*
:
4’
The University of Chicago.
Founded by JoBN D. Rock EFELLER.
Studies in Catalysis.
. The Catalysis of Imidoesters.
º
ſ
A DISSERTATION
SUBMITTED TO THE FACULTIES OF THE GRADUATE
SCHOOLS OF ARTS, LITERATURE, AND SCIENCE,
IN CANDIDACY FOR THE DEGREE OF DOCTOR
OF PHILOSOPHY.
DEPARTMENT OF CHEMISTRY.
By HERMANN I. SCHLESINGER.
;
*
i
;
s
ſº
A.
-**
.
EASTON, PA.:
PRESS OF THE ESCHENBACH PRINTING CO.
!
r
* ..., ".
tº r
§xºgº.
wº.
§§-ºº::s.
f****
****
* *
3 =
*
~gº."
38 &
23.
$º:
* * * * *g
* * * *
* ,


Chemical Library
&\\
S C \
, sº ºvº
The University of Chicago.
Founded by JoHN D. RocKEFELLER.
Studies in Catalysis.
VI. The Catalysis of Imidoesters.
A DISSERTATION
SUBMITTED TO THE FACULTIES OF THE GRADUATE
SCHOOLS OF ARTS, LITERATURE, AND SCIENCE,
IN CANDIDACY FOR THE DEGREE OF DOCTOR
OF PHILOSOPHY.
DEPARTMENT OF CHEMISTRY.
* *
, \ . .
#
* .
\
, -,
By HERMANN I*SCHLESINGER.
EASTON, PA.:
PRESS OF THE ESCHENBACH PRINTING CO,
1908
Studies in Catalysis.
VI. The Catalysis of Imidoesters.’
s-ºr-mº
PART I.
The Velocity of Saponification of the Alkyl Imidobenzoates
tn the Presence of Acids, and Their Degree of Ioniza-
tion.
In the well-known catalysis of ordinary acid esters by means
of acids, according to the equation
RCOOR" -- HOH = RCOOH + HOR' (I),
the “catalytic” functions of the acid are considered by Stieg-
litz” to consist simply in the formation of a salt of the ester
and, thereby, in the increase of the concentration of the posi-
tive ion, which is considered to constitute the “active mass”
of the ester. This increase, on account of the small original
ionization of the ester, is directly proportional to the salt
formed and, hence, to the acid used in the reaction. For the
experimental investigation of the correctness of this theory,
1 Articles I.-V., Am. Chem. J., 39, 29, 166, 402, 437, 586.
2 This article was accepted as a dissertation in July, 1905, and sent to the editor
in September, 1907.
8 Report of the Congress of Arts and Science, 4, 278 (1904); Am. Chem. J., 39,
29 (1908).
4.
the closely related group of imidoesters, RC(NH)OR', was
employed, since for these substances the requisite experimen-
tal and mathematical data could be obtained. These esters
form well defined salts with acids; the rate of their decompo-
sition with water, according to the equation
RC(; NH)OR' + HOH = RCOOR" -- NH, (2),
is accelerated considerably by the addition of an acid. The
investigations carried out in this laboratory by Derby and Mc-
Cracken' showed that, as a matter of fact, the velocity of de-
composition of the imidoesters in this direction is directly
proportional to the concentration of the imidoester salt pres-
ent at any moment. As the experiments were carried out
in dilute solutions (one-tenth to one-fortieth molar) in
which the salts are almost completely ionized, the conclu-
sion was drawn that the “active mass” of the imidoester
would almost surely prove to be, not the total Salt concentra-
tion, but only the ionized portion of the salt. With the pur-
pose of testing rigorously this important point in the conclu-
sion, the following investigation of the relation between the
velocity of saponification of the hydrochlorides of the alkyl
imidobenzoates and their respective degree of ionization
was undertaken by me at the suggestion of Professor Stieg-
litz and carried out under his direction. The results ob-
tained, in solutions varying from half molar to twentieth
molar, not only showed the correctness of the conclusion
that the substance reacting with water according to (2) to
give an acid ester and ammonia actually is the positive ion
of the salt, but, further, brought to notice another most strik-
ing parallelism in the behavior of the imidoesters and ordi-
nary esters towards water. As is well known, the catalysis
of the esters, as well as of cane Sugar and similar substances,
is increased by the presence of neutral salts, which produce
the so-called “salt effect” studied by Spohr,” Arrhenius,”
Euler,” and others. This “salt effect” was found by these
* Loc. cit.
2 J. prakt. Chem. [2], 33, 270.
3 Z. physik. Chem., 4, 226 (1889).
* Ibid., 32, 348 (1900).
5
investigators to be proportionate to the concentration of the
salt added, except in the case of very dilute solutions. The
results obtained by me indicate that the “catalysis” of imido-
ester salts is subject to exactly the same kind of “salt effect,”
except that, in this case, the Salts involved in the reaction,
'viz., the Salt of the imidoesters and the ammonium chloride
formed in the decomposition, themselves produce the accelera-
tion—an “auto effect,” as it might be called.
The “salt effect” is produced most markedly on substances
which are but little ionized," and the change in the speed of
a catalysis in which water is primarily involved is ascribed
particularly to the increase in the “active mass” or the dis-
sociation of water itself in the presence of salts.” This ex-
planation would account perfectly also for the “salt effect”
in the decomposition of the imidoester salts—since the active
mass of the water is one of the factors in the velocity formula
(see equation 2).
In view of the facts established, then, since the decomposi-
tion of an imidoester into an ordinary ester is proportional to
the concentration of the positive ion, the correct equation
for this reaction becomes
chcGNHooch, + HOH = C.H.COOCH, -- NH, (2').
That the analogy between the ordinary esters and the imido-
esters is practically complete will be shown in Part II. by
proving that the imidoesters are decomposed also by means of
alkalis into nitriles and alcohols, with a velocity directly pro-
portional to the concentration of hydroxyl ion—exactly as
is the case in the saponification of esters.”
* Arrhenius: Loc. cit.
2 Euler: Loc. cit.
8 By the action of absolute alcohols on imidoester salts Reitter and Hess (Ber. d.
chem. Ges., 40, 3020) obtained ortho acid esters:
CH3C(: NH2C1)OC2H5 -|- 2HOC2H5 —- CH3C(OC2H5)3 + NH4C1.
The difference in the result of this action and that of water on such salts suggested
that the formation of ortho esters would be found due to the following sequence of
reactions:
-- + ---
CH3C(: NH2) OC2H5 -- HOC2H5 —- CH3C(NH3) (OC2H5)2 —- CH2 : C(OC2H5)2 —- NH4;
and CH2 : C(OC2H5)2 + HOC2H5 —- CH3C(OC3H5)3.
As a matter of fact Reitter has isolated, in the case of cyanacetic ester, a com-
pound C(OC3H5)2 : CH.CO2C2H5 (loc. cit., 40, 3358), either as an intermediate or a final
6
EXPERIMENTAL.
I. The Hydrochloride of Ethyl Imidobenzoate.
The velocity of the decomposition by water of the hydro-
chloride of ethyl imidobenzoate, according to the equation
+ +
C.H.C(: NH)OC, H, 4 H.O = C.H.COOC, H, -- NH, (3).
was determined at 25°, in solutions containing one mole
in 2, 4, 5, IO, and 20 liters, respectively. The solutions
were placed in a bath of 320 liters capacity, whose tempera-
ture was kept at 25° (+o.or) by means of an Ostwald ther-
mostat. The method of analysis used was the one described
by Derby: by means of a pipette, a measured portion (25,
IO, or 5 cc.) of the salt solution was rapidly run into a known
amount of standard sodium hydroxide solution (an excess
of this was used); the free imidoester liberated was removed
by three extractions with carbon tetrachloride (15, Io, and 5
cc., respectively); and the excess of sodium hydroxide in the
aqueous residue determined by titration with standard hy-
drochloric acid. Methyl orange was used as indicator.
A mixture of a benzimidoester hydrochloride and ammo-
nium chloride behaves under this treatment exactly as does
a mixture of free hydrochloric acid and ammonium chloride;
the imidoester salt neutralizes part of the added alkali, while
product. On the basis of this hypothesis of an intermediate product, the reaction
would not work with the imidoester salts of a tertiary acid, like benzimidoesters, and
a reply of Professor Reitter to a private inquiry has confirmed the correctness of this
surmise, as far as preliminary experiments go. Professor Reitter, with Dr. Hess,
lias kindly consented to our taking up this part of the problem as well as the physico-
chemical study of the formation of ortho esters, while he and his collaborators will
develop it from the preparative side and endeavor to isolate the compounds CaFIFCH :
C(OR)2 in the action of alcohols on the imidoester salts of benzyl cyanide. These
facts are mentioned here because the work on saponification and esterification, etc.,
is being taken up by us along these lines with aliphatic esters and imidoesters of
primary, secondary, and tertiary acids, as such reactions may form parallel or even
the main in termediate actions in some cases of ordinary saponification, etc. It may
be recalled that McCracken and I have already found that the velocity of decomposi-
tion of imidoester salts of phenylacetic acid is exceedingly high, very much higher
than that of aromatic salts, and it should be said that Dr. Blunt has obtained a
similar result for the action of ammonia on such salts. These and a large number of
other facts recorded by different investigators, (Cf. Nef, Standinger, Wilsmore) which
will be discussed at the proper time, promise to 1ead to a deeper insight into all the
various sides of these problems than has hitherto been obtained.—JULIUS STIEGLItz
(May, 1908).
* Am. Chem. J., 39, 437 (1908)
7
this is not true in the case of ammonium chloride; unlike the
free imidoester, ammonia in very dilute solution is not ap-
preciably removed by carbon tetrachloride, and hence, since
methyl orange is sensitive to it as well as to the fixed alkali, it
does not affect the titration. Thus, if b cc. of N/ Io sodium
hydroxide are added to v ce. of the salt solution and d ce. of
N/ Io hydrochloric acid are required after the extractions to
Ineutralize the excess of alkali, the concentration of the un-
changed salt is proportional to (b–d): v. The observed
values given below were obtained in this way.
The velocity constants were calculated according to the
equation for a monomolecular reaction,
dx.
... = K, (C-3) (4),
and -
__ I C–3,
K Ti, i. log nata=y (5).
As Derby" has shown, the substance undergoing change is
the salt or its positive ion alone, and not the free base libera-
ted by the hydrolytic action of the water present. Inasmuch
as the degree of hydrolysis is less than one per cent for ordi-
nary dilutions, the free base liberated is a negligible quantity
and the salt may be put equal to the total quantity of unchanged
substance.
For each concentration for which the velocity was deter-
mined, the degree of ionization of the salt was found experi-
mentally. This value, in a given experiment with given con-
centration of salt, may be considered a constant for the ex-
periment; although the imidoester salt used is growing smaller
in concentration during the course of the reaction, each mole-
cule is being replaced by one of ammonium chloride, and the
total electrolytic content in this way remains unchanged.
Under these conditions, according to the laws of isohydric
solutions, the degree of ionization of the imidoester salt may
be considered practically a constant for the whole period of
a velocity determination. The value of the ionization factor
Loc. cit.
8
will be incorporated in the final velocity constant in the man-
ner discussed below (p. 13); for each determination the velocity
constant, independent of the ionization, will, however, first
be calculated simply according to equation (5).
In the following tables, v gives the volume in liters contain-
ing one gram molecule of the imidoester salt, w gives the por-
tion in ce. used for each titration at the time indicated under
t. This line contains the time in minutes, from the time
of the first titration to the moment when the first quantity
(w) of solution was introduced into the excess of alkali; in
line 2, under (C–3), will be found the ce. of N/ Io alkali neu-
tralized by the unchanged imidoester salt contained in the
w ce. used for the titration. The concentration of the un-
changed substance, (C–3), etc., is proportional to this value.
In line 3, the velocity constants calculated according to equa-
tion (5) are given, the values being 43430 times the true con-
stant. All the determinations were made, as mentioned, at
I5° (+O.OI).
As was anticipated, the tables show that the velocity of
decomposition is very markedly dependent on the concentra-
tion of the salt used. To determine rigorously whether this
result is due to the fact that only the positive ion of the salt
is reacting and not the nonionized molecule, the degree
of dissociation of the hydrochloride of ethyl imidobenzoate
in the various concentrations used was determined by the con-
ductivity method. Owing to the fact that the salt is very
rapidly decomposed by water at 25°, the conductivity meas-
urements were made at o’; at this temperature, much less
difficulty was experienced in obtaining satisfactory results,
since the rate of decomposition is slow. The work of H. C.
Jones' on the degree of ionization of chlorides shows that this
value is practically the same at O’ as at 25°; hence the values
obtained at o? for the hydrochloride used may be employed
to calculate the relations desired at 25°. On account of the
hydrolytic dissociation of the hydrochloride according to the
equation
CeBI,C(NH,Cl)OC, H, ºr C.H.C(NH)OC, H, + HC1 (6),
1 Am. Chem. J., 26, 1442 (1901).
The Hydrochloride of Ethyl Imidobenzoate.
Table Ia.
7) = 2C). 'w) = 25 CC. Mean, 4343O Ka = I36.
t O 6O 90 I 2C) I 50 I8O 2 IO
(C–3) II. 77 9. 75 8. 93 8. IO 7.35 6.73 6. O4.
4343O Kw - tº e g tº I36 I 33 I 35 I36 I35 I 38
Table Ib.
‘U = 2C). 70 = 25 CC. Mean, 43430 K = I35.
f O 3O 6O 90 I8O 2 IO 24O -
(C–3) * I 2 ... O I IO. 97 9.98 9. O9 6.79 6. II 5. I8
4343O K tº º & © I3 I I34. I34. I 37 I4O I 35
Table IIa.
7) = [O. 70) = IO CC. Mean, 4343O Ka = I32.
t O 3O 6O 90 I 2G) I 50 286 3I 5 375
(C–3) 9. 6O 8. 74 8. O3 7.35 6. 64 6. O4 4. . I 3 3. 69 3. IO
43430 Kv º e º º I36 I29 I 29 I 34 I35 I 28 I32 I 32
Table IIb.
7) = I.O. 70) = IO CC. Mean, 4343O Ka = I.33.
i O 3O 6O 90 I 20 I5O 24O 27O 3OO
(C–3) 9. 75 8.89 7. 98 7. 35 6.67 6. 20 4. 70 4.32 3.87
43430 Kw . . . . I34 (145) I36 I36 I3 I I 32 I3 I I34.
3.
The Hydrochloride of Ethyl Imidobenzoate. ~s.
Table III a.
‘U = R. 70) = IO CC. Mean, 4343O Ka = I22.
f O 3O 6O 90 I 2G) 24O 285 33O
(C–3) I9. 24. I 7. 57 I6. 24 I5. O4. I3. 6 I 9.84 8. 57 7. 55
4343O Kw * * * (131)* I 23 I IQ I 25 I 2 I I23 I23
Table IIIb. ?
Q) = 5. 70) = 5 CC. Mean, 4343O Ka = I26.
t O 3O 6O 90 I 2G) I 53 I8O 2 IO
(C–3) 9.61 8. 74 8. O5 7. 4. I 6.79. 6. I4. 5. 79 5. I 7
4343O Kw (I37) I 28 I25 I26 I27 I 2.2 I 28
- Table III c.
Q) = 5. ‘U2) = IO CC. Mean, 4343O Ka = I2 I.
t O 25 55 85 II 5 I45 236 265
(C–3) I9. I5 I 7. 79 I6. 48 I5. O9 I 3.95 I2. 77 9.82 9. 27
4343O Kw tº a tº e (128) I IQ I 2.2 I2O I 2 I I23 II9
Table IV a.
7) = 4. 70) = IO CC. Mean, 4343O Ka = I2O.
f O 45 75 IO 5 2IO 24O 27O 3OO 345
(C–3) 24. 27 2 I. 22 IQ. 7I I8. I 7 I 3. 59 I2. 64 II.47 IO. 7 I 9.59
4343O Kw' & (130) I 20 I 2G) I IQ I 20 I 2 I I IQ I23
1 The first titrations at the end of 25 to 30 minutes often gave an abnormal constant, high or low, probably due chiefly to the fact that
a small difference in the titration, (C — ºc), produces a 1arge difference in the value Kzy.
* The temperature of the bath during this determination was about 0.03° too high. This probably accounts for the high value of Kzy'
E
f
(C–3)
43430 Kv
f
(C–3)
43430 Kv
t
(C–3)
43430 Kv
The Hydrochloride of Ethyl Imidobenzoate.
Table IVb.
Q) = 4. 70) = IO CC.
O 3O 6O 9C
24. I7 22.44, 20.50 I8.94
(IO7) II9 II.8
Table Va.
‘U = 2. 70) = 5 CC.
O 3O. 2 6O 90
22.89 2 I. 24. I9.89 I8. 26
IO8 IO3 IO9
Table : Vb.
7) = 2 70 = 5 CC.
O 29 59 . 89
24 .33 22. 74. 2 I. 2 I I9.78
. . . . IOI IOI IO3
Mean, 4343O Ka = II9.
I 2 I 255 285 3I6 345 .
I7. 36 I2. I2 II. I9 IO. O6 9.43
I IQ II.8 II 7 I 2G) I IQ
Mean, 4343O Kv = IO7.
24I. 27O 3OO 33O
I 2.86 II. 86 IO. 90 IO . OO
IO4. IO6 Io 7 IO9
Mean, 4343O Ka = IO5.
II9 244. 269 299 332
I8. I5 I3.67 I 3. O3 II . 95 II. O9
IO7 IO4. IO 5 IO7 IO6
I 2
it was necessary to determine the conductivity by Bredig's
method" in the presence of an excess of the free base suffi-
ciently great to neutralize the acid liberated and small enough
to have no appreciable conductivity of its own in the presence
of its salt. Because of the instability of the salt in aqueous
Solution, the following methods of operation were used: In
the case of the more concentrated solutions (N/Io-N/2) a
small cell of the Kohlrausch type was employed. The cell
had a capacity of about 4 cc.; the electrodes, whose diameter
was about 7 mm., were, when fully immersed, separated by
a column of liquid about 35 mm. in length and 5 mm. in diame-
ter. Solutions of the free base of the approximate concentra-
tions, N/40 and N/80, were made up and kept at —5°. The
Salt was weighed out into a dry graduated 5 cc. flask, the flask
filled with one or the other of the free base solutions, and the
whole well mixed and cooled to zero; the mixture was then
poured rapidly into the small cell, which was kept in a zero
bath, and the resistance of the solution measured by means
of a Kohlrausch bridge and telephone. In the case of the
more dilute solutions (N/20–N/256), an ordinary Arrhenius
cell was used. The solutions of free base were kept in 50 cc.
flasks calibrated for delivery; the salt was weighed out in a
small beaker, the solution of free base poured over it from
the graduated flask, and the whole well mixed and cooled
to about —O’. 5. This solution was then poured into the
Arrhenius cell and its resistance measured in the usual way.
In all the experiments, a calibrated thermometer was used,
and the resistance measured at exactly oº. It was found that
the resistance remained constant for longer than IO minutes,
thus showing that no appreciable decomposition had set in.
In Table VI., the results obtained for the conductivities
are given. Column I contains the volume in which one gram
molecule of salt is dissolved; in columns 2 and 5 the molecu-
lar conductivities (A) in the presence of an excess of one-
eightieth and one-fortieth normal free base, respectively, will
be found; in column 3 are given the differences (A) which
must be added to the molecular conductivities (A,) to obtain
1 Z. physik. Chem., 13, 213 (1894).
I 3
the conductivity (A2) for the salt at infinite dilution. Ac-
cording to Ostwald" and Bredig,” these differences are con-
sidered to have a fairly constant value for the various classes
of salts in solutions varying in concentration from N/ IO24 to
1ower concentrations. The differences (A) used in the fol-
lowing table were calculated from Bredig's values” by multi-
plication, first by the factor 1.065 to reduce reciprocal mer-
cury units to reciprocal ohms, and Second, by the factor o. 45,
which is the approximate correction (within two or three per
cent) for the change of temperature from 25° to oº'." Briefly,
Ao = Asso X 1.065 × 0.45.
O
In columns 4 and 6 we have placed the molecular conduc-
tivities at infinite dilution (A2) as calculated from A, and A.
In column 7, the degree of ionization (a.) of the salt is given
for each concentration.
Ethyl Imidobenzoate H ydrochloride.
Table VI.
I 2 3 4. 5 6 7
W Av.” A. Aco.” Azy." Aco . & v × IOO.
2 30.6 3O. 5 56.6
4 37. O 37. O 68.5
5 38.7 38.7 7I. 7
IO 43.6 . . . . 43.6 . . . . 80. 7
2O . . . . 7.7" . . . . 45.7 53.4 84.6
32 47. O 6.7 53.7 47.3 54. O 87.3
64 48.7 5.3 54. O 49. I 54.4 9I.O
I 28 5O. 4 3. 54.2 50.4 54.2 93.3
Mean, 54. O Mean, 54. O
If the observed velocity constants, K, (column 3, Table
IV.) for the decomposition of the hydrochloride of ethyl
imidobenzoate in varying concentrations be divided by the
1 Z. physik. Chem., 2, 840.
2 Ibid., 13, 198 (1894).
3 Loc. cit. z-
4 Stieglitz and Upson: Am. Chem. J., 31, 469 (1904).
* These are the values for Az, when an excess of N/80 base was used.
6 These are the values for Az, when an excess of N/40 base was used.
7 Found by extrapolation. f
* This column contains the values of Ago calculated from Av in column 2; those
calculated from Az, in column 5 will be found in column 6.
14
corresponding degree of ionization," we obtain as the apparent
velocity of decomposition of the positive ion,
K.,
Kion app. = o, (7).
This gives a series of values representing the velocity con-
stants (column 4, Table VII.), calculated on the assumption
that only the positive ion of the salt is undergoing decom-
position by water. As is well known, however, the speed
of reactions in which water is an active agent is greatly ac-
celerated by the presence of neutral salts.” This effect may
be due to an increase, by the salt, in the active mass of the
water;” but, whatever the cause, the fact remains that neu-
tral salts have this influence and that, except for extreme
dilutions, the effect is fairly proportional to the concentra-
tions of the Salt used. As a general rule, it is true also that
there is comparatively little difference in the effect of
analogous inorganic Salts.” In order, in our case, to determine
this “salt effect” and find the true velocity of decomposition
of the positive ion, Kºº, when, hypothetically, there is no
salt present and hence no “salt effect,” we may consider the
apparent velocities, K., a, (given in Table VII., column 4),
for O. O5, O. I, and O. 20 molar solutions, respectively. In-
spection shows at once that the “salt effect” for every O. OS
mole salt may be taken to be 3 in the terms of the same units
* In the decomposition of the imidoester salt, ammonium chloride replaces the
hydrochloride of the imidoester. According to the laws of isohydric solutions, how-
ever, the degree of ionization of the unchanged imidoester salt will remain practically
unaltered, since the total electrolyte content remains constant. The somewhat greater
degree of ionization of ammonium chloride will have a negligible effect.
* Part I., page 4.
* Euler: Z. physik. Chem., 32, 348 (1900).
* Arrhenius: Loc. cit. This may explain why the change of the imidoester salt into
ammonium chloride does not seem to change the “salt effect” to any great extent. In
Tables I.-V. Kzy does not change much in the six hours during which velocities were
measured. It is peculiar, however, that the “auto effect” of the hydrochloride of
methyl imidobenzoate seems to be twice as great as for the hydrochloride of ethyl
imidobenzoate. A thorough investigation of the effect of different salts is being
carried out by Miss Edith E. Barnard.
NOTE.-Miss Barnard’s work has now been completed. A more reliable method
of determining A2, and consequently the degrees of ionization, employed by us
has given very accurate results, confirming the fact that the salt effect is proportionate
to the ion concentrations, and that the velocity of decomposition is otherwise propor-
tionate to the ionized part of imidoester salts.—J. STIEGLITZ.
I5
as are used for K. Then for the hypothetical case, where
the salt concentration is o, we would have, since here there
is no “salt effect,” …”
K ion
= 16I–3; -
Kºo, = 158 * (8).
1072.
From this we see that the “salt effect” consists in an accelera-
tion of the velocity of 3/158, or practically 2 per cent for every
O. OS gram molecule of salt present. Considering the effect
as proportional to the total salt present, we may write
Kion app. = Kion X (I + cº X O.O2) – -
I58 × (1 + cº × 0.02) (9).
Combining equations (7), (8), and (9), we find that we can
calculate the velocity with which the hydrochloride of ethyl
imidobenzoate should be decomposed by water at any con-
centration. The following equation expresses the relation-
ship: -
K.-K. xa, x (; +***) Go).
O. O5
--—- " " ſº 'm X **) /
= I58 X (X, X ( + o.os (Io').
In this equation, m is the concentration of the imidoester salt
in gram molecules per liter and a, is the degree of ionization.
The values for K.…. are found in column 5, Table VII.
It is very likely, however, that the true “salt effect” is es-
sentially due to the ionized part of the salt present and not
to the total salt; under these circumstances, the following
equation should more correctly represent the facts:
- ºg • m X O.O2 × 0.,
Kcale.— Kion X Cº., X ( -- O. O.5 X & oos ) (II),
- - ſm, X *...) /
= I58 × a., X ( ***. (II'),
1 Arrhenius: Loc. cit, Euler: Loc. cit,
I6
The values calculated in this way are found in the last column
of Table VII. It is obvious that a striking agreement exists
between the theoretical velocities calculated according to
either equation (IO') (Kate, in column 5) or equation (II')
(K'_ale, in column 6) and the velocities actually observed (K,
in column 3). The greatest deviation, in fact, from the theo-
retical values is 3.5 per cent., which is well within the limits
of experimental errors.
Table VII.
I 2 3 4. 5 6
R.
7%. Ioo &v. 43430 A v. 43430 × . 4343OXK'calc. 43430XXcalc.
O. 50 56.6 IO 5 I85.5 IO7 IOI. 5
O. 25 68.5 I2O I 75 II9 II 7
O. 20 7I. 7 I2 I. 5 IG9.5 I22. 5 I 2 I
O. IO 80. 7 I 32.5 I64 I 32.5 I32
O. O5 84.6 I36 I6I I36 I 36
O. OO . . . . . . . . I58% º e - © tº
The “salt effect” may be represented graphically in the
following diagram, in which the velocity constant K, Q,
with no correction for the “salt effect” (column 4), is given
on the horizontal coördinate, and the total concentration of
ionized Salt is given on the ordinate (Cºns - m × 0).
The curve represents an almost straight line, showing an ac-
celeration directly proportional to the concentration of the
ionized salt present. *
0.30-
0.25-
0.204
0./5-
0 /0-
0.05-
j
_j - TI I ſ !
/J 6 /65 /68 /7.5 /73 /85 /38
Ko: Cºv
Fig. 1.-Ethyl Imidobenzoate,
1 This value is hypothetical.






17
Judging by these results, we would decide that the decom-
position of the hydrochloride of ethyl imidobenzoate by water
proceeds with a velocity proportional to the concentration
of the positive ion of the salt, and this velocity is accelera-
ted by the influence of a “salt effect” which is proportionate
to the concentration of the salt ions present." This result is
in perfect agreement with Stieglitz's theory of the catalysis
of esters. -
2. The Hydrochloride of Methyl Imidobenzoate.
In order to test the theory more fully, a series of experi-
ments, exactly similar to those performed with the hydro-
chloride of ethyl imidobenzoate, was carried out with the cor-
responding salt of methyl imidobenzoate. The tables and
the symbols employed correspond to those in the preceding
part of the paper, as do also all the experimental details; no
further discussion, therefore, is required. --
* This conclusion has been confirmed by the extended investigations of Dr.
Darnard.
f
(C–3)
43430 K,
C(–3)
4343O Kw
(C–3)
4343O Kw
The Hydrochloride of Methyl Imidobenzoate.
Table VIIIa.
= 2C). w) = 25 cc. Mean, 4343O Ka = 236.
O 3O 6O 9C ‘I 2C) I 50 255 285
IO. 93 9.40 7. 63 6. 72 5. 7 I 4.8 I 2.8O 2 : 61
g º is e 2 I 8 26O 234. 235 237 23O 2 I 8
Table VIIIb.
2O. 10 = 25 cc. Mean, 4343O Ka = 235.
O 3O 6O 9C I 20 225 255
8. IO 6. 90 5. 67 5. 5 I 4 35 2. 39 2. O7
232 258 (186) 225 235 233
Table IXa.
IO. ‘U) = IO CC. Mean, 4343O Ka = 242.
O 3O 6O 90 I 2G) I5O 27O 3OO
9.38 7. 96 6.47 5. 75 4.88 3.99 2 . 2C) I. 75
238 (269) 236 237 247 233 24.3
3I 5
257
225
33O
236
.85
.85
&
S
f
(C–3)
4343O Kw
43430 Kv
(C–3)
4343O Kw
The Hydrochloride of Methyl Imidobenzoate.
Table IXb.
‘U = IO. 70) = IO CC. Mean, 43430 Kw = 246.
O 3O 6O 90 I 2 O I 50 27I 3OI
9. 47 7. 75 6. 37 5.85 4.79 4. I8 2 . [O I. 82
e & (290) (287) 233 , 247 236 235 283
Table Xa.
'U = 5. 70) = IO CC. Mean, 4343O Ka = 234.
O 3O 6O 90 I 20 2 IO 24O
I6.46 I 3.95 II. 97 9.81 7. 93 5.43 4. 6 I
239 23.I. (249) (264) 23O 236
Table Xb.
‘U = 5. 70) = IO CC. Mean, 4343O Ka = 23.I.
O 3O 60 90 I5O 24O 27O
I8.50 I5.96 I 3. 3O II. 45 8. 27. 5.4O 4. 49
(214) 239 232 23 I 224 228
33O
24. I
255
. 23
237
3I 5
227
. 50
. 56
8
The Hydrochloride of Methyl Imidobenzoate.
Table XI.
7) = 4. 70) = 5 CC. Mean, 4343O Ka = 23O.
f O 3O 60 90 255 27O 285 3OO 315
(C–3) I2. 24. IO .43 8.85 7. 59 3. I 3 2.99 2. 70 2. 53 2. 36
4343O Kw tº e g tº 232 235 23I 232 227 23 I 228 227
Table XIIa. F-
‘U = 2. 70) = 5 CC. Mean, 4343O Ka = 207. *
f O 3O 6O 75 I 2G) I 50 27O 3OO 33O
(C–3) 23. 37 2O. 33 I 7.6 I I 5. 75 I 3. 29 II. 20 6.47 5. 50 4.85
43430 Kv $ tº e is 2O2 2O5 (222) 2O5 (213) 2O7 2O7 2O7
Table XIIb.
7) = 2. * w) = 5 CC. Mean, 43430 K, = 208.
f O 3O 6O IO 5 I 2 O I 50 3OO 33O
C–3. 23.89 2O.7O I7.66 I4.46 I 3-4-3 II.32 5.76 4.99
& & & 2O8 (218) 2O8 2O9 (216) 2O6 2O6
4343O Kw
2I
Table XIII."—The Degree of Ionization of the Hydrochloride
of Methyl Imidobenzoate.
I 2 . 3 4 5 6 7
V. Av.” A. Aoo .2 Av.” Aoo .3 0.7) X IOO.
2 31.6 32.9 58.5
4 38.5 38.6 7O. O
5 4O. I 39.8 72. 5
IO 43.8 . . . . . . 44.7 . . . . 80.3
32 48.2 6, 7 54.9 48.2 54.9 87.5
64 50. I 5.3 55.4 49.9 55. I 90.7
I 28 51.8 3.8 55.6 50. 7 54.5 92.5
ſ Mean, 55.3 Mean, 54.9
Average Aco = 55. I.
Table XIV.”
I 2 3 4. 5 6
A v
7%. IOO &v. 43430 A v. 43430 a. 43430 A calc. 43430 Å"calc.
O. 50 58.5 2O7 3.54 225 2O6
O 25 7O. O 23O 328 23 I 225
O 2C) 72.5 232 32O 23 I 227
O . IO 8o. 3 244 3O4. 238 237
O. O.5 84.6% 236 279 242 242
O. O.5 84.6 24.2% 2867 242 242
O . OO * * * * & G tº 2757 4. & tº
The values in the table above (Table XIV.) give the rela-
tions of the velocity of decomposition of the hydrochloride
of the methyl imidobenzoate (as taken from Tables VIII.-
XIII.) to the degree of ionization of the salt and to the
“salt effect” in any concentration. In Table XIV., as in
Table VII., column 3 gives the velocities actually observed
(K.), and in columns 5 and 6 the theoretical velocity constants
1 Tables XIII. and XIV. are comparable to Tables VI. and VII., respectively.
* Column 2 contains Az) when N/80 base was used in excess. In column 4 the
Aco was calculated from Az, in column 2.
* Column 5 contains Az, when N/40 base was used in excess; column 6, the cor-
responding Age.
* Tables XIII. and XIV. are comparable to Tables VI. and VII., respectively.
5 Taken from the ethyl ester. See Table VI., column 7.
9 Taken from the work of Derby: Loc. cit. Mean of 23 determinations.
11.1
7 The acceleration due to salt is taken from these two values. It is 275 = 0.04
for 0.05 mole of salt.
22
as calculated by equations analogous to (IO') and (II'), respec-
tively, are given;" into this last column “salt effect” and de-
gree of ionization are both incorporated. It is noticeable
at once that the calculated values (K'.a.) for an accelera-
ting effect of the salt ions agree remarkably well (within 2–3
per cent) with the observed velocities (K.). As was expected,
the values calculated for an accelerating effect proportional
to the total salt (Kan.) do not agree so well with the observed
velocities—the maximum deviation being as high as 8.8 per
cent in one case. In both cases, therefore, for the hydrochlor-
ides of methyl as well as of ethyl imidobenzoate, the agree-
ment between the velocities calculated on the basis of an
equation like (II') agree well with the observed veloci-
ties. We may consider then that such an equation actually
gives the best expression of the course of the reaction accord-
ing to which an imidoester salt is decomposed into an ester
and an ammonium salt (equation 2').”
PART II.
The Catalysis of Imidoesters by Means of the Hydroxyl Ion.
A qualitative investigation of the action of alkalis on imido-
esters of the aromatic series showed that the reaction proceeds
in a direction different from that exhibited in acid solution;
instead of an ester and an ammonium Salt (equation 2'),
a nitrile and an alcohol are formed according to the equation
C.H.C(; NH)OR = C.H.C = N + HOR (I2).
This, it should be noted, is a reversal of the synthetic method
of preparing imidoesters from nitriles and alcohols in the
presence of acids.
The quantitative determinations of the velocity of decom-
position, under the influence of varying proportions of barium
hydoxide, show that the velocity is always directly propor-
tional to the concentration of the imidoester and to the con-
centration of the hydroxyl ion, i. e.,
* The acceleration is taken as 4 per cent for 0.05 mole salt.
* Work is being continued along these lines by Mr. W. Hickman, velocities being
measured at 0° as well as at 25°.-J. S.
23
d -
† = KOH. X Cester X CoA. (13).”
The hydroxyl ion may be said to act “catalytically,” in the
old sense of the term, as it does not appear to be involved
in the reaction and its concentration remains constant.
This is true because both the original imidoester used and
the alcohol formed are too weak acids to neutralize apprecia-
bly the alkali in the dilute solutions used. For a given ex-
periment, therefore, we may write
dx
* = Kon XCeste, X. kory (14),
dt
= K, X Caster - (I4')."
-. dºc
Or iſ T K., X (C–3) and
_ ! Or C – 3,
Kº-º-; log nata—, (I5).
The velocity constant of the decomposition, Koº, is, accord-
ing to (14) and (14'),
Koź, — 3 – (16),
in which kozy is the given concentration of the hydroxyl
ion used in an experiment and K, is the observed velocity
constant in that experiment. If the decomposition actually
proceeds according to (13), the same value for Kor, Ought
to be obtained from all the determinations, independent of
the concentration of the imidoester and of the barium hy-
droxide used. The results reported below fully prove that
such is the case.
According to the theory developed in Stieglitz's paper,”
the enormous acceleration of the decomposition of imidoes-
ters into alcohols and nitriles produced by alkalis as catalytic
agents must be due to acid properties of the imidoesters.
Such a character would lead to the formation of metal salts
1 Cester is a function of x,
* Loc. cit,
24.
whose negative ion would be the actual substance undergoing
decomposition and the concentration of whose negative ion
would constantly represent the real “active mass” at any
moment. The imidoesters, consequently, must have an
amphoteric character, ionizing chiefly as bases, yet minimally
also as acids, and forming salts with the alkalis. There are
two ways in which such a salt formation could occur; either
by direct addition of alkali to the imidoester, just as to an
ordinary ester," according to the equation
NH,
C.H.C(: NH)OR + NaOH = C.H.-ONa (17),
OR
or by salt formation in the imido group according to the equa-
tion *
C.H.C(; NH)OR + NaOH = C.H.C(; NNa)OR + H2O (18).
In view of the fact that the very closely related class of bodies,
the amidines, RC(; NH)NH, besides their pronounced basic
properties—they are far stronger bases than the imido esters,
even stronger than ammonia—have been shown to possess
also weak acid properties, forming salts such as RC(: NMe)NH,”
and RC(: NCHE) (NMeCPIs),” it is extremely probable that
equation (18) represents the actual method of Salt forma-
tion.” For the further development of our theory, it is a
matter of indifference whether (17) or (18) is correct, but (18)
gives a rather more simple decomposition of the imidoester
into nitrile and alcohol than does (17), from which one would
expect the formation of an alcohol and an acid amide, RCONH2.
The salts of very weak acids, as the imidoesters must be,
are hydrolyzed in aqueous Solutions according to the formula-
tion
C.H.C(; NMe)OR + HOH -
C.H. (: NH)OR + MeOH (19),
1 Claissen: Ber. d. chem. Ges., 20, 641 (1887); Nef: Ann. Chem. (I_iebig), 270,
322; 287, 279.
2 Bruce: J. Am. Chem. Soc., 26, 419 (1904).
* Bamberger: Ann. Chem. (Liebig), 273, 277 (1893).
* The isolation and analysis of the metal salts of the imidoesters will be attempted
in this laboratory.
25
and we would have for the free weak acid (the imidoester)
Cneg. ion X C# *º-º-º-º:
-a- = racia (20),
7,202.
in which the Cº., represents the concentration, in the
usual units, of the negative ion of the acid, Cºf the concentra-
tion of hydrogen ions, and Cºol that of the nonionized acid,
and kazºº is the affinity (ionization) constant of the imidoester as
an acid. The proportion of ester which is ionized as an acid
is so exceedingly small that we may consider the concentra-
tion of the nonionized ester, Cao, equal to the total concen-
tration of ester, Cº., at any moment and hence write
Cneg. ion X C#. = k
C t acid (20').
€Sté?’
Since
C# X CoA = k Ho & (21),
we have, combining (21) and (20'),
Cnes. ton – AEacid (2 2)
Cester X CoH AEH,o
If, now, according to Stieglitz's theory, the substance which
is actually undergoing decomposition into a nitrile and an
alcohol is the negative ion of the imidoester, in other words,
if we put simply
dx
— = Pº . e 1.
di Kion X Cneg. ZO72 (23),
then we must find the following equation, obtained by com-
bining (23) and (22), to be true:
d AEaci
* = K, x * x Cao, X Cog (24),
di AEH,o
– Kory X Cester X Coff (25).
This expression, however, is identical with equation (13),
which is, on the one hand, in perfect agreement with the ex-
perimental results; and can, on the other hand, it is seen, be
deduced theoretically by the application of the well-known
* Cneg, ion is a function of x.
26
laws of chemical equilibrium to the fundamental conception
of salt and ion formation as the explanation of the catalytic
action of acids and bases.
We may, therefore, properly write the decomposition re-
action in alkaline solution as follows:
amºs-º
/N cºmmem *
C.H., = C.H.C E N + OR (26),
NOR
which corresponds in every way to equation (2') for the de-
composition of the positive imidoester ions into ammonium
ions and esters. In the latter case, the basic properties of
the imidoesters were sufficiently strong to permit a deter-
mination of the actual concentrations of the positive ion and
of the real constant, Kº ; in the case of the decomposition of
1077 )
, the negative ions into nitriles and alcohols, the acid ionization
of the imidoester is so slight that the determination of kaº;
in equations (20) and (22) was not attempted, and consequently
we have been unable to resolve the experimental constant
Kon of equation (25). But from the fact that Koh has
been found to be 1.991 for methyl imidobenzoate (see below),
it is seen that the velocity of decomposition of the negative
ion is very much larger than the velocity of decomposition
of the positive ion of the same ester, whose value has been
found to be o. oO63 (see p. 21, Table XIV.). There again,
the imidoesters show a perfect parallelism to the ordinary
esters, which are decomposed very much faster by alkalis than
by acids.
EXPERIMENTAL.
The qualitative examination of the change of the imido-
esters in alkaline solution was made on two representative
substances, methyl imidobenzoate and methyl imido-m-nitro-
benzoate. The experiments were carried out as follows:
About one gram of the ester was weighed out into a tube of
about 50 cc. capacity and to this was added a measured amount
of N/ Io barium hydroxide. The tube was then attached to
the wheel of a small engine and shaken for ten to twenty
hours. In the case of the methyl imidobenzoate, benzoni-
27
trile separated out as a heavy oil; this was filtered off, dried
in a desiccator, and weighed. In order to establish its iden-
tity, a portion was dissolved in alcohol, mixed with I5 drops
of concentrated ammonia Solution, and treated for sixty hours
with hydrogen sulphide. The resulting thiobenzamid was
dissolved in dilute sodium hydrate and reprecipitated by
means of carbon dioxide. This product, when dry, melted
at I 15°. No very accurate quantitative results were obtained
in this way in the case of methyl imidobenzoate, as benzo-
nitrile is quite soluble in water and there is another loss; it
was necessary to wash the filter with a few drops of alcohol, as
was done with the tube in order to obtain all of the nitrile;
the alcohol was completely evaporated in the desiccator, but
experiment showed that benzonitrile was also slightly vola-
tile under these circumstances. In spite of these difficulties,
o. 9 gram of methyl imidobenzoate, when treated with 25 cc.
of N/Io barium hydroxide, for twelve hours, yielded o. 35
gram of benzonitrile—50 per cent of the theoretical value.
In the case of methyl imido-m-nitrobenzoate, the nitrile,
under the treatment described, separated as a white, crystal-
line substance. It was filtered from the solution, washed
with dilute hydrochloric acid to remove any unchanged solid
ester, dried on a clay plate, and weighed. The nitrile thus
obtained melted sharply at 116°. The filtrate was then heated
to about 80° in the presence of dilute hydrochloric acid to
convert any unchanged dissolved ester into methyl m-nitro-
benzoate and extracted with low boiling ligroin. The ex-
tract was dried with calcium chloride and allowed to evaporate
spontaneously. Only a small quantity of pure m-nitrobenzoni-
trile was left. One gram of ester shaken for 20 hours with
25 cc. of N/Io barium hydroxide yielded o. 5 gram of nitrile—
67 per cent of the theoretical value. It is apparent, there-
fore, that equation (12) represents the chief reaction.
The quantitative study of the accelerating action of hydroxyl
ions on the decomposition of imidoesters into nitrile and alco-
hols included velocity determinations on methyl, ethyl, propyl,
and isopropyl imidobenzoates and methyl imido-m-nitro-
benzoate. The method of analysis used to follow the reac-
28
tion and the conditions of the experiment will be described
in detail for the first one of the series examined, methyl imido-
benzoate, where special precautions had to be used.
I. Methyl Imidobenzoate.
Methyl imidobenzoate was prepared from its hydrochlor-
ide by treatment of the latter with three times the calculated
amount of a cold 30 per cent solution of sodium hydroxide and
rapid extraction with ether. The extract was dried with cal-
cium chloride and the ether distilled off under diminished pres-
sure. The remaining oil was then distilled at a pressure of
2 mm. or less." At this pressure, methyl imidobenzoate (as
well as ethyl, propyl, isopropyl, and isobutyl esters) distil
without visibly boiling and without decomposition. In order
to test the purity of the compound, it was titrated” with N/ Io
hydrochloric acid. Instead of stopping the addition of acid
at the ordinary “neutral” tint, the solution was titrated to
a “standard” tint prepared by adding a few drops of the in-
dicator (methyl Orange) to a N/50 solution of corresponding
pure hydrochloride. The ester was diluted so as to give,
when titrated, a salt solution of the same concentration as
that of the standard tint. The very slight hydrolysis of the
hydrochloride is by this means prevented from affecting the
accuracy of the titrations. -
The velocity of decomposition of the imidoesters in the pres-
ence of barium hydroxide was determined, as in the case of
the hydrochlorides, at 25° (+o°. OI). The required amount
of N/IO barium hydroxide was run into a 500 cc. graduated
flask and the solution made up to the mark. This aqueous
Solution was poured into an ordinary 750 cc. flask and kept,
well stoppered, in the bath until it had acquired the proper
temperature. The ester was kept in a weighing bottle which
contained a little pipette; in this way it was possible to weigh
out the ester directly into the solution of barium hydroxide.
Errors in volume introduced by this method were found to
be negligible. The greatest difficulty, however, lay in find-
1 A mercury pump was employed.
2 The titration was carried out in Nessler tubes.
29
ing a means of analysis by which to follow the course of the
reaction. Finally, a simple method was adopted. A meas-
ured amount of solution was run, by means of a calibrated
pipette, into an Erlenmeyer flask and titrated with N/Io
hydrochloric acid. Alizarin sulphonate was used as indi-
cator since no sharp “end point” could be determined with
methyl orange. Even with this delicate indicator considera-
ble difficulty was experienced; this was true especially at the
beginning of the reaction where the concentration of the
imidoester relative to the amount of barium hydroxide was
fairly large, and where a very small error in titration markedly
affects the constant. The calculation of the amount of un-
changed ester from these titrations is very simple. Experi-
ment showed that the reaction products contained in the
solution do not affect the titration. Thus barium hydroxide
could be sharply titrated in the presence of benzamide, which
might be formed to some extent by the action of baryta on
the first reaction product, benzonitrile; it could likewise be
titrated in the presence of benzonitrile and of both compounds
at the same time. The neutral tint, when it was once ob-
tained, did not change for several hours. Therefore the
amount of acid used in titrating the measured amount of solu-
tion represents the total base contained in the given volume;
subtracting from this the acid required for the known amount
of barium hydroxide, we have, as the remainder, the concen-
tration of unchanged ester. In Tables XV-XIX. these values
are placed in line 2, following (C–3).
The amount used for each titration is 50 cc., unless other-
wise stated under w; the amount of barium hydroxide con-
tained in the volume titrated is given under ba; the time in
minutes is given in line I, and the velocity constant calcula-
ted from equation (15) in line 3. The concentration of
ester for each benzoate is approximately the same through-
out the series of tables; its value appears in the equation as
(C–3) and is given in terms of N/ Io alkali.
3
Methyl Imidobenzoate.
Table XV a.
w) = 50 CC. ba = O. 59. Mean, 4343O Ka = IO2.
O 2O 6O 8O IOO I 20 I4O I6O
6. 60 6. 27 5. 63 5. 44 5. I 8 5. OO 4. 78 4 - 54
(III) (II.5) IO 5 IO 5 IOO IOO IOI
Table X Vb.
ba = O. 58. Mean, 4343O Ka = IOO.
O 2O 4O 6O 8O IOO I 20 I4O I6O
5.94. 5. 66 5. 32 5. I8 4.88 4. 69 4 - 54 4 : 34 4.. I 9
IO 5 I 2G) 99 IO7 IO3 97 97 95
Table XVI a.
ba = I. I.4. Mean, 43. 430 K = 193.
O 2O 4. I 6O 8O IOO I 2G I 4O I6O 18O
6. 56 5.9 I 5. 33 4. 93 4. 52 4. . 2 I 3.84 3. 64 3.3 I 3. O3
(227) (220) 2O8 2O2 I93 I94. I82 I 88 I 86
Table XVIb.
ba = I. 20. Mean, 43430 K, = 203.
O 42 6O. 5 8O IOO I 2C) I4O I6O
I9 4. 72 4 - 4 O 4. OI 3. 64 3. 27 3. OO
5.
(2II) (215) 2O I 2O I 2O2 2O7 2O4.
º
C—x
Kº
ba = 3. I8
O 2 I
5. 33 4. I 7
ba = 3. II
O 2O
5. 4.7 4. 37
. . (487)
ba = 4.66
O
6. 40
ba = 4.80.
O
6. 37
Methyl Imidobenzoate.
Table XVIIa.
Mean, 4343O Ka = 5 II.
4O 6O 8O IOO I 2G) I4O I6O
3. 33 2 : 68 2 . I 8 I 66 I. 29 I . O 2 O
5OO 5 IO 497 485 5I 3 5I 2 5.I.4.
Table XVIIb.
Mean, 4343O Ka = 497.
42 6O 8O IOO I 2G) I4O I6O
3.42 2.76 2 . IO I 77 I. 37 I. I.3 O .
(489) 4.95 5 IQ 49C) 50 I 488 5OI
Table XVIIIa.
Mean, 4343O Ka = 742.
II 2O 3O 4.2 5O 6O. 5 7O
5. 3 I 4. 62 3.90 3. O9 2 . 72. 2. 33 I
736 (707) (7O2) 753 743 (725) 734
Table XVIIIb.
Mean, 4343O Ka = 74.I.
IO 2O 3O 4.O 5O 6 I 7O
5. 4. I 4. 55 3. 80 3.25 2. 7 I 2.24 I
(703) (723) 745 (729) 74. I 743 743
8O
84
96
. 92
I8O
543
I8O
5.I.4.
8O
743
8O
I. 56
(764)
Methyl Imidobenzoate.
- Table XIXa.
ba = 7. O2. Mean, 4343O Ka = IO90.
O . O IO 2O 3O 4O 47 55 62.5 7O
5.82 4.58 3. 52 2. 73 2. O7 I 84 I. 39 I. I 7 O. 89
IO4O IO90 I IOO II 2C) IO7O II 3O II IO II IO
Table XIXb.
ba = 7. O6. Mean, 4343O Ka = IO90.
O ... O IO 2O 3O 4O 48 55 63 7O
6. I6 4... 79 3. 74 2.85 2 . 22 I 82 I. 58 I . 26 O. 99
IO90 IO8O II 2C) II IO I IOO IO7O IO90 II 30
§
33
From the data found in these tables it can be shown that
the velocity of the decomposition of methyl imidobenzoate
in alkaline solution is proportional to the concentration of
the hydroxyl ion,
Kv = Kozy X Cozy and Koh = ;
OH
in which K, is the observed velocity constant and Cory the
concentration of the hydroxyl ion in a given series. Coir"
is determined from the concentration of the barium hydrox-
ide used and its known degree of dissociation as calculated
from its conductivities.” The values obtained in this way
will be found in Table XX. In column I the number of the
experiment is given, in column 2 will be found the amount
of N/ Io barium hydroxide used in 500 cc. of solution, in col-
umn 3 the corresponding degree of ionization of the barium
hydroxide, in column 4 the concentration of the hydroxyl ion
for each cubic centimeter of solution; column 5 contains the
observed velocities, K., and column 6 this value divided by
the hydroxyl concentrations.
Table XX.
I 2 3 4. 5 6
No. Ba(OH)2. IOO ov. 10' X (OH). 43430 Kv. o. 434 × KOH.
XVa. 5. 90 97.5 II.5 I O2 O. 887
XVb 5.80 97.5 II 3 IOO O.885
XVIa II .4O 97 22 I I93 O. 873
XVIb I 2 . OO 97 233 2O3 O. 871
XVIIa 31.80 96 6IO 5 II O. 837
XVIIb 31 . Io 96 597 497 O. 833
XVIIIa 46.60 95 886 742 O. 840
XVIIIb 48. Oo 95 9I2 74 I (o. 812)
XIXa 7o. 20 93 I306 IO90 O. 835
XIXb 70. 60 93 I3 I3 IO90 O. 830
.854
Mean, O. Kory = O. 854. K. = *** = 1.07.
4343 K-OH 54 OH O. 4343 97
1 Neglecting the OH ions due to the ester itself. This is exceedingly small ins
comparison with the OH ions from barium hydroxide since the ionization of a weak bae
is suppressed by the presence of a strong base.
2 Kohlrausch and Holbein: “Leitverinógen der Electrolyte,” p. 167.
34
The values in Table XX. for Koº, indicate that the hypothe-
sis regarding the proportionality between the velocity of de-
composition and the hydroxyl concentration is a correct
representation of the facts. Below will be found similarly
conclusive evidence in the case of the other esters investigated.
2. Ethyl Imidobenzoate.
Tables XXI-XXIII, give the velocity determinations. In
these everything corresponds to the work on the methyl
ester. All the solutions were made up to 500 cc. and 50 cc.
was used for each titration. The symbols used have the same
significance as in the case of the methyl ester.
§
O
6.89
6. 25
2 I 2
35
4. 76
(268)
Ethyl Imidobenzoate.
65
5. I4.
I95
85
4. 64
2O2
Mean, 4343O Ka = 205.
8O
4. O6
2OO
IOO
3.66
2O5
I 2G)
3 : 34
2O4.
I4O
3. OO
2O8
Mean, 4343O Ka = 204.
IO 5
4. . I 7
2O8
ba = 3.59.
ba = 3.57.
ba = 5. 22.
25
4.90
3 I7
O
5. 88
O
5.87
2 I
5 : 33
I99
Table XXI a.
4O
5. 3O
I IO)
Table XXIb.
6O
4.38
2 IO
Table XXIIa.
45
4 - 23
3.18
65
3.81
29O
85
3. 29
296
I 25
3. 84
2O2
IO 5
2.85
3OO
Mean, 4343O Ka = 299.
I 25
2.49
3OO
I45
3.47
2O5
I45
2. I 8
298
I6O
2 I I
3. I 3
I65
2.99
Table XXIIb.
ba = 5.35. Mean, 4343O Ka = 31O.
O 25 4O 5O 90 I IO I3O I 50
6. 22 6. O8 4. 64 4. . 4C) 3. 27 2.85 2.47 2. O7
(40) 3.18 3OI 3IO 3II 308 3I 7
Table XXIIIa.
ba = 7. OI. Mean, 4343O Ka = 40I.
O I5 30 4.5 55 65 75 86
6.4O 5. 58 4.82 4. 2C) 3.87 3.54 3. I9 2.90
398 4 IO 4O7 397 4O3 4O3 4OO
Table XXIIIb. -
ba = 7. O4. Mean, 4343O Ka = 40I.
O 25 75 85 I IO I 3O I5O I 75
6. I 6 4. 9C) 3. O6 2.83 2. 25 I. 79 I. 56 I . II
395 4O4. 399 397 4. I2 394. (425)
IOO
2 . 57
396
I90
O. 8O
(466)
‘ā,
37
Table XXIV.
I 2 3 4. 5 6 *
NO. Ba(OH)2. IOO ov. Io'X (OH). 43430 Kv. o. 4343X Koh.
XXIa 35.90 96 689 2O5 O. 298
YXIb 35. 70 96 685 2O2 O.298
XXIIa 52.2O 95 992 299 O. 3OI
XXIIb 53.50 95 IO7O 3 IO O.29O
XXIIIa 70. IO 93 (1304) 4OI O.307
XXIIIb 70.40 93 (1309) 4O2 O.307
Mean, 0.43430 Ky-o. 300. Koń = = O.69.
It is evident that the agreement between theory and ex-
periment is as good for the ethyl ester as it is for the methyl
ester. -
3. n-Propyl Imidobenzoate.
Tables XXV-XXVII. contain the velocity determinations.
The solutions in each case were made up to 500 cc. and 50 cc.
was used for each titration.
f
C—x
Kz,
2O
226
. 29
Table XXVa.
3O
2.84
(170)
5O
2.63
(168)
Mean, 4343O Ka = I52.
7O
2.47
I6O
Table XXVb.
Table XXVI a.
4O
2.90
25O
6O
2.48
27O
Mean, 4343O Ka = I 71.
8O
2. 27
258
90
2. 39
I44.
IOO
I. 9 I
28I
I3O
I. 99
I6O
Mean, 4343O Ka = 255.
I2O
I. 84
24O
I 50
. 88
I 53
I5
3. O3
I5O
2O
3
I46
. I 6
4O
2.86
I8 I
6O
2.62
I84
7O
2.5I
I85
8O
2.46
I 72
IOO
2. 29
I69
I2O
I 74.
I4O
25O
. I4.
. 64
I6O
264
. 38
%
§
Table XXVIb.
ba = 6.43.
O 2O 4O 6O 8O
3.25 2 . 72 2.38 2 . I 7 I. 92
(387) (340) 292 286
Table XXVIIa.
ba = 4.56.
O 2O 4O 6O - 8O
3. 35 2.95 2. 74 2.50 2. 27
tº; (276) 2 I 8 2 II 2 IO
- Table XXVIIb.
ba = 4.62.
O 2O 4O 6O 8O
3. 28 2.98 2. 57 2.47 2. IQ
2O8 222 2O5 I94
Mean, 43430 K = 291.
IOO I 20 I4O
I 7O I. 43 I.
28 I 3OO 29O
Mean, 4343O Ka = 207.
IOO I 2G) I4. I
2. I 3 I. 89 I
I97 2O7 2OO
Mean, 4343O Ka = 2 IO.
IOO I 2G) I4O
I . 99 I 8O I
217 2 I 7 2O I
28
76
. 66
I6O
I. O8
298
I6O
I. 55
2O9
I6O
I. 47
2 I 8
4O
Table XXVIII.
I 2 3 4 5 6
NO. Ba(OH)2. IOO &v. IO*X(OH). 43430 Áv. o. 4343 × KOH.
XXVa. 3O 50 96 59 I I52 O. 255
XXVb 30.70 96 596 I 71 (o. 285)
XXVIa 61.50 94. I I56 255 O . 22O
XXVIb 64. 30 94. I2O8 29I O. 24.I
XXVIIa 45.60 95 866 2O7 O. 239
XXVIIb 46.2O 95 878 2 IO O. 239
Mean, O. 4343 Koh = O. 239. Koh = O. 55.
4. Isopropyl Imidobenzoate.
Tables XXIX.-XXXI. contain the velocity determina-
tions. The solutions in each case were made up to 500 cc.
and 50 cc. was used for each titration.
e
Mean, 4343O Ka = 67. 6.
6O
4.5C)
7O . 4.
ba = 3. O9.
t O 2O
C—x 4.96 4. 79
Ko tº gº tº & ( 75. 2)
ba = 3. 22
i. * O 4O
C—x 4. 97 4. 55
K.," gº tº e º is is tº gº
ba = 4.62.
t O 2O
C—x 4.69 4.48
Ko . . . . IOO
* Calculated from t = 40, 1 — 2 = 4.55.
Table XXIXa.
3O
4. 73
68.4
4O
4.67
65.4
Table XXIXb.
6O
4.38
(82. O)
Table XXXa.
4O
4 - 23
II 2
8O
4. 27
69. O
6O
4. II
96
Mean, 4343O Ka = 67.5.
IOO
4. II
73.6
8O
3.93
96
8O
4 : 34
72 . 4.
I4 I
3.9 I
65. I
IOO
3.68
IO5
IOO
4. 28
64. O
I62
3. 77
66.9
Mean, 4343O Ka = IOI.
I 2G)
3.59
97
I 2G)
4. I4
65.4
I 75
3. 7 I
64. 9
I4O
3. 39
IOI
I56
3.90
67.5
I90
3.63
65.4
I6O
3. I 8
IO 5
Table XXXb.
ba = 4.65. Mean, 4343O Ka = IO9.
O 2 I 4O 6O 8O IOO I 20 I4O
4.82 4. 55 4. 29 4. I4. 3.94. 3. 75 3.59 3.46
II 5 (126) I IO IO9 IO9 IO7 IO3
~,
Table XXXIa. -
ba = 6. Io. Mean, 4343O Ka = 130.
O 23 42 6O 8O IOO I 20 I4O
4 - 54 4.. I 5 3.93 3. 78 3.5C) 3.4C) 3. I6 2.99
(I 7o) (15O) I 33 I4 I I 26 I3 I I 3O
Table XXXIb.
ba = 6. I6. Mean, 4343O Ka = I28.
O 23 4. I 6O 8O IOI I 2G) I4O
4. 26 3. 92 3.63 3.58 3.36 3. I4 - 3. O I 2 . 88
(16O) (170) I3O I 3 I I 27 I 26 I 2.2
I6O
3. 22
I IO
I6O
2.84
I 27
I6O
2.63
I3 I
+
43
Table XXXII.
I 2 3 4. 5 6
NO. Ba(OH)2. IOO Gºv. Io"X(OH). 4343OXA v. 'O.4343XA OH.
XXIXa 3O. 90 96 593 67. 6 O. II.4
XXIXb 32.2O 96 618 67.5 O. IO9
XXXa 46. 20 95 878 IO I O. II 5
XXXb 46.50 95 884 IO9 O. I23
XXXIa 61.50 94 II56 I 30 O. II 3
XXXIb 6I 6O 94. II 58 I 28 O . I II
Mean, O. 4343 KOH = O. II.4. KOH = O. 265.
5. Methyl Imido-m-nitrobenzoate.
Methyl imido-m-nitrobenzoate was prepared from its hy-
drochloride by treating this with three times the required
amount of a cold 30 per cent solution of sodium hydroxide and
rapidly extracting the base with ether. The extract was dried
and the ether evaporated at reduced pressure, leaving the
ester behind as a yellow solid. Upon crystallization from
low boiling ligroin, this was obtained as a white, crystalline
substance melting at 53°. Dissolved in hydrochloric acid
and heated, it yielded m-nitrobenzoic methyl ester, melting
at 70°. On account of the slight solubility of the ester in
water, one liter of a N/2OO solution was made up for each
velocity determination, and IOO ce. were used for each titra-
tion. It was impossible to titrate these solutions in the ordi-
nary way; the lack of a sharp change of tint, even with the
help of alizarin sulphonate, made it necessary to titrate in
Nessler tubes, using as a standard tint for the end point of the
reaction a faintly alkaline solution of alizarin Sulphonate.
Even under these conditions considerable difficulty was ex-
perienced in making the titrations at the beginning of the re-
action. The indistinct color change, mentioned above," is
not the only difficulty. In the case of the nitrobenzoate a
brownish compound seems to be formed as a result of an ac-
tion of the ester on the indicator. The formation of this
substance produces a color change similar to that produced
by making the acid solution alkaline, and only by very rapid
titration could the confusion resulting from this secondary
reaction be avoided. Below are given the results of the
velocity reactions; the symbols have the same significance
as in the previous experiments.
1 Page 29.
Table XXXIII.
ba = 3.27. Mean, 4343O Ka = 51O.
O I9 - 33 59 7O 83 IOO II 5
2. 57 2. O8 I .86 I. 3 I I. I 7 I. O I O. 79 O. 64
. . . . 5CO (440) 5OO 500 500 52O 53O
Table XXXIV.
ba = 4.37. Mean, 4343O Ka = 690.
O I 2 22 34. 42 55 66 8O 88
2. 32 I. 92 I . 68 I. 35 I. I5 O. 89 O. 78 O. 65 O. 46
* Q & 68O 64O 690 7OO (750) 72O 690 (790)
Table XXXV.
ba = 6. 25. Mean, 43430 K = 930.
O 2O 34 65 75 88 - IOO I [O
3. 50 2. 23 I. 55 O. 9 I
O. 73 O. 57 O. 35 O. 25
f
45
Table XXXVI.
I 2 3 4. 5 6
NO. Ba(OH)2." IOO Gº. IO"> (OH). 43430 A v. o.4343 KOH.
XXXIII 32.70 96 3I4. 5 IO I. 62
XXXIV 43.70 95 4 IS 690 I 66
XXXV 62.50 96 6OO 93O I. 55
Mean, O. 4343 KOH = I. 61. KOH = 3.7.
PART III.
The Velocity of Decomposition of Imidoesters in Aqueous Solu-
tion.
Besides the study of the velocity of decomposition of imido-
esters under the influence of added acids or alkalis, an inves-
tigation of the velocity of decomposition of methyl imidoben-
zoate in aqueous solution was undertaken, since only a few
preliminary determinations had been carried out by Derby.”
The experimental results are given below; they agree perfectly
with the preliminary determinations just mentioned. The
theoretical interpretations which should be placed on these
results is a question of peculiar interest. Qualitatively,
the chief reaction” in aqueous solution has been found to be
the same as in alkaline Solutions—a nitrile and an alcohol are
formed. Quantitatively, the velocity determinations of the
decomposition gave a constant according to the simple loga-
rithmic function,
Kester == log 11at #. (27).
This is the equation for a monomolecular reaction which pro-
ceeds proportionately to the concentration of the total sub-
stance. -
In an aqueous solution of a substance like methyl imido-
benzoate, which is a very weak base and also, as shown in
Part II., a still weaker acid, we must have both its positive
and its negative ions. Since these ions will be present only in
very small amounts, practically all of the imidoester must be
1 Column 2 represents in this table the ce. N| 10 Ba(OH)2 in 1 liter of the mixture.
2 Am. Chem. J., 39, 471.
* It is likely that an acid amide is to a small extent formed in a parallel reaction.
46
in the nonionized form. Calculations made with the aid of
the affinity constant for the base as determined by Derby"
show that in one-twentieth molar solution about o.o.3 per
cent of the free base is ionized and in one-hundredth molar
Solution less than O. I per cent is ionized. It is clear, there-
fore, that the concentration of the nonionized molecule may
be put equal to the concentration of the total substance; and
that consequently the constant obtained by the logarithmic
function (27) would be the true constant giving the velocity
of the reaction if the nonionized molecule forms the chief
reacting substance in these solutions and breaks down into
a nitrile and alcohol according to the equation
C.H.C(: NH)OCHA = C.H.C.; N + HOCH, (28).
It should be noted that although the products in this re-
action are also ultimately a nitrile and an alcohol, as in strongly
alkaline solutions, the reaction equations in the two cases
are different. In the latter instance, we have the reaction
expressed by equation (26),” which shows that the negative
ions are breaking down into nitriles and alcohol ions (OR).
As was to be expected, the characteristic velocity constants
for the two different reactions are different. Kon, the veloc-
ity constant for the decomposition of the negative ion when
one gram molecule of ester is acted upon by one gram ion of
hydroxyl was found to be I. 97 for methyl imidobenzoate,
and K_º, the constant calculated from (27), according to
which the nonionized molecule breaks down, is only O. OOO28.
So even if both the negative ions and the nonionized imido-
esters should give qualitatively the same ultimate products,
it is obvious that the negative ions are by far the more reac-
tive constituents in the solution; and, therefore, the rate of
change of the nonionized molecules would become of moment
only when, by an enormous advantage in concentration, their
decomposition becomes the fundamental one. This view,
of course, is the simplest possible one based on the fact that
the velocity of the decomposition of the methyl imidoben-
1 Loc. cit.
2 P. 26.
47
zoate in aqueous solution is found to be determined by the
logarithmic equation (27).
Before accepting this view, however, of the decomposition
of the nonionized molecule, we shall also consider the fact
that both the positive and the negative ions which must be
present in the solution must also be decomposing as estab-
lished in Parts I, and II., and we shall try to determine whether,
small as their concentrations are, their relatively very great
rates of decomposition cannot account for the actual velocity
of decomposition observed in aqueous solution.
Since the decomposition constant for the negative ion
is very much larger than for the positive imidoester ion, and
since the aqueous solution of the free base is strongly alkaline,
we might very well have a case of auto-catalysis in which
the reaction proceeds exactly as in the presence of hydroxyl
ions from barium hydroxide. On this account we shall first
consider whether the observed velocity of decomposition
can be accounted for as due simply to the decomposition of
the imidoester under the influence of the hydroxyl ions pro-
duced from itself. The fundamental equation which would
apply to such a reaction is
d
... =Koh X Cester X Coff (I3).
Whereas in the presence of barium hydroxide Con is a con-
stant throughout the course of the experiment and variable
only by way of different original concentrations of the hydrox–
ide, in the case under discussion Corf is a variable depending
on the concentration of the reacting ester itself. There are
two possibilities to be considered in endeavoring to establish
the nature of this dependence. First, the concentration of
the hydroxyl ion at any moment could be taken as dependent
essentially on the ionization of the imidoester as a base alone,
on the assumption that the acid character of the imidoester
is so exceedingly weak that formation of more positive ions
from a salt—which in turn would effect the concentration of
the hydroxyl ion—would be so slight as to be negligible in the
application of the general equation.
48
Ceos. ion X CoA. *º-
Cmol. --- Kaff. (29) e
On this assumption we would obtain a maximum concentra-
tion of the hydroxyl ions to be used in connection with equa-
tion (13), and this is quite sufficient for our purpose. On
the basis of such an assumption,
CoH = WK., X Cmol.
and putting Cao. - Cºe, we have
este?’
Con = NKaz. X Cº., (30).
Combining (30) and (13), we find
dº
di
= Koº, X (C–3) × VKaz × (C–2) (31).
This, on integration, becomes
Kory X Cester X VKaz X Cºle, y
=== Nº lºw-wºw
Koff - l, - tº Kaff. lº ..)". (C– 3:1)"? | (32).
The values for Korf, calculated according to this equation
(32), gave, indeed, a tolerably good constant for the whole
course of each individual experiment carried on through a
period of from seventy to eighty hours for each concentra-
tion; they did not, however, give the same constant for the
varying original dilutions of the ester; and the constants were
all very much larger than the same constant, Kon, determined
for the same ester by the action of barium hydroxide, as given
in Part II.
In the following table, the first line, after V, gives the vol.
ume in liters containing a gram molecule of the free imido-
ester; the second line, after K,…, gives the constant calcu-
lated according to equation (27), referring the active mass
of the changing substance to the nonionized molecule, and
the third line, after Kon, gives the average constants calcu-
lated according to (32), on the assumption that the reaction
in aqueous solution is a case of auto-catalysis, the hydroxyl
49
\
ion of the free, base catalyzing the ester in the same way as
barium hydroxide does.
Table XXXVII. i
V 2I. 7 47. O 83.3 I 25
Kmol. O. OOO28 O. OOO26 O. OOO26 O. OOO24.
KOH 22.6 3O . I 4O. 7 45. 2
The constant Koº obtained in alkaline solutions by the ad-
dition of varying amounts of barium hydroxide to the aqueous
solution of the free base is I. 97 (see p. 33, Table XX.).
Since Kory, as determined for the pure aqueous solution, is
very much larger" than I. 97 and is not a constant for different
dilutions, we are led to the conclusion that the reaction is not
a case of auto-catalysis and that the chief reacting substance
must be the nonionized molecule of the free base, breaking
down, very slowly but with a constant velocity, in the course
of seventy hours, into nitrile and alcohol. The important
point in this conclusion is the fact that, having established
in Parts I. and II. that both the positive and the negative
ions of the imidoesters show characteristic, specific reactions
proceeding with constant velocities, we now find that the
third form of the esters, the nonionized form, also is active,
changing exceedingly slowly, compared with the other forms,
but with a measurable velocity” according to the equation
C.H.,(: NH)OR = C.H.C.; N + HOR (33)
In this calculation of Cory, it will be recalled, the salt forma-
tion between the acid and the basic parts of the amphoteric
was considered negligible (in equation 29). This was done
because the basic functions are far more pronounced than the
acid functions, as is shown by the marked alkaline reaction
of the ester’s aqueous solution. Taking the amphoteric char-
acter of imidoesters into account, we would have the case of
1 The calculation was made on the assumption of a maximum possible concentra-
tion of OH ions (equation 29, p. 48), and KOH would be still larger if the concentration
of the OH ion were smaller, consequently the conclusions drawn in the text hold
independently of this assumption.
* The importance of this question of the decomposition of the nonionized mole-
cule is such that Mr. B. B. Freud will continue this part of the investigation to develop
a broader range of experiments.-J. STIEGLITZ.
5O
Salt formation between a weak base and a very weak acid.
Consequent hydrolysis of the salt would establish the condi-
tion of equilibrium expressed by the formulation
C.H.C(NH)OCH, + C.H.C(; NH)OCH, -
loase acid
[C.H.C(: NH)OCH.], (33).
Salt and acid formation take place, of course, at the imide
(: NH) group in the molecule. In such solutions, we would
have for the very weak acid:
Caſ X Curg ton
s – k ci ,
Cester acid (34)
and for the weak base
CoA. X Ceos. foºt 1
Cester - kbase (35) e
Combining (34) and (35), we have, as usual, for the hydroly-
sis of salts of weak acids and weak bases,
Cnes. ion X Ceos. #on -
Cºaster. *H,o
&acid X Æbase = } (36)
If the positive and negative ions are formed chiefly from the
salt, as in the case of aniline acetate," we may consider
Cneg. 70775 -- Cpos. ions,
and we would have
Cºneg. ion
Cºaster = K (37),
Cneg. Ž0775 - Æ'Cester (37').
In a word, if salt formation does occur in the case of such an
amphoteric substance, which is a weak base and a weak acid,
to an extent such as to make the salt, rather than the base or
acid, the chief source of the ions of the organic compound,
then the concentration of the negative ions would be in prac-
tically constant proportion to the concentration of the total
ester. Therefore, a reaction proceeding at a rate proportionate
1 The concentration of the nonionized molecule may be put equal to that of the
ester.
* Arrhenius: Z. physik. Chem., 5, 19 (1890).
5 I
to the total ester would proceed also proportionately to the
concentration of the negative ion. The reaction might also,
on this account, be a function of the concentration of the
negative ion, considered as the “active mass.” This pos-
sibility is actually realized in the experimental results for
the decomposition of methyl imidobenzoate in water solution
—the rate of change is actually proportional to the total ester.
This result in itself has justified us, in the study of
the theory of the case, in not being content with only the
first conclusion, viz., that the nonionized imidoester must
constitute the decomposing “active mass.”
It will be recalled also that the fundamental equation,
d t
. = Koh X Caster X CoA (I3),
is the experimental equation, identical with the one based
on the theory of catalysis in alkaline solution, according to
which negative ions represent the decomposing substance.
This latter equation was written
.dx
d; – Kion X Cneg. ion (23).
Theoretically, therefore, a reaction proceeding with a velocity
directly proportional to the concentration of the total ester
could be satisfied also by a constant representing the rate of
change as proportional to the concentration of the negative
ion of the amphoteric substance (equation 37'). The only
question to be answered for the substance under investiga-
tion is whether the actual rate of change, as experimentally
determined, agrees in value with that deduced from theory.
This test may be made as follows:
Above it was shown that on the basis of an assumption that
Coz, has a maximum value, the decomposition in aqueous so-
lution is far too rapid to be cousidered a case of auto-catalysis.
On the present assumption that the concentration of the posi-
tive ester ion is perceptibly increased by salt formation, Conſ
(equation 29) and consequently also Cº. ºo, (equation 22)"
must have still smaller values: the reaction ought to proceed
1 This was established in Part II. and the whole argument rests on that work.
52
more slowly and this would be in still greater discrepancy with
the experimental fact.
We are obliged to conclude again, therefore, that in aqueous
Solution methyl imidobenzoate decomposes into nitrile and
alcohol with a speed too great to be accounted for on the as-
Sumption that it is proportional to the concentration only
of the negative ion, as was found to be the case in strongly
alkaline solutions.
The decomposition, which occurs simultaneously with that
of the negative ion, of the positive ion into ammonia and ester
(Part I.) is far too slow, in the minimal concentrations in which
it is present, to account for this excessive speed. The
velocity constant for that change is o. oO63 for methyl imido-
benzoate (see Part I., p. 21, Table XIV., column 4). For a
twentieth molar solution, the original concentration of the
positive ion calculated from the affinity constant,
2 *mº –8
C zons T O. 5 X IO X Cºol.
is
Cº., = O.OOOOI5.
10%.
The original rate of change of the ion referred to a whole
gram molecule of the ester would be
O. OOOOI5 X O. OO6 X 20 = O.OOOOOI8,
a quantity which is negligible in accounting for the observed
rate of change of one gram molecule total substance, viz.,
o.OOO28.*
EXPERIMENTAL.
The analytical method of following the decomposition of
methyl imidobenzoate in aqueous solution was practically
the one employed for the study of the decomposition of the
same substance in strongly alkaline Solution (see p. 29).
Since the reaction proceeded very slowly, titrations were made
at long intervals and special care was taken that no varia-
tions should occur in the temperature, the bath being kept
at 25° (+o°. OI) for several days. The logarithmic equation
1 Only traces of ammonia and ester are formed, as a matter of fact.
53
(27) was used for obtaining the constants, Kaoſ, given in the
third columns of the following table, and equation (32) was
used for the determinations of the values of Kon (the fourth
column). For these calculations, equation (32) is best trans-
formed as follows: Calling the original concentration of the
base I gram molecule in V liters and x, the part of a gram
molecule changed in t minutes, we have:
I I — ºcł
C = y, and C — 3 = V
If v, cc. and v, cc. acid were used to neutralize the free base.
present in a measured amount of solution when t = 0 and
t = t (e. g., 4.62 cc. and 4.34 cc., in Table XXXVIII., when
t = 0 and t = 288), then the unconsumed part, (1–3), of the
free base may be expressed by
‘Ut
I — 3: = −.
‘U
O
Equation (32) acquires the very simple form:
...— ... Y. Wi- | f
Kon-i Wº...! 'U/ (32').
Column I, in the following tables, gives the time in minutes;
column 2, the number of cc. o. I normal acid required to neu-
tralize w cc. of the solution.
Table XXXVIII.
V = 21.7. 70) = I O CC.
i. vi. 43430 A mol. KOH. t. vi. 43430 A mol. KOH.
O 4.62 * & E & tº e s & I963 2.60 I 2.7 .22.6
288 4.34 (9. O) (14.6) 2948 2. OI I 2.7 23, 2
338 4. I8 I 2.8 2O. 3 3 I69 I. 8O I2.9 25. 2
438 4. I.3 II. I I 7.5 336O I. 76 I 2.4. 24.6
I 52 I 3. OO I 2.3 2 I. O 429O I. 44 II. 8 24. 5
1684 2.88 I 2 .. 2 20.9 4635 I.33 II .. 8 23. 7
Mean, Knoi. = O.OOO28. Mean, Koh = 22.4.
54
Table XXXIX.
V = 47. w = Io ce. r
i. Z/t. 43430 Kmol. KOH. t. vi. 43430 Å mol. KOH.
O 2. I3 • * * d e tº gº tº I935 I. 24 I 2. O 3I. I
266. I. 92 (17. O) (37.4) 2826 I. O8 IO. 4. 25.8
4I4 - I.87 (13.6) 31.7 3143 O. 99 IO. 6 29. 5
I498 I.46 II ... O 27. I 3338 O. 90 II. 2 3I. 5
I66O I. 39 II . I 28. O 4265 O. 69 II . 4. 34. 7
I765 I. 36 II ... O 27.9 4553 O. 67 II ... O 33. 7
| Mean, Knot. - O.OOO26. Mean, Kofi = 30. I.
Table XL.
V = 83.3. 70 = 25 cc.
t. vt. 43430 A mol. Koh. #. vi. 43430 A mol. A OFI.
O 3. OO e is e s g g º ºs I926 I. 80 II. 5 39. 3
I22 2.90 I 2. O 36. O 2817 I. 42 II . I 4 I. 9
295 2.84 (9.0) 42.2 31.91 I. 34 IO . I 4O .4
464 2.54 (I5.5) 48.7 3327 I. 24 II. 2 43.4
I489 2. OO I 2 ... O 39.3 426O I. OO II. 2 44. 7
I65I I. 94 II . 4. 36.3 4614 O. 95 IO. 8 43.6
Mean, Knoi. = O.OOO26. Mean, Kofi = 40.7.
Table XLI.
V = I25. w = 50 cc.
t. Ot. 43430 A mol. KOH. t. vi. 43430 A mol. A OH.
O 4. OO . . . . tº tº g tº I78o 2.57 IO. 8 44. 7
2O2. 3. 77 II. O 47.6 2793 2. O9 IO . I 44. O
437 3.67 (8.5) (32.25) 3165 I. 91 Io. I 44.9
I46O 2. 90 9. O 38.3 33OI I. 85 IO. I 45.6
I653 2.67 IO. I 43.4 4233 I. 49 IO . I 48. 3
* 4586 I. 35 IO .. 2 5O .4
Mean, Kmol. = O.OOO24. Mean, Koh = 45. 2.
It will be seen that Kºo, does not vary with V, its average
being O.OOO26." The relation existing between V and Kon
as well as between this Kou and the Kou determined from
the velocity in strongly alkaline solutions, has been fully dis-
cussed. | i -
The experimental data of this investigation have justified
us in certain definite conclusions, which shall be briefly stated
here:
1 Derby found 0.00028.
55
I. The speed of the decomposition of the imidoester salts
by water is dependent on their positive ions as is shown in Part I.
The velocity is proportional to the concentration of the ions,
but is also influenced by a “salt effect,” which is due to the
imidoester salt itself. We have here, in other words, an
“auto-salt effect.” The reaction is represented by the equa-
tion
+ +
RC(: NH)OR' + H.O = RCOOR' + NH,
II. The imidoesters are amphoteric in character; the nega-
tive ions decompose into nitriles and alcohol (ions) with a
velocity proportional to their concentration. Since the acid
ionization is very slight, the concentration of negative ions
is proportional to that of the ester and that of the hydroxyl
ions, and therefore the action has the appearance of a SO-
called “catalytic” action of hydroxyl on the molecular ester.
The reaction is represented by the equation
RCGN)OR – RCN + OR'.
III. The nonionized molecules appear to decompose also
into nitriles and alcohols with a much smaller velocity than
that exhibited by the negative ions. The reaction is given by
the equation
RC(; NH)OR = RCN + HOR.
In conclusion, the writer takes pleasure in acknowledging
gratefully his appreciation of Dr. Stieglitz's suggestive guid-
ance of his work, and of the interest he has taken in it.
ºst ºt
; , ; * ****
§
#: * {
IVERSITY OF MICHIGAN
A
i
|
3
-4 & 3 is
$
& 1 &
..;;:,'º. :::
* r * : * * * *
}
i
90.15 O7709 6280
$44,168. ...,b13",

$j.}）;
$！ºcº
$$$$$$$$
{{ſae}}
!, și •
§¶•@
5 #4
*æș**·
* ſ;.&&
*ae
:：-
frº-
r: ---
º:
*ș
！«ºſº•
§§ș
！ſ.$§§
Œ