Product selectivities and third-order rate laws for solvolyses of ethyl phenylphosphonochloridate in aqueous alcohols

Article abstract of DOI:10.1002/poc.425

Rate constants and product selectivities (S) for solvolyses of ethyl phenylphosphonochloridate [PhP(=O)OEtCl] in aqueous ethanol and methanol at 0°C are reported; S = ([ester product]/[acid product]) x ([water]/[alcohol solvent]). The results show trends

Full text of DOI:10.1002/poc.425

JOURNAL OF PHYSICAL ORGANIC CHEMISTRY
J. Phys. Org. Chem. 2001; 14: 759–763
DOI:10.1002/poc.425
Product selectivities and third-order rate laws for
solvolyses of ethyl phenylphosphonochloridate in
†
aqueous alcohols
T. William Bentley* and David N. Ebdon
Department of Chemistry, University of Wales Swansea, Singleton Park, Swansea SA2 8PP, UK
Received 5 April 2001; revised 23 May 2001; accepted 1 June 2001
ABSTRACT: Rate constants and product selectivities (S) for solvolyses of ethyl phenylphosphonochloridate
[
(
PhP(
[water]/[alcohol solvent]). The results show trends very similar to those previously reported for solvolyses of
O)Cl] and p,p'-dichlorodiphenyl phosphorochloridate, much more
hydrophobic substrates. Implications of these results are as follows: (i) as S increases about threefold from 99.8 to
0% alcohol–water and only about the same amount from 90 to 40% alcohol–water, the main cause of variations in S
=
O)OEtCl] in aqueous ethanol and methanol at 0°C are reported; S = ([ester product]/[acid product]) Â
diphenyl phosphorochloridate [(PhO) P(
=
2
9
does not appear to be medium effects of the solvents; (ii) contrary to the general trend to increase, values of S may
decrease slightly (ca 15%) in highly aqueous alcohol–water mixtures, even when (as in this case) there is no evidence
for mechanistic changes; (iii) the deviations from expected third-order rate product correlations previously reported
for solvolyses of diphenyl phosphorochloridate and the p,p'-dichloro derivative do not appear to be due solely to their
high hydrophobicity. Copyright  2001 John Wiley & Sons, Ltd.
KEYWORDS: phosphonochloridate; acid choride; kinetics; solvolysis; selectivity
5
INTRODUCTION
clear, additional evidence was obtained from kinetic
6–8
studies in a wider range of solvents and/or by studies
of product selectivities [S, Eqn. (1)]. Rates of solvolyses
8
Studies of rates and products of solvolyses in alcohol–
water mixtures can provide useful insights into solvo-
of benzoyl chloride can be explained by two competing
mechanisms, and substituents (e.g. p-Cl and p-Me) alter
the solvent composition at which the mechanistic change
occurs. Independent supporting evidence is provided by
values of S [Eqn. (1)] in alcohol–water mixtures, which
reach maxima for the solvent compositions very close to
those predicted from rate-rate profiles as the positions of
1,2
lyses, and recent progress has been made particularly
for reactions of acid halides. Reactions proceeding via
7
cationic intermediates may be ca 10 times slower in
ethanol than in water (based on solvolysis rates for 1-
adamantyl chloride, a model compound for S 1 reactions
used to define the YCl scale of solvent ionizing power).
N
3
8
Consequently, an alternative mechanism may be
favoured in ethanol and in other less aqueous media,
e.g. evidence for a mechanistic change can be seen from
the two distinct linear regions of the rate-rate profiles
observed by comparing rates of solvolyses of p-
dimethylaminobenzoyl fluoride in ethanol–water mix-
mechanistic change.
S ˆ ([ester product]/[acid product])
 ([water]/[alcohol solvent])
ꢀ1†
4
tures with YCl.
Solvolyses of electron-rich benzenesulfonyl chlorides
were interpreted similarly in terms of two competing
mechanisms,
Because evidence for mechanistic changes from
9–11
‘
breaks’ in rate rate profiles for solvolyses of other acid
but the effect of substituents is smaller
chlorides in ethanol–water mixtures may not always be
than for benzoyl chlorides. There appears to be general
agreement that less electron-rich sulfonyl chlorides
including p-methylbenzenesulfonyl chloride solvolyse
via a single reaction channel, probably via an S 2
N
10,12–15
*
Correspondence to: T. W. Bentley, Department of Chemistry,
mechanism.
However, recent alternative propo-
University of Wales Swansea, Singleton Park, Swansea SA2 8PP, UK.
E-mail: cmsolvol@swan.ac.uk
sals for solvolyses of electron-rich sulfonyl chlorides are
(i) that there is only one mechanism, and the non-linear
rate–rate profiles {and by implication possibly also the
†
Dedicated to Dr John Shorter on the occasion of his 75th birthday.
10
Contract/grant sponsor: EPSRC.
Copyright  2001 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2001; 14: 759–763

7
60
T. W. BENTLEY AND D. N. EBDON
decrease in S [Eqn. (1)] in more aqueous alcohols} is due
Table 1. First order rate constants !k) for solvolyses of ethyl
phenylphosphonochloridate !1) in aqueous ethanol and
methanol at 0°C
to a non-linear dependence of solvent nucleophilicity on
a,b
14
Y ; and (ii) that solvolyses occur by S 2 mechanisms
Cl
N
with ‘exploded’ transition states having both bond
formation to the incoming nucleophile and extensive
À3 À1
k (10
s )
16,17
Solvent (%, v/v)
100
Ethanol
Methanol
cleavage of the S—Cl bond.
Solvolyses of the chlorophosphate diphenyl phosphor-
ochloridate [(PhO) P( O)Cl] in alcohol–water mixtures
0.86 Æ 0.03
6.00 Æ 0.02
10.1 Æ 0.1
13.0 Æ 0.1
18.6 Æ 0.3
27.0 Æ 0.2
51.7 Æ 0.2
123 Æ 1
5.65 Æ 0.15
17.9 Æ 0.3
31.0 Æ 1.0
41.4 Æ 0.8
55.3 Æ 1.1
76.2 Æ 0.2
117 Æ 3
9
0
=
2
18
80
react via an S 2(P) mechanism; although there is no
N
7
0
0
evidence for mechanistic changes, S decreases slightly
6
for solvolyses in 30% methanol–or ethanol–water and
50
40
18
‘
pure’ water. However, the diphenyl substrate is not
3
0
183 Æ 4
readily soluble in highly aqueous mixtures, and values of
5,18
20
220 Æ 4
244 Æ 3
S are dependent on the efficiency of mixing.
To
1
0
350 Æ 4
363 Æ 2
investigate this problem further, and also aspects of our
Water
430 Æ 10
430 Æ 10
18
third-order treatment
of observed first-order rate
a
constants (see Discussion), we now report rate and
product data for solvolyses of ethyl phenylphosphono-
chloridate (1) in ethanol–and methanol–water mixtures.
Determined conductimetrically at least in duplicate; errors
shown are average deviations.
Injected 3–15 ml of freshly prepared 1% solutions of ethyl
phenylphosphonochloridate in acetonitrile into 5 ml of solvent.
b
RESULTS
20
chromatographic conditions. Product selectivities (S)
are given in Table 2; because values of S are low, and the
HPLC conditions were investigated in detail, it was
The substrate 1, prepared by reacting phenylphosphonic
dichloride with 1.1 equiv. of ethanol at 0°C, was further
solvolysed in alcohol and in alcohol–water mixtures at
19
0
°C to give an ester (2) and an acid (3) as shown in
Table 2. Product selectivities, S [Eqn. !1)] for solvolyses of
ethyl phenylphosphonochloridate !1) in aqueous ethanol
and methanol at 0°C
Scheme 1. Reaction rates were monitored from the
change in conductivity due to 3 and to the HCl by-
product, and first-order rate constants were calculated by
standard methods (Table 1).
a
S
Product ratios were determined using freshly distilled
Solvent (%, v/v)
99.8
Ethanol
Methanol
substrate 1 by ion pair reversed-phase high-performance
b
c
c
c
c
c
c
b
c
c
c
c
c
c
18
0.052 , 0.048
0.129 , 0.151
liquid chromatography (HPLC) as before, except that
instead of acetic acid in the eluent, triethylamine and a
higher concentration of ion-pairing reagent were added to
the eluent to give pH ꢁ10. Corrections were made for the
small percentage of ester (2) in the starting material, and
control experiments showed that the products were stable
under the acidic reaction conditions and under the basic
b
b
9
9
9
9.5
9
0.063 , 0.058
0.193 , 0.190
b
b
0.066 , 0.065
0.197 , 0.211
b
b
8
0.089 , 0.084
0.229 , 0.229
b
b
95
0.135 , 0.131
0.313 , 0.301
b
b
9
9
8
7
6
5
4
3
2
1
2
0
0
0
0
0
0
0
0
0
0.179 , 0.174
0.434 , 0.411
b
c
d
b
c
d
0.19 , 0.19 , 0.21
0.44 , 0.43 , 0.44
d
d
0.33
0.68
d
d
0.44
0.88
d
d
0.57
1.07
d
d
0.68
1.21
d
d
0.72
1.28
d,e
d
0.7
1.15
d,e
d
0.6
1.11
e
d
—
1.10
a
All values calculated from the ester:acid molar product ratio
determined by duplicate HPLC analyses on at least two
independent samples. Typical errors 1–5%. Exact concentra-
tions of water in 95–99.8% alcohol were determined by Karl
Fischer titration.
b
Typical injection: 2 ml of neat ethyl phenylphosphonochlor-
idate into rapidly stirred solvent (5 ml).
c
Solvolyses repeated using the same conditions as in footnote
b.
d
Typical injection: 20 ml of a 5% solution in acetonitrile into
rapidly stirred solvent (5 ml).
e
Additional experimental uncertainty due to the presence of
Scheme 1
diester (2) in the starting phosphonochloridate (1).
Copyright  2001 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2001; 14: 759–763

SOLVOLYSES OF ETHYL PHENYLPHOSPHONOCHLORIDATE
761
possible to obtain reliable values of S for 99.5% alcohol–
water mixtures, and even for 99.8% ethanol–water.
required to explain any decrease in S in highly aqueous
23
media.
Selectivities for carboxylic acid chorides are compli-
8,23
cated by mechanistic changes,
but solvolyses of p-
DISCUSSION
nitrobenzenesulfonyl chloride strongly support the theory
that values of S may reach a plateau in highly aqueous
24
The substrate 1 was selected because it contained one
aromatic ring, a suitable chromophore for UV detection
in HPLC, and reacted by a single nucleophilic substitu-
tion mechanism at a convenient rate. Preliminary
conductimetric measurements in 95% (v/v) acetone–
media. Selectivities for p-nitrobenzoyl chloroformate
in methanol–water also increase as water is added, but in
ethanol–water there is a decrease from 5.2 in 40% to 4.7
19
25
in 10% ethanol–water. Solvolyses of 1 (Table 2) and
18
(PhO) P(=O)Cl confirm that S values may decrease
2
À4
water at 0°C gave a rate constant of (7.63 Æ 0.14) Â 10
slightly (ca 15%) in highly aqueous media, even when a
mechanistic change is unlikely.
À1
s , in only fair agreement with the published titrimetric
À4 À1 19
value of 9.9 Â 10 s . Solvolyses of phosphono-
To analyse the results (Tables 1 and 2) in more detail,
18,22–25
chloridates having OMe, OEt or OPh groups occur at
we follow previous work
and assume that the
very similar rates in 95% (v/v) acetone–water at 0°C
pseudo-first-order reactions occur by a combination of
four third-order processes in which one molecule of
solvent (water, w, or alcohol, a) acts as a nucleophile and
the other acts as a general base: see Eqn. (2), in which the
first letter of the subscript denotes the nucleophile and the
second letter denotes the general base:
19
(
within a factor of two), and in general are much less
sensitive to substituent effects than solvolyses of
1
9,21
corresponding carbonyl compounds.
an alkyl group by an O-alkyl group only results in a ca
5-fold rate decrease, whereas ethyl chloroformate
Even replacing
1
4
(
EtOCOCl) solvolyses about 10 times more slowly than
21
2
acetyl chloride.
kobs ˆ k ‰waterŠ ‡ ꢀk ‡ k †f‰alcoholŠ‰waterŠ
ww
aw
wa
Rate constants for solvolyses of the chlorophosphonate
(Table 1) are close to 10-fold faster than those for
2
‡
k ‰alcoholŠ g
ꢀ2†
aa
1
18
corresponding solvolyses of [(PhO) P(=O)Cl] over the
2
The third-order rate constant for hydrolysis, kww and kaa,
can be calculated from the rate constants in pure solvents,
and the other two rate constants can be obtained from the
full range of alcohol–water mixtures. Hence, there is no
evidence for significant differences in initial state or other
effects for solvolyses of the more hydrophobic diphenyl
product selectivities as described below.
18
ester in more aqueous solvents. The additional O-aryl
group in the phosphorochloridate [(PhO) P( O)Cl] is
2
When the kaa term (k [alcohol] ) is unimportant, S
aa
=
2
2
9,21
[
Eqn. (1)] simplifies to Eqn. (3), and when k [water] is
1
ww
the main factor lowering the reactivity.
23
negligible S is given by Eqn. (4):
Product selectivities for 1 (Table 2) are greater than but
very similar to (within a factor of two) those for
1
=S ˆ ꢀk =k †ꢀ‰alcoholŠ=‰waterŠ† ‡ kww=k
ꢀ3†
ꢀ4†
17
wa aw
aw
[
(PhO) P(=O)Cl]. For both substrates, S increases
2
about three fold from 99.8 to 90% alcohol–water and
only about the same amount from 90 to 40% alcohol–
water. Consequently, the main cause of variations in S
does not appear to be medium effects of the solvents,
because medium effects usually change more gradually
S ˆ ꢀk =k †ꢀ‰waterŠ=‰alcoholŠ† ‡ k =k
aw wa
aa wa
Two independent measurements of the ratios of the
3
with variations in solvent composition. The similar
trends for both substrates also confirm that S may
decrease slightly in more aqueous media (from ca 40%
alcohol–water towards water), even when there is no
Table 3. Ratios kaw/kwa derived from 1/S and S plots [Eqns !3)
and !4)] for solvolyses of ethyl phenylphosphonochloridate
!1) in alcohol±water mixtures
19
evidence for mechanistic changes.
Plot
Solvent range
Slope
k /k
aw wa
An explanation of the ca 10-fold increase in S from
a
9
9.8% alcohol to 10% alcohol–water solvents (Table 2)
1/S
30–80% EtOH
98–99.8% EtOH
40–80% MeOH
92–99.8% MeOH
1.61 Æ 0.07 0.62 Æ 0.03
0.60 Æ 0.04 0.60 Æ 0.04
0.47 Æ 0.01 2.13 Æ 0.05
1.24 Æ 0.06 1.24 Æ 0.06
b
S
cannot be achieved if the reactions simply involved
second-order reactions, in which either alcohol or water
attacked the substrate, because S would then be expected
c
1
S
/S
d
2
a
to be constant. The results can be explained by third-
Eqn. (3): slope = 1.606 Æ 0.070, intercept = 1.053 Æ 0.044,
order processes, in which one molecule of solvent acts as
a nucleophile and the other acts as a general base.
According to this theory, S should reach a maximum
r = 0.996.
b
Eqn. (4): slope = 0.601 Æ 0.042, intercept = 0.047 Æ 0.002,
r = 0.995.
c
Eqn. (3): slope = 0.474 Æ 0.012, intercept = 0.631 Æ 0.012,
(
plateau) value in highly aqueous media, when water
r = 0.999.
22
d
dominates as the general base; another factor such as
mechanistic change was previously thought to be
Eqn. (4): slope = 1.235 Æ 0.056, intercept = 0.173 Æ 0.006,
r = 0.997.
Copyright  2001 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2001; 14: 759–763

7
62
À1
Table 4. Calculated third-order rate constants !M dm s )
T. W. BENTLEY AND D. N. EBDON
À2
6
CONCLUSIONS
for solvolyses of ethyl phenylphosphonochloridate !1) in
alcohol±water mixtures at 0°C
For solvolyses of phosphorochloridates and phosphono-
chloridates, values of S [Eqn. (1)] decrease slightly (ca
Ethanol–water
Methanol–water
15%) in highly aqueous media (see Ref. 18 and Table 2),
Rate
1/S,
Eqn. (3)
S,
1/S,
S,
for reasons unlikely to be due to mechanistic changes.
Consequently, the decrease in S observed for solvolyses
of electron-rich sulfonyl chlorides is not necessarily
associated with mechanistic changes (as originally
constant
Eqn. (4)
Eqn. (3) Eqn. (4)
a
À4
À4
À4
À4
À5
À5
À6
kww
1.4 Â 10
2.1 Â 10
1.3 Â 10
3.0 Â 10
1.4 Â 10
6.3 Â 10
3.8 Â 10
3.0 Â 10
1.4 Â 10 1.4 Â 10
À4
À4
À6
À5
À5
À6
À4
kwa
1.1 Â 10 5.4 Â 10
À4
kaw
2.2 Â 10 6.6 Â 10
10
b
À6
proposed ). However, electron-rich sulfonyl chlorides
kaa
9.3 Â 10 9.3 Â 10
show larger (ca 50%) changes in S than the phosphorus
a
2
From kobs/[H O] .
substrates, and the solvent composition for a maximum in
2
b
2
1
0
From kobs/[ROH] .
S depends on the structure of the sulfonyl chloride. If
there are two reaction channels, both are expected to be
1
6,17
nucleophilically assisted
and they may respond
1
4
similarly to changes in solvent nucleophilicity.
third-order rate constants k /kwa can been obtained from
aw
the slopes of 1/S plots [Eqn. (3)] and S plots [Eqn. (4)].
The two measurements of the ratios of k /k show
EXPERIMENTAL
aw wa
excellent agreement for solvolyses in ethanol–water
(
k /k = 0.61 Æ 0.01) (see Table 3); this is the best
Materials. Ethyl phenylphosphonochloridate (1), pre-
pared from phenylphosphonic dichloride (Lancaster) and
ethanol (1.1 equiv.) in diethylether at 0°C in the presence
aw wa
agreement obtained so far for investigations of this
type.
18,23–25
The ratio for methanol–water mixtures
19
covers a larger range (k /kwa = 1.7 Æ 0.4), continuing a
aw
of triethylamine (1.1 equiv.), was distilled prior to use
19
general trend that there is better agreement of these ratios
in ethanol–water than in methanol–water (possibly the
(b.p. 82°C at 0.01 mmHg; lit. 103°C at 0.3 mmHg and
26
lit. 120°C at 2 mmHg), and was shown by HPLC
analysis of the ethanolysis product to contain <0.1% of
acid (3), 2.4% of diester (2) and 1% of the dichloride
starting material. Standard solutions for response calibra-
tion of the HPLC system were obtained by reaction of the
acid chloride (1) at 0°C for 10 half-lives with ethanol to
give the diester (2) or with acetonitrile–water to give the
acid (3); acid/ester peak areas were 0.48 Æ 0.03;
similarly, methanolysis of 1 gave a methyl ester with
acid/ester peak area = 0.51 Æ 0.01.
k /k ratios are slightly solvent dependent in methanol–
water mixtures ).
Absolute values for each of the third-order rate
constants, kww, k , k and k , are listed in Table 4;
aw wa
24
wa aw
aa
the third-order rate constants kwa and kaw calculated from
/S plots [Eqn. (3)] are approximately 2–3-fold larger
than those calculated from S plots [Eqn. (4)]. This
difference is very similar to that for (PhO) P( O)Cl
where the kaw and kwa values calculated from the 1/S and
1
=
2
18
S plots vary by 2–4-fold. Substitution of the values of
third-order rate constants (Table 4) into Eqn. (2) leads
erroneously high predictions for observed first-order rate
constants, suggesting that solvolyses of ethyl phenyl-
phosphonochloridate (1) may be affected by additional
solvation effects such as variations in ground-state
Solvents for kinetics were dried and distilled and
2
7
solvolysis media were prepared as described elsewhere.
The accurate water contents of the low-percentage water–
alcohol mixtures, determined by Karl Fischer titration
using a Mettler Toledo DL18 apparatus and Hydranal
titrant (2), were usually within Æ0.2% of the nominal
values given in Table 2 (except the 98% methanol–water
was actually 97.52%); the more exact compositions were
used for calculations.
Materials for chromatography were AR grade metha-
nol and acetic acid, distilled water and tetrabutylammo-
nium bromide (TBAB), (Lancaster).
20
18
stabilization in aqueous organic solvents.
Solvation effects might also explain the sinuosoidal
plots of logk vs YCl for solvolyses of (PhO) P( O)Cl
see Figs 1 and 2 of Ref. 18; plots for solvolyses of
=
2
[
20
ethyl phenylphosphonochloridate (1) are very similar ].
Consequently, a non-linear plot of logk vs YCl is not
‘
proof’ of a mechanistic change, and (as usual) evidence
should ideally be based on several criteria, e.g. (i) for
solvolyses of benzoyl chlorides and substituted deriva-
tives, breaks in rate–rate profiles such as plots of logk vs
Kinetic methods. Conductimetric measurements were
made by the rapid injection method, and calculations
were performed on a PC version of LSKIN.
23
Y , marked maxima in S and a clear dependence on
Cl
substituents strongly support the case for mechanistic
changes; (ii) in contrast, solvolyses of p-toluenesulfonyl
Analytical methods. Four sets of HPLC conditions were
used for product studies; separate methods were used for
solvolyses in aqueous ethanol and methanol. All chroma-
tography was performed on 10 ml aliquots using a
15 cm  1/4 in Spherisorb ODS 2 column, eluted at a
8
chloride do not show breaks in plots of logk vs Y , or
Cl
10
clear maxima in S, consistent with the absence of
mechanistic changes.
Copyright  2001 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2001; 14: 759–763

SOLVOLYSES OF ETHYL PHENYLPHOSPHONOCHLORIDATE
763
À1
flow-rate of 1 ml min , with UV detection at 265 nm.
7. Liu K-T, Chen H-J. J. Chem. Soc. Perkin Trans. 2 2000; 893–898.
8
9
. Bentley TW, Koo IS. J. Chem. Soc. Perkin Trans. 2 1989; 1385–
392.
. Koo IS, Bentley TW, Kang DH, Lee I. J. Chem. Soc. Perkin Trans.
For solvolysis mixtures containing up to 90% ethanol–
water, the absorbance range (AUFS) was 0.01, and the
eluent, 40% (v/v) methanol–water, contained 0.07 M
TBAB and 0.15% (v/v) triethylamine; corresponding
figures for solvolyses up to 90% methanol were 0.03 M
TBAB and 0.1% (v/v) triethylamine. Solvolyses in more
alcoholic media were analysed using 30% methanol–
1
2 1991; 175–179.
1
1
0. Koo IS, Bentley TW, Llewellyn G, Yang K. J. Chem. Soc. Perkin
Trans. 2 1991; 1175–1179.
1. Bentley TW, Koo IS, Llewellyn G, Norman SJ. Croat. Chem. Acta
1
992; 65: 575–583.
12. Gordon IM, Maskill H, Ruasse M-F. Chem. Soc. Rev. 1989; 18:
23–151.
20
1
water as eluent with AUFS = 0.05. The HPLC equip-
ment was as described earlier.
18
13. Bonifaci R, Forbes RM, Henderson R, Maskill H, Palou J,
Richardson HL, Votsmeier M. Gazz. Chim. Ital. 1995; 125: 569–
5
75.
4. Kevill DN, D’Souza, M. Collect. Czech. Chem. Commun. 1999;
4: 1790–1796.
1
1
6
Acknowledgement
5. Lee I, Kim CK, Li HG, Sohn CK, Kim CK, Lee HW, Lee B-S. J.
Am. Chem. Soc. 2000; 122: 11162–11172.
This work was supported by EPSRC, through a student-
ship (to D.N.E.) and equipment grants for HPLC and Karl
Fischer equipment.
16. Forbes, RM, Maskill, H. J. Chem. Soc., Chem. Commun. 1991;
54–855.
7. Thea S, Carpanelli C, Cevasco G. Eur. J. Org. Chem. 2001; 151–
54.
8
1
1
1
8. Bentley TW, Ebdon D, Llewellyn G, Abduljaber MH, Miller B,
Kevill DN. J. Chem. Soc. Dalton Trans. 2 1997; 3819–3825.
9. Hudson RF, Keay L. J. Chem. Soc. 1960; 1859–1864.
1
2
2
2
2
REFERENCES
0. Ebdon, DN. PhD Thesis, University of Wales, 1999.
1. Lucken EAC, Whitehead, MA. J. Chem. Soc. 1961; 2459–2463.
2. Bentley TW, Harris HC. J. Org. Chem. 1988; 53: 724–728.
3. Bentley TW, Jones RO. J. Chem. Soc., Perkin Trans. 2 1993;
1
2
3
. Bentley TW, Schleyer PvR. Adv. Phys. Org. Chem. 1977; 14: 1–
7.
. Ta-Schma R, Rappoport Z. Adv. Phys. Org. Chem. 1992; 27: 238–
91.
. Bentley TW, Llewellyn G. Prog. Phys. Chem. Soc. 1990; 17: 121–
58.
6
2
351–2357.
4. Bentley TW, Jones RO, Koo IS, J. Chem. Soc. Perkin Trans. 2
994; 753–759.
2
2
2
1
1
5. Koo IS, Yang K, Kang K, Lee I, Bentley TW. J. Chem. Soc.,
Perkin Trans. 2 1998; 1179–1183.
4
. Song BD, Jencks WP. J. Am. Chem. Soc. 1989; 111: 8470–8479.
. Bentley TW, Harris HC, Koo IS. J. Chem. Soc., Perkin Trans. 2
5
1
988; 783–789.
. Kevill DN, Rudolph TM, D’Souza MJ. J. Phys. Org. Chem. 2000;
3: 192–196.
26. Ramaswami D, Kirch ER. J. Am. Chem. Soc. 1953; 75: 1763–1764.
27. Bentley TW, Carter, GE, Roberts K. J. Org. Chem. 1984; 49:
5183–5189.
6
1
Copyright  2001 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2001; 14: 759–763