6625
Acid-Catalyzed Oxidative Dimerization of Formaldehyde Dimethylhydrazone to Glyoxal Bisdimethylhydrazone. A Mechanism Study F. E. Condon* and D. Fhcasiu' Contribution from the Department of Chemistry, The City College of the City University of New York, New York, New York. 10031. Received December 22. 1969 Abstract: Formaldehyde dimethylhydrazone (1) gives glyoxal bisdimethylhydrazone (2) when treated with acids, with or without a solvent. The reaction is accompanied by the formation of dimethylamine and tarry by-product. The transformation of 1to 2 was also observed to take place slowly with water as a catalyst, and very slowly in pure, dry samples of 1 at room temperature, with no added catalyst. With acetic acid as catalyst, and 1,6dioxane or l-propanol as solvent, the yield of 2 approached 51.5%, or 7 7 x of theory, with one-third of 1 undergoing reduction. When the reaction was carried out in the presence of 1,l-dimethylhydrazine,and 1,4-dioxane as solvent, polymer formation was reduced, and the yield of 2 approached 77%, or again 77% of theory, since the 1,l-dimethylhydrazine served as a replenisher of the 1 undergoing reduction. The reaction was shown to involve the formation of the simple head-to-head dimer of 1, 2,2-dimethylhydrazinoacetaldehydedimethylhydrazone (8). This intermediate was isolated and was shown to give 2 when heated with acetic acid. Following established pathways, 8 may undergo cleavage to dimethylamine and glyoxal dimethylhydrazone imine (lo), which can give 2 by a transamination reaction with 1,l-dimethylhydrazineor, in its absence, by a transhydrazonation reaction with 1.
D
uring the course of work with hydrazines in this laboratory, a rapid, exothermic reaction of formaldehyde dimethylhydrazone (1) with acids was observed, resulting in the formation of a large amount of glyoxal bisdimethylhydrazone (2). In this paper H+
2Me2NN=CH2 +Me2NN=CHCH=NNMe2 f [2H] (1) 1 2
are described results of experiments designed to shed light on the course and mechanism of this reaction. The reaction bears superficial resemblance to the well-known oxidative dimerization of aldehyde phenylhydrazones,Z eq 2. This is brought about by oxidizing RCH=NNHPh 3
- H.
+[RCH=NNPh
RCHN=NPh
I
RCHN=NPh 4
RtHN=NPh]
t )
- I
RC=NNHPh RC=NNHPh 5
(2)
agents, however, such as manganese dioxide, iodine, and ammoniacal silver nitrate.2a It takes place by a free-radical mechanism2a,band gives a complex mixture of products, including uic-bisphenylazoalkanes (4) and osazones (5); the latter arise by tautomerization of the former, as in eq 2. The free-radical pathway shown in eq 2 is not available to formaldehyde dimethylhydrazone because of the absence of hydrogen on nitrogen. Indeed, it has been shown that only N-monosubstituted hydrazones (like 3) undergo oxidative dimerization to osazones (like 5) by the free-radical mechanism3 of eq 2. * To whom correspondence should be addressed.
(1) On leave from the Institute for Atomic Physics, Bucharest, Rumania. (2) (a) J. Buckingham, Q u a f f .Rea., Chem. Soc., 23, 37 (1969), and references therein; (b) C . Wintner and J. Wiecko, Tetrahedron L e f t . , 1595 (1969), and references therein; (c) the first to report such a reaction were F. R. Japp and F. Klingemann, Justus Liebigs Ann. Chem., 247, 190 (1888). (3) H. Minato, H. Tatena, and H. Yokokawa, Bull. Chem. SOC.Jap., 39, 2724 (1966).
An observation more closely related to that of eq 1 is the finding of glyoxal bisphenylhydrazone (5, R = H) among the products of reaction of formaldehyde and phenylhydrazine in the presence of acetic acid, in an early, unsuccessful attempt t o prepare formaldehyde phenylhydrazone (3, R = H).4 That reaction gives a complex mixture containing at least five nitrogen-containing c ~ m p o u n d s including ,~ a small amount of glyoxal bisphenylhydrazone. 4,5 The formation of glyoxal bisphenylhydrazone under these conditions was assumed4bto be a result of autoxidation like that described earlier.2c Our findings with formaldehyde dimethylhydrazone, however, suggest a different interpretation.
Results and Discussion The transformation 1 --t 2 (eq 1) was first observed when an attempt was made to purify a sample of 1 that was contaminated with about 20% 1,l-dimethylhydrazine, by treatment with phthalic anhydride in an amount sufficient to convert the 1,l-dimethylhydrazine to a nonvolatile hydrazide. After brief refluxing, the reaction mixture was distilled (oil bath to 135') for recovery of 1. Recovery was poor (about 5 0 x ) , and the recovered material still contained about 10% 1,ldimethylhydrazine. The nonvolatile residue was a dark brown mixture of liquid and gummy solid. The liquid portion was taken into ether and then distilled under reduced pressure. The product, 2, was obtained as a yellow oil, bp 90-95' (4 mm), and was identified by comparison of its nmr, ir, and uv spectra with an authentic sample prepared by reaction of aqueous glyoxal with 1,l-dimethylhydra~ine.~ In subsequent experiments, 2 was obtained in a small amount upon attempted reduction of 1 (to tri(4) (a) H. von Pechmann, Ber., 30, 2459 (1897); (b) ibid., 31, 2183 (189 8). ( 5 ) E. Schmitz and R . Ohme, Justus Liebigs Ann. Chem., 635, 82 ( 1960). (6) R. H. Wiley, S. C . Slaymaker, and H. Kraus, J . Org. Chem., 22, 204 (1957).
Condon, Fifrcatiu 1 Glyoxal Bisdimethylhydrazone
6626 Table I. Acid-Catalyzed Oxidative Dimerization of Formaldehyde Dimethylhydrazone (l)=
Run
Catalyst (mol/mol of 1)
1 2 3 4 5 6
AcOH (0. 15)e AcOH (0.60) p-TsOH (0.33)' AcOH (0.56) AcOH (0.69) AcOH (0.69)
Solventb
Time, hr
Temp, "C
Nonee None None MeOH n-PrOH 1,4-Dioxane
1 1.75 2.75 3.5 4.25 4.5
110-135 80-100 80-105 70-78 75-103 75-105
-Anal. yield (glpc), mol % 1-Isolated 2, mol %UnPurity reacted 1 8 2c YieWd (glpc) 3 1.5 2 1 1
0 8 7.25 12.1 0.7 0.6
26 49.5 40 47 51 51.5
50 51
95.5 94.2
a Scale: 0.25 mol of 1. Where a solvent was used, the amount was 80 ml/mol of 1. Based on eq 1. The yield is that of pure 2 in the isolated product. For example. in run 5, the isolated yield corresponded to a 52.5 % yield of 2 of 95.5 purity. e Acetic anhydride, 0.25 mol/mol of 1, was also present. Upon mixing of the reagents, a solid salt appeared; upon being heated, it disappeared; then a new insoluble salt appeared.
'
I. The Acidic Catalysis. Presented in Table I are data from experiments seeking to determine the optimum conditions for production of 2 from 1. While acetic acid was the catalyst in most of those experiments, p-toluenesulfonic acid was also effective (run 3), and the use of formic acid and phthalic acid (or its acid hydrazide) was mentioned above. The possibility that the reaction is thermally induced is eliminated by the fact that it took place as well in methanol (run 4), at a temperature at which pure 1 distilled unchanged after being heated at reflux for a similar time. The use of a solvent had little influence on the yield of 2, but it helped to control the temperature, as the reaction is vigorous and exothermic, once started. The yield of 51-51.5x, achieved in runs 5 and 6, with 1-propanol or 1,4-dioxane as solvent, corresponds to 76-77% of theory, with one-third of 1 undergoing reduction, as in eq 3. In Figure 1 is shown the rate of formation of 2 from 1 at room temperature in the presence of various quantities of acetic acid (runs T1, T2, and T3), in 11 the presence of various quantities of water (runs T4 3Me2NN=CH2 --f Me2NN=CHCH=NNMe2 + and T5), in pure 1 without added catalyst (run T6), 1 2 Me2" + [H2C=NH] (polymer) (3) and in pure 1 stored on solid sodium hydroxide (run T7) or calcium hydride (run T8). The rate is clearly dependent on the concentration of acetic acid (runs The results to be presented indicate that this transTl-T3) or water (runs T4 and T5), which would be formation takes place by a mechanism comprising expected to behave as a weak acid in solution i n 1. at least the following two steps: (i) acid-catalyzed Runs T6-T8 show that 2 is produced from 1 at a head-to-head dimerization of 1 to an isolable intervery slow, nearly constant rate in purified samples mediate, 2,2-dimethylhydrazinoacetaldehyde dimethylof 1 (run T6) and in samples of 1 stored on solid sodium hydrazone (S), pictured as in eq 4a-c; and (ii) dehyhydroxide or calcium hydride (runs T7 and T8, redrogenation of 8, with 1 serving as hydrogen acceptor, spectively). This slow formation of 2 cannot be ascribed as shown in eq 5. to the presence of traces of water, because the rates Me2NN=CH2 + HA =+= were essentially the same, while it is unlikely the con1 centrations of water were the same, in the three samples. [Me2NGH=CH, Me2NNH&H21+ A- (4a) It may be the result of autoxidation or a base-catalyzed 6 process with a mechanism involving anions as intermediate~.~ + Me2NNHCH2+ CH2=N-NMe2 --3 11. The Dimerization Step. The head-to-head di6 1 merization of 1 to 8 contrasts with the behavior of formal+ dehyde imines, which ordinarily undergo head-to-tail Me2NNHCH2CH2N=NMe2 (4b) polymerization and give hexahydro-s-triazines.8 AS 7 pictured in eq 4, of the three potentially nucleophilic 7 + AMe2NNHCH2CH=NNMe2 + HA (4) centers in the hydrazone, namely, dimethylamino ni8 trogen, imino nitrogen, and methylenic ~ a r b o n ,the ~
methylhydrazine) with formic acid, and in about 3 5 2 yield from a reaction of 1 (containing about 2 2 each 1,l -dimethylhydrazine and water) with acetic anhydride in a molar ratio of 3 : l . Presumably acetic acid was the catalyst in the latter experiment. In a search for by-products that would account for the two hydrogens lost in the transformation 1 -+ 2 (eq l), reaction mixtures were examined by gas-liquid partition chromatography (glpc). No trimethylhydrazine and no acettrimethylhydrazide were found (see below); but dimethylamine was identified by its retention time in glpc and by its nmr spectrum in ether solution. Also, when the reaction was carried out in the presence of acetic anhydride, N,N-dimethylacetamide was identified in the mixture by its retention time in glpc and by its nmr and ir spectra. In addition, the reaction is accompanied by formation of some tarry, presumably polymeric, material, which is soluble in water, but only partially soluble in ether. Accordingly, the reaction of eq 1 can be written as a disproportionation, eq 3.
8
+1
--f
2
+ Me2" + [NH=CH2] (polymer)
(5)
Evidence in support of this mechanism and some elaboration of its details are presented below. Journal of the American Chemical Society
1 92:22 1 Nouember
(7) J. S. Walia, H. Singh, M. S. Chatta, and M. Satyanarayana, Tetrahedron Lett., 1959 (1969). (8) J. Graymore, J . Chem. Soc., 1490 (1931). (9) S . F. Nelsen, J . O r g . Chem., 34, 2248 (1969).
4 , 1970
6627 - I
. + I
-~E"al 18
2
4
K
0
z
2
0
E 12
€0.5
-e
0
uL
CY;
6
0
0 0
Time, hr
Time, hr
Fig. I C
Time, hr
Figure 1. Rate of formation of 2 from 100 mol of 1: run T1, with 19.44 mol of AcOH ( 0 ) ; T2, with 9.61 mol of AcOH (+); T3, with 5.1 mol of AcOH (A); T4, with 20.25 mol of HzO (0); T5, with 10.21 mol of H 2 0 (0); T6, without catalyst (0); T7, stored on solid NaOH (X); T8,stored on calcium hydride (0).
imino nitrogen is involved in protonation (eq 4a) and the methylenic carbon in dimerization (eq 4b). The result appears t o be a consequence of two factors: (i) protonation at carbon lo gives a substance, Me2N+= NCH3, which is a poor electrophile; and (ii) electrophilic attack of 6 on 1 at the methylenic carbon is aided by electron release by the dimethylamino group as shown in eq 4b. The behavior is consistent with that observed with other hydrazones, which undergo attack at methylenic carbon by electrophiles,2a such as diazonium ions, l 1 chlorosulfonyl isocyanates,l?" and Vilsmeier reagent.12b Although increasing acidity promoted the dimerization (data of Figure l), it is noteworthy that 1 gives a stable hydrobromide, which can be purified at ambient temperatures without alteration. The dimerization clearly requires the presence of both 1 and 6 in the reaction mixture. 111. The Intermediate. Close examination of the curves in Figure 1 shows that some of them (those showing the very early stages of the reaction) exhibit a point of inflection, which suggests that 2 arises from an intermediate. In fact glpc analysis showed the presence of six compounds with retention times between those of 1 and 2 at various times in the reaction mixtures. These are designated for convenience A-F, in order of increasing retention time in gas chromatog(IO) Probably all three sites are protonated reversibly in varying degrees; cJ G. J. Karabatsos, F. M. Vane, R. A. Taller, and N. Hsi, J . Amer. Chem. Soc., 86,3351 (1964). (11) A. F. Hegarty and F. L. Scott, Chem. Commun., 622 (1966); J . Org. Chem., 32, 1957 (1967); F. A. Neugebauer and H. Trischmann, Justus Liebigs A m . Chem., 706, 107 (1967). (12) (a) R. Brehme and H . J. Nikolajevski, Tetrahedron, 25, 1159 (1969); (b) Z . Chem., 8 , 226 (1968). (13) S. Wawzonek and W. McKillip, J . Org. Chem., 27, 3946 (1962).
1,pI :
y
,082 0 200 400 600 800 0 V
0
0
V
Time, hr
,
,
,
I
2
3
me, hr Time,
Time, hr
Figure 2. Rates of formation of products from 100 mol of 1 with 19.44 mol of AcOH (run Tl).
Figure 3. Rates of formation of products from 100 mol of 1 with 9.61 mol of AcOH (run T2).
raphy. Of these, C was the one present in highest concentration throughout most of the reaction (Figures 2 and 3). The same compound was found in significant amounts in the preparative runs conducted for short times or at low temperatures (Table I, 8). As shown in Figures 2 and 3, the concentration of C rose faster than that of 2 in the initial stages of the reaction, reached a maximum, and decreased rapidly during the last stages of the reaction. This is the behavior that would be expected if C were an intermediate to 2. l 4 , I 5 The substance C was isolated by preparative glpc and was shown to be 8. (14) Of the other components of the reaction mixture, only E accumulated faster than 2 in the early stages of the reaction and therefore could be considered a precursor of 2 (Figures 2b, 3). E accumulated more slowly than C, as it would if it were an intermediate subsequerzr to C. It disappeared completely from the reaction mixture, however, while the concentration of C was still increasing (Figure 2c), as it might if it were an intermediate preceding C. Furthermore, in the runs at low acidity (water as catalyst), E developed later than the final product, 2. (This is not shown by the figures given.) It is concluded, therefore, that E is an intermediate in a side reaction. The same appears to be true of the other components, A, B, D, and F. The small concentrations of these minor components precluded their isolation and identification. (15) Of the runs from Figure 1, only runs T1 and T2 are presented in detail in Figures 2 and 3 . The other runs gave curves with a similar pattern.
Condon, Fircaziu J Glyoxal Bisdimethy [hydrazone
6628 Table 11. Acid-Catalyzed Oxidative Dimerization of Formaldehyde Dimethylhydrazone (1) in the Presence of Other Hydrazine Derivativesa Mol of AcOH/mol Run of 1 7 8
1.12
9 IO
1.12
11e 12' 13 c
1 .38
i.12 1.70 1.68 1 .70
- Added hydrazine deriv -
Solventb
Formula
None None None 1,4-Dioxane 1,4-Dioxane 1,4-Dioxane 1,4-Dioxane
MesNN=CMen MepNN=CMe2 MeyNN=CMe2 MeyN-NMe? Me2N-NH2 Me2N-NH2 Me2N-NH2
Mol/mol Time, of 1 hr 1.0 1.0 0.6 1.0
3.5 3.75 4
1.0 1.0
4 5.5 6
1.0
Temp, "C 75-85 75-100 75-102 75-105 75-102 75-103 75-105
6
-Anal. yield (glpc), mol 1-Isolated 2, mol %UnPurity reacted 1 8 2c Yieldcvd (glpc) 12 6.5 2.8 1 220 12 6
See Table I, footnote a. See Table I. footnote b. See Table I, footnote c. d See Table I, footnote d. runs was dark yellow or light brown and transparent; in all the others. it was black.
The nmr spectrum of C (6 2.27, singlet, 6 H; 2.78, singlet, 6 H ; 3.07, doublet, 2 H ; and 6.63 ppm, triplet, 1 H, J = 5.5 cps, in CDCI, with internal TMS) and its ir spectrum were in agreement with structure 8. The substance C was identical in all respects with a specimen of 8 synthesized by reaction of I , I-dimethylhydrazine with chloroacetaldehyde, eq 6, a method previously
+
3Me2NNH2 CICH2CH0 + Me2NNHCH2CH=-NNMel 8
+ MelNNH,+C1- + H?O
(6)
used for similar compounds.16 Finally, C was transformed into 2 (although i n low yield) by treatment with acetic acid under conditions similar to those used for the conversion of 1 to 2 (see Experimental Section). IV. The Dehydrogenation. For conversion of 8 to the final product 2, several paths were considered, even though the step is like a corresponding step in osazone formation. li a. A mechanism involving hydride transfer from 8 to 6, as in eq 7, appears to be ruled out by the fact Me2NN=CHCHy '
Me2NNH
6
8
+
Me2NN=CHCH=NHNMey 2,H+
-+ Me2NNHMe
+
MeJH
+
[H.X=CH,]+
(7)
(8)
out by the results of experiments in which other hydrazine derivatives were tested as hydrogen acceptors. These results are presented in Table 11. (16) H. Simon, G. Heubnch, and H. Wacker, Chem. Ber., 100, 3101 (1967). (17) H. Simon, G. Heubach, and H. Wacker, ibid., 100, 3106 (1967); H. Simon and W. Moldenhauer, ibid., 102, 1191 (1969). (1s) G. J. Bloink and I