicate ordered replacements of AI for Si.

INTRODUCTION

A STRUCTU~L investigation of the dioctahedral mica polytype 2M 2 is quite expedient not only because it may reveal some new data concerning a peculiar mica, but also because it is closely connected with problems of crystal chemistry and polytypism of micas in general. Indeed, in principle six possible mica poly- types consisting of centrosymme.trical layers and saris- fying the homogeneity condition are subdivided into two distinct groups: 1--1M, 2M 1, 3 T; and 2--2M2, 20, 6H which are distinguished by the relative orientation of adjacent layers, differing by even (group 1) or odd

(group 2) numbers of 2n/6. These groups differ diffrac- tionally by positions and intensities of reflections with k = 3n (indices correspond to axes with b = a x/3).

If the formation and existence of polytypes of group 1 is quiteclear and natural, the same is not so easy to comprehend for polytypes of group 2. As shown by Bailey (1967), the action of octahedral cations is in favor of a "cubic" packing of oxygens in a layer. Such layers have been designated by Franzini (1969) as layers A in contrast to layers B with "hexagonal" pack- ing of oxygens. Orientations of group 1 give nearly a close packing of oxygens and an octahedral coordina- tion of cations in the interlayers. In contrast to this, in polytypes of group 2 favorable arrangements of oxygens both inside the layer and in the interlayers cannot be simultaneously realized. If, say, all the layers are of type A, the oxygens in the interlayers are directly stacked, one over another, forming prismatic poly- hedra. As a consequence, polytypes of group 2 are less probable and only 2M2 has been found experimentally and then quite seldom.

In order to explain the existence of 2M2, Radoslo- rich (1958) supposed a hexagonal geometry of the

tetrahedral sheets for such cases. Franzini (1969) has found another possibility for the formation of 2M 2 polytypes by alternation of layers A and B in the struc- ture. In such a condition the interlayers are the same as for group 1 and the ditrigonal geometry of tetra- hedral sheets is not an obstacle. As it is difficult to un- derstand the reason for differences of adjacent layers in one and the same structure, one may suppose that an alternation of mode packing takes place inside each layer in dependence on the relative orientation of the octahedral sheet with the lower and upper tetrahedral sheets. All layers are thus equivalent and the structure is homogeneous, but polar. This model has the sym- metry Cc in accordance with observed reflection absences, in contrast with symmetry C[ proposed by Franzini.

Independent of these considerations, the proposals of Franzini were acceptable only for trioctahedral micas, because dioctahedral B-layers would have H: bonds between OH and Ob,s not c0nfirmed by i.r.-data. Meanwhile, dioctahedral 2M 2 micas do exist, as has been shown by Drits et al. (1966) and Sokolova (1966). All this underlines the necessity of a detailed structural analysis of such a mica and that is the pur- pose of this work.

STRUCTURE DETERMINATION

For a structural study by means of high-voltage electron diffraction, a sample kingly given us by Mkhitarian (1969) was chosen from a number of diocta- hedral 2M 2 micas of different origin and structural order. The sample gave very high quality oblique tex- ture electron diffraction patterns (see Fig. 1). Unfortun- ately, this sample contained some non-separable foreign material. Therefore, the chemical analysis (Table 1)

465

Page 2: THE CRYSTAL STRUCTURE OF THE DIOCTAHEDRAL MICA …clays.org/journal/archive/volume 21/21-6-465.pdflayers of the 2M2 mica; thus, the above mentioned 'unfavorable' stacking of layers

466 A.P. ZHOUKHLISTOV, B. B. ZVYAGIN, S. V. SOBOLEVA and A. F. FEDOTOV

Fig. 1. Oblique-texture electron diffraction pattern of the mica 2M2 (~o 60~

could not be recalculated directly to give a structural formula. Electron microprobe analysis established that a considerable amount of Si was concentrated on areas greater than 10 #m, and Ca and Fe were related to in- clusions about 1 5/~m in dimensions. Areas with con- stant proportions of K, A1, Si were found, and these areas should be attributed tO the 2M 2 mica. The sur- face of the powdered specimen did not satisfy the requirements for quantitative estimation of the chemi- cal composition.

According to spectral data, the specimen contains no noticeable amount of Li and cannot be a lepidolite. It differs from giimbellite by the absence of Mg. Under such conditions, the structural formula was deduced by using the subsequently established structural data.

The structural analysis was based on high-voltage oblique-texture patterns, which are advantageous in such cases, obtained with multiple exposures at opti- mum and maximum angles (60 and 70~ It is essential that reflections, which would coincide i.n an ideal case when c. cos/3 = -a/3, are distinctly resolved, indicat- ing a deviation of the 2M 2 mica lattice from the ideal greater than the case for 2M2-1epidolites and triocta- hedral micas. This situation helped to provide 504 ref- lections with non-zero intensities distributed over 14 ellipses(h 2 + 3k 2 ~< 124), in correspondence to a 2M 2

lattice. The cell parameters, measured according to the reflection positions, are a = 8.965, b = 5.175, c = 20-31 A,/3 = 10W 40'. The monoclinic lattice deforma- tion is expressed by the relation - c cos/3/c~ = .0"419. Absences 0freflections Okl with l odd indicate a space

group of symmetry C2/c or Cc. The intensities have been estimated by comparison

of patterns with multiple exposures. The most distinct reflections were measured with a photometer. The in- tensities so measured were used as standards for esti- mation of intensity of other reflections. F2-values were calculated by means of experimentally justified local intensity formulas (Vainstein, 1964).

The structure was determined by successive refine- ments of an initially ideal model constructed of three- sheet layers of a muscovite composition KAI2(Si3AI) O10(OH)2 with a hexagonal geometry of tetrahedral sheets, disposed in a sequence ~5 a4 as q4. �9 �9 (Zvyagin, 1967). The relative displacements ai are counted off in a coordinate system with axes a and b = a ~ 3 . There- fore, the structural model has the angle c~ > ~/2. Since the glide plane c passes outside the previously chosen origin in a vacant octahedral site, the transition to a standard setting was realized by an interchange of axes a,b and by transition of the origin in the position 1/4, - 1/4. The initial coordinate system (with b = a x/3), in which all polytypes are considered, is right-handed and the resulting system is left-handed. In order to use a right-handed system, the structure is reflected in the plane bc. The model was also subjected to a homo- geneous shearing deformation in a negative direction along the a-axis, in order to satisfy the real value of/3.

The structure refinement proceeded by means of plane sections of the three-dimensional potential dis- tribution normal to the b-axis for two variants of the initial models: with layers having or not having sym- metry centers (space groups C2/c or Cc). In the second case, the tetrahedra were rotated by 5 ~ but differently for the two tetrahedral sheets of a layer; i.e. according to the "cubic" law in one sheet and the "hexagonal" law in the other. Incidentally, this rotation permitted check of the plausibility of a change in the packing mode of oxygens inside a layer.

It became evident after several stages of refinement that the oxygens were being displaced into centrosym- metrical positions, so the subsequent refinement pro- ceeded within the limits of the space group C2/c.

In view of the complete or partial overlapping of many reflections, only 145 distinct and Well separated reflections of the first four ellipses were used in the first stages of refinement. Later, the intensities of composite reflections were divided in the ratios of Fc, lc 2 corre- sponding to the increasingly refined structure. The final refinement by means of least squares in an iso-

Page 3: THE CRYSTAL STRUCTURE OF THE DIOCTAHEDRAL MICA …clays.org/journal/archive/volume 21/21-6-465.pdflayers of the 2M2 mica; thus, the above mentioned 'unfavorable' stacking of layers

The crystal structure of mica 2M 2 467

- - ~ - 2 3

. . . . 4

. . . . . 5

~ 7 ~

Y " Y ~ �9 b

t "~\ ) ;, ".�9 )

, , . X6 qr Otd

Fig. 2. The normal projection of the structure on the plane ab. (1) upper octahedral bases; (2) lower octahedral bases; (3) upper tetrahedral bases; (4) lower tetrahedral bases; (5) lower tetrahedral bases of the next (upper) mica layer. The initial configuration of (3) and (5) characterized the

structure of the interlayer space.

As the tetrahedral sheets have a ditrigonal geometry, the K-cations are inside slightly sloped trigonal prisms, the average K - O distance being 2.859 A. In the projec- tion 0n ab the edges of the tetrahedral bases are rotated relative to their ideal hexagonal geometry through angles indicated in Table 4, the average rotation angle being 11.5 ~ The oxygens forming the bases of the tetrahedra have different z-coordinates so that these bases are inclined to the plane ab by an aver- age of 5-5 ~ and the surface of the basal oxygens is cor- rugated. The direction of the corrugations is parallel to the reflection planes m of single layers, along which oxygens with the least absolute values of z-coordinates are lying, with indexes [110] and [110] in the layers a S and 64, respectively.

Table 2. Atomic coordinates (with standard deviations) and individual thermal coefficients for the 2M2-structure

tropic approximation has been taken to a value R -~ 11.7 per cent for all reflections. The resulting atomic

coordinates and interatomic distances are given in Tables 2 and 3. The normal projection of the structure on the plane ab is schematically drawn in Fig. 2.

DISCUSSION

Even the first steps of refinement made it clear that a 'cubic' packing of oxygens is preserved inside the layers of the 2M2 mica; thus, the above mentioned 'unfavorable' stacking of layers is the observed mode. This peculiarity, a result of the interaction between tetrahedral oxygens and octahedral cations, has proved to be energetically 'bearable' and did not pre- vent the formation of the dioctahedral 2M 2 mica. This stacking of layers and concomitant repulsion of adja- cent O, results in an increase of the interlayer thickness to a value q = 3'41A. There is also a repulsion between Si-atoms which leads to the displacement of layers in the direction of the a-axis by a value - 0 ' 0 1 8 a = -0"16,~.

Table 1. Chemical analyses of the dioctahedral mica 2M2"

SiO2 85"81% TiO2 0"33 A1203 7"89 Fe203 1.60 FeO 0.28 CaO 0.96

Na20 0.17 K20 2.11 HO + 1.10

100.25~

*Made on IGEM Ac. Sc. U.S.S.R., Analysts C. A. Gorbatsheva and E. L. Borodina.

Atoms x y z B, A 2

K 0.0 0.0921(17) 0.25 0.45 AI 0.0900(10) 0.2468(16) 0.0040(9) 0.40 T~ 0 .1248(7) 0.5670(15) 0.1348(8) 0.34 T 2 0 .2964(7) 0.0986(15) 0.1339(7) 0.37 OH -0.0522(12) 0.0657(23) 0.0524(10) 0.25 O 1 0 .0853(7) 0.5629(23) 0.0540(10) 0.29 0 2 0-2697(16) 0-1313(22) 0.0539(7) 0.28 0 3 0.1941(15) 0.3139(22) 0.1688(8) 0.28 04 0.4788(11) 0.1294(20) 0.1675(7) 0.22 0 5 0.2570(14)-0.1986(26) 0.157l(11) 0.27

As seen from Table 2, the two symmetrically inde- pendent tetrahedra T 1 and T 2 differ in their interatomic distances. The average bond lengths T-O are 1.619 and 1.654 A, indicating that there is an ordered distribution of isomorphous replacement of A1 for Si mainly in the T 2 tetrahedra. Applying the relation d,,, = d s i ~ o ( l - x ) +

dAI_oX, where ds~_o = 1.62 and dAl_O = 1"77, the fol- lowing contents of tetrahedra are deduced: T~ = Si, T2= Sio.7sAlo.25.

The cations T1 are essentially in the centers of tetra- hedra, while cations T2 approach O,p and move away from the bases, the distance T2-Oap being even less than Tj O,p despite the corresponding average T-O distances being greater for T 2 than for T~. This pheno- menon, observed also for other layer silicates, accom- panies the substitution of A1 for Si. As Aleksandrova et al. (1972) have indicated, a weaker cation has to approach closer to O,p in order to saturate its unsatis- fied valence. At the same time, the cation is farther from O b a s and the increasingly unsatisfied valences of the latter oxygens are partially saturated by Si approaching closer to them. These shifts cause con- traction of the bases of Si-tetrahedra and expansion of the bases of Al-tetrahedra. The unsatisfied valences are partially saturated by interlayer cations.

Page 4: THE CRYSTAL STRUCTURE OF THE DIOCTAHEDRAL MICA …clays.org/journal/archive/volume 21/21-6-465.pdflayers of the 2M2 mica; thus, the above mentioned 'unfavorable' stacking of layers

468 A.P. ZHOUKHLISTOV, B. B. ZVYAGIN, S; V. SOBOLEVA and A. F. FEDOTOV

The octahedral sheet has the distortions usual for dioctahedral micas. The octahedra are flattened and the surface of the bases has a ditrigonal geometry. The rotation angles of the edges of the octahedra (Table 4) are in the range 4.5-7 ~ (the average value is 5-5~ Because of this rotation, the shared octahedral edges are shortened, The edge OH-OH (2-569 A) is the lon- gest of the shared edges; in muscovite (Gti'ven, 1971) this edge is the shortest one (2'402 A).

It is essential that, whatever the polyhedra distor- tions, the average O-O and M-O distances be in the same relation as for regular polyhedra with centered cations, in accordance with the conclusions of Drits (1970).

Table 3. Interatornic distances in 2M2--structure (A)

T-Tetrahedron T1-O1 1"617 T2-O 2 I"605

-O~ 1"643 -O3 1"681 -O, 1"589 -O, 1"659

Os 1"626 -O5 1"666

av. T~-O 1.619 av. Tz-O 1-653 O1-O 2:683 O2-O3 2.715 O-O~ 2-677 -O4 2-693

-O5 2-662 -O2 2.721 03-0,, 2.525 O3-O4 2.730

-02 2.605 -02 2-731 O4-O5 2"695 O4-O5 2"593

av. O-O 2-641 av. O-O 2.697-

A1-Octahedron* A1-OH 1.982 O1-OH 2.851 O'I-OH 2.877

-OH' 1"974 -Oz 2.779 -O~ 2.851 -O1 1.930 O2-OH 2.901 O~-OH 2.761 -O'1 2.035 O ~-O'~ 2 .518 O's-OH 2.866 -02 1.839 OH-OH 2,569 Oi-O'2 2"924

O~ 1 .977 O 2 - O ~ 2.948 02 OH 2.819

A1-O 1.956 av. O-O 2.766

K-Prisms inner outer

K-O3 2-849 K-O 3 3:265 -04 2-908 -O, 3-233 -O5 2-821 -Os 3.567

av. K-O 2"859 av. K-O 3.356

* O'1, O~, OH!--Atoms of the lower octahedral base.

The relative positions of the origins of the tetrahedral and octahedral sheets (in the centers of Si- and Al-hex- agons) indicates that the real displacements a,z have components along a,b (in a coordinate system with b = a~/3(0, +1/3), and are expressed as follows: a5 (0.340, -0.346); ~,~ (-0.340, -0.346); z (0,4.018). The deviation of the real displacements from ideal is usual for dioctahedral layer silicates and is the reason for the lattice distortion manifested by the measured values of fl and c. cos fl/a

Table 4. Rotation angles ~t of the edges of octahedral and tetrahedral bases on projection on plane ab

AL-Octahedron Tetrahedron Tetrahedron

T1 T2 Upper base Lower base

03-05 13~ ' 03-02 13~ ' O1 OH 6~ ' Ol-OH 7635 ' 03-04 10~ ' 03-04 11~ ' 02 OH 6~ ' O2-OH 5~ ' 04-02 10025 , O4-Os 9040 ' O1 O2 4~ ' Oi-O2 4045 ̀

ear. ll~ 11~ 5~

The multiplicity values for K and A1, found by least squares, and the indicated tetrahedral compositions correspond to a forrriula Ko.sAl~.94 [Si3.5Ai0.5] Olo.t(OH)l.9. By taking into account the degree of re- placement of A1 for Si, the structural formula may be dedt~ced from the chemical analysis. The calculations have been carried out for three variants: (a) consider- ing Fe also as an octahedral cation; (b) supposing A1 occupies all octahedral positions: and (c) accepting the deficiency of octahedral cations (only A1) indicated by least squares. A lack of O,OH compared to the required (O,OH)I 2 has been obtained in the first case, a surplus in the second case. Only the third variant gave, after a small correction, a satisfactory formula (Ko.6s Nao.o9) (Alz.oa) [Si3.sAlo.s] Olo.o6 (OH)1.94- On this basis only 13"3 per cent SiO2, compared to a total sample silica of 85"8 per cent belongs to the 2M 2 mica. The remaining 72.5 per cent is present as admixtures, mainly quartz.

It is interesting to compare the investigated 2M2 structure tO the structure of lepidolite 2M 2 (Takeda et al., 1971) and muscovite 2M1 (Gfiven, 1971). Lepidolite belongs to the same 2M2 modification, but differs by composition. Muscovite is qualitatively of the same composition, but has another arrangement of layers. This comparison is aided by Table 5, where structural formulas and some other features of these minerals are given.

Lepidolite 2M2 has nearly the same degree of tetra- hedral replacement of A1 for Si, but differs by having Li, as 1/3 of the octahedral cations, randomly distri- buted over all occupied octahedra. Obviously, the low degree of replacement for A1 for Si favors the forma- tion of a 2M2 mica with interlayer trigonat prisms. The anion repulsion between adjacent basal oxygen sur- faces is therefore decreased. The 2M2 lepidolite occu- pies an intermediate position between di- and triocta- hedral micas, In both 2M2 micas, the oxygen of the layers are packed according to the cubic law and form trigonal prisms in the interlayers. They differ by the angle of tetrahedral twist, which is much less in lepido- lite and favors the formation of a 2M2 -structure

Page 5: THE CRYSTAL STRUCTURE OF THE DIOCTAHEDRAL MICA …clays.org/journal/archive/volume 21/21-6-465.pdflayers of the 2M2 mica; thus, the above mentioned 'unfavorable' stacking of layers

The crystal structure of mica 2M2

Table 5. The main structural features of the dioctahedral mica 2M2, lepidolite 2M 2 and muscovite 2Mr

469

( T - O).v. - O cos fl 0 sin fl

Mineral a(b) a cq, av. (K-O)~. ~/(A) Azo, TI-O T2~O Ta-O.~. (T2-O),~.

Dioctahedral mica 2M2 8.965 0.419 i 1'~ ' 2-859 3.413 0.22 (K0.6 sNa0.09)A11.93 Si3.5A10. 5 O 10.o6(OH)1.94 Lepidolite 2M 2 (Takeda, 1971) 9.032 0.378 5020 ' 2.98 3.36 0.09 (K0. svNa0.12)(All.4Lil.05) Si3.4Alo. 6 0 lo(F1.2(OH)o, e3 Muscovite 2M1 (Giiven, 1971) 9.008 0-387 11~ ' 2-855 3-391 0.22 (Ko. 86Nao. 1)(All. 9) (Fe, Mg)o. 1 Si3A1 O10(OH)2

1.619 1-653 1.605 1.668

1.622 1.633 1'604 1.643

1.643 1.643 1.640 1.644

(Radoslovich. 1958 I. The basal oxygens Ot of lepldoli te are m o r e in a plane, which decreases the interlayer th ickness q. In muscovite , q also has a lesser value, bu t for ano the r reason, the closer packing of inter layer oxygens.

Lepidol i te 2M 2 has a grea ter K - O distance despi te the lesser r/-value, obvious ly a result of a lesser angle of te trahedra] twist c~ a n d a greater a-value. K - O dis-

tances in the 2M2 a n d 2M 1 d ioc tahedra l micas are near ly the same. Since the angles ~ are equal, the de- crease of t / m muscovi te 2M 1 is c o m p e n s a t e d by an m- crease in b (2M1) agains t a (2M:). Bo th 2M2 micas have the same k ind of latt ice d is tor t ion as expressed by the ra t io c . c o s fl/a. but this d is tor t ion is greater, as it should be. in the pure d ioc tahedra l case. In bo th 2M2 micas the te t rahedra T~ and T 2 are nonequiva len t . This non-equiva lence is also more p r o n o u n c e d in the dioc- tahedra l case. In b o t h cases the T2 ca t ions are dis- p laced towards O~p.

AcknowledoemenrsThe authors wish to express their thanks to R. G. Mkhitarian for the kind presentation of an unique sample, and to N. V. Troneva for the micro-probe investigation of sample homogeneity.

R E F E R E N C E S

Aleksandrova. V. A.. Drits. V. A. and Sokolova. G. V. (1972J

Structural features of one packet dioctahedral chlorite: Kristallographia 17, 525-532.

Bailey, S. W. (1967) The status of clay mineral structures: Clays and Clay Minerals 14.1-23.

Drits, V. A. (19701 On the relation of average anion-anion and cation-anion distances for the simplest structural polyhedra, tetrahedra and octahedra: Kristallographia 15. 913-917.

Drits. V. A.. Zvyagin, B. B. and Tokmakov. P. P. (1966) Gfimbelite--A dioctahedral mica 2M a : Dokl. AN S.S.S.R. 170, 1390-1393.

Franzini. M. (19691 The A and B mica layers and the crystal structure of sheet silicates: Beitr. Mineral. Petrol. 21,203- 224.

Gffven. N. I19711 The crystal structures of 2M 1 phengite and 2Mj muscovite: Z. fur Kristall. 134. 196-211.

Mkhitarian. R. G.. Achikgesian. S. O. and Nalbandian. E. M. (1969) On the find of the structural modification 2Mz among hydromicas of near-ore metasornatites of some pyrite deposits of North Armenia: Dokl. AN Arm. S.S.R. 49, 38-41

Radoslovich. E. W. (1958~ Structural control of polyrnor- phism in micas: Nature 183, 253.

Sokolova. E. P t1966~ On the structure of gf/mbelite: Zap. Vses. M ineralog. Obshchestva.. 95. 106-107.

Takeda. H.. Haga. N. and Sadanaga, R. (1971) Structural in- vesngation of the polymorphic transition between 2M2, 1M lepidolite and 2M 1 muscovite: Mineral. J. (Japan) 6, 203 215.

Vainstein. B. K. 11964~ Structure Analysis by Electron Dif- fraction: Pergamon Press. Oxford.

Zvyagin, B. B. (1967) Electron-Diffraction Analysis of Clay Mineral Structures: Plenum Press. New York.

Resume---La structure d'un mica diocta6drique 2M 2 a ~t6 d~finie par diffraction 61ectronique/~ haute ten- sion. Les param&res de la maille sont: a - 8.965. b - 5.175. c = 20.31 A. fl - 100 ~ 40', Z - 4. groupe spatial C2/c. En d6pit du caract~re particulier de la disposition des feuillets (a4 a5 a4 as), les atornes d'oxy- g6ne des feuillets sont ernpil6s selon un arrangement cubique. Ainsi. les cations K interfeuillets ont une coordination trigonale prismatique. L'angle de rotation du t6tra6dre est 11 ~ 20'. Les distances interatomi- ques T-O indiquent l'existence de remplacements A1-Si ordonnes.

Kurzreferat--Die Struktur eines dioktaedrischen 2M2-Glimmers wurde durch Hochspannungselek- tronenbeugung bestimmt. Die Zellparameter sind: a = 8.965. b - 5.175. c = 20.31A. fl = 100 ~ 40'. Z = 4.

Raumgruppe C 2 c. Trotz des eigenartigen Charakters der Schichtgliederung (a 4 a5 a4 as) sind die Sauer- stoffe der Schichten nach dem kubischen Gesetz angeordnet Die Zwischenschichtkaliumionen besitzen infolgedessen eine trigonale prisrnatische Zuordnung. Der Winkel der Tetraederdrehung betr~/gt 11 o 20'. Die Atomabstande T O weisen auf geordneten Ersatz von Si durch A1 hin.

Page 6: THE CRYSTAL STRUCTURE OF THE DIOCTAHEDRAL MICA …clays.org/journal/archive/volume 21/21-6-465.pdflayers of the 2M2 mica; thus, the above mentioned 'unfavorable' stacking of layers

470 A.P. ZHOUKHLISTOV, B. B. ZVYAGIN, S. V. SOBOLEVA and A. F. FEDOTOV

PealoMe- ]~HdppaKtt~ell 3BCKTpOIIOB B],ICOKOrO Hanpaxcemm onpe~enrmacb cTpyKTypa ]IHOKTa- 3apanbaoll cnlo~bi 2M2. 1-IapaMCTpSl aqefllm cne~ytoml~e: a=8,965, b~5,175, c=20,31 /~, ~= 100~ ', Z=4, npoerpaacTScHHaa rpylma C2/c. HCCMOTpa ~a cncttrlqbH~eczaB xapazTcp pacnonomcrma cnocB (o '4aso, o's), cnozxerme ZHc.rfopo~oB B C.IIOJtX c.qe]IyeT Ky6H~IeCKOMy 3axoHy. Cne]~oBaTenbao, KaTaOnbI npoMe~yTo~moro cno~ K HMCIOT Tpeyronbriyro IlpH3MftTHtleCZyIO KOOp]/HaaI~rlIO. YFYlOM TeTpa3]IpaYlbHOrO BtITKa ~IB2IJIeTC~I 11~ '. Me)KaTOMHbIe rlpOCTpaHcTBa T - O yi~a3],maroT Ha ynop~]Iosern~yro 3aMeHy Si Ha AI.

Note cnio~a = mica