115

Acta Cryst. (1998). B54, 115-120

fl-KzSO4-Type lsomorphs: Prediction of Structures and Refinement of RbzCrO4

S. ALEKSOVSKA, a S. C. NYBURG, b* LJ. PEJOV a AND V. M. PETRUSEVSKIa

alnstitut za hemija, PMF, Univerzitet 'Sv. Kiril i Metodij', Skopje, Macedonia, and bDepartment of Chemistry, King's
College, London WC2R 2LS, England. E-mail: [email protected]
(Received 26 February 1997; accepted 27 A u g u s t 1997)

Abstract Table 1. Structures isomorphous with (a) f-K2S04,

The crystal structures of members within a group of (b) (NH4)2S04 and TI2S04, and (c) (NH4)2ZnCI 4
isostructural compounds may be successfully predicted. (a)
This is demonstrated for the fl-K2SO 4 group isomorphs fl-K2SO4 McGinnety (1972)
with the general formula M2XO4, which were chosen as K2SeO4 K~lm~n et al. (1970, 1971)
K2CrO4 McGinnety (1972) and Toriumi
a family of very closely related compounds nearly all & Saito (1978)
with accurately refined crystal structures. The unit-cell KzMnO4 Palenik (1967)
parameters and the fractional atomic coordinates are K2FeO4 Hoppe et al. (1982)
shown to exhibit systematic variations with both cation Rb2SO4 Nord (1974) and Weber et al.
and anion size, as well as the Mulliken charge on the O (1989)
RbzSeO4 Takahashi et al. (1987)
atom in the tetrahedral anion. This allows the predic- Rb2CrO4 Smith & Colby (1941)
tion of the crystal structures of members in the series, Cs2SO4 Nord (1976) and Weber et al.
with only the chemical composition of the compound (1989)
being known. The agreement is good, except for an Cs2SeO4 Ztifiiga et al. (1991)
Cs2CrO4 Morris et al. (1981)
early structure determination of Rb2CrO4. The now Cs2FeO4 Mader & Hoppe (1991)
refined structure gives excellent agreement with that Cs2MnO 4 Kopelev et al. (1991)t
predicted.
(b)
(NH4)2504 Schlemper & Hamilton (1966)
(NHa)zBeF4 Garg & Srivastava (1979) and
1. Introduction Onodera & Shiozaki (1981)
(NH4)2MoS4 Lapasset et al. (1976)
A large number of compounds of the general formula (NI-h)2WS4 Sasv~iri (1963)
M 2 X O 4 have structures very closely related to that of fl- TlzSO4 Pannetier & Gaultier (1966)
K 2 S O 4 (Table 1). They are all orthorhombic, space TlzSeO4 Ffibry & Breczewski (1993)
TlzCrO4 Carter & Margulis (1972)
group Pnam, Z = 4. The two cations, atom X and two
independent O atoms of XO4 lie on mirror planes in the (c)
four special equivalent positions x, y, J etc. (Fig. 1). The (NH4)zZnCI4 (phase 1) Matsunaga (1982)
remaining independent O atom lies in eight general RbzZnCl4 Secco & Trotter (1983)
equivalent positions. Cs2BeCI4 Gaebell & Meyer (1984)
RbzZnBr4 Hogervorst & Helmholdt (1988)
Slightly different (although still related to the above) Cs2ZnBr4 Plesko et al. (1980)
are the structures related to (NH4)2SO4 and T12SO4 Cs2CdBr4 Plesko et al. (1980)
(Table lb).The principal difference between the struc- CszHgBr4 Plesko et al. (1980)
ture of fl-K2SO 4 and those of the four ammonium t Not accurately refined. Incomplete powder data are also available
compounds is the existence of hydrogen bonds, thus for RbzMoO4 and C s 2 M o O 4 (Kools et al., 1970).
resulting in appreciably lower coordination numbers (4
and 5 for the two types of ammonium anion) compared
with fl-K2SO4 and its isomorphs.t On the other hand, zewski, 1993). This stereoselectivity seems to induce
the differences between the thallium compounds and additional distortion of the already distorted coordi-
the title compounds arise most probably as a conse- nation polyhedron around TI ~.
quence of stereoselectivity of the 6s 2 lone pair in one Even more pronounced are the differences in struc-
(TI l) of the two different TI ÷ cations (Ffibry & Brec- tures related to (NHn)~ZnC14 (Table lc), which also
crystallize in space group Pnam with Z = 4. In all these
t Further evidence that ( N H 4 ) 2 S O 4 and K 2 S O 4 a r e not strictly
isostructural (isomorphous) may be given by the non-statistical
compounds the tetrahedral anions show large bond and
substitution of K ÷ with NHj, found in the solid solutions of these angular distortion. Rb2ZnCI4, a structure with disorder
two salts (Petrugevski & Sherman, 1993: Srinivasan et al., 1983). in the CI atoms, was also considered (Itoh et al., 1983).

{_) 1998 International Union of Crystallography Acta Crystallographica Section B

Printed in Great Britain - all rights reserved ISSN 0108-7681 ~.,) 1998
116 /~-K2SOn-TYPE I S O M O R P H S

Having all these structural data available, one may Table 2. Values of the independent variables used as
try to systematize them in order to extract some sort of input data in regression analyses
structural information for compounds whose structures Values for effective ionic radii taken from Shannon (1976). OchaTgcis
have either not been determined so far or whose the calculated Mulliken charge of the O atom at the tetrahedron
reliability is for some reason questionable. The purpose vertex.
of this paper is to present a possible way of elucidating
Compound R(M*) R(XO~- ) Ochargc
structural information for members in a group of very
closely related compounds. The important precondition K2SO 4 1.38 1.47 -0.864
is that (i) all compounds in question have to be strictly KzSeO4 1.38 1.63 -1.019
KzCrO4 1.38 1.61 -0.736
isostructural (isomorphous) and (ii) accurately refined KzMnO4 1.38 1.605 -0.689
crystal structures are available for a number of these K2FeO4 1.38 1.60 -0.628
compounds. By strictly isostructural (or isomorphous) is Rb2SO4 1.52 1.47 -0.864
meant compounds that crystallize with the same space Rb2SeO4 1.52 1.63 - 1.019
CszSO4 1.67 1.47 -0.864
group and at least the site symmetry and the coordi- CszSeO 4 1.67 1.63 - 1.019
nation numbers of the corresponding atoms are also the Cs2CrO4 1.67 1.61 -0.736
same. For instance, the sulfates, selenates, chromates, Cs2FeO4 1.67 1.60 -0.628
manganates and ferrates of potassium, rubidium and
caesium are beyond doubt all strictly isostructural (15
compounds in all). The structures of Rb2MnO4 and Rb2CrO4, when this project was commenced, the
Rb2FeO4 have not yet been determined, but those of agreement between prediction and observation was
K2MnO4 and K2FeO4 are known (Table la). We have poor. This we attributed to the fact that the structure
predicted the structures of all four compounds. The was determined some 50 years ago. We have therefore
agreement for the K compounds is very good. The carried out a new structure determination which we
structure of Cs2MnO4 is not accurately known. As to report here. The refined structure agrees well with the
predictions.

2. Statistical m o d e l - survey o f the diffraction data

X
For some time, we have been interested in both the
spectroscopic properties and structural relationships
between the compounds of the fl-K2SO 4 family
(Petru~evski & Sherman, 1990, 1994). Encouraged by
the results obtained previously (Petru~evski & Alek-
sovska, 1991, 1994), where the knowledge of the
effective crystal radii (Shannon & Prewitt, 1969;
Shannon, 1976) of the constituent atoms led to a
successful prediction of the unit-cell parameters, an
attempt was made to implement a simple method to
predict the crystal structures of these isomorphs. This
appeared to be possible by means of statistical analyses.
In fact, previous results proved that the unit-cell para-
meters in Tutton salts and alums vary in a regular and
predictable way with composition (Petru~evski &
Aleksovska, 1991, 1994). In the case of both classes of
compounds, the factors of principal significance were
the effective ionic radii of the constituent atoms. The
correlations were highly significant and led to predic-
tions of unit-cell parameters and of unknown values for
the effective ionic radii (e.g. the radius for hexa-coor-
dinated Ru2*). A possible isomorphism between Tutton
salts and dimethylammonium metal(Ill) sulfate hexa-
hydrates (Gale~ic & Jordanovska, 1992) has also been
discussed (Petru~evski, 1994).
Fig. 1. Crystal structure in y-projection of compounds M2XO 4 In the present study the source data in the statistical
isomorphous with/~-K2SO4. analysis (multiple regression) were the effective crystal
 ALEKSOVSKA, NYBURG, PEJOV AND PETRUSEVSKI 117

Table 3. Regression parameters used in the prediction of dependent variables [cf. (1)]
Variables with t-statistics lower than 3 were rejected in the stepwise selection model.
Dependent vari- m n p q ~Radjusted) 2
" "

ables
a (~) 2.57908 1.18523 0.22567 2.36889 0.997
b (A) 2.75848 2.13064 -0.18438 2.98931 0.986
c (A) 1.47847 1.01167 -0.25069 2.04336 0.983
x(M1) 0.01487 -0.03747 -0.02513 0.68620 0.981
y(M1) -0.01136 0.01353 Rejected 0.40702 0.580
x(M2) -0.00827 Rejected -0.01500 -0.01152 0.783
y(M2) -0.00968 Rejected -0.03111 0.69056 0.986
x(X) 0.02705 -0.03845 0.00862 0.25974 0.962
y(X) -0.00613 0.00868 Rejected 0.41521 0.787
x(O1) 0.08702 -0.15457 -0.00279 0.14242 0.999
y(O1) -0.01797 0.01996 -0.01587 0.39904 0.811
x(O2) Rejected Rejected Rejected 0.29896 0.000
y(O2) -0.03845 0.07496 0.02168 0.51931 0.968
x(O3) 0.01117 Rejected 0.02175 0.30406 0.658
y(O3) 0.01605 -0.02379 Rejected 0.36423 0.924
z(O3) 0.04920 -0.11862 -0.02612 0.12588 0.992

radii of the constituent ions and the atomic charges# of to be o r t h o r h o m b i c requires the d e t e r m i n a t i o n of three
the O atoms in the tetrahedral anions. A t o m i c charges parameters. Within this isostructural group the
were calculated at the ab initio H a r t r e e - F o c k (HF) SCF observed variation in each cell p a r a m e t e r is --q2%. As
level of theory [with the GAUSSIAN94 (Frisch et al., seen above, five atoms are in special positions whose
1995) package, on a pentium IBM compatible PC and a variable parameters are x and y (ten parameters) and
dual pentium-pro Digital workstation], using a 6-311 + one O atom is in a general equivalent position (three
G (triple zeta) basis set. This set includes diffuse parameters). Within the group there is a wide range of
functions for a better description of the anionic wave- observed p a r a m e t e r variation, some such as x (O1)
functions. The Mulliken scheme for the population varying by 0.05, whereas the smallest variation is in y
analysis was applied (Mulliken, 1955a,b). Atomic (O1), only 0.0026. Thus, in all, 16 parameters need to be
charges obtained by fitting to the electrostatic potential, determined.
with point selections according to the CHelp (Chirlian The STA T G R A P H I C S (Statistical Graphics
& Francl, 1987) and C H e l p G ( B r e n e m a n & Wiberg, Corporation, 1989) package on an IBM PC-486 was
1990) scheme, were also calculated, but these electro- used in the statistical calculations. Stepwise variable
static-potential-derived values gave somewhat worse selection was performed. This is a multiple regression
results c o m p a r e d with those obtained using the analysis that includes or rejects an i n d e p e n d e n t variable
Mulliken scheme. A l t h o u g h the potential-derived in the model, depending on the value of the t-statistics.
charges may seem more realistic, one has to be aware The point is to judge whether the inclusion of a given
that, in general, atomic charges are not quantum- i n d e p e n d e n t variable, in the presence of other variables,
mechanical observables, so they cannot be computed is statistically significant.
exactly 'from first principles'. Thus, having in mind the Since the crystal structures of the compounds were
arbitrariness of all charge assignment schemes, the use refined by various groups of authors, there are differ-
of Mulliken charges for this correlation seems to be ences in the choice of cell, orientation of axes and atom
fully justified. It should be mentioned, in addition, that designation. In order to avoid possible confusion, all
characterizations of even rather involved molecular data were transformed to match the choices made in/3-
systems are often performed employing Mulliken K2SO 4 (McGinnety, 1972). It should be mentioned, in
population analysis (Dimitrova, 1995, and references passing, that there is a misprint in one of the fractional
therein). coordinates of Rb2SeO4 (Takahashi et al., 1987) The
The parameters to be determined are the cell paper gives 0.4911 as the x coordinate of Rb(1),
dimensions and atomic coordinates. Asssuming the cell whereas it should be 0.4111.
Each d e p e n d e n t variable d (numerical value of a
t The importance of including the radii of the ions is unquestionable. unit-cell p a r a m e t e r or fractional atomic coordinate) was
The atomic (Mulliken) charge was added to account for the different
bonding ability of the O atoms in various oxo anions. In the presented as a function of the type
calculation it was assumed (for consistency) that the X--O bond
length of an XO]- anion is equal to the sum of Shannon's radii and no d = m.R(M +) + n.R(X--O) + p.Ocharge + q, (1)
further optimization was performed. The package GAUSSIAN94
(Frisch et aL, 1995) was employed for the quantum mechanical
calculations. where R(M+), in ,~, is the effective crystal radius
118 fl-K2SO4-TYPE ISOMORPHS

(Shannon, 1976) for the hexa-coordinatedt univalent Table 4. Experimental details

cation, R(X--O), in ,~, is the sum of the effective Crystal data
crystal radii for the tetra-coordinated central atom X Chemical formula Rb2CrO4
(i.e. for the X 6+ ion) and the O atom ( 0 2- ion with Chemical formula weight 286.93
CN = 2) in the XO 2- group, Ocharg e is the calculated Cell setting Orthorhombic
Space group Pnam
Mulliken charge for the O atom and q is a constant. The
a (A) 7.976 (4)
parameters m, n, p and q for each dependent variable h @) 10.692 (1)
are to be determined by multiple regression. Depen- c (a) 6.057 (2)
dent variables are listed in Table 2 and the regression v (A3) 516.5 (3)
results in Table 3. Z 4
D, (Mg m -3) 3.690
It will be seen that the predictive strength of the Radiation type Mo Ka
regression equations is, in most cases, very high. This is Wavelength (A) 0.70930
shown by the value of the coefficient of determination No. of reflections for cell 8
(Radjusted)2, which is higher than 0.9 for 10 out of the 16 parameters
0 range (°) 20.99-21.01
dependent variables. For one parameter, x(O2), there is /~ (mm -1) 20.29
no correlation between the chosen set of independent Temperature (K) 297
variables and the 11 values of the dependent variable. Crystal form Needle
However, the total variation in the numerical values of Crystal size (ram) 0.60 × 0.13 × 0.10
the dependent variable x(O2) is 0.006 and, therefore, Crystal colour Yellow
may be considered as a parameter that is rather
Data collection
insensitive to the variations of atomic radii/atomic Diffractometer Picker NRC
charges. In such cases the experimental errors in the Data collection method 0/20 scans
determination of the fractional atomic coordinates also Absorption correction Integration
become an important factor in the scatter of the points. Tmin 0.1047
Tmax 0.1908
Consequently, for x(O1), where the range of the sample
No. of measured reflections 1662
is 0.05 (an order of magnitude higher), the coefficient of No. of independent reflections 501
determination reaches the highest value (0.999) in the No. of observed reflections 468
class. Criterion for observed reflections /net > 2.50"(ln~t)
Rim 0.113
0max( ° ) 24.95
Range of h, k, l --9 --* h ---~ 9
-12--~ k ~ 12
0--+1--+7
No. of standard reflections 1
3. Crystal structure of RbzCrO4 Frequency of standard reflections Every 100 reflections
Intensity decay (%) 1.0
3.1. X-ray data
The structure of Rb2CrO4 was determined by Refinement
photographic methods over 50 years ago (Smith & Refinement on F
R 0.051
Colby, 1941). The calculated distances and angles in the wR 0.062
CrOZ4- anion differed appreciably from the values in S 8.46
K2CrO4 and Cs2CrO4 (McGinnety, 1972; Morris et al., No. of reflections used in 468
1981), suggesting rather low accuracy. This notion was refinement
No. of parameters used 41
confirmed by the prediction of the structure (see
Weighting scheme Unit
below). In fact, all calculations showed sufficiently large (A/'r)r,ax 0.002
differences from the early X-ray crystal structure Apmax (e ,X,-3) 2.28
analysis to warrant a refinement. Apmin(e A -3) - 1.98

X-ray intensity data were collected on an automated Extinction method Larson (1970)
Extinction coefficient 1.98 (16)
four-circle Picker diffractometer using Zr-filtered Source of atomic scattering International Tables for X-ray
MoKc~ radiation with pulse-height analysis. The factors Crystallography (1974, Vol. IV)
correction 00 was checked against 12 reflections at 0 --
32 ° from a crystal of sodium chloride. Data reduction Computer programs
and all other computations were carried out using the Data reduction N R C V A X DA TRD2 (Gabe et al.,
N R C V A X package (Gabe et aL, 1989). Crystal data are 1989)
Structure solution N R C V A X S O L V E R (Gabe et aL,
1989)
t Although, as mentioned, the univalent cations in this family of Structure refinement N R C V A X L S T S Q (Gabe et al.,
crystals are known to be 9-, 10- or l 1-coordinated, values for the 1989)
effective crystal radii for M ÷ ions were taken for coordination number Preparation of material for N R C V A X T A B L E S (January 1994
(CN) 6, since they are usually considered to be most reliable. publication Version)
 ALEKSOVSKA, NYBURG, PEJOV AND PETRUSEVSKI 119

Table 5. Unit-cell parameters and fractional atomic T a b l e 6. Selected interatomic distances (A) and angles (°)
coordinates in RbeCr04 in RbzCr04 (1~1 as for Table 5)
The predicted and redetermined structures (present work; standard Distance ( , ~ , ) Predicted (i) Refined (ii) [6ii -il
deviations given in parentheses refer to the least significant digit) are
given together for the purposes of comparison. 161 is the absolute Rbl--O1 2.899 2.876 (16) 0.022
difference between values in the columns indicated. The coordinates Rbl--O2 3.388 3.385 (15) 0.003
z(Rbl), z(Rb2), z(Cr), z(O1) and z(O2) are all necessarily 0.25, and Rbl--O2 3.074 3.051 (2) 0.023
are not listed. Rbl--O3 3.272 3.243 (10) 0.029
Rbl--03 3.312 3.285 (10) 0.027
Parameter Predicted (i) Redetermined [~ii - i[ Rbl--03 3.083 3.071 (10) 0.021
(ii) Rb2--O1 3.062 3.048 (15) 0.014
Rb2--O1 3.292 3.260 (6) 0.032
a (.4,) 8.0314 7.976 (4) 0.0554 Rb2--O2 2.886 2.865 (15) 0.021
b (,&) 10.7482 10.692 (1) 0.0562 Rb2--O2 2.949 2.936 (14) 0.013
c (,&) 6.1039 6.057 (2) 0.0469 Rb2--O3 2.889 2.890 (10) 0.001
x(Rbl) 0.6670 0.66788 (23) 0.0009 Rb2--O3 2.945 2.924 (10) 0.021
y(Rbl) 0.4115 0.41051 (17) 0.0010 Cr--O1 1.645 1.634 (16) 0.017
x(Rb2) -0.0131 -0.01460 (22) 0.0015 Cr--O2 1.654 1.631 (14) 0.023
y(Rb2) 0.6987 0.69781 (16) 0.0009 Cr--O3 1.649 1.632 (10) 0.017
x(Cr) 0.2326 0.2331 (4) 0.0005
y(Cr) 0.4199 0.42117 (25) 0.0013 O 1 - Cr--O2 110.4 111.0 (8) 0.6
x(O1) 0.0279 0.0285 (19) 0.0006 O1--Cr--O3 109.8 109.5 (5) 0.3
y(O1) 0.4155 0.4145 (14) 0.0010 O2--Cr--O3 108.4 108.8 (5) 0.4
x(O2) 0.2990 0.2980 (18) 0.0010 O3-Cr-O3' 109.9 109.0 (5) 0.9
y(O2) 0.5656 0.5658 (13) 0.0002
x(O3) 0.3050 0.3057 (13) 0.0007
y(O3) 0.3503 0.3510 (9) 0.0007
z(O3) 0.0289 0.0306 (16) 0.0017 T a b l e 7. Predicted structures of Rb2Mn04 and Rb2Fe04
Rb2MnO4 Rb2FeO4
a (A,) 8.036t 8.044
given in Table 4. T h e r e d e t e r m i n e d cell d i m e n s i o n s a n d b (A) 10.729 10.707
a t o m i c c o o r d i n a t e s are listed in T a b l e 5.t c (/k) 6.087 6.067
xRbl 0.6660 0.6646
yRbl 0.4115 0.4114
3.2. Predicted structures xRb2 -0.0138 -0.0147
T h e results of the m u l t i p l e r e g r e s s i o n are s u m m a r - yRb2 0.6973 0.6954
xMn/Fe 0.2332 0.2339
ized in Tables 5 a n d 6. It m a y be s e e n t h a t virtually all yMn/Fe 0.4198 0.4198
p a r a m e t e r s of the p r e d i c t e d s t r u c t u r e ( e x c e p t for the a xO1 0.0285 0.0291
a n d b axes) are e q u a l w i t h i n o n e s t a n d a r d d e v i a t i o n to yO1 0.4147 0.4136
the c o r r e s p o n d i n g v a l u e s of t h e r e d e t e r m i n e d s t r u c t u r e . xO2 0.2990 0.2990
yO2 0.5662 0.5672
S e l e c t e d i n t e r a t o m i c d i s t a n c e s a n d i n t e r b o n d angles are
xO3 0.3061 0.3074
set o u t in Table 6. T h e a v e r a g e d e v i a t i o n of R b - - O yO3 0.3504 0.3506
d i s t a n c e s b e t w e e n p r e d i c t e d a n d r e d e t e r m i n e d struc- zO3 0.0283 0.0273
t u r e s is "-'1.8 pm, t h a t of C r - - O d i s t a n c e s is also 1.8 p m
Predicted dimensions of the MnO]- and FeO~- anions (A and °)
a n d the d e v i a t i o n of the O - - C r - - O angles is 0.5 °.
It s e e m s t h e r e f o r e that t h e p r o p o s e d m e t h o d gives MnO~- FeO42-
v e r y g o o d results, at least for a series of strictly
MnFFe-- O1 1.646 1.649
i s o s t r u c t u r a l c o m p o u n d s . T h e real a d v a n t a g e of t h e Mn/Fe -- 02 1.657 1.663
m e t h o d is that it m a y give a c c u r a t e l y p r e d i c t e d struc- Mn/Fe -- 03 1.649 1.650
tures for c o m p o u n d s that can o n l y be o b t a i n e d in
p o w d e r o r m i c r o c r y s t a l l i n e f o r m or are c h e m i c a l l y O1 --Mn/Fe-- O2 110.5 110.7
O 1 -- Mn/Fe -- 03 109.9 109.9
unstable. For instance, the s t r u c t u r e s of R b 2 M n O 4 a n d O2-- Mn/Fe-- 03 108.3 108.3
Rb2FeO4 m a y be p r e d i c t e d in this way, since n o single- 03 -- Mn/Fe -- 03 109.9 109.9
crystal s t u d y is a v a i l a b l e for e i t h e r c o m p o u n d . A n
t Cell dimensions from powder diagram (Herbstein et al., 1971):
a c c u r a t e s t r u c t u r e is k n o w n for K2MnO4 ( P a l e n i k , 8.00 (2), 10.65 (2), 6.08 (2) A.
1967) a n d o u r p r e d i c t i o n s fit the s t r u c t u r e v e r y well, the
largest d i s c r e p a n c y in a n y a t o m i c p a r a m e t e r b e i n g
0.0022. O n l y p o w d e r d a t a are available for R b 2 M n O 4
t Lists of atomic coordinates, anisotropic displacement parameters ( H e r b s t e i n et al., 1971). In v i e w of t h e g o o d a g r e e m e n t
and structure factors have been deposited with the IUCr (Reference:
o b t a i n e d for K z M n O 4 a n d the g o o d a g r e e m e n t b e t w e e n
LI0250). Copies may be obtained through The Managing Editor,
International Union of Crystallography, 5 Abbey Square, Chester the p r e d i c t e d a n d o b s e r v e d unit-cell p a r a m e t e r s for
CH1 2HU, England. Rb2MnO4, we give the p r e d i c t e d a t o m i c c o o r d i n a t e s in
 120 fl-K2SO4-TYPE I S O M O R P H S

Table 7. Our calculations for K2FeO 4 yield equally good K~ilmfin, A., Stephens, J. S. & Cruickshank, D. W. J. (1971).
results (maximum discrepancy in atomic parameters, Acta Cryst. B27, 1070.
0.0026) and we accordingly predict the, as yet unknown, Kools, E X. N. M., Koster, A. S. & Rieck, G. D. (1970). Acta
structure of RbzFeO4 (Table 7). Our goal will be to fully Cryst. B26, 1974-1977.
determine these two structures by crystallographic Kopelev, N. S., Val'kovskii, M. D., Popov, A. I. & Chumaevskii,
methods. N. A. (1991). Russ. J. lnorg. Chem. 36, 1246-1250.
Lapasset, J., Chezeau, N. & Belougne, E (1976). Acta Cryst.
Further work on the kieserite group isomorphs B32, 3087-3088.
(Aleksovska et al., 1998) is also in progress. Larson, A. C. (1970). Crystallographic Computing, edited by
F. R. Ahmed, S. R. Hall & C. P. Huber, pp. 291-294.
The financial support of the Ministry of Science, Copenhagen: Munksgaard.
Republic of Macedonia, is sincerely appreciated. The Mader, K. & Hoppe, R. (1991). Z. Anorg. Allg. Chem. 592, 51-
authors are also indebted to Dr Marian Gu~ev from the 58.
D e p a r t m e n t of C o m p u t e r Sciences, Faculty of Science, Matsunaga, H. (1982). J. Phys. Soc. Jpn, 51, 864-872.
Skopje, for kindly putting the D e p a r t m e n t ' s equipment McGinnety, J. A. (1972). Acta Cryst. B28, 2845-2852.
Morris, A. J., Kennard, K. L., Moore, F. H., Smith, G. &
at our disposal.
Montgomery, H. (1981). Cryst. Struct. Cornmun. 10, 529-
532.
Mulliken, R. S. (1955a). J. Chem. Phys. 23, 1833-1840.
Mulliken, R. S. (1955b). J. Chem. Phys. 23, 2338-2342.
References
Nord, A. G. (1974). Acta Cryst. B30, 1640-1641.
Aleksovska, S., Petru~evski, V. M. & Soptrajanov, B. (1998). Nord, A. G. (1976). Acta Chem. Scand. A, 30, 198-202.
Acta Cryst. Accepted for publication. Onodera, A. & Shiozaki, Y. (1981). Ferroelectrics, 31, 27-36.
Breneman, C. M. & Wiberg, K. B. (1990). J. Comput. Chem. Palenik, G. J. (1967). lnorg. Chem. 6, 507-511.
11, 361-373. Pannetier, G. & Gaultier, M. (1966). Bull. Soc. Chim. Fr. 1966,
Carter, R. L. & Margulis, T. N. (1972). J. Solid State Chem. 5, 3336-3341.
75-78. Petru~evski, V. M. (1994). Acta Cryst. B50, 625-627.
Chirlian, L. E. & Francl, M. M. (1987). J. Comput. Chem. 8, Petru~evski, V. & Aleksovska, S. (1991). Croat. Chem. Acta,
894-905. 64, 577-583.
Dimitrova, Y. (1995). J. Mol. Struct. (Theochem.) 334, 215- Petru~evski, V. & Aleksovska, S. (1994). Croat. Chem. Acta,
222. 67, 221-230.
Ffibry, J. & Breczewski, T. (1993). Acta Cryst. C49, 1724-1727. Petru~evski, V. & Sherman, W. E (1990). J. Mol. Struct. 219,
Frisch, M. J., Trucks, G. W., Schlegel, H. B., Gill, P. M. W., 171-176.
Johnson, B. G., Robb, M. A., Cheeseman, J. R., Keith, T., Petru~evski, V. M. & Sherman, W. F. (1993). J. Mol. Struct.
Petersson, G. A., Montgomery, J. A., Raghavachari, K., A1- 294, 171-174.
Laham, M. A., Zakrzewski, V. G., Ortiz, J. V., Foresman, Petru~evski, V. M. & Sherman, W. E (1994). Bull. Chem.
J. B., Peng, C. Y., Ayala, P. Y., Chen, W., Wong, M. W., Technol. Macedonia, 13, 69-76.
Andres, J. L., Replogle, E. S., Gomperts, R., Martin, R. L., Plesko, S., Kind, R. & Arend, H. (1980). Phys. Status Solidi,
Fox, D. J., Binkley, J. S., Defrees, D. J., Baker, J., Stewart, 61, 87-94.
J. P., Head-Gordon, M., Gonzalez, C. & Pople, J. A. (1995). Sasvfiri, K. (1963). Acta Cryst. 16, 719-724.
GA USSIAN94. Revision B.2. Gaussian, Inc., Pittsburgh, PA, Schlemper, F.. O. & Hamilton, W. C. (1966). J. Chem. Phys. 44,
USA. 4498-4509.
Gabe, E. J., Le Page, Y., Charland, J.-P., Lee, F. L. & White, Secco, A. S. & Trotter, J. (1983). Acta Cryst. C39, 317-318.
P. S. (1989). J. Appl. Cryst. 22, 384-387. Shannon, R. D. (1976). Acta Cryst. A32, 751-767.
Gaebell, H.-C. & Meyer, G. (1984). Z. Anorg. Allg. Chem. 515, Shannon, R. D. & Prewitt, C. T. (1969). Acta Cryst. B25, 925-
133-140. 945.
Gale~ic, N. & Jordanovska, V. (1992). Acta Cryst. C48, 256- Smith, H. W. Jr & Colby, M. Y. (1941). Z. Kristallogr. 103, 90-
258. 95.
Garg, A. &. Srivastava, R. C. (1979). Acta Cryst. B35, 1429- Srinivasan, V., Subramanian, C. K. & Narayanan, P. S. (1983).
1432. Pramana, 21, 41-50.
Herbstein, F. H., Ron, G. & Weissman, A. (1971). J. Chem. Statistical Graphics Corporation (1989). STATGRAPHICS.
Soc. A, pp. 1821-1826. Statistical Graphics Package. Educational Institution ed.
Hogervorst, A. C. R. & Helmholdt, R. B. (1988). Acta Cryst. Version 4.0. Statistical Graphics Corporation.
B44, 120-128. Takahashi, I., Onodera, A. & Shiozaki, Y. (1987). Acta Cryst.
Hoppe, M. L., Schlemper, E. O. & Murmann, R. K. (1982). C43, 179-182.
Acta Cryst. B38, 2237-2239. Toriumi, K. & Saito, Y. (1978). Acta Cryst. B34, 3149-3156.
Itoh, K., Hinasada, A., Matsunaga, H. & Nakamura, E. (1983). Weber, H. J., Schulz, M., Schmitz, S., Granzin, J. & Siegert, H.
J. Phys. Soc. Jpn, 52, 664-670. (1989). J. Phys. 1, 8543-8557.
Kfilmfin, A., Stephens, J. S. & Cruickshank, D. W. J. (1970). Ztifiiga, F. J., Breczewski, T. & Arnaiz, A. (1991). Acta Crvst.
Acta Cryst. B26, 1451-1454. C47, 638-640.