Thermolysis of tert-butyl phenylperacetates: Delicate control of the rates through contributions from translational and rotational entropy

Article abstract of DOI:10.1021/jo010143j

The first-order rate constants (k_y) at several temperatures in CDCl₃ were measured for thermal decompositions of YC₆H₄CH₂CO₃C(CH₃) ₃ with Y being p-OCH₃, p-OPh, p-CH₃, p-Ph, p-H, p-Cl, m-Cl, and p-NO₂. The relative rates (k_y/k_H) exhibit excellent ρ⁺/σ⁺ Hammett correlations with ρ⁺ < 0, indicating a polar TS. Activation parameters (ΔH?_Y and ΔS?_Y) and their differential terms (ΔΔH?_Y-H and ΔΔS?_Y-H) were obtained from the Eyring plot. Differential activation terms (ΔΔH?_Y-H and ΔΔS?_Y-H) disclose an isokinetic relation with p-CH₃, p-Ph, p-H, p-Cl, and m-Cl (isokinetic temp, 230 K). However, p-OCH₃, and p-OPh show negative deviations, and a positive deviation occurs with p-NO₂. Plot of ΔΔH?_Y-H vs σ⁺ exhibits a good linear relation (r = 0.95) with a slope (α₁ = -3.34). A better linear correlation (r = 0.97) and steeper slope (α₂ = -5.22) were observed for TΔΔS?_Y-H vs σ⁺. Negatively larger slope (α₂ = -5.22) may point to entropy control of rates. Differential activation parameters (ΔΔHS?_Y-H and ΔΔS?_Y-H) reflect variations of activation process. Differential activation entropies (ΔΔS?_Y-H) are discussed in terms of contributions of translational and rotational entropies. Similar deviation behaviors of p-OCH₃, p-OPh, and p-NO₂ were again observed for the both plots. p-NO₂ can strongly destabilize the cationic site of the polar TS but serves an eminent spin delocalizer for the homolytic TS.

Full text of DOI:10.1021/jo010143j

4006
J . Org. Chem. 2001, 66, 4006-4011
Th er m olysis of ter t-Bu tyl P h en ylp er a ceta tes: Delica te Con tr ol of
th e Ra tes th r ou gh Con tr ibu tion s fr om Tr a n sla tion a l a n d
Rota tion a l En tr op y†
Sung Soo Kim,* In Sang Baek, Alexey Tuchkin,‡ and Kyoung Moon Go
Department of Chemistry and Center for Chemical Dynamics, Inha University,
Inchon 402-751, South Korea
sungsoo@inha.ac.kr
Received February 6, 2001
The first-order rate constants (kY) at several temperatures in CDCl3 were measured for thermal
decompositions of YC6H4CH2CO3C(CH3)3 with Y being p-OCH3, p-OPh, p-CH3, p-Ph, p-H, p-Cl, m-Cl,
and p-NO2. The relative rates (kY/kH) exhibit excellent F+/σ+ Hammett correlations with F+ < 0,
indicating a polar TS. Activation parameters (∆HqY and ∆SqY) and their differential terms (∆∆Hq
Y-H
and ∆∆SqY-H) were obtained from the Eyring plot. Differential activation terms (∆∆Hq
and
Y-H
∆∆SqY-H) disclose an isokinetic relation with p-CH3, p-Ph, p-H, p-Cl, and m-Cl (isokinetic temp,
230 K). However, p-OCH3, and p-OPh show negative deviations, and a positive deviation occurs
with p-NO2. Plot of ∆∆Hq
vs σ+ exhibits a good linear relation (r ) 0.95) with a slope (R1 )
Y-H
-3.34). A better linear correlation (r ) 0.97) and steeper slope (R2 ) -5.22) were observed for
T∆∆Sq vs σ+. Negatively larger slope (R2 ) -5.22) may point to entropy control of rates.
Y-H
Differential activation parameters (∆∆Hq
and ∆∆SqY-H) reflect variations of activation process.
Y-H
Differential activation entropies (∆∆SqY-H) are discussed in terms of contributions of translational
and rotational entropies. Similar deviation behaviors of p-OCH3, p-OPh, and p-NO2 were again
observed for the both plots. p-NO2 can strongly destabilize the cationic site of the polar TS but
serves an eminent spin delocalizer for the homolytic TS.
In tr od u ction
distinguish the mechanism between one-bond (z) and
two-bond (x and z) homolyses. In the previous work,7
extent of rupture of bond x is shown to be seriously
influenced by substituents, which provokes significant
variations of secondary R-deuterium kinetic isotope ef-
fects.
Thermal decompositions of tert-butyl phenylper-
acetates1-7 occur through simultaneous two-bond scis-
sions to yield a carbon dioxide, benzyl and tert-butoxy
radical. Both activation enthalpy (∆Hq) and activation
entropy (∆Sq) gradually decrease1 when R in RCO3C-
(CH3)3 is varied in the order of C6H5 (∆Hq) 33.5 kcal
mol-1, ∆Sq) 7.8 eu), C6H5CH2 (∆Hq) 28.7 kcal mol-1
,
∆Sq) 3.9 eu),8 (C6H5)2CH (∆Hq) 24.3 kcal mol-1, ∆Sq)
-1.0 eu),8 C6H5-CHdCH-CH2 (∆Hq) 23.5 kcal mol-1
,
∆Sq) -5.9 eu).8 The synchronization of the bond cleav-
ages2 produces remarkably small activation volumes (∆Vq
) 1-3 cm3 mol-1). Accordingly, decrease of the activation
parameters (∆Hq, ∆Sq, ∆Vq) has been recently7 explained
in terms of formation of π-bonds w and y in 1. Viscosity
studies of the solvents4 have introduced a method to
Hydrogen atom abstractions from substituted toluenes9
and cumenes11 and â-scissions of the carbinyloxy radi-
cals10 reveal entropy control of rates via the polar
substituent interactions. Only relative rates were avail-
able with previous studies9-11 thereby providing dif-
ferential activation terms (∆∆HqY-H and ∆∆SqY-H) derived
from the Eyring equation.12 The positive sign of dif-
† The late Professor Glen A. Russell has been a Distinguished
Foreign Adviser for this investigation. We are heavily indebted to his
mentorship.
ferential activation entropies (∆∆Sq
)
9-11 with electron-
Y-H
donating substituents is ascribed to the dominance of the
‡ A visiting scholar from St. Petersburg State University by a grant
from the Science Technology Policy Institute under the Korea-Russia
manpower exchange program (1996).
(8) The rates are accelerated by 65 times at 60 °C when R ) (C6H5)2-
CH takes the place of R ) C6H5CH2 in the decomposition of RCO3C-
(CH3)3. The skeletal variation of R causes a relatively large change of
activation enthalpy (∆∆Hq ) 24.3 - 28.7 ) -4.4 kcal mol-1) that may
well outweigh the entropy decrease (-1 - 3.9 ) -4.9 eu) to achieve
enthalpy control of rates. The negative values of ∆Sq are shown with
R being (C6H5)2CH and C6H5CHdCH-CH2. Extensive charge delocal-
(1) (a) Bartlett, P. D.; Hiatt, R. R. J . Am. Chem. Soc. 1958, 80, 1398.
(b) Bartlett, P. D.; Simons D. M. J . Am. Chem. Soc. 1960, 82, 1753. (c)
Bartlett, P. D.; Ru¨chardt, C. J . Am. Chem. Soc. 1960, 82, 1756.
(2) (a) Neuman, R. C., J r.; Behar, J . V. J . Am. Chem. Soc. 1967, 89,
4549. (b) Neuman, R. C., J r.; Behar, J . V. J . Am. Chem. Soc. 1969, 91,
6024. (c) Neuman, R. C., J r.; Behar, J . V. J . Org. Chem. 1971, 36, 654.
(d) Neuman, R. C., J r.; Behar, J . V. J . Org. Chem. 1971, 36, 657.
(3) (a) Koenig, T.; Wolf, R. J . Am. Chem. Soc. 1969, 91, 2574. (b)
Koenig, T.; Huntington, J .; Cruthoff, R. J . Am. Chem. Soc. 1970, 92,
5413.
izations occurring in (C6H5)2CHδ+ and C6H5CHdCH-CH2 of the
δ+
corresponding TS can prohibit the free rotation of phenyl rings. The
negative figures (∆Sq < 0) should be due to reduction of the rotational
entropy.
(4) Pryor, W. A.; Smith, K. J . Am. Chem. Soc. 1970, 92, 5403.
(5) Pryor, W. A.; Smith, K. Int. J . Chem. Kinet. 1971, 3, 387.
(6) Wolf, R. A.; Trocino, R. J .; Rozich, W. R.; Sabeta, I. C.; Ordway,
R. J ., J r. J . Org. Chem. 1998, 63, 3814.
(9) Kim, S. S.; Choi, S. Y.; Kang, C. H. J . Am. Chem. Soc. 1985,
107, 4234.
(10) Kim, S. S.; Kim, H. R.; Kim, H. B.; Youn, S. J .; Kim, C. J . J .
Am. Chem. Soc. 1994, 116, 2754.
(7) Kim, S. S.; Tuchkin, A. J . Org. Chem. 1999, 64, 3821.
(11) Kim, S. S.; Kim, C. S. J . Org. Chem. 1999, 64, 9261.
10.1021/jo010143j CCC: $20.00 © 2001 American Chemical Society
Published on Web 05/02/2001

Thermolysis of tert-Butyl Phenylperacetates
J . Org. Chem., Vol. 66, No. 11, 2001 4007
Ta ble 1. Kin etic Da ta for Th er m a l Decom p osition s of ter t-Bu tyl P h en ylp er a ceta tes in CDCl3
a
)
kY × 104, s-1 (kY/kH
Hammett correlationsb
temp, °C
60
p-CH3O
p-PhO
p-CH3
p-Ph
H
p-Cl
m-Cl
p-NO2
F
+ (r)
F (r)
1.25
(10.7)
0.465
(3.97)
0.26
(2.22)
0.221
(1.89)
0.117
(1.00)
0.101
(0.86)
0.050d
(0.43)
0.019d
(0.138)
- 1.13
(0.996)
- 1.53
(0.913)
70
80
4.05
(11.04)
1.50
(4.09)
0.817
(2.23)
0.745
(2.03)
0.367
(1.00)
0.319
(0.87)
0.151
(0.41)
0.058d
(0.158)
- 1.15
- 1.56
(0.912)
(0.996)
11.90
(11.23)
4.82
(4.55)
2.63
(2.48)
2.29
(2.16)
1.06
(1.00)
0.84
(0.79)
0.424
(0.40)
0.165
(0.156)
- 1.18
(0.996)
- 1.596
(0.911)
80c
90
9.207
(12.89)
3.04
(4.26)
1.89
(2.65)
1.76
(2.46)
0.714
(1.00)
0.584
(0.818)
0.289
(0.404)
0.128
(0.179)
- 1.17
- 1.631
(0.992)
(0.905)
35.67
(11.7)
13.9
(4.56)
8.36
(2.74)
7.17
(2.35)
3.05
(1.00)
2.79
(0.91)
1.26
(0.41)
0.44
(0.144)
- 1.20
(0.997)
- 1.642
(0.919)
100
111.8d
(11.95)
45.38d
(4.85)
24.98
(2.67)
19.54
(2.09)
9.36
(1.00)
7.60
(0.81)
3.61
(0.39)
1.438
(0.154)
- 1.21
(0.997)
-1.629
(0.910)
110
333.9d
(12.25)
138.2d
(5.07)
78.85d
(2.89)
72.06d
(2.64)
27.27
(1.00)
22.48
(0.82)
9.77
(0.36)
3.855
(0.141)
- 1.26
(0.996)
-1.701
(0.912)
a
The values of kY were derived from the plot of ln C/C0 ) kYt. The linearities of the plot carry correlation coefficient r g 0.9990 in most
cases. Refer to Supporting Information. b Hammett substituent constants were taken from ref 14. c CCl4 was used for solvent. The peresters
d
decomposed either too quickly or too slowly at the temperature specified to measure the accurate rates. Therefore, the figures were
obtained by the extrapolations utilizing the relative rates.
Sch em e 1
bond-breaking (x) that augments the translational de-
grees of freedom.12 The bond cleavage is inevitably
accompanied by formation of a π bond (w) causing
restricted rotations. Increment of translational entropy
may well overwhelm9-11 diminution of the rotational
entropy. Fortunately, homolysis of present peresters
allows us to measure the first-order rate constants and
corresponding activation parameters(∆Hq and ∆Sq
)
Y
Y
absolutely. We’d like to herein emphasize that substit-
uents can exert profound influences in controlling the TS
structure.
Resu lts a n d Discu ssion
P r ep a r a t ion a n d Th er m olysis of t h e P er est er s.
Various ring-substituted tert-butyl phenylperacetates 2
were synthesized according to the known methods.1,7
Reactions at 80 °C for 5 half-lives in degassed and sealed
Pyrex ampules containing 2 (0.05 M), I2 (0.2-0.4 M), and
CH3CN (0.02 M, internal standard) in CDCl3 gave rise
to benzyl iodide (90.8), tert-butyl benzyl ether (4.5), tert-
butyl alcohol (54.3), methyl iodide (43.3), and acetone
(43.3) (Scheme 1). The figures in the parenthesis cor-
respond to mol % of products based on the amount of a
perester decomposed. The products were identified by
GC-MS and NMR and quantified by NMR. The material
balance indicates 95.3% conservation of protons of benzyl
moiety of 2.
Deter m in a tion of Absolu te Ra tes, Ha m m ett Cor -
r ela tion s, a n d Activa tion P a r a m eter s. The rates of
decomposition of 2 in CDCl3 were obtained at the tem-
peratures by measuring periodically its depletion. The
consumption was monitored by disappearance of the
proton peak of benzyl (δ ) 3.6-3.7). The rate constants
(kY) were measured with ln(C0/Ct) ) kYt. The magnitude
of peak of benzylic protons was used for C0 and Ct. The
plots13 of ln(C0/Ct) vs t revealed excellent linearities with
correlation coefficients r g 0.9990 in most cases thus
making it possible to assign reliable values of kY. Ham-
mett reaction constants were obtained with σ+ and σ. The
better correlations have been found with F+/σ+. The
values of F+ become more negative with higher temper-
atures. These kinetic data are tabulated in Table 1.
Activation parameters and their differential terms were
obtained from the Eyring plot12 (Figure 1). Table 2 reveals
the activation terms and their differentials. An isokinetic
relation is attained with p-CH3, p-Ph, p-H, p-Cl, and m-Cl
(Figure 2; isokinetic temp Tk, 230 K). Plots of ∆∆Hq
Y-H
vs σ+ and T∆∆Sq
vs σ+ are given in Figures 3 and 4,
Y-H
respectively. p-OCH3 and p-OPh persistently show nega-
tive deviations, whereas positive deviations occur with
p-NO2 for Figures 2, 3, and 4, respectively.
High ligh t of th e Hom olytic Rea ction s. The con-
certed bond-breaking yields a polar transition state (TS,
1) where the positive charge is delocalized into the phenyl
ring. Virtually the same values of F+ are obtained with
various solvents (F+ ) -1.18 in CDCl3, F+ ) -1.17 in
CCl4, F+ ) -1.10 in cumene2, and F+ ) -1.0 in C6H5Cl1
(12) (a) Eyring, H. J . Chem. Phys. 1935, 3, 107. (b) Absolute rate
theory states that vibration/rotation of the bond being broken should
be replaced by a translational mode. The translation could be also
termed loose vibration/rotation. The replacement of vibration/rotation
by translation can trigger increase of activation entropy. The trans-
lational motion concerns three dimensions whereas vibration and
rotation occur with one and two dimensions, respectively.
(13) Refer to Supporting Information.

4008 J . Org. Chem., Vol. 66, No. 11, 2001
Kim et al.
Ta ble 2. Activa tion P a r a m eter s for Th er m a l Decom p osition of ter t-Bu tyl P h en ylp er a ceta tes in CDCl3
activation
parameter
p-CH3O
p-PhO
p-CH3
p-Ph
H
p-Cl
m-Cl
p-NO2
a
∆Hq
27.52 ( 0.54 28.08 ( 0.53 28.24 ( 0.50 28.16 ( 0.70 26.82 ( 0.53 26.61 ( 0.58 26.08 ( 0.40 26.20 ( 0.59
Y
b
∆Sq
5.87 ( 1.50
0.70
5.58 ( 1.49
1.26
4.90 ( 1.41
1.42
4.37 ( 1.96 -0.92 ( 1.48 -1.84 ( 1.61 -4.83 ( 1.13 -6.39 ( 1.65
Y
∆∆Hq
Y-H
T∆∆Sq
1.33
5.28
1.87
0
0
0
-0.22
-0.92
-0.33
-0.73
-3.91
-1.38
-0.62
-5.47
-1.93
a,c
Y-H
∆∆Sq
6.79
2.40
6.50
2.30
5.82
2.05
b,d
a,e
Y-H
a
b
d
In kcal mol-1
.
In eu. c ∆∆Hq
) ∆Hq - ∆Hq
.
∆∆Sq
) ∆Sq - ∆Sq
.
e T ) 353 K.
Y-H
Y
H
Y-H
Y
H
F igu r e 1. Erying plot of ln(kY/T) vs 1000/T.
F igu r e 3. Perturbation of activation enthalpy by substituent
effects.
F igu r e 2. Isokinetic relationship by enthalpy/entropy com-
pensation effects.
F igu r e 4. Perturbation of activation entropy by substituent
effects.
at 80 °C). The values of activation enthalpy (∆HqY) appear
comparable in CDCl3 (Table 2) and in C6H5Cl (Table 1 of
ref 1c) despite of different solvents and unidentical
methods of measurement. These similar figures (F+ and
∆HqY) may suggest little solvent interference with 1.
Activation parameters should thus be determined mainly
by substituent interactions. Locations of bonds y and z
are too far separated from the substituent for the
interactions to take place. Therefore, 1 can experience
quite similar substituent interactions influencing reac-
tivities of photobrominations of substituted toluenes9 and
cumenes11 and â-scissions of the carbinyloxy radicals.10
The deviation behaviors of p-OCH3, p-OPh, and p-NO2
may suggest that the substituents engender activation
processes that are, more or less, different from those
triggered by the rest of the substituents.
cationic moiety of 1, which can be mainly effected through
delocalization of positive charge accompanied by forma-
tion of double bond w (factor a). Factor a provokes to
decrease (increase) magnitude of ∆Hq for the donators
Y
(attractors). The stabilization (destabilization) may then
lengthen (shorten) the extent of cleavage of x (factor b).
Factor b tends to counteract the effect of factor a on the
magnitude of ∆HqY. Cleavage of x occurs with C-C sigma
bond that is accompanied by formation of C-C pi bond
of w. Therefore, the donators (attractors) could render
∆HqY larger (smaller) than in the case of the substituent
being hydrogen atom, i.e. Y ) H. Accordingly, b over-
whelms a to determine the size of differential activation
enthalpy (∆∆HqY-H). The secondary R-deuterium kinetic
isotope effects7 also suggest the pivotal role of b for the
rates. Figure 3 shows a good linearity (slope -3.34; r
Con tr ibu tion of Activa tion En th a lp y. Electron
donators (attractors)7,9-11 tend to stabilize (destabilize) the

Thermolysis of tert-Butyl Phenylperacetates
J . Org. Chem., Vol. 66, No. 11, 2001 4009
Ta ble 3. Activa tion P a r a m eter s of Hom olysis of Va r iou s
In itia tor s
0.950) with five substituents (Y ) p-CH3, p-Ph, p-H, p-Cl,
and m-Cl). Differential activation enthalpy (∆∆Hq
)
Y-H
∆Hq
∆Sq
(eu)
could be the net change derived from the compensating
effects of a and b when substituent Y replaces the
initiators
(kcal mol-1
)
hydrogen atom on the phenyl ring of 1. p-CH3 (∆∆Hq
a
c
C6H5C(O)OOC(CH3b)3
(CH3)3COOC(CH3)3
CH3C(O)OO(O)CCH3
37.5
37.8
31.3
25.4
26.8
24.3
23.5
16.1
13.8
11.6
-1.1
-0.9
-1.0
-5.9
Y-H
) 1.42 kcal mol-1) and p-Ph (∆∆Hq
) 1.33 kcal mol-1
)
Y-H
exhibit ∆∆Hq
> 0 because increment of ∆Hq per-
Y-H
Y
d
(C6H5)2CdCdNCH2C6H5
turbed by b is larger than the reduction of ∆Hq caused
e
Y
C6H5CH2CO3C(CH3)3
f
by a. p-Cl and m-Cl are the attractors and show ∆∆Hq
(C6H5)2CHCO3C(CH3)3
Y-H
f
C6H5CHdCHCH2CO3C(CH3)3
< 0, i.e., -0.22 kcal mol-1 for p-Cl and -0.73 kcal mol-1
for m-Cl. The negative figures can occur when a is again
excelled by b during the activation. A similar switchover
a
b
d
Ref 21. Ref 22. c Ref 23. Ref 19. e Table 2 of this paper.
f Ref 1.
of the sign from ∆∆Hq
> 0 to ∆∆Hq
< 0 has been
Y-H
Y-H
kcal mol-1, which is far below ∆Hq ) 37.5 kcal mol-1 for
tert-butyl perbenzoate undergoing one-bond homolysis
(Table 3). Thermal decomposition of tert-butyl 1-arylcy-
cloalkanepercarboxylates6 also shows characteristic val-
ues of activation enthalpy for one-bond (∆HqY ) 34.4 kcal
mol-1) and two-bond (∆HqY ) 27.0 kcal mol-1) homolysis
involving p-NO2 (refer to Table 3 of ref 6). The photobro-
minations of 4-substituted 3-cyanotoluenes15 proceed
through a TS assuming reduced polarity. The cyano
group with σm ) 0.6214 may destabilize the cationic site
to yield a less polar TS that could be compensated by
contribution of spin delocalizations. p-NO2 is a notorious
spin delocalizer16-19 and carries a spin delocalization
constant σC• ) 0.57.16 The plot7 of kYH/kYD vs σ+ also shows
a remarkable positive deviation (refer to Figure 1 of ref
7). Accordingly, p-NO2 may induce the concerted bond
cleavages traversing homolytic TS 3. Thermolysis of
already observed with the photobrominations of substi-
tuted toluenes.9 On the other hand, activation enthalpies
(∆HqY) are almost immune to substituent effects to render
corresponding differential terms close to nil (i.e., ∆∆Hq
Y-H
≈ 0) for â-scissions of the carbinyloxy radicals10 and
photobrominations of substituted cumenes.11 This phe-
nomenon10,11 has been rationalized in terms of the cancel-
lation of the opposite effects invoked by a and b.
p-OCH3 (σ+ -0.78)14 and p-OPh (σ+ -0.50)14 share
negative deviations with Figure 3. Both substituents can
disperse positive charge on 1 into their own oxygen atoms
forming a quinoid-like TS. The extra charge dispersion
helps a reduce size of ∆HqY yielding ∆∆Hq
) 0.70 and
Y-H
1.26 kcal mol-1 for p-OCH3 and p-OPh, respectively.
Photobrominations of substituted toluenes9 also show a
very similar negative deviation with p-OCH3 carrying
∆∆Hq
) 0.73 kcal mol-1. The effect of a becomes
Y-H
•
p-NO2 produces three entities, i.e. p-NO2PhCH2 , CO2,
significant enough to give the negative deviation but
and (CH3)3CO•. The three cage species can undergo
recombination if viscosity of the solvents retards the
rotational diffusion in order to keep the coplanarity
mentioned by Bartlett and Hiatt.1 The reactions of other
peresters20 involving 1 may not experience similar cage
reactions because the electronic reorganization should be
preceded. Therefore, only the rate of thermal decomposi-
tion of p-NO2 involving 3 can diminish with increase of
solvent viscosity.
hardly excels contribution of b for ∆∆Hq
> 0. The
Y-H
25
disturbance of activation enthalpy by p-OCH3 again
reveals a positive value, i.e., ∆∆Hq ) 1.2 kcal mol-1
,
Y-H
when tert-butoxy radical abstracts hydrogen atom from
p-CH3OC6H4OH.
Figure 3 shows a positive deviation for p-NO2. Homoly-
sis of the perester bearing p-NO2 has shown viscosity-
dependent rates4 and exhibits very low secondary R-deu-
terium kinetic isotope effect (compare kYH/kYD ) 1.051
for p-NO2 with kYH/kYD ) 1.293 for p-OCH3).7 Further-
more, p-NO2 delivers σ+ ) 0.79,14 which can drastically
destabilize the cationic site of 1. These phenomena,
altogether, might have suggested that p-NO2 prevents
cleavage of x and promotes scission of z only. However,
our activation enthalpy for p-NO2 marks ∆Hq ) 26.20
Con tr ibu tion of Activa tion En tr op y. One-bond
homolysis21-23 reveals positive activation entropy (∆Sq >
10 eu) with the exception of isomerizations of keten-
imines19 (Table 3). The reactions of a ketenimine19
Y
(14) March, J . Advanced Organic Chemistry, 4th ed.; Wiley: New
York, 1992; Chapt 9.
(15) (a) Fisher, T. H.; Meierhoefer, A. W. J . Org. Chem. 1978, 43,
220 and 224. (b) Fisher, T. H.; Dershem, S. M.; Prewitt, M. L. J . Org.
Chem. 1990, 55, 1040.
(16) Creary, X.; Mehrsheikh-Mohammadi, M. E.; McDonald, S. J .
Org. Chem. 1987, 52, 3254.
(17) Dust, J . M.; Arnold, D. R. J . Am. Chem. Soc. 1983, 105, 1221.
(18) (a) J iang, X.-K.; J i, G.-Z. J . Org. Chem. 1992, 57, 6051. (b) J iang,
X.-K. Acc. Chem. Res. 1997, 30, 283.
(19) Kim, S. S.; Zhu, Y.; Lee, K. H. J . Org. Chem. 2000, 65, 2919.
(20) The TS assumes a polar structure. However, the products are
neutral species, i.e., YPhCH2•, (CH3)3CO•, and CO2.
(21) Blomquist, A. T.; Berstein, I. A. J . Am. Chem. Soc. 1951, 73,
5546.
(22) Raley, J . H.; Rust, F. F.; Vaughan, W. E. J . Am. Chem. Soc.
1948, 70, 1337.
discloses ∆Sq ) -1.1 eu, which has been attributed to
Y
loss of degrees of freedom of the TS derived from the
restricted rotations of the phenyl ring.24 Other negative
activation entropies due to such hindered rotations
occurring in two-bond homolysis are shown in Table 3.
Our values of activation entropy (∆SqY) seem to be
relatively small being ∆Sq < 10 eu (Table 2). Small
Y
figures can be primarily attributed to decrease the
rotational degree of freedom derived from the formation
of two π bonds.
The electron donators such as p-CH3 and p-Ph can
make a stronger π bond (w) through a and bring about
reduction of rotational entropy. On the contrary, the
influence of b derived from bond cleavage (x) provokes to
increase the translational entropy. The positive sign of
(23) Smid, J .; Rembaum, A.; Szwarc, M. J . Am. Chem. Soc. 1956,
78, 3315.
(24) The idea of negative entropy due to the restricted rotations of
the phenyl ring has been kindly proposed by Professor T. T. Tidwell.
(25) (a) Das, P. K.; Encinas, M. V.; Steeken, S.; Scaiano, J .C. J . Am.
Chem. Soc. 1981, 103, 4162. (b) The negative sign of activation
entropies (∆SqY < 0) can be understood in terms of TS structure of the
hydrogen abstractions.
∆∆Sq
for p-CH3 (5.82 eu) and p-Ph (5.28 eu) can occur
Y-H
when increment of translational entropy outweighs dec-

4010 J . Org. Chem., Vol. 66, No. 11, 2001
Kim et al.
rement of rotational entropy. Hydrogen abstractions from
eters (∆∆Hq
and T∆∆SqY-H). The linear trends can be
Y-H
substituted phenols by tert-butoxy radical25 exhibit ∆Sq
attained when the substituents systematically control the
activation process. The differential activation entropy
multiplied by the temperature (T∆∆SqY-H) exceeds its
enthalpic mate (∆∆HqY-H) to realize entropy control of
rates. The differential entropy terms (∆∆SqY-H) are
determined by counteracting contributions of transla-
tional and rotational entropies. p-OCH3 and p-OPh are
strong electron-donators and trigger an abnormally large
loss of rotational entropy to give the negative deviations.
However, the rates are still influenced by entropy when
b outweighs a. p-NO2 is such a powerful spin-delocalizer
that the concerted mechanism may alternatively choose
a radical-like TS. The spin dispersions can engender a
weaker π bond (less hindered rotations) in 3 to afford less
negative rotational entropy, which is responsible for the
positive deviations. The solvent viscosity studies for
p-NO2 can be consistent with two-bond homolysis. The
entropy scale is sensitive enough to detect relatively weak
molecular interactions involving differential activation
Y
) -5.10 eu for p-OCH3 and ∆Sq ) -11.72 eu for p-H.
The reaction of p-OCH3 indicates the differential entropy
Y
term of ∆∆Sq
) 6.62 eu, which can be again attributed
Y-H
to dominance of b against a when p-OCH3 takes place of
p-H. Utilizing electron-withdrawing ability, p-Cl and
m-Cl may provide weaker π bond (w) and less bond
cleavage (x). Therefore, the rotation could become less
frozen via a and the translation should be curtailed by b.
The negative differential entropy values (p-Cl ∆∆Sq
Y-H
) -0.92 eu; m-Cl ∆∆Sq
) -3.91 eu) can be produced
Y-H
when b excels a. Accordingly, plot of T∆∆Sq
vs σ+
Y-H
(Figure 4) shows a good linear relationship with p-CH3,
p-Ph, p-H, p-Cl, and m-Cl (R2 ) -5.22, r ) 0.97).
The negative deviations of p-OCH3 (∆∆Sq
) 6.79
Y-H
eu) and p-OPh (∆∆Sq
) 6.50 eu) in Figure 4 can be
Y-H
due to the quinoid-like TS of 1 that prohibits another free
rotation. However, such further entropy reduction cannot
still nullify the increase of translational entropy derived
free energy ∆∆Gq
< 2 kcal mol-1. Activation enthal-
from b!! Therefore, p-OCH3 and p-OPh reveal ∆∆Sq
Y-H
Y-H
pies (∆H*Y) exhibit far larger magnitudes than their
differential terms (∆∆H*Y-H). However, differential ac-
tivation entropy terms (∆∆SqY-H) display figures that are
comparable to or even bigger than corresponding activa-
tion entropies (∆SqY).
> 0 to maintain entropic dominance. On the contrary,
p-NO2 exhibits a remarkable positive deviation with
∆∆Sq
) -5.47 eu. A large negative value of activation
Y-H
entropy (∆SqY ) -6.39 eu) may also strongly support two-
bond homolysis (refer to Table 3). The positive deviation
could be due to more flexible TS structure. The spin
delocalizations occurring in 3 may engender a weaker pi
bond so that the rotations can be less hindered than in
case of 1.
Exp er im en ta l Section
Ma ter ia ls a n d Meth od s. Substituted phenyl acetic acid,
N,N′-carbonyldiimidazole, tert-butyl hydroperoxide, and other
reagents were purchased from the major suppliers. Liquids
were distilled with center-cut collection, and solids were
purified according to standard procedures.27 A Varian Gemini
2000 NMR spectrometer was used for the quantitative analysis
of the reaction mixtures. Identifications of the products were
performed with a Shimadzu GC MS-QP505A.
Rin g-su bstitu ted ter t-bu tyl p h en ylp er a ceta tes (YC6-
H4CH2CO3Bu t, Y) p-OCH3, p-OP h , p-CH3, p-P h , p-H, p-Cl,
a n d m -Cl)7 were prepared by the reactions of the correspond-
ing phenylacetic acid, N,N′-carbonyl diimidazole, and tert-butyl
hydroperoxide.
F ea tu r e of En tr op y Con tr ol of Ra tes. p-CH3, p-Ph,
H, p-Cl, and m-Cl show good linearities as shown with
Figures 2, 3, and 4. The parallelism may indicate that
similar mode of substituent interactions take place in the
bond cleavage (x) and formation (w) in 1. The slope of
isokinetic relation of Figure 2 indicates an isokinetic
temperature (TK ) 230 K), which is well below present
reaction temperatures, indicating entropy control of rates.
The entropic change multiplied by the temperature
against variations of substituents (Figure 4) retains a
slope of R2 ) -5.22. The corresponding slope for the
perturbation of enthalpy indicates R1 ) -3.34 from
Figure 3. The steeper slope (R2 ) -5.22) may again
suggest that the substituents influence the rates through
entropic variations. The relative rates of the five sub-
stituents either slightly increase or stay constant with
higher temperatures (a sign arguing for entropy control
of rates).9-11 Such pattern of relative rates violates
reactivity/selectivity principle (RSP) and can be rational-
ized with eq 7. Drastic skeletal alterations of R in RCO3C-
(CH3)3 may engender rates of the thermolysis to be
influenced by an enthalpy factor. On the contrary, the
variation of substituents on the phenyl ring can cause
much weaker interactions26 which can be properly moni-
tored with entropy scale only.
ter t-Bu tyl p-n itr op h en yl p er a ceta te (p-NO2C6H4CH2-
CO3Bu t)1 was synthesized according to a previous method. All
the peresters show IR stretching at 1760-1770 cm-1 1H NMR
.
data for the peresters (CDCl3 with 0.03% TMS): p-P h O 7.4-
6.9 (m, 9H), 3.62 (s, 2H), 1.26 (s, 9H); p-P h 7.6-7.3 (m, 9H),
3.65 (s, 2H), 1.29 (s, 9H); p-Cl 7.3-7.2 (m, 4H), 3.62 (s, 2H),
1.27 (s, 9H); m -Cl 7.3-7.1 (m, 4H), 3.62 (s, 2H), 1.28 (s, 9H).
The NMR data of other peresters not given here were already
provided in the previous paper.7
Th er m a l Rea ction s of ter t-Bu tyl P h en ylp er a ceta tes.
Weighed samples of a perester (50 mg), acetonitrile (7 mg,
internal standard), and iodine (100-150 mg) were dissolved
in CDCl3 (5 mL). The solutions were divided into several Pyrex
ampules, which were degassed and sealed by a freeze-pump-
thaw method. The ampules (i.d. 4 mm, length 3 cm, 2/3 full)
were immersed in a constant-temperature bath for at least
200 s for thermal equilibration. Less than 20 s were required
for complete thermal equilibration, which was measured by a
Copper-Constantan thermocouple. At various intervals, the
tubes were removed from the bath, quenched in ice-water,
and opened for NMR analysis.
Q ) ∆∆SqY-H/R - ∆∆HqY-H/RT, where Q ) ln kY/kH
(7)
The integrations of the benzylic peak of a perester at δ )
3.6-3.7 ppm and the peak of acetonitrile (δ ) 2.0 ppm) have
been made. When integrated values of the perester and
acetonitrile are designated as IP and IS, respectively, the
Con clu sion
p-CH3, p-Ph, p-H, p-Cl, and m-Cl cooperate to reveal
linear relations employing differential activation param-
2
following equation can be obtained: IP/IS ) /3C/CS. C and CS
(26) Inoue, Y.; Ikeda, H.; Kaneda, M.; Sumimura, T.; Everitt, S. R.
L.; Wada, T. J . Am. Chem. Soc. 2000, 122, 406
(27) Perrin, D. D.; Armarego, W. L. F.; Perrin, D. F. Purification of
Laboratory Chemicals, 2nd ed.; Pergamon Press: Oxford, 1980.

Thermolysis of tert-Butyl Phenylperacetates
J . Org. Chem., Vol. 66, No. 11, 2001 4011
represent the concentrations of the perester and acetonitrile,
respectively. Rate constants were then produced by the method
of least squares from ln(C0/Ct) ) kYt. C0 and Ct are concentra-
tions of the perester at time 0 and t, respectively. C0 has been
fixed as a concentration of a perester that already underwent
homolysis for more than thermal equilibration time. Most of
the plots of ln(C0/Ct) vs t showed excellent linear relations
(r g 0.9990). However, the reactions of p-methoxyphenyl and
p-phenoxyphenyl peresters did not show simple first-order
kinetics. The product analysis suggested that corresponding
benzyl iodides are unstable under the reaction conditions,
which could interfere with homolysis of the perester. Pyridine
has been added to trap the benzyl iodides so as to observe first-
order kinetics.
Ack n ow led gm en t. We warmly thank Korea Re-
search Foundation (KRF) for a Non-Directed Research
Fund (1996-1999). Technical assistance was provided
by Mr. J i Young Song and Miss Eun J ung Kim. We are
also indebted to Brain Korea 21 provided from Ministry
of Education through KRF.
Su p p or tin g In for m a tion Ava ila ble: Graphical plots of
lnC0/Ct vs t have been provided to show the accurate measure-
ment of rate constant (kY). This material is available free of
charge via the Internet at http://pubs.acs.org.
J O010143J