VoL 7, No. 6 , June 1968
STRUCTURE O F
The intramolecular 0---0 distance (2.38 8) is the shortest yet reported but within experimental error cannot be considered significantly shorter than that observed for nickel dimethylglyoxime (2.40 A) .l It does however appear to be significantly shorter than that observed for bis(glyoximato)nickel(II) (2.45 8). 2 This shortening might be expected since the constraint imposed by the presence of two intramolecular hydrogen bonds is absent. There is no significant deviation from 120’ bond angles around N(1) and N(2). However, owing presumably to distortion resulting from coordination, the bond angles around C(l) and C(2) are significantly different from 120’. The bond angles around N(3), N(4), C(3), and C(4) are within experimental error identical and normal tetrahedral angles. Since the distances involving hydrogen atoms are not accurately known, they are not included in Table VII. The four N-H distances range from 0.70 to 0.92 A (average 0.84 8); the 18 C-H distances, from 0.62 to 1.26 A (average 0.92 8); and the two 0-H distances for the water molecule are 0.62 and 0.64 8 . The final hydrogen is 1.24 A from 0(1) and 1.22 8 from O(2). This hydrogen is not on the line connecting the two oxygen atoms, but rather the 0-H-0 angle is 157’. From the shape of the difference Fourier peak (Figure
CONTRIBUTION FROM
THE
AND THE
TIN(IV) CHLORIDE GLUTARONITRILE 1135
4) i t is not possible to decide whether this hydrogen atom has only one equilibrium position, but i t does appear to be approximately a t the symmetrical position. Definite location of the hydrogen atoms will be carried out by neutron diffraction.
p“ Figure 4.-A portion of the difference Fourier synthesis in the plane of the oxime nitrogen and oxygen atoms showing the hydrogen atom in the intramolecular hydrogen bond. The cyntours are drawn a t 0.1 e - / b 3 intervals from 0.2 to 0.5 e-/A3.
Acknowledgment.-The author gratefully acknowledges R. Kent Murmann for making available the crystals used in this study. Acknowledgment is made t o the donors of the Petroleum Research Fund, administered by the American Chemical Society, for partial support of this research.
DEPARTMENT OF PHYSICS, EASTERNMONTANA COLLEGE, BILLINGS,MONTANA, DEPARTMENT OR CHEMISTRY, MONTANA STATE UNIVERSITY, BOZEMAN, MONTANA
The Crystal and Molecular Structure of Tin(1V) Chloride Glutaronitrile (SnCl,NC(CH,),CN) BY DAVID M. BARNHART,” CHARLES N. CAUGHLAN,lb AND MAZHAR UL-HAQUElb
Received August 15, 1967 The structure of SnClaNC(CH&CN, tin(1V) chloride gluta!onitrile, has been determined by an X-ray diffraction study. The crystals are orthorhomic, a = 6.11, b = 12.90, c = 14.57A, space group Pbcm; D(obsd) was 2.03 and D(ca1cd) was 2.05 on the basis of four asymmetric units per unit cell. The intensities of 656 independent reflections were recorded with a diffractometer. The structure was solved from the Patterson map and has been refined with a full-matrix least-squares refinement to a conventional R of 5.9%. The structure is in the form of a one-dimensional polymer with the glutaronitrile acting as a bridging ligand between the SnC14 units. The glutaronitrile is in the TT conformation. The tin is octahedrally coordinated with the nitrogen atoms in the cis configuration. The Sn-Cl distance is 2.35 =k 0.01 b; the Sn-N distance is 2.29 f 0.02 A. The glutaronitrile distances appear normal.
Introduction Complexes of dinitriles with tin(IV) have been described and characterized by Kubota and Schulze.2a The complexes were believed to be polymers, with the dinitrile acting as a bridging ligand. This conclusion was arrived a t as a result of the general properties, which were characteristic of polymeric substances, and by a consideratiorl of the infrared spectra, ~h~ infrared spectra of coiliplexes of glutaronitrile with a number ( 1 ) (a) Eastern Montana College; (b) Montana S t a t e University.
( 2 ) (a) M. Kubota and S. R. Schulze, I m r g . Chenz., 3, 853 (1964); M. Kubota a n d L).L. Johnston, J . A m . Chem. Soc., 88. 2461 (1966).
(b)
of metal ions were studied by Kubota and Johnston.2b According to these authors, the spectra indicated the rotational conformer TT3 for the &taronitrile in the tin(IV) complex, but spectral assignments for the TT conformer disagreed markedly with those designated by (3) This notation, t v a n s ~ r a i t s applies , t o rotational isomers of saturated hydrocarbons. One carbon a t o m is fixed ill space a n d the directions of t h e four bonds around i t at-? designated by a, h , c , rl. A unique designation of the carbon skeleton can be obtained b y denoting t h e dii-ections of C-C bonds in or-der a s a , b, c , o r d. TT, tvans,li.aizs, is designated as a b a b ; TG, bans,gauche, is a b a c ; GG is abcd, etc.: S. Mizushima, “Structure of Molecules and Internal Rotation,” Academic Press Inc., New York, iY.Y.,1954> P a r t I , Chapter V.
c.
1136 D. M. BARNHART, N.
CAUGHLAN, AND
1%.UL-HAQCE
Matsubara4 in the pure glutaronitrile. This seems unusual since succinonitrile in complexes shows only slight shifts of the spectra from that of the pure compound. This X-ray crystal structure investigation was undertaken in order to determine with certainty the conformation of glutaronitrile in the tin(1V) complex, to aid in the assignment of the infrared frequencies, as well as to study further the coordination of nitrogen to tin(1V).
were calculated before refinement was started. Refinement was carried out using the full-matrix leastsquares program of Busing, Martin, and Levy8 refining positional and anisotropic temperature factors minimizing Zw(lF.1 - IF,l)2. Each observed reflection was given unit weight for refinement. Two cycles of reTABLEI ATOMICPARAMETERS AXD THEIR STASDARD DEVIATIONS
Experimental Section SnC14KC(CH:)sCIT was prepared according to the method described by Kubota and Johnston.2b This consists of a dropwise addition of a dilute solution of gn (hereafter glutaronitrile is designated as gn) in benzene to a dilute solution of SnC14 in benzene. I n very dilute solution crystallization proceeds slowly enough to yield small single crystals. Since the crystals are extremely hygroscopic, the reaction was carried out under a dry nitrogen atmosphere. Several crystals were transferred under dry nitrogen t o capillaries, which were evacuated and sealed for microscopic and X-ray examination. A single crystal of approximate dimensions 0.1 X 0.1 X 0.2 mm was selected for X-ray study. The lattice parameters were a = 6.11 =k 0.01, 5 = 12.90 =k 0.06, c = 14.57 zt 0.05 8. Space groups Pbcm or PbcZl were indicated by systematic extinctions; k # 272 for OkZ zone, I # 2n for h01 zone. Interpretation of the Patterson map and final refinement has indicated Pbcm. Each lattice parameter was calculated from measurements of diffraction angles on 8-10 reflections on the GE diffractometer, the uncertainty indicating standard deviations from these measurements. The density was measured on another crystal by flotation and was 2.03 g/cm3, compared with 2.05 g/cm3 calculated for four molecules per unit cell. The linear absorption coefficient for Cu K a radiation which was used in collection of intensity data is 262.3 cm-1. The data were collected using a General Electric single-crystal orienter and spectrogoniometer equipped with a scintillation counter and pulse height analyzer for detector. The n axis was parallel to the axis of the single-crystal orienter. Independent reflections were collected b y 8-28 scan technique (moving-crystalmoving-counter method) using equal background and scan counting time. The background was counted on each side of the peak and averaged for calculation of intensities. The scan rate in 28 was 1.8O0/min, the take-off angle was 4",and the counting time was 100 sec/scan for each background count. total of 1449 reKickel-filtered copper radiation was used. flections were scanned and 656 were found to be significantly above the statistical fluctuations of the background count. The data were corrected for Lorentz-polarization using a n I B M 1620 corx1puter.j For structure factor calculations, form factors used were from the literature6 including the anomalous terms of the form factors in final cycles of refinement. Absorption corrections were made using a table look up program of Duane Swank and Roger Willet.7 S o extinction corrections were made since examination of structure factors did not indicate serious errors due to extinction. C I#
A~
Structure Determination and Results From a three-dimensional Patterson map, correct tin and chlorine positions were found on the Harker section and a partial structure factor calculation provided phases for the first electron density map. Several structure factor calculations and electron density maps (4) 1. Matsubara, .I. C h e l ~ l .P h > s . , 36, 373 (1Y61); 84, 1719 ( i 9 c i l ) . ( 5 ) All programs used except the least-squares refinement are from t h e hlontana State tlnivetsity CI->stallogi-aphic P,-ogl-ani l i b r a r y written for the 1B51 1620 bI(i(le1 I 1 I), C . N . C:iughlan, C . 7'.I.i, G . W. Svctlch, K. U. Watenpaugh, and R. 1). Wittei-s. (6) J. il. Ihei-s, "International Tables for Crystallugraphy," Vol. 111, The Kynoch Press, Birmingham, England, 1962, Table 3 . 3 1A. ( 7 ) W. L. Bond, "Intesnational Talrles for C Y )stallori.ai~hy,"Vol. 11, 'The Kynoch Press, Birmingham, England, 195'3, T a b l e 5 . 8 . 5 B , pi, 295-2118,
1norganic Chemistry
Atom
Y
X
2
-0.1604(3)a 0.5259(1) 0.2500(-) 0.0102 (17) 0.3626 (6) 0.2500 (-) -0.4217 (12) 0.6599 ( 5 ) 0.2500 (-) 0.0407 (8) 0,5900 (4) 0,3761 (3) -0.3983 (23) 0.4473 (12) 0.3495 (10) -0,5125 (33) 0.3944(15) 0.3921 (11) C(7) -0.3522 (34) 0.3276 (13) 0.4484 (11) C(8) - 0.4987 (49) 0,2500 (-) 0,5000 (-) The number in parentheses is the standard deviation aiid refers to the least significant digit. Sn(1) C1(2) Cl(3) cu4) K(5) C(6)
TABLE I1 THERMAL PARAMETERS A X D MEAX-SQUARE DISPLACEXEN~ (Thaee are given a s B i J ' a rather than bills)
Man square Displacement A 2
B(l1)
B(22)
B(331
B(12)
8(13)
B(23)
Sn(l) Cl(2) Cl(3) Cl(4)
2.5
2.8
1.3
-0.1
0.0
0.0
S(5)
EQ 0.03
urn 0.02
5.9
3.8
2.3
0.0
0.0
0.09
0.05
3.5
3.0
3.6
0.5
0.0
0.05
0.03
6.4
2.2
-0.5
0.6
0.02
4.3
2.8
0.08 0.06 0.07 0.06 0.05
0.05
3.1
0.0 1.0 -1.2
0.05
3.4
0.04 0.03 0.04
0.02 0.02 0.02 0.02
4.3
3.1
4.8
1.8
-1.1
C(7)
3.3
3.4
2.8
0.1
1.0 0.1 -0.1
c(8)
2.9
3.3
2.6
0.0
0.0
C(6)
MX 0.04
0.1
-0.5
-1.3
-1.7
0.01
0.03
TABLE I11 ISTERATOVIC DISTAXCES Atoms
Bond,
Sn(1)-C1(2)
2 . 3 5 (1)a 2 . 3 5 (1) Sn (1)-C1(3) Sn(l)-C1(4) 2 . 3 6 (1) Sn(l)-S(5) a .29 (2) N(5)-C(6) 1.16 (4) C(6)-C (7) 1 . 5 4 (4) C(7)-C(8) 1.54 (4) The iiumber in parentheses is the standard deviation aiid refers to the least significant digit.
TABLEIV BOSDANGLES Atoms
Angle, deg
C1(2)-Sn (1)-Cl(3) 163.6 (2)c~ C1(2)-Sn( 1)-C1(4) 9 4 . 8 (1) C1(2)-Sn( l ) - S i Z ) 8 3 . 3 (1) C1(3)-Sn( 1)-C1(4) 95.5(2) 8 4 , O (2) C1(3)-Sn ( 1) - 5 ( 5 ) 8 9 . 5 (2) C1(4)-Sn i 1)-N(5) Sn(1)-S( 5)-Ci6) 1 6 9 . 8 (11) 103.5 (13) s (5)-C( 6)-C(7 ) 104.8 i l l ) C (6)-C (7)-C (8) C(7)-C( 8)-C (7') 108.9 (8) a The number in parentheses is the standard deviation and refers to the least significant digit.
finement, refining positional parameters and isotropic
teiiiperature factors, and two cycles, refining positional parameters and anisotropic temperature factors, including the anomalous terms of the scattering factors in the (8) W. 11. Busing, K. 1). RIa>tirl, and H. A. Levy, tional Laborator), Oak Ridge, Tenn, 1082.
ORVLS,
Oak Ridge
Sa-
Vol. 7, No. 6 , June 1968
STRUCTURE OF TIN(IV) CHLORIDE GLUTARONITRILE 1137 TABLE V OBSERVED A N D CALCULATED STRUCTURE FACTORS& - 2l5e
218 ,
0
3
b 57
3 61
I
eo
2
b5
5 b 7 8 9
48
-"I LO
0 , -55
99
I"1
1IO LOO
-111 18 -69
19
62
5k
IO7 60
.1
3b
37 2,
31 1
6 91
I03 62 78 53
.. .
51 51
37 4 40 6 -I! B -37 I O 1 1 2 -90 101 I 62 3 -81 6 I, 5
35
+I 27
3 Ill
2
.I
. 65
-68
31
-33
65
48 46
68
-*a
20
I5
5
39
-46 0
I
-b5
I T
I 42 I O
3b
33 I P
b5
7i 67 5b
Ok
-57 52
7
33
0 36 -39 9 < I 3 L 10 10 2 9 11 3b 31 12
97 30 711
50
26
2v 26
-is 1 3 -,3 1 L 1 5 16 29 -I8 u
71
Y
73 '6
'3
9
.I5
I 2
31
-28
0
I1
-'I
a I
-26
3
L
. 0
2
+
h -99 IU 1 2 12 -$2
- 3 3 1' -"I 31 76 -57
51 .2 b4 IP
0
.I b3
30 21 20 211
d9
3
21 -28 -28
5 1
211 t II Ob -36 b1 0 -73 I 45 ro 3 78 70 5 4, -47 b 23 21 8 I 1 11 b 53 4, 1 3 I28 125 l b 34 -31
.I
b5 81
6 h1 7 31
7
I01
-65 ,e -d5 .(I9
8
65
'I2
P
b
62 6 0
6
,I
3
Y
34
L?
1
4
63 5 L b -32 7 3b 33 8 3 1 0 1 9 78 73 1 1 I1 -70 13 61 59IL 69 -09 1 5 55 5b 3 4 01 b -,IL 1u
3
.I -,5
14
-., 711
I I 3 5
-SI
5 55
0 I 2
-135 35 98
In
3
5 7 9
, 39 -26
105
6
37
T h e columns reading from left to right contain the value of I ,
i
61
I7
a
-72
$7
31 51 35 22
-90
br
4
- 328 8 I 8O i 1, -28 12
E L I 3 -(I
33 I1
38
7
30
80
-74
108 $8
+O 51
72
-01
161
3 C
1 1 1 1 3b 34
3
I
2
69
37
1
61
1
0
L
5 7
3
21 28
-LO 3.
-b3 50 I O 34 11 -3b I2 -49 I 3 r5 -27 1 -+I 1 33 5 22 7 3 1 e
3
35 a9
44
b2 Y3
63
re
ae
$9
L'-
2
37 35
34 -3b
4
,9
-42
1
5b l o 3* 3 3 12 33 -33 35 3b I 20 -20 3 PI -2. 5 . L 7 re -73 9 b5 b2 28 -26 I L, +I 5b 57 0 58 -50 2 6 . 3, 31 33 + 5 L 8
.
*I
62
b3 3.
32
+'
Bb 2' 4
5
L 1 8
11. 31 32 92
33
-57 '15
42
31 'I
-32 -,1
I 5 b
52 10
24
2 5 11
38 35
3
3 25
27 b I I L I7
7
22 b 17
L 30 P
5
L
24 '0
3 27
L I
7
39 *l
32 -38
5 k
4 3
1
.o
11 23 22
I
0
1
2
.*
-49
43 a1
10 -I* 70 7
0
I 35
30
I8 7
I7 b
'35
7
5'
0
69 -65
2
72 55
I5 -53
-10 -22
L
21
43
L 'I -59 53 46
30
-38
56
l 16
-25
23
'9
8
I"
1
25
i
-20 b
1'
, -211
I1 19
.o
L 29 -19 -2'
5
7
. .'
5'
$3
3
I8 I9 7
9
bl
2
22
;
52
5 3 5 3 - 3L2
7
4
22 -21
91
d1
0 2
59 -58
19
7
0
b
- 231 ?
L
L
b3 59
n
21
18
O
$1
20 I5
3
'I '27 '3 -31 II
5
LS 6
;
20 35 31
5
b
35 -b8 61 b -45 b 26 10 - 3 7 11 '2
*3
;
I9 23 e1
L
++
37 37 ' I O L
1
37
'3 4 51 -52 12 '3 30 -30 0 4 P L 2 35 50 63 b3 26
L
>'Le
8 6 1 2 56 -5,
1 I5
L
3I 72
1
b9
3
I!
h
44
5 I 9
b l l I
3
57
3
32 -63 38 -25
-bl
57
38 I O
2 8 12 - 5 e I1 61 -37 1
52 59
k
'
31
I3
I9
b
Y 9
'I 11 -be 1 3 -50 83 0
+o
b
57
bo 30
67 I , .
7 67 51
b 11
22 L
20 4,
b 5
39
llb
..
50
+
+
3
60
-bO
-2' L -101 31
-69
3
5
0
-31
76 '0 92 73 79 3b bb
.I
3b
6.
33
I18
a
33
0 2
0
-91 -12, Pb
111 9b
62
L 3
I.
-51 1 -57 28
-29 11
120 13
121
LOO
2
0
3
-5, 3,
'9
2
13 2b
0
5'
+" I
+I
3
51
-56
31
I, I5
-63 -511 57 3'
5k
59 I 1 -35 I 3 1 L I4 3 30 26 I 5
11
77 62 47
43 59 55
12
L -5,
2
F,, F,.
I
CI (3) I igure 2.-Coordination and bond angles around tin in tin chloride glutaronitrile.
ON
.C
Figure 1.-Projection of the structure of tin chloride glutaronitrile projected down the a axis,
last cycle, have reduced the conventional X factor to 5.9%. Inclusion of the anomalous terms made an insignificant improvement. A final difference map showed no significant features. Final positional parameters with standard deviations are listed in Table I. Table I1 gives anisotropic temperature factors and mean-square displacement. Table I11 shows the interatomic distances, and Table IV gives the bond angles. Table V shows calculated and observed structure factors. Figure 1 shows the projection down the a axis of
the structure, and Figure 2 indicates in more detail the coordination around the tin atom.
Discussion of the Structure It can be seen from Figure 1 that the glutaronitrile molecule acts as a bridging ligand between the tin chloride molecules ; this results in a one-dimensional polymer in the direction of the c axis. The glutaronitrile is in the TT conformation and is nearly planar. The best leastsquares plane is described by the equation 0.0000X 0.6457Y 0.76362 = 7.6269.9 The maximum devia-
+
+
(9) All least-squares planes were calculated according t o the method de'. Schomaker, 5. Waser, R. E. Marsh, and G. Bergman, A c l a scribed b y 1 Cryst., 12, 600 (1959). T h e equation aA' bY C Z = D is t h e nornial equation, D beinx the origin-to-plane distance in hngstroms. All atoms were given equal weight. T h e program for t h e IBRS 1620 is part of the M o n t a n a State University Crystallographic Library of Programs.
+
+
1138 DENISFORSTER, KURTMOEDRITZER, AND JOHN R VAN WAZEK tion from this plane is 0.07A. It should be noted that C(8) sits on the twofold axis so that half of the glutaronitrile molecule is related to the other half by the twofold symmetry. The tin is octahedrally coordinated, the nitrogen atoms being in the cis configuration. Examination of Figure 2 indicates that the octahedron is somewhat, although not greatly, distorted. Sn, N(5), N(5'), C1(4), and Cl(4') are nearly planar, the best leastsquares plane being described by the equation -0.3240X 0.9461Y 0.00002 = 7.2514; maximum deviations of any atoms from this plane are only about 0.01 A. Also, Sn, Cl(2), C1(3), C1(4), and N(5) are close to being planar, the equation being 0.3362X 0.4679Y 0.81732 = 5.2602, with Sn being about 0.2 A from this plane. The Sn-C1 distance of 2.35 f 0.01 agrees well with those in SnC14(OSeC14)2,;.e., 2.36 and 2.41 A.1o In SnCl4(OSeCl4)2the octahedron around tin is slightly more distorted than in SnC14(gn). In benzaldeminium stannichloride" the SnClC2-ion shows three distances, 2.42, 2.50, 2.54 8, all of which are somewhat longer
+
+
+
+
(10) Y. Hermodsson, Acta C ~ y s t . ,13, 656 (1960).
Inorganic Chemistry
than those in SnC14(gn). The distance shown in SnC14 gas by electron diffraction is 2.31 f. 0.01 A,11. this of course being a tetrahedral molecule. No Sn-S distances are available for comparison. Thus the conformation of the glutaronitrile as TT has been confirmed, and it has been shown that this complex is, in fact, a one-dimensional polymer. There is no immediately obvious explanation of the difference in spectral assignments for the glutaronitrile in the tin(IV)complex from those in the pure glutaronitrile.
Acknowledgment.-D. 31. B. wishes to acknowledge a postdoctoral fellowship granted to him by Harvey Mudd College, Claremont, Calif., during which period the experimental data were collected. We acknowledge a grant from the Montana State University Computing Center without which it would not have been possible to perform the calculations, and we also wish to thank the University of IVashington Computing Center for time on their IBM-7094and Keith Watenpaugh for assistance in this computation. (11) "Tables of Interatomic Distances and Configuration in Molecules and Ions," Supplement Special Publication No. 18, T h e Chemical Society, London, 1965, p M22S.
RESEARCH DEPARTMEST, MOXSANTO COMPANY, ST.LOUIS,MISSOURI 63166
C O N T R I B U T I O X FROM 'THE C E S T R a L
Ligand Preferences of Transition Metals in Complexes as Determined from Nuclear Magnetic Resonance Contact Shifts BY DENIS FORSTER, KURT MOEDRITZER, AND J O H S R. V A S WAZER Received Febi.ueiy 2, 1968 The Fermi contact term causes a large change in the nuclear magnetic resonance chemical shift of a nucleus in a ligand bonded t o a paramagnetic transition metal atom as compared to the same ligand in the free state. When there is rapid exchange of ligand between the free state and the metal complex(es), the variation in the position (depending on the ratio of free to coordinated ligand) of the resonance is large so t h a t for ligand mixtures a reasonably precise measurement may be obtained of the fraction for each kind of ligand which is coordinated. From data of this type, the competitive equilibria for coordination of nickel by pyrid'ne 11s. a-, p-, or */-picoline, picoline oxide, triphenylphosphine, or triphenylphosphine oxide were investigated. An equilibrium constant for treating these data is derived and applied to the appropriate systems.
The literature1 dealing with labile replacement of one ligand by another on transition metals has been essentially limited to spectral studies involving mixtures of molecules which exhibit appreciably different electronic transitions. Thus these studies, based on electronic spectra as well as on a few other techniques such as nuclear magnetic resonance (nmr) , are generally limited to pairs of ligands involving coordination through two different elements. By the technique described herein, i t is possible to make reasonably precise measurements on systems in which the rapidly exchanging, competing ligands are very similar. For diamagnetic systems, it is generally found that (1) E.E., A. E. Harvey and 12. L.M a n n i n g , J . Aiiier. Chem. Soc., 1 3 , 4488 (1S50).
the nmr chemical shift of a magnetically active nucleus which is part of a ligand is not very different in the complex as compared to the free state, except when the nucleus is directly bonded to the acceptor atom. When the acceptor is a paramagnetic atom,2 delocalization of the unpaired electron(s) into the ligand orbitals leads to a large difference in shift between the coordinated and free ligand, with delocalization into the ligand u orbitals causing downfield shifts as large as several hundred parts per million and delocalization into the T orbitals usually giving severalfold smaller shifts which alternate in sign in a conjugated system. This led us to consider the applicability to paramagnetic systems of a quantitative treatment of the shifts of the (2) 1). I
