Kinetics and mechanism of the cleavage of phthalic anhydride in glacial acetic acid solvent containing aniline

Article abstract of DOI:10.1021/jo8011072

(Chemical Equation Presented) Apparent second-order rate constants (k _n^app) for the nucleophilic reaction of aniline (Ani) with phthalic anhydride (PAn) vary from 6.30 to 7.56 M^-1 s^-1 with the increase of temperature from 30 to 50°C in pure glacial acetic acid (AcOH). However, the values of pseudo-first-order rate constants (k _s) for the acetolysis of PAn in pure AcOH increase from 16.5 × 10^-4 to 10.7 × 10^-3 s^-1 with the increase of temperature from 30 to 50°C. The values of k_n^app and k_s vary from 5.84 to 7.56 M^-1 s^-1 and from 35.1 × 10^-4 to 12.4 × 10^-4 s^-1, respectively, with the increase of CH₃CN content from 1% to 80% v/v in mixed AcOH solvents at 35°C. The plot of k_s versus CH ₃CN content shows a minimum (with 10⁴ k_s = 4.40 s^-1) at 50% v/v CH₃CN. Similarly, the variations of k_n^app and k_s with the increasing content of tetrahydrofuran (THF) in mixed AcOH solvent reveal respective a maximum (with k_n^app = 17.5-15.6 M^-1 s^-1) at 40-60% v/v THF and a minimum (with k_s = ～ 0-1.2 ～ 10^-4 s^-1) at 60-70% v/v THF. The respective values of ΔH* and Δ S * are 15.3 ± 1.2 kcal mol^-1 and -20.1 ± 3.8 cal K^-1 mo^-1 for k_s and 1.1 ± 0.5 kcal mol^-1 and -51.2 ± 1.7 cal K^-1 mol^-1 for k_n^app, while the values of k_n (= k _n^app/f_b with f_b representing the fraction of free aniline base) are almost independent of temperature within the range 30-50°C. A spectrophotometric approach has been described to determine f_b in AcOH as well as mixed AcOH-CH₃CN and AcOH-THF solvents. Thus, the observed data, obtained under different reaction conditions, have been explained quantitatively. An optimum reaction condition, within the domain of present reaction conditions, has been suggested for the maximum yield of the desired product, N-phenylphthalamic acid.

Full text of DOI:10.1021/jo8011072

Kinetics and Mechanism of the Cleavage of Phthalic Anhydride in
Glacial Acetic Acid Solvent Containing Aniline
Emmy Fadhiza Damit, Azhar Ariffin, and M. Niyaz Khan*
Department of Chemistry, Faculty of Science, UniVersity of Malaya, 50603 Kuala Lumpur, Malaysia
niyaz@um.edu.my
ReceiVed May 22, 2008
Apparent second-order rate constants (knapp) for the nucleophilic reaction of aniline (Ani) with phthalic
anhydride (PAn) vary from 6.30 to 7.56 M-1 s-1 with the increase of temperature from 30 to 50 °C in
pure glacial acetic acid (AcOH). However, the values of pseudo-first-order rate constants (ks) for the
acetolysis of PAn in pure AcOH increase from 16.5 × 10-4 to 10.7 × 10-3 s-1 with the increase of
temperature from 30 to 50 °C. The values of kn and ks vary from 5.84 to 7.56 M-1 s-1 and from 35.1
app
× 10-4 to 12.4 × 10-4 s-1, respectively, with the increase of CH3CN content from 1% to 80% v/v in
mixed AcOH solvents at 35 °C. The plot of ks versus CH3CN content shows a minimum (with 104 ks )
4.40 s-1) at 50% v/v CH3CN. Similarly, the variations of kn and ks with the increasing content of
app
app
tetrahydrofuran (THF) in mixed AcOH solvent reveal respective a maximum (with kn ) 17.5-15.6
M-1 s-1) at 40-60% v/v THF and a minimum (with ks ) ∼ 0-1.2 × 10-4
s
-1) at 60-70% v/v THF.
The respective values of ∆H* and ∆ S * are 15.3 ( 1.2 kcal mol-1 and -20.1 ( 3.8 cal K-1 mol-1 for
ks and 1.1 ( 0.5 kcal mol-1 and -51.2 ( 1.7 cal K-1 mol-1 for knapp, while the values of kn () knapp/fb
with fb representing the fraction of free aniline base) are almost independent of temperature within the
range 30-50 °C. A spectrophotometric approach has been described to determine fb in AcOH as well as
mixed AcOH-CH3CN and AcOH-THF solvents. Thus, the observed data, obtained under different
reaction conditions, have been explained quantitatively. An optimum reaction condition, within the domain
of present reaction conditions, has been suggested for the maximum yield of the desired product,
N-phenylphthalamic acid.
Introduction
reaction conditions for maximum yields of the desired organic
products. Recently, Perry and Parveen2 carried out a systematic
kinetic study of the cyclization of substituted phthalanilic acids
to N-arylphthalimide derivatives in glacial acetic acid solution.
These authors have provided mechanistic details of these
Diagnosis of fine molecular details of mechanisms of reac-
tions carried out in mixed aqueous-organic cosolvents and pure
organic solvents is complicated by the uncertainty of the
measurements of pH of buffers and pKa or pKb of acids or bases
in such solvents. Partly because of this and related reasons, very
limited reports could be found in this area of research.1
However, organic and mixed organic solvents are generally the
best reaction solvents for the synthesis of organic compounds.
The kinetic study provides the mechanistic details of a reaction
at the molecular level and hence such kinetic studies in organic
and mixed organic solvents are expected to give the optimum
1
reaction processes where kinetic and H NMR spectrometric
evidence have been presented for the presence of reactive
intermediate (1) on the reaction path. Perhaps it is important to
note that the intermediate similar to 1 could not be detected in
several studies on the cleavage of phthalanilic and substituted
phthalanilic acids in pure and mixed acidic aqueous-organic
cosolvents.3-6 N-Alkyl and N-arylphthalimides have been
conveniently synthesized in good yield by heating or refluxing
(1) Godenschwager, P. F.; Collum, D. B. J. Am. Chem. Soc. 2007, 129, 12023,
and references therein.
(2) Perry, C. J.; Parveen, Z. J. Chem. Soc., Perkin Trans. 2001, 2, 512.
(3) Hawkins, D. M. J. Chem. Soc., Perkin Trans. 1976, 2, 642.
10.1021/jo8011072 CCC: $40.75
Published on Web 10/28/2008
2008 American Chemical Society
J. Org. Chem. 2008, 73, 9303–9313 9303

Damit et al.
SCHEME 1
the reaction mixture of phthalic anhydride and alkyl or arylamine
in glacial acetic acid.7 In view of the published related studies,
the expected reaction scheme for the formation of N-substituted
phthalimide from the reaction of phthalic anhydride (PAn) with
a primary amine, RNH2, in acidic solution is shown by Scheme
1 where the solid lines are drawn through the least-squares
calculated data points using eq 1 and kinetic parameters, ks
and knapp, listed in Table 1.
It is evident from eq 1 as well as product characterization
study that a brief reaction scheme for the cleavage of PAn in
glacial acetic acid containing Ani may be expressed by Scheme
2 where PAn and Ani represent phthalic anhydride and
nonprotonated aniline, respectively.
1
1. Mechanistic details of the k2 -step for the cleavage of
N-phenylphthalamic acid in glacial acetic acid at 85 °C are
already reported.2
Determination of Fraction of Free Aniline Base (fb) in
Glacial Acetic Acid Solution. It is practically almost impossible
to determine accurately the pKa of an acid such as anilinium in
an organic solvent (such as glacial acidic acid) of dielectric
constant (ε) ∼6.8 Under such circumstances, a quantitative
interpretation of kinetic data of Figure 1 is apparently difficult
to achieve. The linearity of the plots of Figure 1 shows that the
fraction of free base, [Ani]/[Ani]T, (where Ani represents
nonprotonated aniline) remained unchanged with the change of
[Ani]T within the range of [Ani]T attained in the study. The
linearity of the plots in Figure 1 would be necessary but not a
sufficient condition to state that the fraction of free base as a
function of [Ani]T remains constant. The aqueous pKa values
The present study was initiated with the aims (i) to find an
experimental approach to determine the fraction of free amine
base (fb) in nonaqueous solvents where conventional technique
involving pH measurements becomes unable to provide such
information (i.e., fb), (ii) to find the mechanistic details of the
reaction of phthalic anhydride with aniline in glacial acetic acid
1
solution (i.e., k2 -step of scheme 1), and (iii) to determine an
optimum condition for maximum yield of product N-phe-
nylphthalamic acid. The observed data and their plausible
explanation(s) are described in this manuscript.
Results
Effects of [Aniline] on the Cleavage of Phthalic Anhy-
dride (PAn) in Glacial Acetic Acid Solution. Several runs were
carried within the total aniline concentration [Ani]T, range 5.0
× 10-4 - 4.0 × 10-3 M at a constant temperature in glacial
acetic acid solution containing 1.5 × 10-4 M phthalic anhydride.
Pseudofirst-order rate constant (kobs) are shown graphically
(Figure 1) at different temperatures ranging from 30 to 50 °C.
The values of kobs at different [Ani]T and at a constant
temperature were found to fit to eq 1
kobs ) ks + knapp[Ani]T
(1)
where ks and knapp represent respective pseudo-first-order rate
constant for solvent-assisted and second-order rate constant
for aniline-assisted cleavage of PAn. The least-squares
app
calculated values of ks and kn at different temperature are
summarized in Table 1. the reliability of the observed data
fit to eq 1 is evident from the standard deviations associated
with ks and knapp values (Table 1) and from the plots of Figure
(4) Blackburn, R. A. M.; Capon, B.; Mcritchie, A. C. Biorg. Chem. 1977, 6,
71.
(5) Grandos, A. M.; de Rossi, R. H. J. Org. Chem. 2001, 66, 1548.
(6) (a) Leng, S.-Y.; Ariffin, A.; Khan, M. N. Int. J. Chem. Kinet. 2004, 36,
316. (b) Sim, Y.-L.; Ariffin, A.; Khan, M. N. Int. J. Chem. Kinet. 2006, 38,
746.
(7) (a) Sim, Y.-L.; Ariffin, A.; Khan, M. N. J. Org. Chem. 2007, 72, 2392.
(b) Cheong, M.-Y.; Ariffin, A.; Khan, M. N. Indian J. Chem. 2005, 44A, 2055.
(c) Ahmad, W. H. W.; Ariffin, A.; Khan, M. N. Indian J. Chem. 2007, 46A,
416.
FIGURE 1. Plots showing the dependence of kobs versus [Ani]T at
different temperature: 30 ([), 35 (∆), 35 (*), 40 (O), 45 (2), and
50 °C (0) in AcOH. The solid lines are drawn through the calculated
data points using eq 1 and the kinetic parameters, ks and knapp listed
in Table 1.
9304 J. Org. Chem. Vol. 73, No. 23, 2008

CleaVage of Phthalic Anhydride
TABLE 1. Values of Rate Constants, ks and knapp, Calculated from Equation 1 for the Cleavage of PAn in Pure and Mixed AcOH-CS
Solvents Containing Ania
d
e
f
CSb (% v/v)
104 ks (s-1
)
10 kn (M-1 s-1
)
Nc
fb
10 kn (M-1 s-1
)
[Ani]T range
app
AcOH (% v/v)
T (°C)
100
100
100
100
100
100
100
100
99
95
90
85
80
75
70
60
50
40
30
20
10
95
90
85
80
75
70
60
50
40
40
35
30
30
25
20
20
15
10
10
30
30
35
35
40
40
45
50
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
0
0
0
0
0
16.5 ( 3.0g
13.8 ( 2.1
38.2 ( 2.2
29.9 ( 3.6
54.2 ( 4.0
51.5 ( 2.6
81.4 ( 7.3
107 ( 6
63.0 ( 1.3g
62.2 ( 0.9
61.0 ( 0.9
66.9 ( 1.5
62.3 ( 1.8
71.6 ( 1.1
69.2 ( 3.0
75.6 ( 3.0
58.4 ( 1.7
59.3 ( 0.5
59.5 ( 1.4
63.9 ( 1.1
61.7 ( 1.0
61.8 ( 1.9
60.7 ( 1.3
69.1 ( 0.7
75.6 ( 1.6
77.1 ( 1.6
80.5 ( 1.4
75.6 ( 1.7
54.7 ( 1.3
78.6 ( 1.3
86.8 ( 2.7
101 ( 5
6
6
6
6
6
6
6
6
10
10
9
10
9
10
10
9
9
8
10
7
6
10
10
10
8
8
7
7
7
7
7
6
6
6
6
6
6
6
6
0.146
0.132
0.158
0.163
0.18
432
471
386
410
346
398
348
327
408
368
358
323
292
281
264
234
204
166
145
104
70.2
385
384
365
369
349
329
285
232
195
190
153
124
133
98.8
79.5
81.6
58.6
39.3
38.5
5.0 × 10-4-4.0 × 10-3
5.0 × 10-4-4.0 × 10-3
5.0 × 10-4-4.0 × 10-3
5.0 × 10-4-4.0 × 10-3
5.0 × 10-4-4.0 × 10-3
5.0 × 10-4-4.0 × 10-3
5.0 × 10-4-4.0 × 10-3
5.0 × 10-4-3.0 × 10-3
5.0 × 10-4-4.0 × 10-3
5.0 × 10-4-4.0 × 10-3
5.0 × 10-4-4.0 × 10-3
5.0 × 10-4-4.0 × 10-3
5.0 × 10-4-4.0 × 10-3
5.0 × 10-4-4.0 × 10-3
5.0 × 10-4-4.0 × 10-3
5.0 × 10-4-3.0 × 10-3
5.0 × 10-4-3.0 × 10-3
5.0 × 10-4-2.5 × 10-3
3.0 × 10-4-2.0 × 10-3
3.0 × 10-4-1.5 × 10-3
3.0 × 10-4-1.3 × 10-3
5.0 × 10-4-4.0 × 10-3
5.0 × 10-4-4.0 × 10-3
5.0 × 10-4-3.0 × 10-3
5.0 × 10-4-3.0 × 10-3
5.0 × 10-4-2.5 × 10-3
3.0 × 10-4-2.0 × 10-3
3.0 × 10-4-1.5 × 10-3
3.0 × 10-4-1.3 × 10-3
3.0 × 10-4-1.3 × 10-3
3.0 × 10-4-1.3 × 10-3
3.0 × 10-4-1.0 × 10-3
3.0 × 10-4-1.0 × 10-3
3.0 × 10-4-1.0 × 10-3
3.0 × 10-4-1.0 × 10-3
3.0 × 10-4-1.0 × 10-3
3.0 × 10-4-1.0 × 10-3
3.0 × 10-4-1.0 × 10-3
3.0 × 10-4-1.0 × 10-3
3.0 × 10-4-1.0 × 10-3
0
0
0.18
0.199
0.231
0.143
0.161
0.166
0.198
0.211
0.22
0
1h
35.1 ( 3.6
32.5 ( 3.6
32.5 ( 1.1
31.7 ( 2.9
23.9 ( 2.4
20.4 ( 4.0
18.3 ( 2.8
6.94 ( 0.12
4.40 ( 2.80
9.70 ( 2.50
7.50 ( 1.60
12.4 ( 1.6
5.30 ( 1.00
30.9 ( 2.8
37.3 ( 5.6
30.9 ( 9.3
17.2 ( 2.5
13.2 ( 2.8
13.2 ( 2.9
8.1 ( 1.3
6.9 ( 2.0
-3.0 ( 1.8
0
5h
10h
15h
20h
25h
30h
40h
50h
60h
70h
80h
90h
5i
0.23
0.295
0.371
0.465
0.554
0.729
0.779
0.204
0.226
0.277
0.328
0.404
0.459
0.613
0.737
0.809
0.811
0.849
0.889
0.837
0.904
0.786
0.828
0.836
0.815
0.794
10i
15i
20i
25i
30i
40i
50i
60i
60i
65i
70i
70i
75i
80i
80i
85i
90i
90i
121 ( 2
141 ( 2
151 ( 2
175 ( 1
171 ( 3
158 ( 2
154 ( 3
3.8 ( 1.3
1.2 ( 32
130 ( 2
110 ( 5
2.7 ( 0.7
2.5 ( 0.7
4.3 ( 1.5
1.7 ( 0.8
1.2 ( 0.5
0.4 ( 0.6
1.0 ( 0.4
111 ( 1
89.3 ( 1.1
62.5 ( 14.8
67.6 ( 1.3
49.0 ( 0.7
32.0 ( 0.9
30.6 ( 0.7
6
a [PAn0] ) 1.5 × 10-4 M. b CS ) cosolvent. c Total number of kinetic runs. d The values of fb were calculated from eq 6 or eq 10 with δAni ) 1175
M-1 s-1 and δob values listed in Table 2. e The values of kn were calculated from eq 18 (i.e., kn ) knapp/fb). f Total concentration range of aniline. g Error
limits are standard deviations. h CS ) CH3CN. i CS ) THF.
of benzoic acid, acetic acid, and anilinium ion are 4.20,8 4.76,9
and 4.59,8 respectively. The pKa values of benzoic acid and
acetic acid in dimethyl sulfoxide, DMSO (an aprotic organic
solvent of ε about 48) are 11.0 and 12.3, respectively,8 but the
pKa value of morpholinium ion in DMSO and aqueous solution
are 8.78 and 8.36,9 8.82,10 respectively. The ionization reactions
of anilinium ion and acetic as well as benzoic acid are termed
isoelectric and nonisoelectric ionization reactions, respectively.
The increase in acetonitrile content from 2% to 70% v/v in
mixed aqueous solvents increased the pKa of benzylammonium
and phenol from 9.16 to 9.26 and 10.17 to 13.38, respectively.11
In view of these reports,8-11 it is certain that the pKa values of
acetic acid and anilinium ion must be higher in a solvent of ε
of ∼6 (glacial acetic acid) than that of ∼78.5 (water solvent).
Significant reactivity of aniline with PAn in glacial acetic acid
solution, where [acetic acid]T/[aniline]T is in the range 3.4 ×
104-4.3 × 103 M, indicates that the pKa of anilinium ion is
lower than that of acetic acid under such conditions.
Since the dielectric constant (ε) of glacial acetic acid solution
is ∼6 and according to a very rough rule, solutions free of ion-
pairing or other ionic aggregates at ionic concentrations in
millimolar range require solvents with dielectric constants
greater than 30,8 it is not possible to determine the pKa of
anilinium ion and acetic acid in a solvent of ε of ∼6. However,
the values of fb under a variety of reaction conditions have been
determined as follows. The values of absorbance (A0) at the
reaction time t ) 0 were obtained at 275 nm from eq 2
Aob ) δapp[R0][1 - exp(-kobst)] + A0
(2)
where Aob is the absorbance at 275 nm and any reaction time t
and [R0], δapp, kobs, and A0 represent the initial concentration of
PAn, apparent molar extinction coefficient of the reaction
mixture, pseudo-first-order rate constant and initial absorbance,
respectively. Beer’s law (i.e., eq 3) has been used to calculate
the molar extinction coefficient (δob) of aniline at 275 nm in
96% v/v water solvent containing 4% v/v CH3CN and 0.1 M
in both NaOH and HCl. The values of δob turned out to be ∼0
and 1175 ( 42 M-1 cm-1 at 0.1 M HCl and 0.1 M NaOH,
(8) Rithchie, C. D. Physical Organic Chemistry, 2nd ed.; Marcel Decker,
Inc.: New York, Basel, 1990.
(9) Cordes, E. H.; Jenks, W. P. J. Am. Chem. Soc. 1962, 84, 4319.
(10) Knier, B. L.; Jencks, W. P. J. Am. Chem. Soc. 1980, 102, 6789.
(11) Khan, M. N.; Ariffin, Z.; George, A.; Wahab, I. A. Int. J. Chem. Kinet.
2000, 32, 146.
J. Org. Chem. Vol. 73, No. 23, 2008 9305

Damit et al.
TABLE 2. Values of r and δob, Calculated from Equation 3 Using A0
d
AcOH (% v/v)
CSa (% v/v)
104 PAnb (M)
102 R
δob M-1 cm-1
Nc
T (°C)
[Ani]T range (M)
0
0
0
0
0
96e
96e
96g
100h
100i
100i
0
0
0
0
0
0
0
0
0
0
0
1175 ( 42f
1176 ( 51
0.8 ( 0.6
955 ( 3
1008 ( 3
1022 ( 10
197 ( 1
214 ( 2
236 ( 1
258 ( 2
285 ( 2
171 ( 2
155 ( 6
186 ( 6
191 ( 2
212 ( 4
211 ( 6
234 ( 5
272 ( 12
168 ( 2
189 ( 1
195 ( 12
233 ( 4
248 ( 13
258 ( 3
270 ( 1
347 ( 13
436 ( 12
546 ( 13
651 ( 13
857 ( 8
915 ( 17
240 ( 3
66 ( 7
325 ( 10
385 ( 3
475 ( 7
539 ( 2
720 ( 6
866 ( 5
950 ( 8
953 ( 7
998 ( 10
1045 ( 28
984 ( 13
1062 ( 22
923 ( 48
973 ( 10
982 ( 43
958 ( 42
933 ( 27
4
3
6
9
8
8
6
6
6
6
6
6
6
6
6
6
6
35
35
35
35
35
35
30
35
40
45
50
30
30
35
35
40
40
45
50
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
5.0×10-4-2.0×10-3
5.0×10-4-1.5×10-3
5.0×10-4-4.0×10-3
3.0×10-4-2.0×10-3
3.0×10-4-1.8×10-3
3.0×10-4-1.8×10-3
5.0×10-4-4.0×10-3
5.0×10-4-4.0×10-3
5.0×10-4-4.0×10-3
5.0×10-4-4.0×10-3
5.0×10-4-4.0×10-3
5.0×10-4-4.0×10-3
5.0×10-4-4.0×10-3
5.0×10-4-4.0×10-3
5.0×10-4-4.0×10-3
5.0×10-4-4.0×10-3
5.0×10-4-4.0×10-3
5.0×10-4-4.0×10-3
5.0×10-4-3.0×10-3
5.0×10-4-4.0×10-3
5.0×10-4-4.0×10-3
5.0×10-4-4.0×10-3
5.0×10-4-4.0×10-3
5.0×10-4-4.0×10-3
5.0×10-4-4.0×10-3
5.0×10-4-4.0×10-3
5.0×10-4-3.0×10-3
5.0×10-4-3.0×10-3
5.0×10-4-2.5×10-3
3.0×10-4-2.0×10-3
3.0×10-4-1.5×10-3
3.0×10-4-1.3×10-3
5.0×10-4-4.0×10-3
5.0×10-4-4.0×10-3
5.0×10-4-3.0×10-3
5.0×10-4-3.0×10-3
5.0×10-4-2.5×10-3
5.0×10-4-2.0×10-3
5.0×10-4-1.5×10-3
3.0×10-4-1.3×10-3
3.0×10-4-1.0×10-3
3.0×10-4-1.3×10-3
3.0×10-4-1.0×10-3
3.0×10-4-1.0×10-3
3.0×10-4-1.0×10-3
3.0×10-4-1.0×10-3
3.0×10-4-1.0×10-3
3.0×10-4-1.0×10-3
3.0×10-4-1.0×10-3
3.0×10-4-1.0×10-3
3.0×10-4-1.0×10-3
1.04 ( 0.14f
4.1 ( 0.4
0.9 ( 0.4
0
0
100
100
100
100
100
100
100
100
100
100
100
100
100
99
95
90
85
80
75
70
60
50
40
30
20
10
95
90
85
80
75
70
60
50
40
40
35
30
30
25
20
20
15
10
10
4.9 ( 0.2
3.0 ( 0.4
3.3 ( 0.2
3.7 ( 0.4
3.3 ( 0.5
34.3 ( 0.4
34.3 ( 1.5
41.1 ( 1.3
32.1 ( 0.6
37.1 ( 0.9
31.5 ( 1.4
29.9 ( 1.1
26.7 ( 2.4
20.8 ( 0.4
20.5 ( 0.3
20.8 ( 2.6
17.9 ( 0.7
17.7 ( 2.8
18.1 ( 0.7
17.2 ( 2.9
10.3 ( 2.2
8.2 ( 2.1
9.4 ( 2.0
12.6 ( 1.5
13.4 ( 0.8
24.5 ( 1.3
20.5 ( 0.5
22.6 ( 1.4
21.3 ( 1.8
20.8 ( 0.5
19.5 ( 1.1
21.0 ( 0.3
21.0 ( 0.5
21.2 ( 0.4
22.5 ( 0.6
21.6 ( 0.6
21.0 ( 0.7
19.2 ( 1.9
23.7 ( 0.8
26.1 ( 1.4
27.5 ( 3.1
28.3 ( 0.6
28.1 ( 2.8
24.2 ( 2.7
28.1 ( 1.7
0
0
0
0
0
0
0
0
0
0
0
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
0
0
0
6
5
0
1j
10
10
10
10
10
10
10
9
9
8
10
7
6
10
10
9
8
8
7
7
7
7
5j
10j
15j
20j
25j
30j
40j
50j
60j
70j
80j
90j
5k
10k
15k
20k
25k
30k
40k
50k
60k
60k
65k
70k
70k
75k
80k
80k
85k
90k
90k
6
6
5
6
6
6
6
6
6
6
a CS ) cosolvent. b PAn ) phthalic anhydride. c Total number of kinetic runs. d Total concentration range of aniline. e Water cosolvent contains 4%
v/v CH3CN and 0.1 M NaOH. f Error limits are standard deviations. g Water cosolvent contains 4% v/v CH3CN and 0.1 M HCl. h Pure THF solvent.
i Pure CH3CN solvent. j CS ) CH3CN cosolvent. k CS ) THF.
SCHEME 2
[(A0i s Acaldi)/A0i] where A0i and Acalcdi represent, respectively,
the observed and calculated absorbance at ith [Ani]T) of e0.7%.
Similar observations were also obtained in 100% THF solvent.
In eq 3, R represents absorbance due to solvent and other
additives (if present). The least-squares calculated values of R
and δob at 275nm and 100% CH3CN and THF are summarized
in Table 2.
A0 ) R + δob[Ani]T
(3)
respectively. These results show that the values of molar
extinction coefficient of free aniline base (δAni) and protonated
It is known from the literature that the absorption spectra
may or may not change in different solvents. In order to find
out the effects of mixed H2O-CH3CN and H2O-THF solvents
on the absorption spectra of aniline (Ani), a series of UV
absorption spectra was obtained at 35 °C for the mixed aqueous
aniline (δAniH+) at 275 nm are 1175 and ∼0 M-1 cm-1
,
respectively. The values of A0, obtained at 275 nm and within
[Ani]T in the range 3.0 × 10-4-1.8 × 10-3 M in 100% CH3CN
solvent fit to eq 3 with percent residual error (RE ) 100 ×
9306 J. Org. Chem. Vol. 73, No. 23, 2008

CleaVage of Phthalic Anhydride
solutions containing 1.0 × 10-3 M Ani and CH3CN content
ranging from 1% to 60% v/v at 0.1 M NaOH and 1% to 90%
v/v at 0.1 M HCl. Similar observations were obtained in
H2O-THF solvents. The values of Aob at 275 nm as a function
of the content of CH3CN and THF are summarized in Table 1
in Supporting Information. It is evident from these results that
there is ∼8.4% decrease in Aob with the increase in CH3CN
content from 1% to 60% at 0.1 M NaOH. Similarly, Aob values
decrease by ∼4.6% with the increase in THF content from 30%
to 50% v/v, while the change in THF content from 1% to 30%
v/v has no effect on Aob at 0.1 M NaOH. The values of Aob
remained almost unchanged at nearly zero with change in the
contents of both CH3CN and THF from 1% to 90% v/v at 0.1
M HCl (CH3CN) and 1.0 M HCl (THF).
By definition, fb ) [Ani]/[Ani]T with [Ani]T ) [Ani] +
[AniH+]. The absorbance (Aob) of aqueous solution containing
Ani and AniH+ only (where absorbance due to solvent is zero)
may be expressed as
Aob ) δAni[Ani] + δAnH[AniH+]
FIGURE 2. Plots showing the dependence of A0 versus [Ani]T at
different temperatures: 30 (]), 35 (2), 40 (O), and 50 °C ([) in AcOH
and absence of phthalic anhydride. The solid lines are drawn through
the calculated data points using eq 3 and the parameters R and δob,
listed in Table 2.
) {fbδAni + (1 - fb)δAniH+}[Ani]T
(4)
(5)
Equation 4 can be rearranged to give eq 5
fb ) (δob - δAniH+) ⁄ (δAni - δAniH
)
+
and Table 2, Supporting Information). Although these differ-
ences in δob values are small, they exist beyond the limits of
experimental uncertainty and these observations may be ex-
plained in the following text.
where δob ) Aob/[Ani]T. Since δAniH+ ≈ 0 at 275 nm in both
pure and mixed H2O-CH3CN and H2O-THF solvents, eq 5 is
reduced to eq 6:
It is apparent from Scheme 2 and eq 2 that
fb ) δob ⁄ δAni
(6)
A0 ) δsol[Sol] + δAni[Ani] + δAniH+[AniH+] + δPAn[PAn0]
The values of δAni, obtained in 96% v/v water, for pure water
(ε ) 78) and pure acetonitrile (ε ) 37) as well as THF (ε ) 8)
(Table 2), differ from each other by <19%. The values of δAni
at 275 nm of 1010 and 955 M-1 cm-1 for pure CH3CN and
THF, respectively, differ by <6%. Almost all δob values in Table
2 were obtained in pure AcOH and mixed AcOH-CH3CN and
AcOH-THF solvents. However, since the values of δAni
decreases by less than 19% with change of solvent from 96%
v/v H2O to 100% CH3CN or 100% THF, the values of fb were
approximately determined from eq 6 considering the δAni value
() 1175 M-1 cm-1) obtained at 0.1 M NaOH in mixed
H2O-CH3CN solvent containing 96% v/v H2O.
(7)
where δ represents molar extinction coefficient, Sol is solvent,
and [PAn0] is the concentration of PAn at reaction time, t ) 0.
+
The fact that δAniH ≈ 0 and fb ) [Ani]/ [Ani]T (where [Ani]T
) [Ani] + [AniH+]) reduces eq 7 to eq 8:
A0 ) δsol[Sol] + δAni fb[Ani]T + δPAn[PAn0]
which is similar to eq 3 with
(8)
R0 ) δsol[Sol] + δPAn[PAn0]
(9)
The values of A0, obtained at 275 nm, 30 °C, and within
[Ani]T range 5.0 × 10-4-4.0 × 10-3 M, fit to eq 3, and the
least-squares calculated values of R and δob at different
temperatures are shown in Table 2. These results show that R
values are almost independent of temperature, but δob values
increase slightly with the increase of temperature, which could
be attributed to a modest increase in fb with increasing
temperature. The values of A0 and A∞ calculated from eq 2 where
A∞ ) A0 + δapp[R0] (with [R0] representing the initial concen-
tration of PAn), at a constant temperature and different [Ani]T
were also found to fit to eq 3 with A0 or A∞. The least-squares
calculated values of R and δob at different temperature for A0
and A∞ are shown in Table 2 and Table 2 in Supporting
Information, respectively. The reliable fit of the data to eq 3 is
evident from the standard deviations associated with the
calculated parameters, R and δob, and from the few typical plots
of Figure 2 where the solid lines are drawn through the
calculated data points. The values of R corresponding A0 reveal
significant absorption due to PAn at 275 nm. The values of δob
for A∞ are larger than the corresponding δob values for A0 by
values ranging within 25-73 M-1 cm-1 at 30-50 °C (Table 2
and
δob ) δAni fb
(10)
In view of eq 2 and Scheme 2,
A∞ ) δsol[Sol] + δP1[P1] + δP2[P2] + δAni[Ani′] (11)
where [Ani′] represents the concentration of unreacted free
aniline base. The concentration of reacted free aniline base,
[Ani′′], is equal to [P2] (i.e., [Ani′′] ) [P2]). If F1 ) ks/(ks +
knapp[Ani]T) and F2 ) knapp [Ani]T/(ks + knapp[Ani]T) then F1 +
F2 ) 1 with F1 ) [P1]/[PAn0] and F2 ) [P2]/ [PAn0] and
application of these relationships to eq 11 gives eq 12:
A∞ ) δsol[Sol] + δP1F1[PAn0] + δP2F2[PAn0] + δAni[Ani′]
(12)
In a typical kinetic run, the reaction mixture containing 100%
v/v AcOH and 1.5 × 10-4 M PAn was allowed to progress at
30 °C for the reaction period of >7 half-lives (based upon ks
value, Table 1). The absorbance, Aob, at 275 nm remained almost
unchanged within the reaction period of >7 half-lives which
J. Org. Chem. Vol. 73, No. 23, 2008 9307

Damit et al.
reveals that δP1 ≈ δPAn at 275 nm. The values of δP2 and δPAn
are 7.77 × 103 and 2.24 × 103 M-1 cm-1 at 275 nm,
respectively.
The values of F1 and F2 decrease and increase, respectively,
with the increase in [Ani]T at the constant concentration and
composition of mixed solvent components. The values of F2
were found to follow the linear relationship: F2 ) c + m[Ani]T
within [Ani]T range 1.0 × 10-3-4.0 × 10-3 M at a constant
temperature within its range 30-45 °C and within 1.0 ×
10-3-3.0 × 10-3 M at 50 °C. The respective calculated values
of intercept (c) and slope (m) are 0.79 ( 0.02 and 43.3 ( 8.0
M-1 at 30 °C, 0.62 ( 0.04 and 71.9 ( 14.2 M-1 at 35 °C, 0.40
( 0.04 and 99.7 ( 14.9 M-1 at 45 °C, and 0.34 ( 0.04 and
106 ( 15 M-1 at 50 °C. Although a perfect linearity is
impossible to achieve, maximum residual error (MRE ) 100
× (F2 - F2cald)/F2 where F2cald values were calculated from the
relationship F2calcd ) c + m[Ani]T with least-squares calculated
parameters c and m) turned out to be -2.6%, -5.9%, -8.4%,
and -9.3% at respective temperatures 30, 35, 45, and 50 °C
and the lowest [Ani]T () 1.0 × 10-3 M). The combination of
the relationships F2 ) c + m[Ani]T, F1 + F2 ) 1, and eq 12
gives eq 13:
A∞ ) δsol[Sol] + {δP1 + c(δP2 - δP1)}[PAn0] +
{m(δP2 - δP1)[PAn0] + δAni fb′}[Ani]T (13)
FIGURE 3. Plots showing the dependence of kobs versus [Ani]T at 1%
([), 25% (2), 50% (*), and 80% (∆) v/v CH3CN in mixed
CH3CN-AcOH solvents. The solid lines are drawn through the
calculated data points using eq 1 and the kinetic parameters, ks and
knapp, listed in Table 1.
where fb′ ) [Ani′]/[Ani]T. Equation 13 is similar to eq 3 with
R∞ ) δsol[Sol] + {δP1 + c(δP2 - δP1)}[PAn0]
(14)
and
from a few typical plots of A0 versus [Ani]T of Figure II
(Supporting Information) where solid lines are drawn through
the least-squares calculated data points. The calculated values
of δob were used to calculate fb from eq 10 with δAni ) 1175
M-1 cm-1, and these values of fb are shown in Table 2. The
calculated values of A∞ (A∞ ) A0 + δapp[R0], eq 2) were also
treated with eq 3 where A0, R, and δob were replaced by A∞,
R∞, and δob∞, respectively, and the least-squares calculated
δob∞ ) m(δP2 - δP1)[PAn0] + δAni fb′
(15)
where superscript “∞” merely indicates that these values are
obtained from eq 3 with A0 replaced by A∞. It may be noted
that the observed data point at [Ani]T ) 5.0 × 10-4 M showed
negative deviation from linearity of the plot of A∞ versus [Ani]T
by 16%, 17%, 17%, and 19% at 30, 35, 45 and 50 °C,
respectively. Such significant negative deviations are due to
relatively large negative deviations of observed data points from
values of R∞ and δob are summarized in Table 2 (Supporting
∞
Information).
linearity due to F ) c + m[Ani]T under such conditions.
Effects of [Aniline] on the Cleavage of PAn in mixed
AcOH-THF Solvents at 35 °C. Despite the fact that both THF
and CH3CN are aprotic solvents and useful for organic synthesis,
their various physicochemical properties are significantly dif-
ferent from each other. Furthermore, the value of ε of THF is
similar to that of AcOH, and hence change in the composition
of mixed solvent components, THF and AcOH, is not expected
to change the ε values of mixed AcOH-THF solvents. Thus,
the mixed AcOH-THF system is particularly useful for reaction
rate study where one of the rate-affecting solvent parameters
(i.e., ε) could be kept nearly constant with the change in the
composition of mixed solvents. A series of kinetic runs was
carried out for the cleavage of PAn at different values of [Ani]T
ranging from g3.0 × 10-4 to e4.0 × 10-3 M at a constant
content of THF in mixed AcOH solvent. The values of kobs fit
to eq 1 and the least-squares calculated values of ks and knapp at
different contents of THF are summarized in Table 1. The data
fit to eq 1 is satisfactory as evident from the standard deviations
associated with the calculated parameters and from a few typical
plots of Figure 4 where the observed data points are not
significantly different from the corresponding predicted data
points based upon eq 1.
2
∞
Equations 10 and 15 predict larger value of δob than that of
δob for the same set of kinetic runs because δP2/δP1 g 3.5 and
m g 43 M-1 at g30 °C.
Effects of [Aniline] on the Cleavage of PAn in Mixed
AcOH-CH3CN Solvents at 35 °C. In order to discover the
effects of varying concentration of [Ani] and AcOH on the rate
constants for the reactions of PAn with AcOH and aniline, a
few kinetic runs were carried at a constant composition of mixed
AcOH-CH3CN solvent and within total aniline concentration
([Ani]T) range g2.5 × 10-4 to e 4.0 × 10-3 M. The rate
constants, kobs, were found to fit to eq 1 and the least-squares
calculated values of ks and knapp at different contents of AcOH
are summarized in Table 1. The reliable fit of observed data to
eq 1 is evident from the standard deviations associated with
the calculated kinetic parameters, ks and knapp, and from some
typical plots of Figure 3 where solid lines are drawn through
the calculated data points.
The values of A0, calculated from eq 2, showed satisfactory
fit to eq 3. The least-squares calculated values of R and δob at
different contents of CH3CN ranging from 1-90% v/v are
summarized in Table 2. The extent of reliable fit of the data
(A0 versus [Ani]T) to eq 3 is evident from the standard deviations
associated with calculated parameters, R and δob (Table 2), and
The values of fb at different contents of THF were obtained
from the Beer’s plot of A0, calculated from eq 2, versus [Ani]T.
9308 J. Org. Chem. Vol. 73, No. 23, 2008

CleaVage of Phthalic Anhydride
FIGURE 5. Plots showing the dependence of Aob (at 275 nm) versus
percent content of organic cosolvent, X (v/v) for X ) CH3CN (∆) and
X ) THF ([) in mixed X-AcOH solvents containing 8.0 × 10-4
M
aniline.
TABLE 3. Dipole Moment (µ), Dielectric Constant (ε), Cohesive
Energy Density (Dce), and Hdrogen Bonding Energy (EHB) of Water
and Some Organic Solvents (25°C)a
FIGURE 4. Plots showing the dependence of kobs versus [Ani]T at 5%
([), 25% (∆), 50% (O), and 85% (]) v/v THF in mixed THF-AcOH
solvents. The solid lines are drawn through the calculated data points
using eq 1 and the kinetic parameters, ks, and least-squares calculated
solvent
µ (debye)
ε
Dce (cal cm-3
)
EHB (kcal mol-1
)
H2O
1.855
78.5
6.13b
58.5
37b
550.2
app
values of ks and kn listed in Table 1.
AcOH
HCOOH
CH3CN
THF
1.82
3.44
139.2
86.9
94.7
2.3
0
The values of A0 at a constant content of THF and different
values of [Ani]T fit to eq 3. The least-squares calculated values
of R and δob at different contents of THF ranging from 5-90%
v/v are shown in Table 2. The standard deviations, associated
with the calculated parameters (R and δob), and a few typical
plots of A0 versus [Ani]T of Figure II (Supporting Information),
where solid lines are drawn through the calculated data points,
reveal the extent of reliability of the fit of the data to eq 3. The
calculated values of δob were used to calculate fb from eq 10
with δAni ) 1175 M-1 cm-1. These values of fb are summarized
in Table 2. The calculated values A∞ (A∞ ) A0 + δapp[R0], eq
2) were also attempted to fit to eq 3 with replacement of A0, R,
and δob by respective A∞, R∞ and δob∞, and the least-squares
7.6c
2.21
dioxane
0.45
a Reference 13. b Reference 14. c Reference 8.
AcOH-THF solvents containing a constant concentration of
aniline is expected to increase the pKa of AcOH more strongly
compared to that of anilinium ion. This characteristic effect of
mixed acetic acid-aprotic organic cosolvent predicts that the
increase in the content of organic solvent CH3CN and THF in
mixed acetic acid solvent should increase the values of fb. It is
evident from Figure 5 that the values of Aob
(
fb) increase
almost linearly with the increase in the content of THF from
10 to 60% v/v followed by an independent region within ∼70%
to 100% v/v THF, but the values of Aob remained almost
independent of the content of CH3CN within its range 10-20%
v/v followed by a nonlinear increase within ∼40-100% v/v
CH3CN. The shape of the plot of Figure 5 for CH3CN is not
exactly similar to that for THF, which could be attributed to
some characteristic different structural and solution properties
of CH3CN and THF.
Alkanols have been shown to exist in monomeric, dimeric,
trimeric,..., and n-meric forms due to intermolecular hydrogen
bonding.12 Acetic acid, containing both carbonyl oxygen
(hydrogen bond acceptor site) and hydroxyl oxygen (both
hydrogen bond donor and acceptor site), may also be expected
to exist in monomeric AcOH, dimeric (AcOH)2, trimeric
(AcOH)3,.., and n-meric (AcOH)n forms through both intra- and
intermolecular hydrogen-bonded network. The monomeric acetic
acid, AcOH, is apparently most reactive toward free aniline base,
Ani. Although both acetonitrile and THF are aprotic organic
solvent, they differ significantly in various physicochemical
properties (Table 3). The aqueous mixtures of acetonitrile and
THF have been characterized as typically nonaqueous (TNA)
calculated values of R∞ and δob are shown in Table 2
∞
(Supporting Information).
Discussion
Effects of Mixed AcOH-CH3CN and AcOH-THF on
the Fraction of Free Aniline Base (fb). In order to discuss the
effects of aprotic organic cosolvent, such as acetonitrile and
tetrahydrofuran (THF), the absorbance (Aob) of mixed solvent
AcOH-CH3CN containing 8.0 × 10-4 M aniline (stock solution
of aniline was prepared in pure CH3CN) was measured at 275
nm. The values of Aob at different % v/v content of CH3CN are
shown graphically in Figure 5. Similar observations were
obtained in mixed AcOH-THF (the stock solution of aniline
was prepared in THF) and these results are also shown
graphically in Figure 5.
The value of Aob is related to fb through the relationship Aob
) Asol + fbδAni[Ani]T where Asol represents absorbance due to
solvent. As discussed earlier, the values of δAni are ap-
proximately independent of % v/v content (X) of CH3CN and
THF in mixed aqueous solvents (Table 2 and Table 2,
Supporting Information). Thus, at a constant [Ani]T () 8.0 ×
10-4 M), Aob is expected to be directly proportional to fb because
Asol ≈ 0 at 275 nm. In view of the facts mentioned earlier in
the text, the increase in the content of aprotic organic cosolvent
(such as CH3CN and THF), in mixed AcOH-CH3CN and
(12) (a) Grunwald, E.; Pan, K.-C.; Effio, A. J. Phys. Chem. 1976, 80, 2937.
(b) Khan, M. N. Int. J. Chem. Kinet. 1987, 19, 757. (c) Khan, M. N.; Ariffin, Z.
Langmuir 1996, 12, 261. (d) Khan, M. N.; Ariffin, Z. Langmuir 1997, 13, 6626.
(e) Khan, M. N. J. Phys. Chem. 1998, 92, 6273. (f) Khan, M. N.; Audu, A. A.
Int. J. Chem. Kinet. 1990, 22, 37.
J. Org. Chem. Vol. 73, No. 23, 2008 9309

Damit et al.
and typically aqueous (TA) solutions, respectively.13 The
increase in the content of CH3CN or THF in mixed glacial acetic
acid solvent (i) decreases the total concentration of glacial acetic
acid, [AcOH]T, which in turn decreases [AcOH] (concentration
of monomeric AcOH) and (ii) increases the difference of pKa
(∆pKa) of acetic acid and anilinium ion. Both the decrease in
[AcOH] and increase in ∆pKa should increase fb and conse-
quently it should increase Aob.
The values of Aob are independent of the content of CH3CN
within its range 10-20% v/v, while such observations are not
detectable in the presence of THF (Figure 5). These observations
may qualitatively be explained as follows. It is well-known that
TNA organic cosolvent, such as acetonitrile, diminishes the
degree of long-range order in highly aqueous TNA solvents
relative to water and as a consequence, acetonitrile is a water-
structure-breaking cosolvent. Similarly, TA organic cosolvent,
such as THF, is known as a water-structure-forming cosolvent
in water-rich region.13 If we assume that such solution properties
of CH3CN and THF remain unchanged with the change from
water to glacial acetic acid, then the acetic acid-structure-
breaking ability of CH3CN is expected to increase [AcOH] in
the acetic acid-rich region (i.e., ca. e 20% v/v CH3CN) which
is expected to decrease fb. It is now apparent that the increase
in the content of CH3CN in mixed acetic acid solvent causes
two opposite effects on fb: (i) an increase in fb due to decrease
in [AcOH]T as well as increase in ∆pKa and (ii) a decrease in
fb due to acetic acid-structure-breaking ability of CH3CN. It
seems that these two opposite effects of CH3CN increasing
content counterbalance to each other and as a consequence fb
value remain unchanged at e20% v/v CH3CN. However, the
factors that increase fb become increasingly dominant at g40%
v/v CH3CN. The plot for THF in Figure 5 reveals that the effect
of acetic acid-structure-forming ability of THF on fb remains
insignificant compared to the effects on fb of (i) increase in ∆pKa
and (ii) decrease in [AcOH]T due to increase in the content of
THF in mixed acetic acid solvents. These effects have caused
fb ≈ 1 at g70% v/v THF.
Mechanistic Explanations of Acetolysis and Anilinolysis of
PAn in Mixed AcOH-CH3CN and AcOH-THF Solvents. The
maximum contributions of ks compared to knapp[Ani]T, obtained
at the lowest values of [Ani]T, are <25% at g10% v/v AcOH
and <50% at <95% v/v AcOH for mixed AcOH-CH3CN
solvent system (Table 1). Because of the considerably low
contribution of ks compared to knapp[Ani]T in eq 1, the calculated
values of ks may be considered as less reliable.15 However,
although the values of ks are not very reliable, ks values seem
to be independent of [AcOH]T within AcOH content range
10-60% v/v followed by an almost linear increase with
[AcOH]T within g60-99% v/v AcOH (Figure 6a). Similar
observations were obtained in the reaction of H2O with ionized
phenyl salicylate in mixed H2O-CH3CN solvent.12b A quantita-
tive interpretation of pseudo-first-order rate constants for sol-
volytic reactions in mixed aqueous-organic cosolvent or mixed
organic solvent-organic cosolvent is extremely difficult due to
lack of general theoretical model or models caused by our lack
of understanding about highly dynamic and complex structural
network at the molecular level of these solvent systems. As a
FIGURE 6. (a) The plots showing the dependence of k versus [AcOH]T
for k ) kn M-1 s-1([) and k ) 104 ks s-1(*) in mixed CH3CN-AcOH
solvents. (b) The plots showing the dependence of k versus [AcOH]T
for k ) kn M-1 s-1 ([) and k ) 104 ks s-1 (*) in mixed THF-AcOH
solvents.
consequence, qualitative and empirical approaches are often used
to rationalize such kinetic data.1
A qualitative explanation for the nonlinear variation of ks with
the percent content of AcOH in mixed acetonitrile solvent may
be described in terms of the concept of polarization of
monomeric AcOH molecules such as 2 where B is the basic
site of either RB ) CH3CN or THF. The nucleophilicity of the
polarized AcOH molecule (2) is expected to be sufficiently
increased compared to that of nonpolarized AcOH. Thus, as
the percent content of RB increase, the content or concentration
of 2 also increases, which causes an increase in the rate of
acetolysis. At the same time, the total concentration of acetic
acid [AcOH]T, decreases, which in turn decreases the rate of
(13) Engberts, J. B. F. N. In Water, a ComprehensiVe Treatise; Franks, F.,
Ed.; Plenum: New York, 1979; Vol. 6, Chapter 4.
(14) Anantakrishnan, S. V. J. Sci. Indust. Res. 1971, 30, 319.
(15) Khan, M. N. Micellar Catalysis. In Surfactant Science Series; CRC Press,
Taylor & Francis Group: Boca Raton, FL. 2006; Vol. 133, p 418.
9310 J. Org. Chem. Vol. 73, No. 23, 2008

CleaVage of Phthalic Anhydride
SCHEME 3a
acetolysis. These two opposing effects probably counterbalance
each other in the RB-rich region, i.e., within CH3CN content
range of ∼40-90% v/v, in the acetolysis of PAn and thus leads
to an almost [AcOH]T-independent rate of acetolysis (Figure
6a). Nearly linear increase in ks with the increase in the content
of AcOH at >60% v/v AcOH is merely due to increasing effect
of [AcOH]T on [AcOH].
The free base of Ani corrected nucleophilic second-order rate
constants, kn () knapp/fb), for the nucleophilic reaction of Ani
with PAn within [AcOH]T range 1.75-16.6 M in mixed
AcOH-CH3CN solvents, fit to eq 16
kn ) knun + kga[AcOH]T
(16)
where knun and kga represent respective uncatalyzed and general
acid-catalyzed second-order and third-order rate constant for
un
anilinolysis of PAn. The least-squares calculated values of kn
and kga for AcOH-CH3CN solvent system are 3.39 ( 0.64 M-1
s-1 and 1.95 ( 0.06 M-2 s-1, respectively. The observed data
fit to eq 16 seems satisfactory as evident from the standard
a An ) ArNH2; Ac ) CH3CO.
un
deviations associated with the calculated parameters, kn and
un
kga, and from maximum percent residual error of 6.0% and the
plot of Figure 6a where solid line is drawn through the calculated
data points.
from eq 16 with kn as an unknown parameter. It is apparent
from Figure 6b that all of the observed data points at [AcOH]T
e 5.25 are slightly (e13%) negatively deviated from the
predicted linearity of the plot. In order to test if these negative
deviations have significant effect on the values of knun and kga,
the values of kn, obtained within [AcOH]T range 1.75-6.13 M,
The satisfactory fit of kobs values to eq 1 as evident from the
standard deviations associated with the values of calculated
app
parameters, ks and kn (Table 1) as well as from the typical
un
plots of Figures 1, 3, and 4 reveals the absence of kinetic terms
were used to calculate kn and kga from eq 16, and such
kgb[Ani]2[PAn] and kga [Ani][AniH+][PAn] in the rate law. The
′
respective calculated values are -0.9 ( 0.4 M-1 s-1 and 2.59
( 0.09 M-2 s-1. Again the negative value of knun with standard
deviation of more than 40% reveals its insignificance compared
kga[AcOH]T in eq 16 with knun ) 0, and such a calculated value
of kga is 2.32 ( 0.11 M-2 s-1, which is ∼6% lower than kga ()
2.48 M-2 s-1) obtained by the use of kn values within [AcOH]T
range 1.75-14.9 M. Nearly 6% change in the magnitude of
the slopes of two solid lines of Figure 6b is similar to the
standard deviations associated with the values of these slopes.
In view of experimental results shown by Figure 6a and b, a
plausible reaction mechanism is shown in Scheme 3. Although
the present data are not sufficient to support or oppose the
presence of reactive intermediates T1 and T2 on the reaction
paths, these intermediates are shown in Scheme 3 for the
following reason.
observed rate law, rate ) kobs[PAn], and eqs 1 and 16 can lead
to eqs 17 and 18:
kobs ) ks + (knun + kga[AcOH]T) fb[Ani]T
knapp ) kn fb
(17)
(18)
The values of kn, for nucleophilic reaction of Ani with PAn
in mixed AcOH-THF solvent, show an almost linear increase
with the increase in [AcOH]T until [AcOH]T ≈ 14.0 M (Figure
6b).
It may be considered that the horizontal axes of Figure 6a
and b should indicate activity or molar ratio rather than [AcOH]T
from the viewpoint of thermodynamics, but in solution-phase
kinetics, second-order rate constants are expressed in molarity
instead of activity or molar ratio. Thus, the values of ks and kga
in terms of molarity may be practically more useful. For
instance, second-order rate constants for water reaction with
substrates, based upon 55.5 M for pure water concentration,
are used to determine the effective molarity of the analogous
intramolecular reactions. The values of kn are almost independent
within [AcOH]T range 15.8-16.6 M. The values of kn, obtained
within [AcOH]T range 1.75-14.90 M, were treated with eq 16
and the least-squares calculated values of knun and kga are -0.9
( 0.5 M-1 s-1 and 2.69 ( 0.06 M-2 s-1, respectively. Negative
value of knun with standard deviation of more than 50% reveals
The reactive intermediates T1 and T2 are expected to be more
stable in acetic acid than in water because of the fact that
intermediate 1 was detected spectrophotometrically in acetic
acid, but such an intermediate could not be detected in mixed
acidic aqueous-organic cosolvents.3-6 Since the increase in the
contents of protic and aprotic organic cosolvents in mixed
aqueous-organic cosolvent decrease more strongly the pKa of
acidic groups of RCO2H, ArCO2H, ROH, and ArOH than those
of RNH3+, R2NH2+, and R3NH+ and therefore it is plausible to
assume that the pKa value of leaving group in the k-1-step is
larger than that in the k2-step and k3-step. Thus, based upon the
better leaving ability of the leaving group in the k-1-step and
the release of five-membered ring strain energy in the k2-step
as well as the k3-step, the inequality k-1 . k2 and k-1 . k3
[AcOH] should exist, and as a consequence the k2-step and k3-
un
insignificant contribution of kn compared to kga[AcOH]T in
eq 16. Thus, the value of kga was also calculated from eq 16
with knun ) 0, and such a calculated value of kga is 2.48 ( 0.20
M-2 s-1 which is reduced by <8% compared to kga obtained
J. Org. Chem. Vol. 73, No. 23, 2008 9311

Damit et al.
Arrhenius equation. Although the interesting finding in this paper
is the unusually very low values of ∆H* () 1-2 kcal mol-1
)
and ∆S * () -51.2 cal K-1mol-1) for knapp, its plausible
physicochemical explanations are difficult to provide at the
moment. The values of ∆H* and ∆S* for the hydrolysis of PAn
at 0.01 M HCl are 8.6 ( 1.2 kcal mol-1 and -39.1 ( 3.9 cal
K-1 mol-1, respectively, and the reaction mechanism for
hydrolysis under such conditions involves the presence of a
reactive complex formed between PAn and water molecules.16
Conclusions
It is apparent from the present study that the expected
maximum yield of the product (P2) from the reaction of PAn
with Ani could be obtained under the following reaction
conditions: (i) The synthesis should be carried out at lower
temperature because the value of knapp/ks increases sharply with
decreasing temperature at [Ani]T/[PAn0] g 2.5 in pure glacial
acetic acid solvent. (ii) The presence of 50% v/v CH3CN in
mixed AcOH solvent should give maximum yield of P2 within
relatively shorter reaction period because the knapp/ks value at
35 °C is larger under such condition than under pure AcOH
solvent. (iii) The value of knapp/ks at 60-70% v/v THF are
significantly larger than at 50% v/v CH3CN in mixed AcOH
solvent. Thus, under such solvent system, the maximum yield
of P2 could be achieved at 35 °C within a significantly shorter
reaction period compared to those under conditions of (i) and
(ii).
FIGURE 7. (a) Plot showing the dependence of ln ks versus 1/T for
acetolysis of PAn in AcOH. The solid line is drawn through the
calculated data points using the Arrhenius equation with ln A and Ea
mentioned in the text. (b) Plot showing the dependence of ln knapp versus
1/T for anilinolysis of PAn in AcOH. The solid line is drawn through
the calculated data points using the Arrhenius equation with ln A and
Ea mentioned in the text.
Experimental Section
Materials. Reagent grade phthalic anhydride (PAn), aniline,
glacial acetic acid (AcOH), and tetrahydrofuran (THF) were
commercial products of the highest commercially available purity.
HPLC grade acetonitrile was used throughout the study. AcOH was
purified by distillation. Dry THF (500 mL) containing 5 g Na and
10 g benzophenone was distilled under nitrogen. N-Phenylphtha-
lamic acid (P2) was synthesized using the literature procedure
involving the reaction of aniline with PAn in THF. The observed
spectroscopic data are in agreement with the corresponding reported
data.2
step are rate-determining steps. Similar arguments may lead to
the k4-step as the rate-determining step in the acetolysis of PAn.
There is a likelihood that the k3-step involves a transition state
(TS) where the AcOH molecule acts as a bifunctional catalyst
(i.e., both intermolecular general acid and general base).
Kinetic Measurements. Reaction rates of the cleavage of
phthalic anhydride (PAn) in glacial acetic acid (AcOH) and mixed
AcOH-CH3CN and THF-AcOH solvents containing aniline were
studied spectrophotometrically by monitoring the appearance of
products at 275 nm and at a constant temperature. The stock
solutions of PAn (0.015 M) for kinetic runs carried out in 100%
AcOH were freshly prepared just before the start of the kinetic
runs. For kinetic runs, carried out in mixed AcOH-CH3CN and
THF-AcOH solvents, the stock solutions of PAn were prepared
in 100% CH3CN or THF. The details of the kinetic procedure are
essentially same as described elsewhere.17
All kinetic runs were carried out under pseudo-first-order kinetic
conditions. The observed data, absorbance (Aob) versus reaction time
(t) for each kinetic run were found to fit to eq 2. Pseudo-first-order
rate constant (kobs), apparent molar extinction coefficient (δapp), and
initial absorbance (A0) were calculated from eq 2 with the nonlinear
least-squares technique. The reactions were carried out for up to
6-9 half-lifes of the reactions, and the observed data fitted well to
eq 2.
Activation Parameters for Acetolysis and Anilinolysis
of PAn. The rate constants ks and knapp, obtained within the
temperature range 30-50 °C at 100% v/v AcOH (Table 1), were
used to calculate the activation parameters, ∆H* and ∆S *, from
the Eyring equation as well as Ea and ln A from the Arrhenius
equation. The nonlinear least-squares calculated values of
activation parameters are as follows: ∆H* ) 15.3 ( 1.2 kcal
mol-1 and ∆S * ) -20.1 ( 3.8 cal K-1mol-1 for ks (i.e.,
acetolysis of PAn) as well as ∆H* ) 1.1 ( 0.5 kcal mol-1 and
∆S * ) -51.2 ( 1.7 cal K-1mol-1 for kn (i.e., anilinolysis
app
of PAn). Similarly, linear least-squares calculated respective
values of Ea and ln A from Arrhenius equation are 19.3 ( 1.8
kcal mol-1 and 25.78 ( 2.87 s-1 for ks as well as 1.7 ( 0.5
kcal mol-1 and 4.63 ( 0.87 M-1 s-1 for knapp. A satisfactory
data fit to the Arrhenius equation is evident from the standard
deviations associated with the calculated values of the activation
parameters and from the plots of Figure 7 where solid lines are
drawn through the least-squares calculated data points using the
Product Characerization. In order to characterize the products
of the cleavage of PAn in pure AcOH, and mixed AcOH-X, (X
) CH3CN, THF), solvents containing aniline, three typical kinetic
(16) Khan, M. N. Indian J. Chem. 1993, 32A, 387.
(17) Cheong, M. Y.; Ariffin, A.; Khan, M. N. J. Phys. Chem. B 2007, 111,
12185.
9312 J. Org. Chem. Vol. 73, No. 23, 2008

CleaVage of Phthalic Anhydride
TABLE 4. Values of Observed Absorbance (Aob∞) and Estimated
Absorbance (Aobest) of the Products of the Cleavage of PAn in pure
AcOH and mixed AcOH-X, Solvents Containing Ania
presence of Ani on δP2. As described earlier in the text δP1 ≈ δPAn
at 275 nm in pure AcOH solvent, and as consequence, the values
of δP1 were obtained from R values listed in Table 2.
It is evident from Scheme 2 that the absorbance (Aobest) of the
b
c
AcOH
X
10-3 δP1
10-3 δP2
d
e
f
est
(% v/v) (% v/v) Asol (M-1 cm-1) (M-1 cm-1) FP1 fb Aob Aob
reaction mixture containing 1.5 × 10-4 M PAn and 1.0 × 10-3
M
∞
0.03g
0.03
2.24
2.24
0.41
0.0
2.24
1.21
0.0
7.77
7.77
4.72
4.72
4.72
6.80
6.80
6.80
0.31 0.16 1.04h 1.08
0.39 0.16 1.04 1.01
0.06 0.37 0.90k 1.01
0.06 0.37 0.90 1.00
0.06 0.37 0.90 1.02
0.04 0.74 1.57n 1.63
0.04 0.74 1.57 1.62
0.04 0.74 1.57 1.64
Ani at t ) ∞ (i.e., t equivalent to reaction period of g7 half-lifes
100
100
50
50
50
50
50
50
0
0
of the reactions) may be estimated from eq 19:
50i
50
50
50l
50
50
0.03j
0.03
Aobest)Asol + δP1[P1] + δP2[P2] + δAni[Ani′]
(19)
0.03
0.03m
where Asol represents the absorbance due to solvent, δZ is the molar
extinction coefficient of Z (where Z ) P1, P2, and Ani) and [Ani′]
is the concentration of unreacted free Ani base. If FP1 and FP2
2.24
a [PAn0] ) 1.5 × 10-4 M, [Ani]T ) 0.001 M, 35 °C, λ ) 275 nm,
represent the respective fractions of product P1 and P2, then FP1 )
b
and δAni ) 1175 M-1 cm-1 at 275 nm. δP1 ≈ δPAn ) R/[PAn0] with R
[P1]/([P1] + [P2]) ) ks/(ks + knapp), FP2 ) 1 s FP1, and [P1] + [P2]
) [PAn0]. By introducing these defined parameters into eq 19, one
gets
values listed in Table 2. c The values of δP2 were obtained by the use of
d
app
authentic sample of P2 as described in the text. FP1 ) ks/(ks + kn
)
where ks and kn values are summarized in Table 1. e Values of fb are
app
listed in Table 1. f The values of Aob were calculated from eq 20 as
est
Aobest)Asol+[PAn0][δP1FP1 + δP2(1 - FP1)] +
g
described in the text. Asol ) R at 35 °C, 100% v/v AcOH, and [PAn0]
) 0 (Table 2). h The value of Aob was obtained at 7 half-lifes of the
∞
δAni fb([Ani]T-[PAn0][2 - FP1]) (20)
reaction. i X ) CH3CN. Asol ) R ) 0.01 at 35 °C, 100% v/v CH3CN
j
and [PAn0] ) 0 (Table 2). k The value of Aob was obtained at 8
∞
est
where fb represents fraction of free Ani base. The values of Aob
were calculated from eq 20 with the known values of Asol, [PAn0],
δP1, δP2, FP1, δAni, fb, and [Ani]T as summarized in Table 4. The
half-lifes of the reaction. l X ) THF. Asol ) R ) 0.04 at 35 °C, 100%
m
v/v THF, and [PAn0] ) 0 (Table 2). n The value of Aob was obtained at
∞
12 half-lifes of the reaction.
est
values of Aob are almost similar to the corresponding values of
Aob (Table 4) within the limits of experimental uncertainties and
∞
runs were carried out at 35 °C as well as 0 and 50% v/v X (X )
CH3CN, THF). The values of absorbance of these reaction mixtures
(Aob∞) at 275 nm and reaction time t g 7 half-lifes are shown in
Table 4. N-Phenylphthalamic acid (P2) is known to undergo rapid
cyclization to form 1 in AcOH in the absence of aniline.2 Thus,
the value of δP2, as shown in Table 4, was calculated from the A0
value obtained from eq 2 by the use of Aob versus t data (at 275
nm) of a kinetic run for the cleavage of authentic P2 at 35 °C in
pure AcOH containing 1.5 × 10-4 M P2. In order to find out the
effect of the presence of aniline (Ani) on A0 value, another kinetic
run was carried out at 35 °C in pure AcOH containing 1.5 × 10-4
M P2 and 0.002 M Ani, and the value of A0 was turned to be 1.51,
which on applying the correction of absorption due to free Ani
base (AobAni ) 0.35) resulted in an absorbance value (AobP2) due to
P2 of 1.16 which gave δP2 ) 7.73 × 103 M-1 cm-1. These
observations showed the absence of any detectable effect of the
thus affirming P1 and P2 as the products of the cleavage of PAn in
pure AcOH and mixed AcOH-X (X ) CH3CN, THF) solvents
containing aniline.
Acknowledgment. The authors thank the Ministry of Science,
Technology and Innovation for ScienceFund (Project 14-02-
03-4014) and the University of Malaya for financial support.
The authors also thank one of the reviewers for very extensive
and constructive comments/suggestions.
Supporting Information Available: Tables I and II; Figures
I and II. This material is available free of charge via the Internet
at http://pubs.acs.org.
JO8011072
J. Org. Chem. Vol. 73, No. 23, 2008 9313