1878
nARNETT COIIEN, ERVIN
Ti. VAN ARTSDALENAND JOSEPII HARRIS
VOI.
74
I
remained constant for the hydrolysis of the first ethylenitiionium ions of ilie compounds studiecl ’Therefore we have no evidence for the participatioii (Jf hydroxyl ions in these cases; although we did find this to be apparently the case for the second 2 .4! E1 ion of inethyl-bis-( P-chloroethy1)-amine. Hence, 2 4.0 we must dwmie generally a direct interaction of me I HzO through the polarized C-N linkage to produce a substituted ethanolamine and a proton. SThere the reaction is exceedingly slow, or the concentration of hydroxyl ions is appreciable, or both, then I 4.2 the effect of the latter ion is superposed. It has been pointed out that the high reactivity of the ti [I 6.2 6.4 E1 ion is the combined result of the strain of the 3PK‘.. membered ring and the polarizing effect of the Fig. 1.-Relation of pK’, of honiologous parent amines to the hydrolysis rate of their cyclic ethylenimoniuni ions a t positive charge on the C-N linkage.3b The ion is thus very reactive toward any nucleophilic agent I 37’: RR’+?;CH&Hz, R = CHsCH2C1; R’ = Me = capable of attacking the carbon of the 3-membered methyl; E t = ethyl; %-BU= n-butyl; n-Pr = 12-propyl; ring. The energy of activation of the hydrolytic proci-Pi- = isopropyl. ess lies from 13.4 to 21.4 kcal. (Table I), values From the slope shown in Fig. 1, they should be rather less than that for cyclization, which is about + i l -33 kcal.? expected for the formation of a polar, RR‘NCHzCH2 R’ = CHZCHzCl solvated intermediate, the entropies of activation R = Et n-Bu X-Pr i-Pr were all large negative values7 This accounts for APK.’ = 0.03 0.15 0.20 the relative slowness of the reaction, despite the Mechanism of Hydrolysis of the Ethylenimon- inore favorable energy of reaction. From the data presented here, one can readily ium Ion.-The E1 ion may be regarded as an intramolecularly stabilized carbonium ion and, as such, select a series of 4 compounds with increasing Pit can undergo the reactions characteristic of the chloroethyl substitution and observe that there is C+ ion. Among these are reactions with anions, a considerable increase in the hydrolysis rate e.g., C1-, OH-, ,5207, as well as with uncharged associated with a decrease in the electron density electron-donors such as RDN, etc. The polar of the ring carbon under attack. nature of the E1 ion indicates that it is undoubtedly ( 7 ) S Glasstone, K J Laidler and H Eyring, ”The Theory of Rate Processes,” lfcGran-IIi11 Rook L o , Inc , New York, N Y , 1911 highly solvated in water. A t pH levels betweerr 3.0 and 8.5, the rate values BALTI3IORE 5 , hfARYL.+Nn RECEIVFD SFPTEVBER 17, 3%1 3.8
I
I
I
I
I
-
[CONTRIBUTION FROM
THE
DEPARTMENT O F
CHEMISTRY, SCHOOL UNIVERSITY ]
PHYSIOLOGICAL
OF
MEDICINE, THEJOHNS Hoperrs
Reaction Kinetics of Aliphatic Tertiary 0-Chloroethylamines in Dilute Aqueous Solution. 111. Reactions of the Ethylenimonium Ion with Certain Anions1 B Y BARNETT COHEN,
ERVINR.
ITAN XRTSDALEN AND JOSEPH
HARRIS
The kinetics of the reactions between thiosulfate, bicarbonate, chloride, propionate and benzoate ions with the first ethylenimonium ions of several aliphatic tertiary 8-chloroethylamines was studied and the rate constants of these reactions established. Variation of ionic strength causes pronounced changes in rates which can be explained quantitatively by the Bronsted-Debye relationship only if the ethylenimonium ion is a singly charged cation. Rate of the reversal of the initial cyclization of the parent amine by reactions of the ethylenimonium ion with chloride ion was investigated, and from this i t was shown that this initial cyclization in dilute aqueous solution is over 99.5Yc complete.
The ethylenimonium ion produced in the cyclization of an aliphatic tertiary P-chloroethylamine acts like an internally stabilized carbonium ion, consequently a great variety of reactions is possible. In this, the final paper of the series, we report studies on the kinetics of the interaction with SzOrl HCOr, C1-, propionate and benzoate t o form the corresponding esters. By suitable manipulations, we were able to isolate the reactions and examine many of them in considerable detail. (I) The work described in this report was performed in part under a contract recommended by the National Defense Research Committee, between thc‘Office’of Scientific Research and Development and the Johns Hopkins University.
After the initial reversible cyclization , 2 the kl
RR%”iI,cTI.ei
2 rm‘xcrT:cIr2+ ci+
1
-
7
(1 I
k i
I
I1
(EI’I
reactions of I1 in water may be represented schematically as3 (2) B. Cohen, E. R. Van -1rtsdalen and J Harris, THIS J O U R N A L 70.. 281 (1948). . ~, (3) C. Golumbic, J. S.Fruton and M. Bergmann, J . Org. Chcm., 11, 518, 543, $50 (1916); (bl P. D. Bartlett, S. D. Rose and C. G. Swain. ’ m S J O U R N A L , 69, 2971 (1917).
April 20, 1952
REACTIONS OF
THE
ETHYLENIMONIUM IONWITH CERTAINIONS
1879
where kP is the rate constant a t zero ionic strength, A is the Debye-Hiickel constant5 (0.4977 a t 2 5 O ) , I11 zi and zzare the charges on the reacting species, and ks I’/2 is the ionic strength. For the reaction between 11 + 111 --+RR~NCH~CH~~(RR’)CH~CH~OH (3) S20z and EI, the slope of the plot relating log I1
+ Ha0 + RR’NCHzCHzOH + H + kh
(2)
to fla t 15’ should be -1.991 if the E1 is indeed a singly charged positive ion as postulated, and it is clear that this is the case as seen from Table (5) (kx) 111, last two columns. To save space, we dispense with a diagram of log Hydrolysis (2) with evolution of acid, represents an example of the reaction with compounds of the kobs vs. which shows that the data of Table I11 type HX, e.g., HzO, RSH, and was considered in conform almost exactly to the relations involved in the preceding paper4; equations (3) and (4) are the reaction of thiosulfate ion with a singly charged representative of the general reaction with electron cation (ethylenimonium ion) and none other. Table I1 gives the data for the E1 of ethyl-bisdonors of the type RX, e.g., RIN, R2S; and these processes can be avoided by operating at sufficiently (P-chloroethy1)-amine and shows the good conlow PH where the proton density interferes. At cordance with theory. On the last line of Table present, we are concerned with equation (5), repre- I1 is included an example in which I’/2 was built sentative of a general reaction with anions. As the up with Na2S04. It was not the best choice because reaction with C1- is the reverse of equation (l), of specific effects but its effect in bringing KO, within this enables us to compute a theoretical equilibrium fair approximation to the theoretical value serves as additional evidence for the predicted influence constant for this reversible process. upon the reaction process. Results and Discussion TABLE I1 The Thiosulfate Reaction.-A typical series of EFFECT OF IONIC STRENGTH ON REACTION RATEOF SO3= + observations is given for the E1 of isopropyl-bisP-CHLOROETHYL)-AMINE; PH 4.00, 15’ (8-chloroethy1)-amine in Table I. Similar deter- E1 OF ETHYL-BIS-( Initial concn., Average minations were made on the first ethylenimonium koba. kP mmole/liter ions of the n-propyl, ethyl, methyl homologs and of &OrE1 l./mole min. tris- (0-chloroethy1)-amine. 83.3 116 0.99 7.03 0.99 I1
+ I +linear and cyclic dimers I1 + C1- K-! RR’NCH~CHZCI kd
kobs
(4)
K G
111 8.83 74.1 1.98 1.00 TABLE I 8.90 76.3 115 1.98 0.99 FIRSTE1 OF ISOPROPYL-BIS-( ~~-CHLOROETHYL)-AMINE 4.96 2.50 13.85 63.0 119 SO; 60.8 116 2.50 14.07 4.96 (Rate of formation of the monothiosulfate) 0.252 m o l e of 15.11 1.98 1.00 61.9 (124) amineHC1 4- 0.277 mmole NaOH in 95 ml. of H20 held a t Av. 116 PH 9.0 for 25 min. Adjusted t o pH 4.0 with 0.0082 ml. of 2.473 N HCI. At zero time added 5.0 ml., 0.0991 M Nay Table I11 summarizes the data on the thiosulfate &OS;initial [EI] = 0 2 2 5 3 7 M; initial [SO,-] = 0.004955 reaction with the E1 of the 5 compounds examined M;pH4.00, 15’; dr/2 = 0.139; (at O’, kt = 38.8; E on a uniform basis of temperature, pH and zero 13,300 cal. at = 0).
+
-
Time, min.
2.473 N HCl ml. X 100
0.37 .67 .95 1.32 1.63 2.04 2.67 3.43 3.98 4.37 73.0 130.0
2.19 3.47 4.47 5.43 6.10 6.87 7.60 8.29 8.63 8.88 11.57 11.57
Reaction,
%
21.4 33.8 43.6 52.9
kt
l./mol;min.
139 131 139 137 137 137 139 135 135 133
59.5 67.0 74.1 80.8 84.1 86.5 100
...
...
100 Av.
136
Effect of Ionic Strength.-The rates of reaction between ions are strongly influenced by the ionic environment. This primary salt effect which involves activity coefficients can be described for dilute solutions by the well known equation (5a) derived from the Debye-Hiickel theory of electrolytes. log kob. = log k: 4-2 A Z I Z ~ ~ (W~ (4) B. Cohen, E. R. Van Artsdalen and J. Harris, THIS 72, 1875 (1952).
JOURNAL,
ionic strength. It should be noted that the rate values for the E1 of tris-( ,&chloroethyl)-amine are approximate; nevertheless, its reaction rate stands out as very much greater than those of the other compounds studied. TABLE I11
SUMMARYOF RATEDATAREFERRED TO Reaction: E1 E1 from 8-chloroethylamine
+ &Or
monothiosulfate;
Initial concn. mmole/liter EI S~G
=0
PH 4.00, 15’ k:,
koba,
Average
di72
liter/ mole min.
liter/ mple min.
1.28 1.98 0.101 ... 2500“ 1.33 1.98 .lo0 ... Isopropyl2.51 4.96 .139 138.9 260 bis2.54 4.96 ,139 136.4 n-Propyl-bis- 2.50 4.96 ,139 84.9 160 2.49 4.96 .139 84.2 140 Methyl-bis2.40 4.96 ,139 71.0 2.44 4.98 ,139 76.6 Ethyl-bis2.50 4.96 .139 63.0 118 .141 60.8 2.50 4.96 a Extrapolated from Oo, assuming 13,300 cal. for activation energy (see Table I). Tris-
(6) H.S. Harned and B. B. Owen, “The Physical Chemistry of Elec. trolytic Solutions,” Reinhold Publ. Corp., New York, N. Y.,1944, p. 119.
BARNETT COHEN, ERVINR. VANARTSDALENAND JOSEPH HARRIS
1880
Vol. 74
Reaction of E1 with CI-.-In dilute solutions of the E1 of the bis-(8-chloroethy1)-amines (0,001 to 0.005 M), the back reaction is negligibly small as judged by the great uniformity of the hydrolysis constants obtained in such solutions. To secure an observable effect, it was necessary to operate with added NaCl a t high levels. An example of a determination is given in Table VI which shows the kh 4kL. data obtained at [Cl-] = 0.340 M with the E1 of TABLE IV methyl-bis-( P-chloroethy1)-amine, the only comHCOI- f E1 OF METHYL-BIS-( 8-CHLOROETHYL)-AMINE pound studied in sufficient detail. Initial [EI] = 0.0025 N; 0.0002% phenol red; initial [HCOa-] = 0.025 M ; NaOH to pH 7.6; PH 7.6; 37";
The Bicarbonate Reaction.-The rates of disappearance of E1 in the presence of bicarbonate were followed in five of the homologous compounds. The concentration of bicarbonate, 25 mmolar, was selected because i t is comparable to the concentration in blood serum. An example of the results obtainedlis given in Table IV, where k;bs =
TABLE VI
average I'/2 = 0.027 EL
Time, min.
70 89.7 76.0 63.8 49.4 38.3 27.0 19.8 12.3 9.4
25 50 80 123 165 215 275 3 53 391
REACTION OF E1
dbs
IO' X m h - 1
...
6.7 5.8 5.9 6.1 7.0 5.2 6.2 7 0 Slope 6.2 Replicates, 6.2; 6 . 4
A control experiment in which NaN08 was substituted for NaHCOa gave khbs = 0.0052, within the experimental error of 0.00503 for kh in water alone as previously4 reported. Therefore, the observed difference is not due to a simple salt effect. I n 10 m M bicarbonate, under the same conditions kAbs was found to be 0.0054 inin.-'. The results obtained with the five compounds are given in Table V. The data in Table V are concerned with one level of ionic strength (0.027) only; as the lack of materials prevented further study. The last column in the table gives the second order rate constants, kb. A calculation of the ratio kL/kl: for each of the compounds gives values of the same order of magnitude. This wou1d:'seem to be an indication that the mechanisms of the hydrolysis and bicarbonate reactions may probably be similar dynamically. TABLE V
SUMMARY OF RATEDATAIN 25 M M BICARBONATE pH 7.6; 37'; average I'/2 = 0.027
Isopropyl-
10.00
6.80 n-Propyl6.25 Methyl5.09 n-ButylEthyl4.87 * See Table I of ref. 4 .
7.57 5.33 5.03 4.02 3.73
2.43 1.47 1.22 1.07 1.14
9.72 5.88 4.88 4.28 4.56
The apparent instability of the carbonic ester suggests that the presence of bicarbonate may provide a more rapid by-pass for the decomposition of E1 than is offered by the much slower reaction with HzO. That is, hydrolysis of the ester would accomplish the same end result much more rapidly. Our knowledge of these compounds suggests this possibility.
O F METHYL-BIS-( 8-CHLOROETHYL)-AMINE
WITH C1- AT pH 3.6, 37" 0.3755 mmole amine.HC1 f 123.9 ml. of H20 containing 0.38 mmole of NaOH. Aged 15.2 min., during which 0.034 mmole of NaOH was required t o hold PH a t 9.0. During the next 3.2 min., the mixture was acidified to pH 3.6 with 0.069 mmole of HCI; and 50 mmoles of NaCl added ( = zero time); total vol. = 149.6 ml.; [Cl-] = 0.340 M .
Time, Illlll.
k-i
ELa
Amine,
..
3.69 6.78 11.04 12.60
70
%
IO9 X min.-i
Time, min.
18.9 22.3 .. 26.3 .. 30.0 72.3 . . . . . 33.5 . . 16.24 1.77 38.1 . . 20.42 1.77 43.1 . . 22.83 1.77 48.5 slope, log % DS. ,time (Eq. 6) 0.0175 = 0.0036 = kh. 2.2 4.1 6.7 7.7 8.5 10.2 13.2 15.1
..
1.72 1.73 1.77 1.77
59.1
.. ...
46.6
... ...
...
..
31.14 1.75 35.18 1.74
...
..
41.47 1.73 45.11 1.72 35.05 Average 1.75 32.13 = 0.0211; 0.0211 -
In the concentrated NaCl solutions a t pH 3.6, the hydrolysis rates of E1 are lower than in the absence of NaCl and the values of ki, appear to pass through a minimum, presumably the result of an effect of increasing ionic strength upon the activity of the water. These hydrolysis rate data, obtained as differences between large numbers are not sufficiently precise for further analysis. The effect of ionic strength on this reaction is best described by the following approximation derived from the Debye-Huckel equation for the activities of ions in solution log
K-1
log
k0-1
-2 S m 2 f
Br/2
where B is an empirical constant and S = 0.518g6 a t 37'. Values of B were calculated from the simultaneous equations by substituting the five experimental values in the above equation. The average value was B = 0.553 =k 0.048. This value was substituted into each of five possible equations corresponding to experimental points to give log kLl = -0.875 f 0.008. This corresponds to k!Ll = 0.133 f 0.003 for the bimolecular rate constant a t r/2 = 0. A plot of log k-l against ( 2 . ~ ~ 5 7 3 B r / 2 ) shows a good linear fit of the experimental data to the theoretical line. I n Table VII, attention is directed to the progressive rise in first order rate constants, KI,with increasing [CI-] and to the progressive fall in the second order constants, k-1 as accounted for by the effect of increasing ionic strength.
+
April 20, 1952
REACTIONS O F THE ETHYLENIMONIUM ION
WITH
CERTAIN IONS
1881
TABLE VI1 TABLE VI11 METHYL-BIS-( p-CHLOROETHYL)-AMINE REACTIONS OF PROPIONATE HzO E1 OF METHYL-BISWITH CL( 8-CHLOROETHYL)-AMINE, PH 7.4, 37" Rate constants a t different [CI-1, PH 3.6, 37' Ratio k'obs., pooPk
REACTIONO F E I
[a-I
r/2
+
OF
kL 10: X m h - 1
Reaction with C1k- I IO: X min.-l l./mole min.
..
0 .. 0.005 5.02 .072 (3.0?) .139 3.4 .206 3.7 .340 3.6 .520 4.4 a Computed, see text.
*.
5.5 9.4 11.7 17.5 24.0
0.133" .114' .076 .068 .057 .052 .046
Practically the same value for kW1 in Table VI1 calculated for I'/2 = 0.005, namely, 0.11, can be obtained by introducing the experimental values k1 = 0.42, t = 1.65, a = 0.005, x = 0.0025 together with trial and error substitutions for k-1 in the following unwieldy theoretical equation for the E1 course of the reversible process, Amine
c1-
'
E--=x
2
'-' A coth kl
k-1
lEIl/[Pro~.l,
M
Average r/2
0.00125/0.0251 .001235/0.0248 .0025/0.0499 ,00235/0.0235 .00251/0.0149
0.166 .171 .234 .167 .140
kL 1
tA, where A'
+ k = 2 ( a - &) k-1
It has been shown3b in acetone-water solutions that k-1 of this E1 and of its ethyl homolog are sensitive to changes in ionic strength of the medium. The value of K-1 = 1.4 in acetone solution a t 25' is of a higher order of magnitude than that here recorded (0.13 to 0.05) for aqueous solutions at 37'. Theoretical Equilibrium.-For the initial reversC1-, we may now calible process: Amine @ E1 culate a theoretical equilibrium constant to show the extent of the process. This represents a largely theoretical calculation, because usually the conditions favor the further decomposition of E1 to shift the process to completion. Assuming an upper limit of 0.005 M for a "dilute" solution of methyl-bis-( P-chloroethy1)-amine a t 37' with the appropriate experimental constants, (Table VI1 and ref. 2) , we have
+
+
liter per mole min.
min. '0' 1'
7.55 6.87 9.43 6.72 6.28 Av.
1.51 1.13 1.55 1.09 1.20 1.3
TABLE IX REACTIONS OF BENZOATE HzO E1 OF METHYL-BISPH 7.4, 37" (&CHLOROETHYL)-AMINE,
+
+
Ratio [EIII Benz.1,
L
0.00249/0.0250 .0025/0.0219 .00257/0.0250
0.173 .170 .175
1.2
6.96 6.94 7.11
1.2 1.2
mechanism of these reactions. The initial acid production in the reactions with propionate and benzoate ran according to prediction, but the later stages indicated rather definitely that, certainly in the case of propionate, the monoester was disappearing and, indeed, experiments with the dipropionic ester showed that this was the case. Experimental Materials.-The ethylenimonium compounds studied were produced under proper conditions from the pure tertiary p-chloroethylamines by methods already described. Unless otherwise specified other reagents were purified or A.R. grade materials. Rate Measurements with &O,'.-Consider the following series of reactions of a dilute aqueous solution of a bis-(Bchloroethy1)-amine, [R' = fkhloroethyl]
+
RR'NH.CHtCHzCI Ia
I
RR'NCHzCHzCI
+H+
A
I
ki +n 1 'RR'NCHzCHr + C1-
B
E1
from which one finds that a t equilibrium, [Amine] = 6.8 X M, corresponding to 99.86% conversion to E1 and C1-. Estimates on the other homologous members of the series of aliphatic t-bis- (P-chloroethy1)-amines indicate that they all go essentially to completion in this reaction. Reactions with Propionate and B e n z o a t e . 4 n l y one of the compounds was studied in these processes, namely, the E1 from methyl-bis-(P-chloroethyl)-amine, the object being to establish a tentative base line. At PH 7.4 and the relatively low concentrations of sodium propionate and benzoate there was no detectable effect on the hydrolysis rate, KL = 0.0050. The observations are condensed and summarized in Tables VI11 and IX. It will be observed that the second order reaction rates, K O , of this E1 with C1-, propionate or benzoate a t zero ionic strength are substantially alike, namely, 0.133, 0.13 and 0.12 liter per mole min. This would indicate that steric factors among these various anions seem to play a negligible part in the
+ HsO k h RR'NCHzCHeOH + H + kt E1 + &Or + RR'NCHZCHISZO;
E1
---f
C1 Cl
IV
IV
+H+
RR'HI;CHzCHzSzOi D V Starting with the amine hydrochloride, I., suitable adjustment of the PH of its solution will produce the free amine, I . The latter, on sufficient ageing, canTbe converted quickly and practically completely to its first ethylenimonium ion, EI. Hydrolysis can be estimated from previously determined constants.' If now a known excess amount of thiosulfate is added t o the solution, the monothiosulfate, IV, will be formed a t a bimolecular rate and if, further, the solution is sufficiently acid, the free base, IV, will be converted almost instantly t o the stable ammonium ion, V, thereby blocking subsequent cyclization of the second (R') 6chloroethyl group. Hence, successive determinations of the acid uptake in processes Q and D a t a suitable fixed PH can be used as a measure of the rate of formation of the thiosulfate ester in G. Process Cl is negligibly small by comparison in the compounds under study. The above principle presupposes a knowledge of the PK: of the monothiosulfate formed. It is obvious that the mono-
1882
BARNETT COHEN, ERVINR.
VAN
thiosulfate must be more basic than the parent p-chloroethylamine, and knowing the latter, one can fix on a level of pH a t which the monothiosulfates can be completely titrated acidimetrically. Hence pH 4.0 was selected because the p K i of all these bis-( 6-chloroethy1)-amines is greater than 6.0. Tris-( 6-chloroethy1)-amine involves several complications and will be considered separately. The solution of bis-( /3-chloroethy1)-amine was aged at constant temperature and sufficiently alkaline pH to permit 98-100% conversion to its first ethylenimonium ion. pH was measured by means of a glass electrode and controlled by the addition of 2 N NaOH from a microburet until conversion was completed (processes A and B, above). A microburet containing 2.5 N HC1 replaced the one with SaOH, and the pH was accurately adjusted to 4.00 & 0.03. Then with vigorous stirring, a predetermined amount of thiosulfate solution, a t the same temp., was added from a calibrated syringe, the time of half-delivery (about 0.03 min.) of the thiosulfate being taken as zero time. Hydrochloric acid was then added a t short successive intervals to maintain pH 4.00, and the times and acid volumes were recorded. The validity of the method is shown in Table I . Here the reaction was allowed to go to completion and the pH was found to remain constant until the run was terminated an hour later. The errors in these runs was estimated to be between 5 and IO%, though reproducibility of rate constants was about 2 or 3%% The reaction of bis-( p-chloroethy1)-ethylenimonium ion [from tris-( p-chloroethy1)-amine] with SZOS" is very rapid. It cannot be isolated completely owing t o a combination of unfavorable properties associated with the pK: and the reactions of the parent amine, the cyclical imonium ions and their thiosulfate esters. However, the initial part of the process can be followed to an extent sufficient to give a fair estimate of the magnitude of the reaction rate. The experimental procedure follows: 0.211 mmole of the hydrochloride of tris-( 8-chloroethy1)-amine 0.219 mmole of NaOH in 98 ml. of H20 were aged a t 25', p H 7.6 (tib = 2.4 min.) for 17 min. The solution was then iced t o 0 (28 min.), after which it was adjusted t o pH 4.0 with 0.0155 ml. of 2.473 N HC1. At zero time 2.00 ml. of 0.0991 M Na&03 was added, and the solution titrated to #H 4.0 with 2.473 N HC1. Table X gives the results. The initial concentration of the bis-( P-chloroethy1)-ethylenimonium ion, E I , was computed on the simple, but not accurate, assumption that its hydrolysis was the only interfering factor. It is emphasized t h a t the data of Table X are an approximation (probably too low) of the magnitude of this reaction rate. TABLE X FIRSTE1 OF TRIS-( @-CHLOROETHYL)-AMINE (Rate of formation of the monothiosulfate); initial [EI] = 0.001283 M; pH 4.0, 0'; initial [ & 0 3 " ] = 0.001982 M; r/* = 0.101; (see text for comment).
+
ruin.
2.473 .?i HC1, ml. X 100
Reaction, %
0.36 .83 1.19 1.68 2.18 2.57 3.15 3.94
1.28 2.33 2.90 3.48 3.90 4.16 4.47 4.76
24.7 44.9 55.9 67.2 75.2 80.3 86.2 91.8
Time,
kt I./mole'min.
435 436 445 462 477 (495) (528) (581) Av. 451 Dupl. 443 447
Reaction with HCOc, Propionate or Benzoate.-The principle employed was substantially the same as for the re-
ARTSDALEN AND
JOSEPH
HARRIS
Vol 74
action with & O r , with several changes in detail. With HCOa-, the product would be the monocarbonic ester which is presumably insufficiently stable for isolation.8 Other things being equal, when the simultaneous reactions with HzO and with HCOa- are independent and of the same order, the proportion of E1 consumed in the hydrolysis will be kh/(kl: k i ) , and that reacting with HCOa- will be k6/(k; kb). (Primes on the k's indicate first order rates, with water and bicarbonate in sufficient excess.) Then the ratio k(/kl: represents the relative reactivities. Observations were made a t p H 7.6, 37". The initial concentration of ( @-chloroethy1)-aminewas 0.0025 M; that of the anion varied from 0.015 to 0.05 M . The atmosphere over the reaction mixture with HCOa- was not equilibrated. Measurements of the chloride evolution a t pH 7.4 or 7.6 from the second rapid cyclization were taken as indications of the progress of the sum of the very much slower reactions with HIO and anion. NaOH was added from a microburet t o hold the pH as indicated and to measure the apparent rate of acid formation. The samples were quenched in 0.014 N HNOz and the C1- then determined potentiometrically.B The Back Reaction with Cl-.-Consider the simultaneous reactions C1 and CZ above, and substitute C1- for &Or in CZ. In dilute water solution at pH 3.6, the E1 from process B will produce the chlorohydrin and an equivalent amount of H f t o block its further cyclization; and with NaC1 in relatively large excess, titratable free base will be produced as the parent (@-chloroethyl)-amineis reformed. Then a determination (by SO,' titration) of the E1 concentration a t time t will give a - x , where
+ +
x = a(1
- e-(Ei
+ kilt)
(6)
( a = initial concn. of EI, and x = amount decomposed) from
+
which the sum k i t kb can be computed. Also the amount of parent amine a t any time can be determined acidimetrically. Since the two simultaneous processes are of the same order, the proportions of the products remain constant, and the theoretical yield of parent amine will be
.-. Hence [Amine]$ = ( k L l / k i l
+ kh) a(1 - e - W I + k L ) i )
(7) Therefore, kI-1 and k,,' can be computed from equations (6) and (7). The experiments were performed only with methyl-bis(8-chloroethy1)-amine at 37", PH 3.6. Stock solution of the amine hydrochloride was added to water containing 1 equivalent of NaOH. The pH was held near 9 (glass electrode control) until practically all of the amine had cyclized. The solution was then carefully acidified with HCl to PH 3.6, after which NaCl solution was added ( = zero time) to fix the [Cl-] a t a high, predetermined level. The hydrolysis rate of E1 a t PH 3.6 in the absence of added C1- was determined separately and found t o be 0.00502 min.-l, the same as at pH 7.4. Thereafter, careful maintenance of the pH a t 3.6 by successive small additions of HC1 from a microburet gave the data for calculating the amounts of parent amine produced. At intervals, 25-ml. samples of the solution were removed and analyzed for residual EI. This was done by allowing the sample t o react for 15 min. with excess thiosulfate a t #H 3.6 and determining excess SzOa' iodometrically. The iodometric titration was reproducible t o 1 part in 400. The reaction between E1 and SOa' reached the same endpoint when the mixture was held as long as 45 minutes. The amounts of E1 thus determined yielded a uniform first order rate curve. BALTIMORE 5, MARYLANDRECEIVED SEPTEMBER 21,1951 (6) D. A. MacInnes and M. Dole, THISJOURNAL 61, 1119 (1929).
