MERCURY (11) COMPLEXES OF GUANIDINE AND AMMONIA
Oct. 20, 1964
soluble in 1 F HClOd than in 1 F NaC104 lends credence to these possibilities. If either of the explanations is correct, the lower values of KSI correspond most closely to the conventional equilibrium constant. I n the absence of more accurate data or independent evidence, however, i t seems preferable to propose the average value. \Ve have determined, then, t h a t the equilibrium quotient of reaction 1 is two orders of magnitude greater than that of reaction 2. Therefore, equilibrium in the reaction CjHjX-Hg+
+ CsH6 e CsHsSH' + CsH,Hg'
THE
i4
a
i_;l
.
m
* * ,
. a
,
,
a
f
,
,
,
200
1
v
(8) 00 -20
favors the products m'hile solvation effects have a strong influence on equilibria in aqueous solution, i t seems likely the position of this equilibrium reflects relative strengths of aromatic carbon and aromatic nitrogen bonds with Hg(I1) From a practical point of view, this work points out t h a t in aqueous solution, nonacidic aromatic hydrogens may in fact be replaced by Hg+2. This is important in the study of the interaction of Hg(I1) with complex nitrogen heterocycles, including those of biological significance. [CONTRIBUTION S O . 3117 FROM
4325
-I 0
-00
+IO
LOG [ U t 1
Fig. 3.-Variation of the equilibrium quotient for the mercuration of p-methoxyanisole with [ H + ]' solid points, forward reaction, open points, reverse reaction, circles, squares, and triangles represent solutions of decreasing [Hg'*] (1 X 10-'-5 X lo-' M )
Acknowledgment.-This research has been supported by the Atomic Energy Commission under Contract AT(ll-1)-188. T. H . W. was a Woodrow Wilson Fellow in 1959-1960.
GATES,CRELLIN, AND CHURCH LABORATORIES O F CHEMISTRY, PASADENA, CALIFORSIA]
Mercury(I1) Complexes of Guanidine and Ammonia, and a General Discussion of the Complexing of Mercury(I1) by Nitrogen Bases' BY THOMAS H. WIRTHA N D NORMAN DAVIDSON RECEIVEDMAY27, 1964 T h e stability constants of guanidine and ammonia complexes of Hg(I1) have been measured potentiometrically. For guanidine, the reactions and equilibrium constants a t 27", AI = 1.0 M (NaC104), are: 2 R H + Hg+2 = 2H+, KIH = 2.7 X IO-' mole/l. R p H g + 2 2 H + , K z = 7.4 x 10-3; R H + H g + 2 HzO = RHgOH [R = ( H a N ) z C N H ] . The pKA of the guanidinium ion was determined pectrophotometrically to be 13.54 a t ' NaClOa)], 2"4+ H g + 2 = ("3)227", w = 1.0 M . For complexing of H g f 2 by ammonia [26", M = 1.0 h H g f 2 + 2 H + , K 2 = 0.08 f 0.02. T h e ammonia results are in good agreement with the classic measurements of Bjerrum. An extensive tabulation of the binding constants of nitrogen bases for H g f Zand H + is presented and the approximate linear free energy relationship is discussed.
+
+
+
+
+
+
Introduction For many organic nitrogen ligands and metal ions, a parallel exists between the basicity of the ligand and the stability of the metal-ligand comp l e ~ . ~Guanidine ,~ is among the strongest organic bases known.4 It is therefore of interest to study its binding with metal ions in order to test this relationship over the widest possible range. Very little is known about the characteristics of guanidine as a ligand, however. Few solid guanidine complexes have been prepared. Wormserj studied the formation of the guanidineAg(1) complex in solution, but calculated no stability constants. In this work, the potentiometric method was employed to study complexes of guanidine with Hg(I1). The 2 : 1 complex was found to be stable enough that HgO does not precipitate in basic solutions containing guanidinium ion in excess. Evidence for a mixed complex, RHgOH, was obtained (R = guanidine). (1) T h i s is t h e f o u r t h a n d final paper in t h e series "Studies of t h e C h e m istry of Mercury in Aqueous Solution." (2) J . Bjerrum, Chem. Rev., 48, 381 (1950). (3) R. J. Bruehlman a n d F . H. Verhoek, J . A m . Chem. Soc., 7 0 , 1401 (1948). (4) D. J. C r a m a n d G. S . H a m m o n d , "Organic Chemistry," 1st E d . , McGraw-Hill Book C o . , Inc., N e w York, N. Y . , 1959, p p , 173-174. (5) Y. Wormser, J . Chim. P h y s . , 46, 658 (1949).
The complexes of Hg(I1) with ammonia were also studied potentiometrically. This system has already been studied by Bjerrum6 in his classical investigations of ammine complexes. Bjeruum himself pointed o u t , however, that his glass electrode method is less suitable for studying this system than the many other systems he investigated. Hg(I1) is so strongly bound by amines-including ammonia-that the pH is not a sensitive measure of complex formation a t the high acidities necessary to partially dissociate the complexes. Experimental Materials.-Guanidine perchlorate and nitrate were prepared by neutralizing a slurry of guanidine carbonate (Eastman Organic Chemicals) with the appropriate concentrated acid. The resulting mixtures were cooled and filtered. The perchlorate was recrystallized twice from absolute ethanol and dried in a vacuum desiccator using a Dry Ice-acetone trap. Heating a t 105' for 1 hr. produced no weight loss. T h e melting point was 25% 256"; lit. m . p . 240O.' The nitrate was recrystallized from water, dried, and dissolved in hot absolute methanol. The solution was evaporated t o half its original volume and filtered. After drying in a vacuum desiccator over CaS04, the compound melted at 216-218', which compared with literature values of ( 6 ) J , Bjerrum, "Metal Ammine Formation in Aqueous Solution," Thesis, P. Haase a n d Son, Copenhagen, 1941, p p , 164-173. (7) A. Glasner a n d A. M a k o v s k y , J . Chem. Soc., 182 (1953).
THOMAS H. WIRTH
4326
4ND NORMAN
214.208 and 217".9 One formal (1 F ) stock solutions were prepared by dissolving weighed amounts of these salts in water and diluting t o volume. Ammonium perchlorate and ammonium nitrate stock solutions were prepared by roughly neutralizing aliquots of standSolutions ardized acid with ammonium hydroxide ( D u Pont). (1.0 F ) were prepared by appropriate dilution. Aliquots of these solutions were titrated with acid using a glass electrode and a 1 F XaN03 salt bridge. From the results, the 1 F solutions remaining were accurately neutralized with standard acid. They were then analyzed by evaporating aliquots of solution and weighing the residues. The preparation and standardization of Hg( C10a)2, NaC104, S a N 0 3 , HC104, and NaOH solutions have already been described .lo Potentiometric Measurements.-With minor changes, the technique was the same as in part I. In any one run, solutions containing known amounts of Hg(II), acid, and amine were titrated with acid or base and the p H and free H g + 2concentrations measured with glass and mercury electrodes. The initial concentrations of Hg(I1) and amine were varied in different runs. Beckman Type E-2 glass electrodes were used. Before each separate ammonia experiment, the p H meter was recalibrated by titrating the XaC104-NHaC104 solution to be used in the experiment with standard acid. This was just prior to the addition of Hg(C104)z and the Hg(0) electrode. Sitrogen gas was bubbled through the solutions except when free ammonia was present. Guanidine experiments were carried out at 27'; ammonia experiments a t 25". For the latter series, the apparent standard potential of the Hg(0)-Hg(I1) couple was redetermined a t 25' as described in part I ( E " = -0.6054 v . v s . s.c.e.). All measurements were performed a t ionic strength 1.O ( SaC104). Determination of pK* Values.-In order to correlate basicity with complex stability, it was necessary to determine the basicity of ammonia and guanidine under the conditions employed here. The pK.4 of ammonia was determined by titrating h7H4C104 solutions of different concentrations with base after the pH meter had been calibrated as described above. The salt bridge, a solution of SaNO3 and N H 4 N 0 8 ,was prepared to match the S H 4 +concentration in the reaction vessel at half-equivalence. A spectrophotometric method was chosen to determine the basicity of guanidine in order to avoid the difficulties of p H measurement in strong base. The intensity of the absorption of the guanidinium ion between 220 and 250 mp increases greatly in alkali. The absorption of the hydroxide ion itself makes observation of the spectrum of free guanidine at wave lengths less than 225 mp very difficult. The concentration of free guanidine can be determined from the absorbance in the 225235 mp region, however, even though the absorbance does not reach a maximum. Solutions containing guanidine perchlorate (0.0064.02 F ) , sodium perchlorate (0.03-0.87 F ) , and sodium hydroxide (0.120.96 F ) a t unit ionic strength were prepared and brought to constant temperature. The spectra of three successive samples of each solution were recorded on a Cary Model 14 spectrophotometer with a thermostated cell compartment with air in the reference beam. The spectra of fresh samples were identical; however, absorption of COS changed the spectra of solutions exposed to the air. Blank solutions containing only S a O H were prepared and their spectra recorded; all unnecessary exposure to the air was avoided. The concentration of free guanidine in each solution was determined from the absorbance a t 227.0, 229.0, and 231.0 mp. At the concentrations used, protonated guanidine does not absorb at' these wave lengths. The absorbances were corrected for S a O H background. The conceqtration of guanidine never exceeded 5y0of that of S a O H ; therefore, corrections to [IiaOH] for reaction-were not made. Under these conditions
W + I o - 1
X
EX
+""(-) EX
1 [OH-]
DAVIDSON
VOl. S6
dine concentration, and
KB
=
[ R H + ][OH-] [R 1
R H and R represent C( NHz)3+ and ( H2N)2C=SH, respectively. The value of pKw is necessary in order to find the K.4 of the guanidinium ion from this spectrophotometric K B . pKw was determined from the apparent p H of solutions of known [OH -1 after calibration of the p H meter on solutions of known [H+]. +
Results Basicity of Guanidine.-The average K B ,from data a t 231.0, 229.0, and 227.0 mp, is K B = 0.72
f
0.06 -11(27.2',
p =
1.0 -\! (3)I)
The molar absorptivity of free guanidine a t each wave length is 54, 92, and 150 l./mole-cm., respectively. Kw,defined here as the concentration product [H +]. [OH-], was determined as pKw = 13.68
f
0.04 (27.2',
p =
1.0 M )
(4)
From eq. 3 and 4, the ~ K ofAthe guanidinium ion is 0.06 (27.2', p = 1.0 -Lf) (5) Using a hydrogen electrode, Hall and Sprinkle1] found the pK* to be 13.59 =t0.05 in 2 N KCl a t 24.2'. Basicity of Ammonia.-The pKa of the ammonium ion was measured as 9.47 f 0.02 a t p = 1.0 31 (XaC104), 25.0O. Guanidine-Hg(I1) Complexes.-The quantitative analysis of the results indicates that, in addition to the ion Hg(NHC(NH2)2)2+2, there is a species, the formation of which involves release of two protons and the addition of only one guanidine to Hg+'. A likely species of this type is (H2N)zCh"HgOH+, written here as RHgOH+. The equilibria considered are pK*
=
13.54
f
K = 170 = [ H g z + 2 ] / [ H g + 2(at ] the e l e ~ t r o d e ) ' ? ~ ' ~ ~ (6) K2H = = [Hg(OH)'] [H+]'/ [Hg+2]'3b (7)
Kz
=
[HgRzf'] [H+]'/ [Hg+'] [ R H + ] '
KIH = [RHgOH+][ H + ] ' / [Hg+'] [ R H + ]
(8) (9)
These equilibria lead to the equation
KIH
W+I2
(10)
from which
The quantity on the left of eq. 11 will be designated
Q1. In Fig. 1, the experimental results are presented (I)
where X is the corrected absorbance, C Y the molar absorptivity of (HZX)ZC=NHa t that wave length, [ R H '10 the formal guani(8) C. R . Witschonke, Anal. Chem., 26, 562 (1954). (9) C. B. L. Smith, "Inorganic Syntheses," Vol. I , J. C. Booth, E d . , McCraw-Hill Book Co., Inc., N e w York, N. Y.,1939, p. 96. (10) Paper I of this serie:: T. H. Wirth a n d N . Davidson, J. A m . Chem. Soc., 86, 4314 (1964).
in the form of a plot of Q1 v s . l / [ R H + ] . The results represent variation over wide concentration ranges ([RH+]o = 0.015-0.38 F; [Hg+']o = 1 X 10-4-6 X 10-3 F; PH 2.6-5.1). (11) N. F . Hall and M . R . Sprinkle, i b i d . , 54, 3469 (1932). (12) T. H. Wirth, Ph.D. Thesis, California Institute of Technology, 1964, p. 20.
(13) (a) S . Hietanen and L. G. Sillen, A r k i v Kern$, 10, 103 (1956), (b) Acta Chem. Scand., 6 , 747 (1952).
MERCTTRY (11) COMPLEXES OF GUANIDINE AND AMMONIA
Oct. 20, 1964
4327
Points for which [Hg+2]o/[Hg+2]was less than 300 were excluded. Cnder conditions where this ratio is less than 300, [Hgzf2] represents a large fraction of [Hg+2]o,and little information about Hg(I1) complexes can be obtained.14 The experiments in Fig. 1 were conducted in acidic solution. The pH-e.m.f. behavior is essentially similar in basic solution; there is no evidence for the formation of additional complexes. Irreversible changes in the solutions occur a t high basicity, however. Since the results in basic solution were also less precise, they were not included in the equilibrium constant calculations. From Fig. 1
K~ = 7.4 x 10-3, K~~ = 2.7 x 10-4 mole/l. a t 27.2', k = 1.0 IM From K z and K A
(12)
It is of interest to consider the equilibrium constant of the reaction HgR*+'
+ Hg(OH)*
2RHgOH'
(14) 0.6I 0
I
I
1
I
I
I
I
IO
20
30
40
50
60
70
I
I/CRH+l
This is greater than the statistical value of 4, but i t corresponds to an equilibrium value of [RHgOH+] only about twice as great as the statistical. Marcus15 observed a similar stabilization of mixed complexes in his study of the mixed halides of mercury(I1). Ammonia-Hg(I1) Complexes.-&, defined as in eq. 8, with R E' "3, can be computed from eq. 16 if no other ammine complexes are present
(16) As is true with guanidine, this value of K Z is not constant as the total ligand concentration is changed. However, the relationship between [NH4+] and the calculated Kz is not simple; indeed, K Zgoes through a minimum. In Table I, the average value of K2 from each of six VALUESFOR IT,,
6 1 2 4 5 7
TABLEI STABILITY CONSTANT OF Hg( S H Z ) ~COMPLEX +~ THE
INHI']
0.74 0.38-0.42 0.185-0 189 0,075-0.078 0.019-0.023 0.0066-0.0072
IHg(Cl0i)zla
2.9 X 5 X 1.2 X 6.1 X 1 . 2 X lo-' 1 . 2 X lo-'
THE
Average Kz 0 . 0 9 8 d= 0 005 0.089 f 0.012 0 . 0 6 6 f 0.006 0.080f 0 007 0 . 1 0 1 f 0.006 0 15 i 0 . 0 1
separate titrations is given. Values where [Hgf2Io/ [Hg+'] < 300 are excluded. Also excluded are values where [NH4+]/[H+] > 10'. From the stability constants determined by Bjerrum, it is possible to calculate that if this ratio exceeds lo7, Hg(NH3)3+2begins to form in significant concentration. We observed that the apparent K2 does indeed increase above this point. Each average value of K z represents a wide variation of pH (between 3 and 7) a t approximately constant (14) Reference 12, p. 135. (15) Y.Marcus, A d a Chem. S c a n d . . 11, 611 (1957)
Fig. 1.-Guanidine-Hg(II),
variation of QI with l / [ R H 1'
(NH4+]. The variation in Kz with [NH4+] is small but significant. I t is evidently related to [NH4C104],not to [Hg(C104)2]o. The increase in K Z as [NH4C104] decreases below 0.2 M might be attributed to a HOHgNH8+ complex similar to that formed in the Hg(I1)-guanidine system. The contribution of such a complex to the mercury binding increases as NH4+ decreases; thus run 7 is the most important one in this respect. We felt that it was not practical to push conditions to the point where the possible contribution of HOHgNH3+ to the binding was more pronounced because of the precipitation of basic salts; the trend in Kz in Table I is not sufficient to warrant a confident conclusion that HOHgNH3+ exists. The small increase in K z a t very high N H 4 + concentration can be attributed to activity coefficient effects due to the replacement of N a + by NH4+ in large amounts. Rather than obtaining a value of Kz by extrapolating a graph of (31v s . l / [ R H + ] to l / [ R H + ] = 0 and ignoring the experiments where [",+I > 0.2, it seems preferable to adopt an appropriate mean. The median of the average K Z values seems suitable. Experiment 7 can be excluded from the group of acceptable measures of Kz, however. We have, then
Kz
0.08 f 0.02 (25.0') From the pK* of ammonia =
1.0)
(17)
K ~ / = K 1~0 1 7 . ~ 8 * 0.1 (I./rnole)'
(18)
I.(
=
The value obtained by Bjerrum was log P2 = 17.5 in 2 M NH4N03 a t approximately 22'. In more basic solution, the complexes Hg(NH3)3+2 and Hg(NH3)4+zform. The presence of free ammonia in such solutions makes a number of changes in procedure necessary in order to obtain accurate constants
THOMAS H. WIRTHA N D NORMAN DAVIDSON
4328
TABLE I1 . ~ S S ~ C I A T I O XCOXSTANTS O F ~ - I T R O G E S - ~ O N T A I N I N GLIGANDS WITH
KH' 9 61 9 47 10 7 2 10 71 10 18 10 00
Ligand
log
Ammonia ( L ) Methylamine ( L ) %-Butylamine( L ) Ethylenediamine ( L )
10 00 7 9 9 9 9 5 4 11 9 10 7 13 4
2-Methoxyethylamine ( L ) Aminoacetic acid ( H L ) Ethanolamine ( L ) 2-;11nino-2'-hydroxydiethyl sulfide ( L ) Pyridine ( L ) Aniline ( L ) Piperidine ( L ) Diethanolamine ( L ) Triethylamine ( L ) Triethanolamine ( L ) Guanidine ( L ) d-Cycloserine ( H L )
T
7
Imidazole ( L ) Histidine ( H L )
6 9 3 4 9 9
Adenosine ( L ) Cytosine ( L ) Thymidine ( H L ) Thymine ( H L )
20 30 76 60 18 21 78 12 00 77 90 54 57 40 12 08 12 5 5 6 8
jog XHg'
75
Temp., ' C ,
9 95 05 71 65 59
22 25 25 25 25 25 25
6 45 8 92 9 6 8 66 9 13 5 0 4 60 8 72 7 83 7 80 6 54 12 48 7 04 8 77 8 37 7 50 10 61 4 25 5 45 10 6 10 6
25 30 20 25 30 25 27 25 25 25 25 27 25 25 27 27 27 27 27 27 27
8 8 8 9 11 11 11
Hg( 11)
hlethod
Kef.
b (-,
d d d e
.f e L? ii
(i
r: d z
d j d d C
k k ! ! ! m
m m m
R v f 2. If the complex is HgR2,K H 3 [HK']/[H'] [R], 82 E [ H g R ? + z ] / [ H g +[RI2, 2 j K E ~ .\/& See ref. 6. c Present work J . I . ivatters and J . G . Mason, 1.:lm Chem. Soc., 78, 285 (1956). f C. J . S y m a n , D. I(.Roe, and D . B. Masson, zhid., 77, 1191 (1955). 0 J . Lotz, unpublished; from G. H . McIntyre, Dissertation, Pennsylvania State College, 1952. * H . Y.Fluod a n d 1.. Loras, Tzdskr. Kjemi Bergvesen M e t . , 5 , 83 (1945). ' Part I of this series. 1 J . BjCrrum and S. Refn, Suoinen Keinzstilehti, B29, 68 (1956). P. Brooks and S . Davidson, J . A m . Chem. S n c . , 82, 2118 (1960). K Ferreira, E . k S.G . Lordi, J . Pherm. Sci., 52, 397 (1963). Ben-Zvi, T . Yamane, J . Yasilevskis, and S.Davidson, "Advances in the Chemistry of the Coordination Compounds," S. Kirscliner, E d , Macmillan, S e w York, X. Y., 1961, pp. 457-462. 6
Titration with acid or base is no longer useful, because large changes in [NH4+]result. The method of Bjerrum is suitable for study of these complexes; therefore, the constants were not redetermined. These results confirm those of Bjerrum. It is somewhat disappointing t h a t K z is so sensitive to the nature of the medium. Such variations make it difficult to define a constant of general validity. A constant useful in all 1 F perchlorate solutions evidently cannot be more precisely determined than the Kz given here. Complex Stability and Ligand Basicity The base strength of an electron donor can be discussed in terms of its tendency to bind protons, metal ions, or any other Lewis acid. For a series of sufficiently similar donors, one expects a correlation to exist between basicities measured on different scales. Bruehlman and Verhoek3showed that the equation
+
log KAg = u log KH b 119) is obeyed for several amines. Kag is the equilibrium constant of the reaction Xg'
and KH that of H'
+ R e A4gRC
(20)
+R
(21)
e HR'
R being an amine. In order to investigate the corresponding correlation, we have compiled the known stability constants of Hg(I1) complexes with nitrogen bases in Table 11.
The listed value of each KHg is a "mean complexity constant"-the square root of the over-all association constantsof the 2 : 1 complex (HgL2+?). The first and second stepwise formation constants are usually similar.2 Therefore, K H ~approximates PI, the association constant of the 1: 1 complex, which is not always measured separately from the over-all constant. T h e PI of many mercury complexes is particularly ditEcult to determine. In Fig. 2, log K H g is displayed as a function of log K H . One must be cautious in interpreting plots such a s that presented here. There is likely to be some scatter due to different conditions of measurement and experimental error. It appears, however, that there is a rough correlation, but only a rough correlation, between log K Hand log K H ~ . The most striking and obvious feature of this rough correlation, as displayed in Fig. 2, is the fact t h a t the slopes of the lines in the equation log K H g= a log K H b are relatively large; a = 0.93 (lower line) and n = 1.15 (upper line). For silver, the corresponding slope is about 0.3. We have not made a careful study of other metal ions, but for CuT2,considering ammonia and pyridine as ligands, we find a = 0.63. The large value of the slope for H g f 2 correlates with the high strength of the Hg-N bond compared to other metal to nitri gen bonds. Several additional qualitative generalizations emerge from Fig. 2 , however. Alkylamines, primary, second-
+
Oct. 20, 1964
1,2-DICYANOETHYLENE-1,2-DITHIOLATECHELATES
ary, and tertiary, bind Hg(I1) less well relative to protons than do other nitrogen bases. Bruehlman and Verhoek3 noted a similar effect with Ag(1) and secondary and tertiary amines. Heterocyclic nitrogen compounds such as the purines, pyrimidines, and imidazole derivatives bind particularly strongly. However, other a-electron systems, such as aniline and pyridine, do not. The strong binding by the heterocycles may be due to donation of d-electrons from into the a-electron system of the ring to place negative charge on the other nitrogen atoms. A similar but smaller effect would have been expected, but is not observed, for guanidine. The slightly high value for ethylenediamine may be due to a small contribution from the formation of third and fourth bonds to Hg+*. I t is clear, however, that such chelate effects are very weak for Hg+2 as compared to, say C U + ~because , of the tendency of H g + 2 to form two strong collinear bonds. The case of cycloserine may be complicated because there are two nonequivalent basic nitrogens in this molecule (in any case, the complex surely contains mercury-nitrogen bonds rather than mercury-oxygen bonds as suggested in ref. k of Table 11). \Ye have drawn two lines in Fig. 2. The lower line is our suggestion as to the best correlation for typical amines without bulky alkyl substituents, and the upper line is for heterocyclic nitrogen compounds. Acknowledgment.-This research has been supported by the Atomic Energy Commission under Contract [COSTRIBUTION No. 974
FROM THE
4329
12II
-
IO
-
HISTIDINE THYMIDINE
I" 9 -
Y
0
s
D-CYCLOSERINE
'1 // IMIDAZOLE
8-
0-CYCLOSERINEH
TRIETHANOLAMINE
ETHYLENEDIAMINE
'i$ CYTOSINE
-1
?D :;N IE
ADENOSINE l
3
4
5
6
7
~
9
8
~
1 0 1 1
I
1 2 1 3 1 4
LOG K H
Fig. 2.- Complexes of Hg( 11) with nitrogen bases; mean association constant (KH.)zls. proton affinity of the ligand ( K H ) (AHDS = 2-amino-2'-hydroxydiethylsulfide). Note t h a t one of the points labeled ethylenediamine is for the singly protonated form. The two straight lines represent our suggestions for the best correlations for typical amines without bulky alkyl substituents (lower line) and for heterocyclic nitrogen compounds (upper line).
AT(ll-1)-188. T. H . W. wasa Woodrow Wilson Fellow in 1959-1960. We are indebted to Professor George S . Hammond for his optimistic attitude and helpful advice with regard to the interpretation of Fig. 2
CESTRALRESEARCH DEPARTMENT, EXPERIMENTAL STATION, E. I.
DU
PONTDE NEMOURS AND
COMPASY, WILMINGTON, DELAWARE]
1,2-Dicyanoethylene-l,2-dithiolateChelates BY J. F.WEIHER,L. R. MELBY,AND R. E. BENSON RECEIVED J U N E 15, 1964 Chelates derived from 1,2-dicyanoethylene-1,2-dithiol and certain transition metal ions exhibit singlettriplet magnetic behavior which is attributed to spin interactions between pairs of metal ions through the sulfur atoms of adjacent ligands
Dithiolate chelates 1 have been the subjects of
[ix:;M
