TRINITROTRIAMMINECOB
ALT( 111) Inorganic Chemistry, Vol. 10,No. 5, 1971 1057
CONTRIBUTION FROM THE DEPARTMENT O F CHEMISTRY,
UNIVERSITY
OF NATAL,DURBAN, REPUBLICOF SOUTH AFRICA
The Crystal Structure of Trinitrotriamminecobalt(II1). A Redetermination
BY MICHAEL LAING,* STAFFORD BAINES, AND POLLY SOMMERVILLE
Received September 14, 1970
The crystal structure of trinitrotriamminecobalt(III), Co(NOz)a(NH8)3,has been redetermined and shows that a previously
published structure was incorrect. The final value of the conventional R was 0.090 for 670 data collected by normal film
techniques. The compound is orthorhombic, of space group P212121; a = 10.14, b = 11.68, c = 6.82 A; dobsd = 2.00, doalod
= 2.01 g/cm3; 2 = 4. The molecule is the meridial isomer (mer) with one pair of trans nitro groups and one pair of trans
ammine groups. The average bond lengths are as follows: Co-NH3, 1.96 A; Co-N02, 1.92 A (esd 0.01 A); N-0, 1.24 +&
(esd 0.02 A). There are a large number of intermolecular hydrogen bonds and nonbonded contacts, with the 0 . * * N separa-
tions ranging between 2.95 and 3.22 A.
The present redetermination is a result of a project Using zirconium-filtered Mo Kor radiation, unintegrated data
to study the preparations, structures, and infrared were collected by the precession method for layers Ok13kl and
h01-h3Z, and integrated data were collected by the equiinclination
spectra of the various compounds in the cobalt-nitro- Weissenberg method for layers hkO-hk7. To obtain a reasonable
ammine series CO(NO&~-to Co(NH3)e3f. Although range of intensities, six exposures were made for each layer of
two forms of the neutral compound Co(NH3)3(NO2)3 precession data: 10, 30, and 90 min; 5 , 12, and 36 hr. For the
had been reported' with different infrared absorption Weissenberg data single 24-hr exposures were made using packs
of four sheets of film, interleaved with three pieces of 0.0025-mm
spectra and X-ray powder patterns, only one could be brass shimstock. The intensities were estimated visually by
prepared. X-Ray photographs showed the material comparison with suitable time-calibrated intensity strips. Lor-
to be identical with that studied by Tanito, et aL2 entz and polarization corrections were calculated and applied to
Their structure was determined from two projections the data. The three sets of data were scaled together by com-
using a total of 142 data. The coordinates of the cobalt paring common reflections. The agreement between the values
obtained and those previously published2 for the OkZ and hkO data
atom were found from the Patterson projections and left no doubt that the compound was the same in both cases.
those of the nitrogen atoms from the subsequent Fourier
maps. However, the coordinates of the oxygen atoms Determination of the Structure
had to be calculated by assuming reasonable values for A three-dimensional Patterson synthesis (calculated
N-0 bond lengths and 0-N-0 angles. Then (presum- with program MIFR 1, on an IBM 704 a t the CSIR
ably) the acceptable appearance of the final Fourier facility, Pretoria, South Africa) confirmed that the ap-
projection, the correspondence of F , with Fo, and rea- proximate coordinates of the cobalt atom were 0.317,
sonable intermolecular atom-atom distances were taken 0.25, 0.25. Structure factor calculations and Fourier
to confirni the correctness of the oxygen atom positions. projections yielded nitrogen atom coordinates similar to
It was decided to collect full three-dimensional X-ray those published previously, but i t soon became evident
data and to refine the published structure. Most un- that the reported positions of the oxygen atoms were
expectedly, i t was found that the new observed data did not correct. After many trials the three projections
not fit the published results and hence i t was necessary yielded a self-consistent set of coordinates for all
to redetermine the structure. atoms. Structure factor calculations using an overall
B = 3.0 k2gave a discrepancy index R (R= 2[(iFol-
Experimental Section
Co(N02)3(NHs)a crystallizes in the orthorhombic form with
lFo\)/\Fol])of 0.32 compared with R = 0.42 from the
a = 10.14 =t0.02, b = 11.68 =I= 0.02, c = 6.82 rt 0.02A; space coordinates of Tanito, et aL2
group P212121; dobed = 2.00 Zk 0.02, doslod = 2.01 g/cm3; 2 = 4; All calculations had been done on an IBM 1620 com-
X(MO Kor) 0.7107, X(Cu Kor) 1.5418 A; p(Mo Kor) = 22, ~ ( C U puter, using the set of programs of R. Shiono and D.
K a ) = 163cm-l. Hall, IBM library numbers 8.4.004 to 8.4.007; it was
The material was prepared by Miss B. Wise by a variation of
the method of Duva14using charcoal as a catalyst.6 Recrystal-
now impossible to proceed by conventional three-dimen-
lization from water, slightly acidified with acetic acid, yielded sional Fourier methods, because of the severe limita-
well-formed parallelepipeds elongated parallel to c (in contrast tions of the computational facilities available, and so the
to the plates obtained by Tanito2). The density, measured by work was terminated a t this stage.3
flotation in aqueous zinc bromide solution, was found to be 2.00 f An attempt was made t o refine the new postulated
0.02 g/cm*.
Oscillation, Weissenberg, and precession photographs showed structure by least squares, using program ORFLS on an
the crystal to be orthorhombic and extinction conditions h00 IBM 360/40 computer a t the CSIR, Pretoria, South
+ +
when12 = 2% 1,OkOwhenk = 2% 1, and00ZwhenZ = 2% 1 + Africa. R dropped only to 0.25 while the isotropic
showed the space group to be P212121. Accurate unit cell di- B's of four of the oxygen atoms increased markedly.
mensions were obtained from a- and c-axis zero-layer Weissen-
berg photographs which had been calibrated with CeOz powder
It was obvious t h a t the structure still was not correct.
patterns (a, = 5.411 A). The crystal chosen for the initial data At this stage an 8k-core, single-disk IBM 1130 com-
collection measured 0.2 X 0.2 X 0.5 mm. puter became available in Durban, South Africa, en-
* TOwhom correspondence should be addressed. abling three-dimensional Fourier summations t o be
(1) R. Duval, C. R.Acad. Sci., 206, 1652 (1938). done. All subsequent calculations were done on this
(2) Y.Tanito, Y.Saito, and H. Kuroya, Bull. Chem. SOC.Jap., 26, 188 machine using the local set of programs6
(1952).
(3) C . G . S. Baines, M.Sc. Thesis, University of Natal, Durban, South A series of three-dimensional observed and difference
Africa, 1967. Fourier maps showed that atoms 0(3), 0(4), 0(5), and
(4) C. Duval, C. R. Acad. Sci., 182,362, (1926).
(5) E. G . Rochow, Inorg. S y n . , 6, 189 (1960). (6) M. Laing and A. Sale, Acla Crystallogr., Sect. B , 25, 1674 (1969).
1058 Inorganic Chemistry, Vol. 10, No. 5, 1971 M. LAING,S. BAINES,AND P. SOMMERVILLE
TABLE I
Final Positional and Thermal Parametersa
X Y 2 X Y 2
0.3171 (2) 0.2432 (2) 0.2597 (7) om 0.173 (1) 0.409 (1) 0.064 (3)
0.488(1) 0.314 (1) 0.288 (3) O(6) 0.363 (2) 0.368 (1) -0.084 (3)
0,361 (2) 0.136 (2) 0.459 (4) H(14) 0.14 0.09 0.20
0.278 (2) 0.350 (2) 0.053 (4) H(24) 0.09 0.21 0.14
0.143 (1) 0.171 (1) 0.223 (3) H(34) 0.10 0.16 0.35
0.243 (2) 0,349 (2) 0.448 (4) 0.21 0.31 0.57
0.384 ( 2 ) 0.135 (2) 0.061 (4) H'15)
H(25) 0.16 0.39 0.39
0.586 (1) 0.271 (1) 0.214 (3) H(35) 0.30 0.41 0.48
0.499 (1) 0.402 (1) 0.385 (3) 0.32 0.12 -0.04
0.319 (2) 0.034(1) 0.453 (3) H(26) 0.46 0.16 0.01
0.429 (2) 0.159(1) 0.608 (3) H(36) 0.41 0.06 0.12
Anisotropic Thermal Parameters*
811 822 Paa 812 PIS 0x3
0.0025 (2) 0.0019 (1) 0.0125 (8) 0.0003 (4) -0.0011 (9) 0.000 (1)
0.002 (1) 0.004 (1) 0.016 (7) 0.000 (2) 0.004 (5) 0.009 (5)
0.005 (2) 0.002 (1) 0.022 (8) -0.001 (2) -0.003 (6) 0.011 (5)
0.005 (2) 0.001 (1) 0.027 (8) 0.000 (2) -0.011 (6) 0.000 (4)
0.003 (1) 0.004 (1) 0.006 (5) 0.003 (2) 0.001 (6) 0.000 ( 5 )
0.006 (2) 0 006 (2)I 0.007 (6) 0.000 (3) -0.005 ( 5 ) -0.002 (5)
0.004 (2) 0.006 (2) 0.020 (7) 0.001 (3) 0.000 (6) -0.003 (6)
0.004 (1) 0.005 (1) 0.030 (7) 0,002 (2) -0.006 ( 5 ) 0.000 (5)
0.006 (1) 0.006 (1) 0.031 (6) -0,009 (2) 0.008 (6) -0.002 ( 5 )
0.011 (2) 0.002 (1) 0.029 (6) 0 , 0 0 1 (2) -0.012 (6) 0.004 (5)
0.010 (2) 0.006 (1) 0.027 (7) -0.003 (3) -0.011 (6) 0.010 ( 5 )
0.006 (1) 0.004 (1) 0.033 (6) 0.004 (2) 0,000 (6) 0.007 (4)
0.010 (2) 0.007 (1) 0.015 (6) 0.005 (3) 0.014 (6) 0.005 (5)
a The estimated standard deviations are given in parentheses. The approximate errors for the hydrogen atoms are &0.01 in x and y
and 1 0 . 0 2 in z. All hydrogen atoms were assigned an isotropic B = 4.0 A2. The general form of the anisotropic temperature factor is
+ +
exp[ - (Pllh2 Bt2kz P33Z2 PI& + +
Pi3hZ Pz3k2)1. +
O(6) were incorrectly placed, and adjustment of their dispersion correction. Neither extinction nor absorp-
positional coordinates reduced R to 0.21. However the tion corrections were applied so the values of the aniso-
final Fourier map still showed small ghost peaks related tropic thermal motion parameters, especially for the
to the four oxygens by pseudomirror planes a t y = 0.25 cobalt atom, are not of great physical significance,
and z = 0.25 (presumably due to the cobalt atom having When R had dropped to 0.094, a difference Fourier was
these coordinates). In addition, there were only rela- calculated to check whether the ammonia hydrogens
tively few data with large h and k indices, so it was de- were observable. Areas of positive electron density
cided to re-collect full three-dimensional data for the were found in reasonable positions, and the nine hydro-
compound using Ni-filtered Cu KCYradiation, in spite gen atoms were therefore included in the calculations,
of the absorption and fluorescence problems. but their parameters were not refined. After two fur-
A crystal, of dimensions 0 . 1 X 0 . 1 X 0.3 mm, was ther cycles of least-squares refinement, the structure
mounted parallel to c and multiple-film nonintegrated converged to a conventional R = 0.090.
intensity data were collected by the equiinclination The final positional and thermal parameters are given
Weissenberg method for layers hk0-hk5. The data in Table I . The values for the observed and calcu-
were estimated visually and, after Lorentz and polariza- lated structure factors are given in Table 11. Intra-
tion corrections had been applied, were used in all sub- molecular bond lengths are given in Table I11 and bond
sequent calculations. A three-dimensional observed angles are given in Table IV. The numbering system
Fourier map, phased on the cobalt and six nitrogen and more important bond lengths and angles are illus-
atoms, still showed the mirror planes, but once oxygens trated in Figures 1 and 2. Intermolecular nonbonded
O(1) and O(2) were included in the structure factor contacts are listed in Table V and the nitro oxygen to
calculations, it was possible to destroy the pseudosym-
metry and correctly locate the four remaining oxygen
atoms. The atom parameters were adjusted by the
method of Booth,7 yielding R = 0.185.
The structure was then refined anisotropically by
least-squares using the CRYM system a t the California
Institute of Technology, A Hughes weighting scheme8
was used, with 4Fm,, = 20.0. Neutral atom scattering
factors for 0, N, and H were taken from ref 9; the scat-
tering factor for cobalt was taken from the tabulation of
Hanson, et al.,10and adjusted for the real part of the
(7) A. D. Booth, "Fourier Techniques in X-Ray Organic Structure
Analysis," Cambridge University Press, London, 1948.
(8) E. W. Hughes, J . Amev. Chem. SOL.,63, 1737 (1941)
(9) "International Tables for X-Ray Crystallography," Vol. 3, Kynoch
Press, Biimingham, England, 1962,pp 202,204, 214.
(10) H P. Hanson, F. Herman, J. D. Lea, a n d S. Skillman, Acta C ~ y s t a l -
logy., 17, 1040 (1964). Figure 1.-Intramolecular nonbonded distances.
111)
TRINITROTRIAMMINECOBALT( Inorganic Chemistry, Vol. 10, No. 5, 1971 1059
TABLE I1
FINAL
VALUESOF THE OBSERVED STRUCTURE
AND CALCULATED FACTORS
( X IO)"
a
192
LI1 -69
'I
LIL .x, I *
The numbers in the four columns are k index, F,, F,, and phase angle.
TABLEI11
INTRAMOLECULAR
BONDLENGTHS AND
NONBONDEDDISTANCES (A)"
Co-N(l) 1.93 N(4)...N(6) 2.71
Co-N(2) 1.90 N(1)**.0(4) 2.90
Co-N(3) 1.92 N(1)..*0(6) 2.95
Co-N(4) 1.97 N(2)..*0(1) 3.24
CO-N(5) 1.94 N(3).**0(2) 3.24
Co-N(6) , 1.97 K(4)...0(3) 2.86
N(1)-0(1) 1.22 K(4). . O(5) 3.00
NU ) - 0 ( 2 ) 1.22 N(5)**+0(2) 2.71
N(2)-0(3) 1.26 N(5).* * 0 ( 4 ) 3.11
~ (1-0 2 (4) 1.25 K(5).**0(5) 2.80
1.26 iV(6)..*0(1) 2.79
N(l).-*N(3) 2.70 0(1).'.0(2) 2.11
N(1). * sN(5) 2.75 0 ( 3 ) . * * 0 ( 4 ) 2.12
N(l)...N(6) 2.81 0(5)...0(6) 2.14 Figure 2. Bond lengths and bond angles.
.
N(2) * . N(4) 2.76 0 ( 2 ) * . * 0 ( 4 ) 3.29
N(2)...N(5) 2.77 Co*..O(l) 2.76
N ( 3 ) . * * N ( 5 ) 2.72 C0.+*0(4) 2.80
N(3)...N(6) 2.73 c0*..0(5) 2.77
N(4) * * * N(5) 2.78 C0.**0(6) 2.79
Bond lengths are not corrected for thermal motion; estimated
standard deviations are 0.01 A for all Co-N bonds and 0.02 A for
all N-0 bonds.
TABLE IV
BONDANGLES( D E G ) ~
N( 1)- Co-N(4) 179 N(3)-Co-N(6) 89
N(Z)-Co-N(3) 178 N(4)- Co-N(t5) 91
N(5)-Co-N(6) 177 N(4)- Co-N(6) 87
N(l)-Co-N(2) 90 O(l)-N(1)-0(2) 119
N(l)-CO-N(3) 89 0(3)-N(2)-0(4) 114
N(l)-Co-N(5) 90 0(5)-N(3)-0(6) 118
N(l)-Co-N(6) 92 O( 1)-N( l)-Co 121
N(2)-Co- N(4) 91 0(2)-N(l)-Co 120
N(2)-Co-N(5) 92 0(3)-N(2)-C0 121
N(2)-Co-N(6) 90 0(4)-N(2)-Co 125
N(3)-Co-N(4) 90 0(5)-N(3)-C0 119
N(3)-Co-N(5) 89 0(6)-N(3)-C0 123
a The estimated standard deviations are 1 ' for all angles
involving the cobalt atom and 2' for the O N 0 angles.
1060 Inorganic Chemistry, Vol. 10,No. 5, 1971 M. LAING,S.BAINES,AND P. SOMMERVILLE
TABLEVI
PLANEPARA METERS^
LEAST-SQUARES
Direction cosines Origin to plane
of normal to plane distance,
Plane atoms co N(1) N(4) O(1) O(2) 0.155, -542,0.826 0.41
Dev, A 0.01 -0.01 -0.01 0.00 0.00
Plane atoms co N(2) N(3) O(5) O(6) 0.470,O. 718,O.514 4.47
Dev, A -0.01 0.01 -0.02 0.01 0.01
Plane atoms co N(2) N(3) O(3) O(4) 0.826, -0.269,0.495 1.04
Dev, A -0.02 0.01 0.01 0.00 0.00
Plane ?toms co NO) N(2) N(3) N(4) -0.385,0.624,0.680 1.72
Dev, A 0.02 0.00 -0.01 -0.01 0.00
Plane atoms co N(2) N(3) N(5) N(6) 0.906,0.411,0.098 4.25
Dev, A 0.01 0.03 0.03 -0.03 -0.03
Plane ?toms co NU) N(4) N(5) N(6) 0.216, -0.645,0.733 0.14
Dev, A 0.02 0.00 0.00 -0.01 -0.01
a The equation of the plane is of the form A x +
By + +
Cz D = 0 where A , B , and C are the direction cosines referred to the real
crystallographic axes and D is the perpendicular distance to the origin.
@ TABLE VI1
BONDLENGTHS,
ANGLES, AND NONBONDED SEPARATIONS
INVOLVING HYDROGEN ATOMS'
Atoms Length, A Atoms Angle, deg
K(4)-H( 14) 1.0 Co-N(4)-H( 14) 118
N(4)- H(24) 0.9 Co-N(4)-H(24) 113
N(4)-H(34) 1.0 CO-N (4)-H (34) 110
N(5)-H(15) 1.0 Co-N(5)-H( 15) 112
N(5)-H(25) 1.0 Co-N(5)-H(25) 110
N(5)-H(35) 1.0 Co-N(5)-H(35) 111
N(6)-H( 16) 1.0 Co-N(6)-H(16) 113
N(6)-H(26) 0.9 Co-X(6)-H(26) 108
N(6)-H(36) 1.0 Co-N(6)-H(36) 109
[n positionb &. Distance,
5 2.4
6 2.2
3 2.1
6 2.2
1 2.2
4 2.4
5 2.6
2 2.3
I 2 2.3
I 6 2.5
a The approximate error in the bond lengths is 0.1 A and in the
&,I
bond angles is 5". See Table V.
Figure 3.-Intermolecular nonbonded distances.
I. A full list of these data is available from the au-
ammonia nitrogen distances are illustrated in Figure 3. thors on request.
Parameters for various least-squares planes are given
in Table VI. Tnteratomic distances and bond angles Discussion of the Structure
involving hydrogen atoms are given in Table VII. The bond lengths are similar to those found in re-
The thermal vibration ellipsoids are illustrated in lated compounds; the average Co-NH3 length of 1.96 A
Figure 4. (Subsequent refinement of the structure compares with the average values of 1.96 and 1.95 A
found in Coz(NH3)loOZ4+l1 and C O Z ( N H ~ ) I O O Zre- ~T,'~
spectively. The average Co-NO2 length of 1.92 A can
be compared with the rather imprecise values of 1.96
A
and 1.97 found in C O ( N O ~ ) ~ ( N H ~ ) Z - .Although
'~,'~
the difference between the average Co-NH3 and Co-
NOn bond lengths in C O ( N O ~ ) ~ ( N HisSbarely
)~ signifi-
cant, the pattern of shorter Co-NO2 and longer Co-NH3
OUI seems chemically reasonable. The average N-0 bond
length of 1.24 A can be compared with the range of
values from 1.21 to 1.26 A found in Co(N02)4-
(NH3)~-.13J4The difference between the average N-0
bond length in the trans pair of NO2 groups and that in
the unique NOz is not significant.
Figure 4.-Thermal vibration ellipsoids; a projection down y,
(11) W P Schaefer, I n o r g . Chem., 7, 725 (1468).
using the 740 original molybdenum-radiation data (12) W. P. Schaefer and R. E Marsh, Acta Crystallogr, 21, 735 (1966).
(13) Y Komiyama, BulZ. Chem. Soc. Jap., SO, 13 (1957).
yielded a final R = 0.107, with positional parameters (14) G B. Bokii and E. A. Gilinskaya, Dokl. Akad Nauk S S S R , 88, 461
within one standard deviation of those given in Table (1953).
STRUCTURE
OF [Co(N2C2H&]2 [Cu2CleIC&..2H20 Inorganic Chemistry, Vol. 10,No. 5, 1971 1061
The Bix nitrogens are bonded to the cobalt in the ex- separations of 2.74 and 2.78 A and especially an 0 .O e
pected octahedral arrangement; the angular distortions separation of 2.76 k mentioned in ref 2 hint that the
from ideality do not seem to follow any particular pat- structure may be incorrect. (A three-dimensional
tern. The molecule is the meridial (mer) isomer,15 as Fourier map, using the new observed data, and phased
previously reported by Tanito, et aL2 All intramolec- on the Tanito model, shows a number of spurious peaks
ular nonbonded separations are normal, ranging up- and deep negative "holes" as well as ghost peaks in what
ward from 2.70 A (Table 111, Figure 1). Nitro group eventually proved to be the correct positions for the
O(1)-N(1)-0(2) is twisted 80" from the plane of N(1)- oxygen atoms. R for this calculation was 0.35.)
Co-N(3)-N(2)-N(4), while nitro group 0(3)-N(2)- I n the Baines modelJ3an intermolecular 0 . - 0 sep- -
O(4) is twisted 34" and nitro group 0(5)-N(3)-0(6) aration of 2.44 A shows conclusively that the model was
is twisted 52" from this plane. The planes of nitro incorrect and this conclusion is confirmed by two im-
groups 0(3)-N(2)-0(4) and 0(5)-N(3)-0(6) thus make possibly short intramolecular nonbonded N . * 0 and -
an angle of 86" with each other. 0 . -0separations of 2.53 and 2.09 8 , respectively, and
The molecules are involved in a complicated three- the generally rather poor bond lengths and angles.
dimensional network of hydrogen bonds, with inter- Recent work17 confirms the existence of the two iso-
molecular ammine N to nitro 0 distances ranging be- mers of Co(N02)a(NH3)3and in fact the assignment of
tween 2.95 and 3.16 k (Figure 3, Table V). There are the mer stereochemistry to the product prepared by
also several 0 . * eand nitro N . -0contacts of about
0 the method of Jorgensen18was made on the basis of the
3 k, which share in making it a very closely packed and structure determination by Tanito, et al. Fortunately,
compact structure. the redetermination shows that this was indeed the mer
Comments on the Two Incorrect Methods isomer.
It is of interest (and rather sobering!) t o find that two
incorrect structures could have seemingly correct bond Acknowledgments. We are grateful to the South
lengths and angles, reasonable looking Fourier maps, African Council for Scientific and Industrial Research
and relatively low values for R. I n fact, consideration for a grant to cover running expenses and for the use of
of nonbonded distances between atoms shows immedi- their computational facilities and to the Chemistry De-
ately that the structures must be incorrect in spite of partment of California Institute of Technology for a
apparently satisfactory bond lengths and angles2 or a grant to cover the cost of the final refinement. M. L.
reasonably low R.3 wishes t o thank the faculty of the Chemistry Depart-
The Tanito model gives rise to impossibly short ment of San Fernando Valley State College and espe-
intramolecular ammonia Ne separations of 2.20
e 0
cially Dr. R. E. Marsh of the California Institute of
and 2.09 k, while there are four other intramolecular Technology for their generosity and hospitality during
contacts about 0.3 k shorter than is usual.16 The rela- his stay in Los Angeles, Calif., in 1970. The diagrams
tively short intermolecular ammonia N * nitro 0 - - were prepared by Miss Lillian Casler.
(16) R. F.Trimble, J . Chem. Educ., 81, 176 (1954).
(16) A. Kitaigorodskii, "Organic Chemical Crystallography," Consultants (17) R. B. Hagel and L. F. Druding, Inovg. t h e m . , 9, 1496 (1970).
Bureau, New York, N. Y., 1961. (18) S . M.Jorgensen, Z.Anovg. Allg. Chem., 17, 475 (1898).
CONTRIBUTION FROM THE DEPARTMENT
O F CHEMISTRY,
UNIVERSITY
OF NORTHCAROLINA, CHAPELHILL,NORTHCAROLINA27514
The Crystal and Molecular Structure of Tris(ethylenediamine)cobalt(III)
Di-p-chloro-bis(trichlorocuprate(I1)) Dichloride Dihydrate,
[ Co(N2C2&)3]z[ CWA&IC12 2H2O
BY DEREK J. HODGSON,* POLLEY K. HALE, AND WILLIAM E. HATFIELD
Received August 19, 1970
The crystal and molecular structure of tris(ethylenediamine)cobalt(III) di-p-chloro-bis(trichlorocuprate(I1)) dichloride
dihydrate, [C~(en)~]z[Cu~Cls] Clz '2Hz0, has been determined from three-dimensional X-ray data collected by counter
methods. The material crystallizes in space group Dzhl6-Pbcu of the orthorhombic system with four molecules in a cell of
dimensions u = 13.560 (9), b = 14.569 (9), and c = 17.885 (12) A. The observed and calculated densities are 1.86 and 1.862
g ~ m - respectively.
~ , Least-squares refinement of the structure has yielded a final value of the conventional R factor (on F)
of 0.065 for 1275 independent reflections having Fz > 3 n ( F ) . The CuzCls4-anion consists of two distorted trigonal bipyra-
mids sharing a common edge, the bridging Cu-C1 distances being 2.325 (5) and 2.703 ( 5 ) A while the Cu-Cu separation is
3.722 (5) A. The Co(en)aa+cation is found to have the relatively high energy configuration A(A88) (szA(8AA)) as a result of
intermolecular interactions within the crystal.
Introduction sumed that the anionic moiety consisted of either the
A complex of formulation C 0 ( e n ) ~ C u C l ~ ( H(where
~0) ~ - as found2 in [Cr-
trigonal-bipyramidal C U C ~ ~ion,
en = ethylenediamine, NH2CH2CH2NH2) was first re-
(2) (a) K.N. Raymond, D. W. Meek, and J. A. Ibers, Inorg. Chem., 7 ,
ported by Kurnakowl in 1898, and it has long been as- 1111 (1968); (b) M.Mori, Y.Saito, and T. Watanabe, Bzrll. Chem. Soc. Jag.,
(1) N. S.Kurnakow, Z.Anovg. Chem., 17,225 (1898). 84,296 (1961).
