440
ROBERTEARLDAVISAND ARNOLD COHEN
In like manner, other hydrazides were prepared, and melting points and yields are summarized in Table 11. 2. 1,2-Propylene Diisocyanate.-A 2-1. beaker equipped with a thermometer and a mechanical stirrer and externally cooled in a n ice bath was charged with 600 g. of ice, 100 ml. of carbon tetrachloride, 70 ml. of concentrated hydrochloric acid, and 57 g. (0.36 mole) of 2-methylsuccinic dihydrazide. A solution of 50 gd (0.73 mole) of sodium nitrite in 100 ml. of water pre-cooled to 0 was added dropwise to the stirred mixture over a period of 20 min. The temperature of the reaction mixture was maintained below 8’ by addition of pieces of ice, and stirring was continued for 0.5 hr. after completion of the addition. The layers were separated, and the azide solution combined with two benzene extractions of the aqueous layer. The final volumes amounted to about 400 ml. The solution was dried overnight over anhydrous calcium chloride and then refluxed for 4 hr. t o complete decomposition of the azide. The refluxing solution developed a blue color which disappeared when the decomposition was essentially complete. Solvent was removed and the diisocyanate was then distilled a t reduced pressure. After a short forerun, 19.6 g. (42Yc) of pure 1,Z-propylene diisocyanate was collected as a lachrymatory liquid which had b.p. 83.5” ( 2 5 mm.) and n*‘D 1.4398. Treatment with aniline gave a bis-phenylurea, m.p. 238240 O . ~~
Vol. 86
Anal. Calcd. for C1,H2~Nn02: C, 65.36; H , 6.45; S , l i . 9 4 . Found: C, 65.41; H , 6.45; N, 17.65. Table I11 summarizes the properties and yields of the other isocyanates prepared in this work. All the isocyanates were extremely moisture-sensitive, so no elemental analyses were performed on the monomers themselves. Several monomers were derivatized with aniline, and the bis-phenylureas serve well to identify the isocyanates. 1,Z-Propylene diisocyanate, 1,2cyclohexylene diisocyanate, and glyceryl triisocyanate apparently have not been prepared previously. 111. Infrared Analysis of Phenyl Isocyanate Dimer ( 1,3Diphenyluretidinedione).-Although dimers of aromatic isocyanates (uretidinediones) have been known for man> years,g the carbonyl absorption of the four-membered heterocycle has not apparently been reported in the literature. The infrared absorption spectrum determined on a chloroform solution of phenyl isocyanate dimer, m.p. 177-178” (lit.lo 175-176’), showed a single sharp band at 5.60 p due to carbonyl.
Acknowledgments.-The author is indebted to Dr. R. Zbinden and Miss Mary Schick for interpretation of the infrared spectra, and to Mrs. Raila 31. Mason for determination of molecular weight by osmometry.
~
(8) I t is essential t h a t t h e solution be dilute a t this stage, so t h a t t h e h e a t generated b y t h e slowly decomposing azide m a y be safely dissipated. More concentrated solutions retain sufficient h e a t to cause sudden a n d violent d t compositiop of t h e remaining azide.
[CONTRIBUTION FROM THE
Displacement Reactions. V.
DEPARTMENT OF
(9) J. H. Saunders and K. J. Slocombe, Chem. Rea., 43, 211 (1948). (10) J. S. Blair a n d G. E. P. S m i t h , Jr., J . A m . Chem. Soc., 6 6 , 909 (1934). We are indebted t o Dr. K . C. Smeltz for a sample of pure phenyl isocyanate dimer.
CHEMISTRY, PURDUE UNIVERSITY,
LAFAYETTE, IND.]
The Reaction of Cyanide Ion with Tetramethylthiuram Disulfide
B Y ROBERTEARLD A V I SAND ~ ARNOLD COHEN3 RECEIVED JULY 17, 1963 The reaction of tetramethylthiuram disulfide with cyanide ion has been found to be a two-step process in which the dimethyldithiocarbamate intermediate rises to concentrations allowing spectra! analysis or isolation. The two S Ndisplacement ~ reactions are discussed in terms of the oxibase scale relating oxidation electrode potential and basicity of the nucleophile t o its reactivity. An estimate of 2 12 f 0.04 A. is given to the S S bond length in the disulfide by the correlation of activation energy with bond distance.
Introduction The displacement reactions on a divalent sulfur atom divide into three stoichiometric classes. 4 , 5
+sz
x-Y
+ 2--:
x-s-Y
xs
x-s-2
(a) (b)
+ YZ +Y
(C!
Numerous examples of kinetic and mechanistic studies have been reported for type c While the reaction of cyanide ion with tetramethylthiuram disulfide has the same stoichiometry7 as the type a , CH,
S CH3 lr/ XCSSCN
\ll
S
/
CHn
\
+ -CN+
CH, CHI
\I1
/
CHa
S
NC,
S
II/
s
/cx \
CHa
+ -SCN
(1)
CHI
i t is unlikely from the work of Cambrona that reaction 1 represents the mechanism as well as the over-all
reaction.
Indeed
Cambron’s work and suggested
(1) Paper I V : R E Davis, A . Cohen. and J Louis, J. A m . Chem. S o c , 8 5 , 3050 (1963). ( 2 ) Alfred P. Sloan Fellow, 1962-1964. (3) Taken in p a r t f r o m t h e M Sc. thesis, January, 1963. (4) R . E . Davis, Vol 11, Chapter 1, of “Organic Sulfur Compounds,” N. Kharasch, Ed , Pergamon Press, New York, N. Y., in press. ( 5 ) R E. Davis, Val. I 1 in ”Survey of Progress in Chemistry,” A. Scott, E d . . Academic Press, Inc., iXew York, Pi. Y , in press. ( 6 ) Reviewed also in W. A . Pryor, ‘‘Mechanisms of Sulfur Reactions.” McGraw--Hill Book C o . , I n c . , N e w York, N. Y . ,1962. (7) J. Von Braun and F Stechele, Be? , 36, 2275 (1903) (8) A Carnbron. Can. J. Research, 1,341 (1930).
mechanism have been found to be correct in the present study. In continuation of a series of studies on the S N ~ reaction a t divalent sulfur atoms, we wish to present a detailed kinetic and mechanistic investigation of this reaction of an excellent thiophilicg reagent, cyanide ion, with a compound containing a thiuram disulfide sulfursulfur bond. Results Ultraviolet Spectra.-The ultraviolet spectrum of the nearly colorless tetramethylthiuram disulfide (TMTD) is characterized by strong absorption increasing below 300 mp with a plateau between 280 and 240 mp and finally increasing toward 200 mp.lo The monosulfide (TMTM) is yellow and has one strong maximum near 275 mp in methanol and a minimum near 240 m,u.lo In Fig. 1 the spectrum of a mixture of T M T D and T M T M each a t 5 X 10-5 Af in methanol is reproduced Also in this figure is the spectrum of a solution containing initially 10 X M in T M T D and some cyanide ion after ,SO~. reaction. From this figure one concludes, in accord with Cambron’s results, that the reaction is not a onestep process but that intermediates are present in measurable concentrations -41~0presented in Fig 1 is the spectrum of the dimethyldithiocarbamate with its characteristic double camel hump near 285 and 250 mp. Isolation of an Intermediate.-Cambron* suggested and demonstrated that a dithiocarbamate could react with a thiocarbamyl thiocyanate to produce a thiuram monosulfide and thiocyanate ion. I t seemed most reasonably to suggest that reaction 1 has a two-step (9) P D Bartlett a n d R E Davis J A m Chem Soc , 80, 2513 (1958) ( 1 0 ) R E Davis and C Perrin, ibrd , 8 S , 1500 (1960)
Feb. 5, 1964
CYANIDE IONWITH TETRAMETHYLTHIURAM DISULFIDE
44 1
TABLE I REACTION OF TETRAMETHYLTHIURAM DISULFIDE WITH POTASSIUM CYASIDE ki,
TMTD,M
x
106
KCN,M
x
T , "C.
Salt, 41 X 108
Solution
104
sec
25.0 4.46 12.6 MeOH 25.0 3.20 7.5 MeOH 3.03 15.3 MeOH-HtO" 50% phosphate* 25.0 1.00 2.50 MeOHc 21.6 1.02 5.60 MeOHC 21.6 2.0 43.6 MeOH 21.6 9.98 5.60 MeOH' 21.6 12.1 10.9 MeOH" 21 6 2.02 10.0 MeOH KSCK 6.00 21.6 2.02 10.9 MeOH KSCN 1 20 21.6 2.02 10.9 MeOH KC1 2 . 5 0 21.6 2.02 10.9 MeOH KCl 5 0 21.6 2.02 10.9 MeOH KCI 1 2 . 0 21.6 4.37 5.97 MeOH 15.0 4.57 6.54 MeOH-H20" 15.0 4.95 7.38 MeOH-H20 40% 15.0 70 % 15.0 1.01 5.75 MeOH-H2O 2.02 1.09 MeOH 11.4 2.03 1.08 MeOH 11.4 2.05 10.0 MeOH' 5.0 70% 5.0 2.53 9.80 MeOH-H20 2.40 10.7 MeOH-HZO 10% 5.0 10% 5.0 MeOH-H20C 10.0 1.30 A small amount of sodium methoxide was added to suppress the hydrolysis a Volume yo. * p H 5.6. vent the decomposition of the dithiocarbamate.
S
S
il
il
ki
+ CN - + (CH8)zSCS- + (CH8)zNCSCS
TMTD
I
S
I1
(CHs)zNCS-
-1
M , -1 X 10-1
1.5 1 5 0.7 1.3 1.2 1.2 1.1 1.2 1.1 1.3 1 2 1.1 1.0 0.83 .80 .72 .70 .60 .63 .47 .36 .43
ki, M-1 X 10-1
sec-1
0.7 0.7 0.6 0.4
0.14
.40 0.10 of the cyanide and to pre-
I.0C
(2)
I1
S
II + (CH3)zXC-SCN
ki
---+TMTM
+ -SCX
(3)
mechanism. An attempt t o isolate the thiocarbamyl thiocyanate was made by mixing T M T D and potassium cyanide in methanol followed by the addition of hydrochloric acid in a good hood t o destroy the dithiocarbamate" and remove cyanide ion. An unstable impure fraction was obtained that had an infrared band near 4.5 p which might be the stretching frequency of the carbon-nitrogen triple bond in 11. However, the zinc salt of the dithiocarbamate was isolated in the reaction of zinc cyanide with T M T D . In this case the zinc salt was obtained in high yield. Another unstable, impure fraction was obtained that might have been 11. It produced thiocyanate ion upon standing in numerous solvents. Thus there is good evidence both spectroscopic and chemical for the two-step processes 2 and 3. Kinetics-A detailed analysis of the time behavior of T M T D and cyanide supports the two-step mechanism 2 and 3. Some experiments were run with equivalent concentrations and the data fitted with the equations of the appendix. It was more convenient to run the reaction with a large excess of cyanide. The data are presented in Table I. All kinetic data were processed on a computer. See Fig. 2 for a typical experiment. The reaction is first order in T M T D over a 12-fold variation in initial concentration and nearly first order in cyanide ion. Control of cyanide ion with low pH means that the dithiocarbamate will decompose. The rate constant kl is depressed by the addition of inert salts and by dilution of the methanol with (11) (a) G . D T h o r n a n d R A 1.udwig. " T h e Dithiocarbamates a n d Related Compounds," Elsevier Press, K e w Vork, N . Y , 1962; (b) M Hallaway, Biochiwt B i o p h y s A c t a , 36, 538 f1959), fc) P Zuman a n d R . Zahradnik, Z p h y s i k Chem., 208, 135 (19.571, id) R Zahradnik a n d P Zuman, Chem List?, 6 2 , 2:31 (1938).
2
05c
50% TMTD L
I
Fig. 1.-Ultraviolet spectra in methanol of T M T D , the synthetic mixture of one part T M T D and one part T M T M , and the K C S a t 50% reaction. Also inobserved spectrum of T M T D cluded is the spectrum of dimethyldithiocarbamate. Concentrations are 5.0 X M.
+
water. Both effects are consistent with the reaction of a charge ion with a neutral molecule with charge (CHajzKCSZ-
+ H'
ki
-3 (CHaj2NtI
fast
ks
=
+ CSA
(4)
(CHa)?XH t H i (CHa)rNH>rate = k3[Hf][(CH,),SCS,-] 1.6 X l o c * M-' see.-' a t 25" in water ( p H range 5.9 to 8.4)
442
ROBERTEARLDAVISAND ARNOLDCOHEN
1
TMTO
4.46r165
M
VOl. 86
changes result in a favorable free energy change dnd process is spontaneous process 2
C N - __ + (CH3)2 "2% AF
0
(CH3)pNCSS-
(6)
--+- 2 9 increase + 11 to 13 increase
the oxidative dimerization potential of this polysulfide4j ion would place E near 2 9 An estimate of the pK, would place H between 11 and 13 Thus the E increases and the H probably increases As most of the effect4 is in the change of E-values, this process 7 has an unfavorable A F and would not occur S S TMTD
I, + (CH2)zSCSS-
/I (CH#)*NCClj
+ C N - unfavorable ---------+ >0 AF
(7)
The thiocarbamate will easily displace the thiocyanate leganion In this case the changes of both E and H a r e quite favorable. process 3
(CHa)zNCSz-
--+ -SCX
AF < 0 E 2 . 3 9 +1.83 decrease H 5 . 7 4 + 1 decrease
1960, Davis,lo The Energetics of the System.-In discussing the thermochromism properties of T M T D , reported the heat of reaction 1 in methanol using a TMTD
+ NaCN --+ T M T M + NaSCPi
(1)
A H = - 1 9 . 8 i 0 . 4 kcal./mole
simple calorimeter. Using this datum and the heats of .formation of sodium thiocyanate and sodium cyanide, an estimatelowas given for the A H of TMTD AH
+
--+ =
TMTM '/a Ss - 2 . 4 kcal./mole
(8)
Recently McCullough'j measured the heats of combustion and of formation of T M T D and T M T M using a precision rotating bomb calorimeter. Using their values of AHrO (solid) of 11.44 f 0.30 kcal./mole for T M T M and 9.74 f 0.41 kcal./mole for T M T D , we have calculated the AHof (1) to be -17.20 f 0.70 and the AHof (8) to be - 1.70 f 0.71 kcal./mole. Theagreement of our data, simply obtained, with more exacting data is quite good. Some correction would be needed as the A H f O ' s refer to the solids and our values refer to homogeneous solution in methanol. An Estimate of the Sulfur-Sulfur Bond Length by Kinetic Methods.-The activation energy for the initial reaction 2 is low. Using the empirical correlation' between the activation energy and the sulfur-sulfur bond length, an estimate of 2.12 0.04 A. is placed on the sulfur-sulfur bond length in ThZITD.l6 Such a value is consistent with the general reactivity and thermal
*
(15) W.D. Good, J I, Lacina. and J . P . McCullough. J P h y s . C h r m , 66, 860 (1961). (16) An X-ray analysis of T M T D has been suggested t o Prof Olav Foss, personal communication, spring, 1963.
CYANIDE IONWITH TETRAMETHYLTHIURAM DISULFIDE
Feb. 5 , 1964
instability of TMTD,10,17-20 and the resonance stability of the dithiocarbamyl free radical, RZNCSZ.. Experimental
Then
Materials.-Eastman Kodak Yellow Label tetramethylthiuram disulfide was recrystallized.1° The monosulfide was isolated from the reaction of T M T D (Yellow Label, not purified) with potassium cyanide in methanol. The product was recrystallize: from aqueous methanol, isolated in 88% yield ( m . p . 105-106 , corrected). Methanol was refluxed over magnesium turning and distilled through a n 8-ft. packed column. T h e fraction boiling at 64.5’ (750 mm.) was stored in the dark. Potassium cyanide (Baker reagent) was standardized against silver nitrate. T h e cyanide solutions were prepared daily. Potassium chloride and thiocyanate (Baker reagents) were dried a t 110’. Kinetics.--A Cary Model 14 spectrometer or a thermostated Beckmen D U with 1-cm. cells were used. -%lI kinetic d a t a were treated on a Royal McBee RPC-4000 digital computer using a program written by -4.C. An estimate of the standard deviation was also provided. Isolation.-Zinc cyanide (0.10 mole) was suspended in 600 ml. of methanol. The solvent was then allowed t o reflux slowly through the thimble of a Soxhlet extractor containing 0.10 mole of T M T D . After the T M T D had been added, the solution was partially concentrated. The white solid removed by filtration was washed with acetone. From the acetone solution 0.09 mole of the zinc dithiocarbamate was obtained, identical in all respects with a known sample in the infrared, n.m.r., ultraviolet, and melting point. Further concentration of the original methanol solution rewlted in an unstable oil having strong absorption near 4.5 p . Numerous attempts failed t o win a pure product.
Let
Appendix The differential equations for the kinetic system
443
1 du kiu dt
c
=--
and es =
(vi)
7
and kn/ki
=
k
-
ku
(vii)
Then iv becomes d2u du ..- - .ds2 ds
=
0
Solution of viii using the differential operator D ds gives
with g l and gzconstants.
(viii) =
d/
Defining dl
X =
+ 4k
and evaluation of g l and gz with the boundary conditions 7 = 1 and C = 0 a t t = 0
ki
A + B + C + D
Then as
ks
C + D + E + F
have not been solved in closed form. If one uses equivalent concentrations, then A = B, C = D, E = F. In this case ki
2A
--+
2C
(i)
2c
+2E
(ii)
kg
and solution is possible. The solution of system (i and ii) has not been reported in the literature. From i (iii) (17) J . C . D . Brand and J . R . Davidson, J . Chem. Soc., 15 (1956) (18) C . Walling, ‘Tree Radicals in Solutions,’’John Wiley and Sons, Inc., N e w York, N . Y . , 1957, pp, 335-341. (19) T . E . Ferington and A, V. Tobolsky, J . A m . Chem. SOL.,7 7 , 4 5 1 0 ( 1 9 5 5 ) . 8 0 , 3215 (1958) (20) D . Craig, Rubber Chem. Tech., 30, 1291 (1957).
dE
=
kzC’dt
E=Ao
c
=
k i A 2 dl
31
1--
- dC
- C
(xi;)
The value of kl can best be obtained from expression iii obtained from a time study of the concentration of A . The value of kz can be estimated by plotting the concentrations of A , C, and E vs. 7 - l ; C reaches a maximum a t tc max when dC/dt and d2C/dt2 = 0. An analytical expression then relates k l , t max, and k2. These operations are quite rapid on a computer. Acknowledgment.-We wish to thank the Walter Reed Army Institute of Research, the National Science Foundation, the Research Corporation, and the National Institutes of Health for grants and aids providing the equipment and support for these projects. Some experiments were preformed by Mr. Harry Rubins and Mr. Steve Olin. Mr. Nakshbendi obtained the kinetic data reported in eq. 4.
