The radical-induced decomposition of 2-methoxyphenol

Article abstract of DOI:10.1039/a809619h

The thermolysis of 2-methoxyphenol has been studied between 680 and 790 K, applying cumene as a radical scavenger. Two pathways were observed: a homolytic route involving the cleavage of the methoxyl O-C bond, leading to methane and 1,2-dihydroxybenzene, obeying k_uni (s^-1) = 10^{15.2 ± 0.2} exp(-239 ± 8 (kJ mol^-1)/RT), and an induced route starting with abstraction of the phenolic hydrogen by cumyl radicals. After intramolecular hydrogen transfer a cascade of reactions yields phenol, 2-hydroxybenzaldehyde, and 2-hydroxybenzyl alcohol. The latter compound decomposes instantaneously into o-quinone methide (o-QM) which is reduced to o-cresol.

Full text of DOI:10.1039/a809619h

View Article Online / Journal Homepage / Table of Contents for this issue
The radical-induced decomposition of 2-methoxyphenol
Edwin Dorrestijn and Peter Mulder*
Leiden Institute of Chemistry, Leiden University, PO Box 9502, 2300 RA Leiden,
The Netherlands. E-mail: P.Mulder@Chem.LeidenUniv.nl; Fax: ϩ31-71-5274492
Received (in Cambridge) 9th December 1998, Accepted 20th January 1999
The thermolysis of 2-methoxyphenol has been studied between 680 and 790 K, applying cumene as a radical
scavenger. Two pathways were observed: a homolytic route involving the cleavage of the methoxyl O–C bond, leading
Ϫ1
15.2 ± 0.2
Ϫ1
to methane and 1,2-dihydroxybenzene, obeying kuni (s ) = 10
exp(Ϫ239 ± 8 (kJ mol )/RT), and an induced
route starting with abstraction of the phenolic hydrogen by cumyl radicals. After intramolecular hydrogen transfer
a cascade of reactions yields phenol, 2-hydroxybenzaldehyde, and 2-hydroxybenzyl alcohol. The latter compound
decomposes instantaneously into o-quinone methide (o-QM) which is reduced to o-cresol.
Introduction
Lignin contains numerous methoxyl substituted phenolic
structures, stemming from the precursors p-coumaryl alcohol,
coniferyl alcohol, and sinapyl alcohol. The conversion of lignin
into useful (feedstock) chemical compounds has been a chal-
lenge for a long time. During pyrolysis and hydrogenolysis
complex product mixtures are obtained not only containing
various substituted 2-methoxy- and 2,6-dimethoxyphenols, but
also o-cresol and derivatives, which are believed to originate
1
from the degradation of these methoxyphenols. Besides, a large
1
amount of tarry material is formed, often more than 50%. In
order to develop a process that is commercially feasible, it is
required that the pyrolysate only contains a limited number of
different compounds. Thus, secondary desubstitution reactions
may be beneficial to reduce the product spectrum.
Fig. 1 Conversion and product yields from the thermolysis of
Ϫ3
2
-methoxyphenol in cumene. Conditions: [cumene] = 2.2 × 10 M, [2-
Ϫ4
methoxyphenol] = 1.5 × 10 M, t = 9.5 s (at 740 K). The methane yield
(
Substituted) 2-methoxyphenols have often been used as
is corrected for the contribution stemming from the cumene pyrolysis
model compounds for lignin, e.g. in investigations of the effect
(
styrene and tert-butylbenzene).
2
of solvent, substituents, and processing conditions. However,
most of the studies do not detail mechanistic features. Besides
homolysis of the weak phenoxyl–methyl bond, a radical-
induced pathway appears to be important with 2-hydroxy-
methanol and the hot solution was decanted. 9,10-Dihydro-
anthracene (Janssen, 97%) was crystallized twice from
boiling methanol before use. All other compounds were
obtained from commercial sources in the highest purity and
used as received.
2a
benzyl alcohol as the reactive intermediate.
To investigate the contribution of a free radical degradation
mechanism in the decomposition of 2-methoxyphenol, the
thermolysis of this lignin model compound has been studied in
the gas phase between 680 and 790 K, using cumene as the
radical scavenger in a well-defined chemical environment.
Results and discussion
Products and mechanism
In Fig. 1, the 2-methoxyphenol conversion and product
yields between 680 and 790 K are presented. Major products
were 1,2-dihydroxybenzene, 2-hydroxybenzaldehyde, o-cresol,
methane, and carbon monoxide. Small amounts of 2-ethyl-
phenol and phenol were also found (1% and 3% at 784 K,
respectively). The mass balance decreased to 89% at 91%
conversion of 2-methoxyphenol. Assuming a pseudo-first-
order behavior for 2-methoxyphenol, the Arrhenius para-
Experimental
The thermolysis of 2-methoxyphenol was studied with a
pyrolysis instrument with off-line GC and GC-MS analysis.
Experimental details have been described before. In short: a
mixture of 2-methoxyphenol, chlorobenzene (as internal
standard), and cumene (molar ratio 6:3:91) was injected into a
nitrogen gas stream and led through a heated plug-flow quartz
reactor (effective volume 43.5 ml). The residence times were
about 9 s. After 1 h of equilibration, samples were collected at
about 195 K and analyzed.
Additional experiments in the liquid phase at elevated
temperatures were carried out in pressurized sealed Pyrex
tubes. The experimental procedure has been described before.
3
meters for the overall decomposition were derived: k
overall
Ϫ1
12.2
Ϫ1
(
s ) = 10
exp(Ϫ191 (kJ mol )/RT). Clearly, these param-
eters are indicative for a mixed homolytic and induced
decomposition route.
The homolytic pathway for 2-methoxyphenol starts with the
cleavage of the weak phenoxyl–methyl bond. The methyl and
the 2-hydroxyphenoxyl radicals abstract a hydrogen atom from
cumene to give methane and 1,2-dihydroxybenzene [eqn. (1)].
4
Naphthalene was used as a standard for the quantification of
the compounds in the product mixture.
Chemicals
Solid anthracene (Aldrich, >95%) was washed with boiling
J. Chem. Soc., Perkin Trans. 2, 1999, 777–780
777

View Article Online
The first step in the induced decomposition is the hydrogen
abstraction by cumyl radicals, abundantly present in the
pyrolysis mixture, from the hydroxy group of 2-methoxyphenol
Remarkably, in the thermolysis experiments the CO yield
surmounts (3–4 times) the phenol yield (see Fig. 1), while
according to Scheme 1 a 1:1 ratio would be expected. The
6
(
which is the weakest H–R bond in the molecule). Sub-
thermal decomposition of o-QM into benzene and CO is too
sequently, the phenoxyl moiety shuttles a hydrogen from the
slow at these temperatures,§ and also CO elimination from the
2-hydroxyphenoxyl radical can be neglected since no CO has
been observed in the thermolysis of anisole (to yield phenoxyl
adjacent methoxyl group to oxygen, leading to intermediate A
ؒ
(
Scheme 1). Starting from A, a rearrangement from –OCH2
9
radicals) under similar conditions: hydrogen abstraction from
cumene is much faster than CO elimination. However, a frac-
tion of the 2-hydroxyphenoxyl radicals can disproportionate
with the abundantly present cumyl radicals, yielding o-quinone.
CO elimination from o-quinone is fast and yields cyclopenta-
dienone, which dimerizes into an array of compounds during
3
condensation of the effluent stream. Unimolecular hydrogen
elimination from the 2-hydroxyphenoxyl radical to give
o-quinone can be disregarded as well because of the relatively
Ϫ1
strong O–H bond in this radical (258 kJ mol ).¶
Note that the 2-hydroxyphenoxyl radical is not only formed
after homolysis of the O–CH bond in 2-methoxyphenol, but
3
also after hydrogen abstraction by cumyl radicals from the
product 1,2-dihydroxybenzene, similar to the induced decom-
position of 2-methoxyphenol. Based on reaction enthalpy
assessment, 1,2-dihydroxybenzene may be even more reactive
(
1
vide infra). The ratio 1,2-dihydroxybenzene:COexcess is around
0, indicating that the hydrogen abstraction by the 2-hydroxy-
phenoxyl radical from cumene is faster than the elimination of
CO. Anyway, the additional CO stems from the homolytic
degradation of 2-methoxyphenol.
Scheme 1
Kinetics
ؒ
towards –CH O via a three-membered ring (1,2 aryl shift) gives
2
When 1,2-dihydroxybenzene and the excess CO are considered
as the products of the homolysis, the contribution of the uni-
molecular decomposition to the overall rate of disappearance
for 2-methoxyphenol increases with the temperature, from 29%
at 702 K to 68% at 784 K. The rate constant for bond homolysis
intermediate B. Elimination of a hydrogen atom from this
alkoxyl radical leads to 2-hydroxybenzaldehyde. As a compet-
ing reaction, elimination of formaldehyde from A provides a
2
-hydroxyphenyl radical which, after hydrogen abstraction
from the carrier, is converted into phenol. Under these high
Ϫ1
15.2 ± 0.2
Ϫ1
obeys kuni (s ) = 10
exp(Ϫ239 ± 8 (kJ mol )/RT). After
temperature conditions, formaldehyde eliminates H ,† resulting
2
temperature correction (assuming ∆C ≈ 0: BDE = E ϩ RT
in the formation of carbon monoxide.
The pathway for the formation of o-cresol starts from
p
a
m
where Tm is the medium temperature and C is the heat
capacity), the methoxyl BDE(O–CH ) can be derived as 245
2
-hydroxybenzyl alcohol. This intermediate is not detected
3
Ϫ1
kJ mol , which is in good agreement with the literature values
Note that the weakening effect on the
BDE(O–CH ) by the ortho-hydroxy group of 27 kJ mol
since, at the applied temperature, it readily eliminates water to
Ϫ1 2b,c
(
ca. 243 kJ mol ).
result in o-quinone methide (o-QM) which in turn is reduced
Ϫ1
6
to o-cresol (vide infra). Retention of configuration for the
3
is markedly higher than by an ortho-methoxy group (16 kJ
o-cresol formation was verified by performing the thermolysis
of 2-methoxy-4-methylphenol. Now the main phenolic product
is 2,4-dimethylphenol showing that intramolecular radical
addition in A (to give a four-membered ring, with subsequent
loss of hydroxyl and finally leading to 2,5-dimethylphenol) does
not occur.
Ϫ1
3
mol ).
The induced decomposition is set in by cumyl radicals, and
by using the yield of α-methylstyrene to deduce the cumyl rad-
ical concentration,|| the rate constant for hydrogen abstraction
5
Ϫ1 Ϫ1
6
can be estimated:** 4 × 10 M
s
at 702 K and 1.5 × 10
Ϫ1 Ϫ1
M
s
at 784 K.
The transformation of o-QM into o-cresol consists of a
reverse radical disproportionation (RRD) mechanism: the
transfer of hydrogen from cumene. For example, at 784 K the
RRD rate constant for reaction of o-QM with cumene amounts
The Arrhenius parameters for the rearrangement A→B have
Ϫ1
14.8
Ϫ1
6a
§
From k (s ) = 10 exp(Ϫ281 (kJ mol )/RT) for the unimolecular
Ϫ1
to 87 s .‡ This rate constant is apparently high enough to pre-
decomposition of o-QM it follows that the conversion is 0.1% at 784 K.
6a
vent oligomerization of o-QM.
¶ The O–H bond strength in substituted phenols can be estimated
1
0
ϩ
o
ϩ
ϩ
Ϫ1
using: ∆BDE(O–H) = 30.63 {Σ(σ ϩ σm ϩ σp )} Ϫ 2.68 kJ mol
;
The expulsion of formaldehyde from A gains in importance
with temperature and hence the ratio (2-hydroxybenzaldehyde
ϩ o-cresol):phenol decreased from 42 to 6 between 729 and 784
K. The increase in phenol may also be due to the conversion of
ϩ
ϩ
11
Ϫ1
1
and σo = 0.66 σp
. With BDE(O–H) = 366 kJ mol in phenol, and
ϩ
12
σp (OH) = Ϫ0.92, BDE(O–H) in 1,2-dihydroxybenzene becomes 345
Ϫ1
Ϫ1
kJ mol . Thus, for 2-hydroxyphenoxyl ∆ H = Ϫ145 kJ mol (using
f
ؒ
Ϫ1
∆fH(H ) = 218 kJ mol and ∆ H(1,2-dihydroxybenzene) = Ϫ272 kJ
f
Ϫ1
8
Ϫ1 8
2
-hydroxybenzaldehyde into phenol via hydrogen abstraction
mol ). With ∆ H(o-quinone) = Ϫ105 kJ mol , the BDE(O–H) in the
f
Ϫ1
and CO elimination. When related to lignin chemistry (vide
supra), high reaction temperatures seem to be beneficial to
enhance the formation of base chemicals (e.g. phenol).
2-hydroxyphenoxyl radical can be estimated as 258 kJ mol . With
1
5
Ϫ1
E = BDE Ϫ RT (when neglecting ∆ C ) and A = 10
s
, the rate for
a
r
p
homolysis of the O–H bond in the 2-hydroxyphenoxyl radical becomes
Ϫ2 Ϫ1
2
|
× 10
s , about 10 times slower than the CO elimination.
2
| As an assumption: [α-methylstyrene] = kdisproportionationt[cumyl] . With
t being the residence time (9.9 s at 702 K, 8.5 s at 784 K),
1
0.5 Ϫ1
†
The Arrhenius parameters for this elimination are: A = 10 Ϫ1s and
Ϫ1
5
9
Ϫ1 Ϫ1
Ϫ6
E = 146 kJ mol . The rate constant at 750 K becomes 2 s , and a
kdisproportionation = 10 M
s
, and [α-methylstyrene] = 2 × 10 M at 702
a
Ϫ5
Ϫ8
conversion of 99% is reached within 2 s.
K and 5 × 10 M at 784 K. As a result [cumyl] = 1.4 × 10 M at 702 K
Ϫ8
‡
For RRD the general Arrhenius parameters are log A = 8.6 (per H)
and 7.7 × 10 at 784 K.
Ϫ1
7
and E = E ϩ 0.82 × ∆ H (with E = 37.66 kJ mol ). With cumene,
∆
exp(Ϫ100 (kJ mol )/RT).
** From kuni = Q × kind × [cumyl], with Q is the fraction of homolytic
a
0
r
0
Ϫ1 6a,8
Ϫ1 Ϫ1
8.6
H becomes 76 kJ mol
,
and thus kRRD,cumene (M
s
) = 10
decomposition divided by the fraction of induced decomposition.
r
Ϫ1
Ϫ3 Ϫ1
kuni = 1.7 × 10
s
at 702 K and 0.175 at 784 K.
7
78
J. Chem. Soc., Perkin Trans. 2, 1999, 777–780

View Article Online
13
Ϫ1
12.5
Ϫ1
been reported as kuni (s ) = 10
exp(Ϫ88 (kJ mol )/RT)
with abstraction of a hydride ion from the solvent and as a
and thus, from the ratio (2-hydroxybenzaldehyde ϩ o-cresol):
consequence the overall mass balance comprising semi-volatile
products is reduced.
The induced decomposition is clearly more important in the
liquid phase due to the relatively high anthracenyl concen-
phenol it can be derived that the rate constant for elimination
5
Ϫ1
of formaldehyde from A is around 5 × 10 s at 784 K, in good
3
agreement with earlier results. The product ratio 2-hydroxy-
benzaldehyde:o-cresol equals 2.3 at 784 K. Most likely, the
route to 2-hydroxybenzyl alcohol (and thus o-cresol) starts with
intramolecular hydrogen abstraction in B to give a phenoxyl
radical (see Scheme 1). The estimated Arrhenius parameters
that have been reported for the hydrogen atom elimination from
tration. From the equilibrium relation for eqn. (2), K (1) = 1.3
eq
ؒ
Ϫ1
14.1
Ϫ1
14
PhCH O , k (s ) = 10 exp(Ϫ4.6 (kJ mol )/RT), predict a
2
13
Ϫ1
rate constant of 6 × 10
s
for the formation of 2-hydroxy-
Ϫ1
17
ؒ
Ϫ6
benzaldehyde. This means that the rate constant for the intra-
exp(Ϫ132 (kJ mol )), it follows that [AnH ] = 8 × 10 M,
predicting a complete conversion of 2-methoxyphenol if the
induced decomposition were solely determined by the hydrogen
13
Ϫ1
molecular hydrogen abstraction in B amounts to 3 × 10 s ,
which is in accordance with the fact that alkoxyl radicals react
rapidly with phenolic hydrogens.
4
Ϫ1 Ϫ1
abstraction (with kind = 4 × 10 M
s
at 625 K).†† However,
in the presence of a high concentration of a hydrogen donor,
the hydrogen abstraction is an equilibrium process: hydrogen
Intramolecular hydrogen bonding
abstraction from AnH by the phenoxyl radical is in competi-
2
The reagent 2-methoxyphenol and the products 1,2-dihydroxy-
benzene and 2-hydroxybenzaldehyde exist in two entities: the
open structure and the intramolecularly hydrogen bonded
form. The latter one is the most stable conformer of 2-methoxy-
phenol (Q, see Scheme 2).
tion with the intramolecular hydrogen abstraction from the
methoxyl substituent.
In retrospect, lignin liquefaction can only be applied if the
oligomerization can be prevented by applying highly dilute
solutions or appropriate solvents (e.g. phenol) which lead to
more uniform product distributions.
Acknowledgements
The authors wish to thank Carmen Santos Gonzalez and
Wojciech Le s´ niak for experimental assistance.
Scheme 2
†
† From the rate constants for induced decomposition in cumene it
Ϫ1
8.5
Ϫ1 Ϫ1
5
follows that E is ca. 37 kJ mol , assuming A = 10
the difference in BDE(C–H) between cumene and AnH
derived that the E for hydrogen abstraction by AnH is around 47 kJ
mol and thus kind = 4 × 10 M
M
s
. From
it can be
a
By considering the changes in enthalpy (∆H(P→Q) = Ϫ16 kJ
1
8
Ϫ1
Ϫ1
Ϫ1 15
2
mol ) and entropy (∆S(P→Q) = 0 kJ mol K ), even at 750
K, 93% exists as Q. When the homolytic cleavage takes place
only from Q, the rate for disappearance of 2-methoxyphenol
ؒ
a
Ϫ1
4
Ϫ1 Ϫ1
s
at 625 K.
(
Q ϩ P) is given by kexp = khom(K/(K ϩ 1)). If this is the main
References
pathway (Q→P→P*) the experimental activation enthalpy also
includes the difference in enthalpy between P and Q. Our
kinetic data show that this is not the case. Thus, the oxygen–
carbon bond cleavage occurs without disturbing the hydrogen
bond. Interestingly, computational studies have revealed that
the enthalpy difference between P and Q is identical to that
1
E. Dorrestijn, L. J. J. Laarhoven, I. W. C. E. Arends and P. Mulder,
J. Anal. Appl. Pyrolysis, in the press.
2
(a) A. Vuori, Fuel, 1986, 65, 1575; (b) M. M. Suryan, S. A. Kafafi
and S. E. Stein, J. Am. Chem. Soc., 1989, 111, 1423; (c) M. M.
Suryan, S. A. Kafafi and S. E. Stein, J. Am. Chem. Soc., 1989, 111,
4594; (d) C. P. Masuku, Acta Polytech. Scand., Chem. Technol.
Metall. Ser., 1991, 196, 1; (e) A. Vuori and J. B. Bredenberg,
Holzforschung, 1984, 38, 133; ( f ) I. A. Shkrob, M. C. Depew and
J. K. S. Wan, Res. Chem. Intermed., 1992, 17, 271; (g) L. A. Edye
and G. N. Richards, Environ. Sci. Technol., 1991, 25, 1133; (h) A. I.
Vuori and J. B. Bredenberg, Prepr. Pap. — Am. Chem. Soc., Div. Fuel
Chem., 1981, 30, 366; (i) M. T. Klein and P. S. Virk, Prepr. Pap. —
Am. Chem. Soc., Div. Fuel Chem., 1980, 25, 180; (j) A. I. Vuori and
J. B. Bredenberg, Ind. Eng. Chem. Res., 1987, 26, 359; (k) S. J. Hurff
and M. T. Klein, Ind. Eng. Chem., Fundam., 1983, 22, 426; (l) J. R.
Lawson and M. T. Klein, Ind. Eng. Chem., Fundam., 1985, 24, 203;
(m) J. B. Bredenberg, M. Huuska and P. Toropainen, J. Catal., 1989,
Ϫ1
between P* and Q* (16 kJ mol ) in favour of the hydrogen
15,16
bonded entity. . Hydrogen bonding has no influence on the
weakening effect of the hydroxyl group and thus also not on the
oxygen–methyl bond strength.
Liquid-phase experiments
Since lignin pyrolysis can only be carried out in the liquid
phase, additional experiments with 2-methoxyphenol (0.3 M)
have been performed under typical coal liquefaction conditions.
Thus, a 1:1 mixture of 9,10-dihydroanthracene (AnH , as the
hydrogen donor) and anthracene (An), both 2.4 M, was applied
as a solvent.
Typically, after a reaction time of 5 h at 625 K, a conversion
of 38% was reached. In contrast with the experiments in the gas
phase, the mass balance is surprisingly low: the total product
yield amounted to only 7%, based on the intake of 2-methoxy-
phenol: 4% of 1,2-dihydroxybenzene, 2% of o-cresol, 1% of
phenol, and 0.1% of 2-hydroxybenzaldehyde. The low relative
yield of the latter is the result of a highly efficient scavenging of
the radicals by the hydrogen donor to give 2-hydroxybenzyl
alcohol. This compound is converted quantitatively under these
conditions. At high substrate concentrations and high temper-
1
20, 401; (n) J. B. Bredenberg and R. Ceylan, Fuel, 1983, 62, 342; (o)
2
S. E. Stein and M. M. Suryan, in 1987 International Conference on
Coal Science, ed. J. A. Moulijn, Elsevier Sci. Publ., Amsterdam,
1
1
377.
987, pp. 749–756; (p) M. Funaoka and I. Abe, Wood Sci. Technol.,
987, 21, 261; (q) R. Ceylan and J. B. Bredenberg, Fuel, 1982, 61,
3 G. J. Schraa, I. W. C. E. Arends and P. Mulder, J. Chem. Soc., Perkin
Trans. 1, 1994, 2, 189.
4
(a) E. Dorrestijn, M. Kranenburg, M. V. Ciriano and P. Mulder,
J. Org. Chem., in the press; (b) E. Dorrestijn, S. Hemmink,
G. Hulstman, L. Monnier, W. van Scheppingen and P. Mulder,
Eur. J. Org. Chem., in the press.
5 W. G. Mallard, F. Westley, J. T. Herron and R. F. Hampson, NIST
Chemical Kinetics Database, version 5.0, NIST Standard Reference
Data, National Institute of Standards and Technology: Gaithers-
burg, MD, 1993.
4a
atures in the liquid phase we have shown before that the
formation of o-QM from this compound does not take place.
Instead, a benzylic cation is formed of which the oligomeriz-
ation via electrophilic aromatic substitution is in competition
6
(a) E. Dorrestijn, R. Pugin, M. V. Ciriano Nogales and P. Mulder,
J. Org. Chem., 1997, 62, 4804; (b) E. Dorrestijn, O. J. Epema, W. B.
van Scheppingen and P. Mulder, J. Chem. Soc., Perkin Trans. 2,
1998, 1173.
J. Chem. Soc., Perkin Trans. 2, 1999, 777–780
779

View Article Online
7
8
P. E. Savage, Energy Fuels, 1995, 9, 590.
13 M. F. R. Mulcahy, B. G. Tucker, D. J. Williams and J. R.
Wilmshurst, Aust. J. Chem., 1967, 20, 1155.
14 K. Brezinsky, T. A. Litzinger and I. Glassman, Int. J. Chem. Kinet.,
1984, 16, 1053.
15 M. I. de Heer, H.-G. Korth and P. Mulder, J. Org. Chem., submitted.
16 A. M. Mebel and M. C. Lin, J. Am. Chem. Soc., 1994, 116, 9577.
17 I. W. C. E. Arends and P. Mulder, Energy Fuels, 1996, 10, 235.
18 D. Santoro, H.-G. Korth and P. Mulder, J. Am. Chem. Soc.,
submitted.
S. E. Stein, J. M. Rukkers and R. L. Brown, NIST Structures and
Properties Database, version 2.0, NIST Standard Reference Data,
National Institute of Standards and Technology: Gaithersburg,
MD, 1994.
9
I. W. C. E. Arends, R. Louw and P. Mulder, J. Phys. Chem., 1993,
9
7, 7914.
1
1
1
0 D. D. M. Wayner, E. Lusztyk, K. U. Ingold and P. Mulder, J. Org.
Chem., 1996, 61, 6430.
1 M. Jonsson, J. Lind, T. E. Eriksen and G. Merrényi, J. Chem. Soc.,
Perkin Trans. 2, 1993, 1567.
2 C. Hansch, A. Leo and R. W. Taft, Chem. Rev., 1991, 91, 165.
Paper 8/09619H
7
80
J. Chem. Soc., Perkin Trans. 2, 1999, 777–780