A kinetic model for water reactivity (avoiding activities) for hydrolyses in aqueous mixtures - Selectivities for solvolyses of 4-substituted benzyl derivatives in alcohol-water mixtures

Article abstract of DOI:10.1002/poc.1308

For solvolyses of various benzyl substrates in ethanol-water (EW) and methanol-water (MW) mixtures, product selectivities (S) are reported for chlorides at 75°C defined as follows using molar concentrations: S= ([ether product]/[alcohol product]) × ([water]/[alcohol solvent]). The results support earlier evidence that solvolyses of 4-nitrobenzyl substrates are S _N2 processes, which are not susceptible to mechanistic changes over the whole range of solvents from water to alcohol. S values at 25 and/or 45°C in EW and MW, and additional kinetic data including kinetic solvent isotope effects (KSIE) are reported for solvolyses of 4-nitrobenzyl mesylate and tosylate. A kinetic model, explaining both rates and product, is proposed; a general medium effect due to solvent polarity is combined in one parameter with solvent effects on the nucleophilicity of the water and alcohol molecules acting as nucleophiles in S_N2 reactions. According to this model, as alcohol is added to water the rate of reaction decreases due to a decrease in solvent polarity, but the nucleophilicity of water increases relative to alcohol. The availability of experimental rate and product data over the whole range of solvent compositions from alcohol to water, reveals limitations of alternative approaches using activities. Copyright

Full text of DOI:10.1002/poc.1308

Research Article
Received: 20 September 2007,
Revised: 31 October 2007,
Accepted: 9 November 2007,
Published online in Wiley InterScience: 15 January 2008
(www.interscience.wiley.com) DOI 10.1002/poc.1308
A kinetic model for water reactivity (avoiding
activities) for hydrolyses in aqueous mixtures
– selectivities for solvolyses of 4-substituted
benzyl derivatives in alcohol–water mixtures
a
b
b
a
*
T. William Bentley , In Sun Koo , Hojune Choi and Gareth Llewellyn
For solvolyses of various benzyl substrates in ethanol–water (EW) and methanol–water (MW) mixtures, product
selectivities (S) are reported for chlorides at 75 -C defined as follows using molar concentrations: S ¼ ([ether product]/
[alcohol product]) T ([water]/[alcohol solvent]). The results support earlier evidence that solvolyses of 4-nitrobenzyl
substrates are SN2 processes, which are not susceptible to mechanistic changes over the whole range of solvents from
water to alcohol. S values at 25 and/or 45 -C in EW and MW, and additional kinetic data including kinetic solvent
isotope effects (KSIE) are reported for solvolyses of 4-nitrobenzyl mesylate and tosylate. A kinetic model, explaining
both rates and product, is proposed; a general medium effect due to solvent polarity is combined in one parameter
with solvent effects on the nucleophilicity of the water and alcohol molecules acting as nucleophiles in SN2 reactions.
According to this model, as alcohol is added to water the rate of reaction decreases due to a decrease in solvent
polarity, but the nucleophilicity of water increases relative to alcohol. The availability of experimental rate and
product data over the whole range of solvent compositions from alcohol to water, reveals limitations of alternative
approaches using activities. Copyright ß 2008 John Wiley & Sons, Ltd.
Keywords: kinetics; solvolysis; solvent effects; alcohols; water activity; solvent nucleophilicity; rate-product correlation
Our approach is based on the well-established, qualitatively
INTRODUCTION
predictive concepts of Ingold;[10] when a neutral nucleophile such
as water attacks a neutral substrate in an SN2 reaction, charge
develops in the transition state and the rate of reaction increases
if the ionising power (polarity) of the solvent (Y, or YX where X is
the leaving group) increases. The rates of typical solvolytic
reactions, can be explained quantitatively by predictable changes
in substrate sensitivities to Y and solvent nucleophilicity (N). SN1
reactions, usually have a high sensitivity to Y (or YX) and are
insensitive to N. As nucleophilic solvent assistance increases (i.e.
substrate sensitivities to N increases), the sensitivity to Y (or YX)
decreases because the positive charge is more delocalised (e.g.
by changing from a hindered secondary alkyl gradually to
methyl,[11] or by gradually increasing electron withdrawal in
substituted benzyl substrates[12,13]). Similarly, we have related
rates of hydrolysis of sulphonate esters in aqueous sulphuric acid
quantitatively to YOTs and NOTs, in an alternative to the approach
based on activities and acidity functions.[14]
The factors influencing the reactivity of water in aqueous solution
are vital for an understanding of many chemical and biochemical
processes. Environmentally significant reactions occur in dilute
aqueous solution, and biochemically important processes occur
in the cytoplasm (a relatively concentrated aqueous solution
containing protein and other solutes). The thermodynamic
activity of water is often quoted as a key factor influencing the
reactivity of water in biology and in food chemistry,[1] and is also
included in a thermodynamic analysis of kinetic data for
hydrolyses.[2]
Using transition state theory, rate constants can be related to
the activities of initial states (e.g. for hydrolyses of any substrate,
the activities of the reagent water and the substrate) and the
transition state.[3] Electrolytes in water often give activity
coefficients (g ¼ activity/concentration) less than unity, and
these may be considered as reductions in effective concen-
trations. Solute activity coefficients much greater than unity are
observed for water as solvent (e.g. for alkanes in water, g
exceeds 10[8] for decane!);[4] large g values can be calculated from
solvent effects,[4] and may be interpreted as an indication of
destabilised states,[3] but there is a conflict with the convenient
picture that activity refers to effective concentration.
In typical studies of rates of hydrolyses, reactions in aqueous
solutions containing added electrolyes and/or aprotic solvents
*
Department of Chemistry, Swansea University, Singleton Park, Swansea SA2
8PP, Wales, UK.
E-mail: cmsolvol@swansea.ac.uk
Alternative approaches include structural studies,[5] and
spectroscopic investigations,[6] which have the potential to
provide an understanding of the reactivity of aqueous solutions.
Addition of solutes may ‘make’ or ‘break’ water structure,[7] and/
or lead to a change in the number of free lone pairs or free OH
groups.[8] Some of these ideas have been applied to solvolyses,[8]
including hydrolyses in competition with alcoholyses.[9]
a
T. W. Bentley, G. Llewellyn
Department of Chemistry, Swansea University, Singleton Park, Swansea SA2
8PP, Wales, UK
b
I. S. Koo, H. Choi
Department of Chemistry Education and Research Institute of Natural Science,
Gyeongsang National University, Jinju, 660-701, Korea
J. Phys. Org. Chem. 2008, 21 251–256
Copyright ß 2008 John Wiley & Sons, Ltd.

T. W. BENTLEY ET AL.
lead to a single product. If an alcohol is present in an aqueous
solution, two products are usually formed and rate-product
correlations can be investigated (e.g. for solvolyses of acid
chlorides[15–18]). We now extend our work to classical SN2
reactions, and provide new equations linking observed pseudo-
first order rate constants and product compositions.
Table 1. Product selectivities (S, Eqn (2)) for solvolyses of 4-Z
substituted benzyl chlorides (1, X ¼Cl) in alcohol–water
mixtures at 75 8Ca
Substituent Z
Initially we investigated solvolyses of 4-substituted benzyl
chlorides to check for the possibility of mechanistic changes
arising from a change in solvent from alcohol to water. Solvolyses
of 4-nitrobenzyl substrates were then selected for further study,
with a change in leaving group to tosylate (increasing reactivity)
and mesylate (increasing reactivity whilst maintaining solubility
in highly aqueous systems). Another important feature of this
work is that the whole range of alcohol–water mixtures is
examined.
Alcohol (% v/v)
NO2
Cl
H
Me
OMe
Ethanol
90
80
60
40
20c
10c
5c
Methanol
90
80
60
40
20c
10c
5c
1.3
1.7
2.4
3.2
1.8b
2.4b
2.9b
3.5b
4.0b
2.2
2.6
2.9
3.3
3.1
2.9
2.6
1.7
1.9
1.9
1.8
1.6
1.5
1.6
1.7
1.9
2.6
3.4
3.6
3.9 Æ 0.2
4.1d
4.4d
4.5d
4.0 Æ 0.4
RESULTS
2.3
2.6
3.5
4.1
4.4
4.7
4.8
2.5
2.9
3.4
3.6
3.5
3.0
3.1
3.2
3.5
3.6
4.0
4.1
3.7
3.7
2.5
2.5
2.4
2.3
2.2
2.1
2.2
3.2
3.5
4.2
5.3
6.2
6.3
6.4
If a neutral substrate (e.g. ArCH2X) reacts with a mixture of alcohol
and water solvent (e.g. ROH/H2O – refer to Eqn (1)), two organic
products are formed (ArCH2OR and ArCH2OH). Unlike results from
many studies involving binary mixtures of water and an aprotic
solvent (e.g. acetonitrile), additional information is then available
from the selectivity (S, Eqn (2)), in which square brackets refer
to molar concentrations (mol dmÀ3).[9] Numerical values of S
(Eqn (2)) would be unaffected by expressing the ratio of solvent
concentrations as mole fractions, but would be altered if solvent
activities were incorporated.
a Determined by HPLC from the ratio of product areas; typical
error in S (Æ0.1), although the [ether]/[alcohol] response was
assumed to be 1.0 (Æ5%, based on experimental data); typi-
cally 25 ml of a 10% or 20% solution of the chloride in
acetonitrile was injected into solvolysis solvent (5 ml), so the
solvent also contained 0.5% acetonitrile.
ArCH2X þ ROH=H2O ¼ ArCH2OR=ArCH2OH þ HX
S ¼ ð½ArCH2ORŠ=½ArCH2OHŠÞ  ð½H2OŠ=½ROHŠÞ
(1)
(2)
b The ethyl ether product was not separated sufficiently by
HPLC from the chloride starting material, so reactions were
carried out for 10 half-lives; our data agree within 0.08–0.11
with published values for 90%, 80% and 60% ethanol–water
(Reference [21]).
Selectivites (S) for solvolyses of benzyl chloride and four
4-substituted derivatives (1) are given in Table 1. The main aim of
the experiments was to determine accurately the yields of
product of kinetically controlled reactions in homogeneous
solutions. Product compositions are usually examined after
reacting the substrate at the specified temperature for about 10
half-lives, but shorter reaction times (1–4 half-lives) were used to
obtain S values (Eqn (2)) for solvolyses of derivatives of 1 in
ethanol–water (EW) and methanol–water (MW) at 75 8C (Table 1).
Addition of 2,6-lutidine to remove the HCl by-product did not
significantly affect product compositions, but products were not
always formed by kinetic control; oxidation of alcohol products to
acids occurred, perhaps because chloride ion may be oxidised to
chlorine (Deacon process[19]), followed by the oxidation of
alcohols. Oxidation was minimised by reducing reaction times to
less than one half-life and/or by heating the solvents in a stream
of nitrogen immediately prior to product studies. Reactions of
4-methoxybenzyl chloride at 75 8C are rapid in comparison with
mixing times (ca. 1 s), but almost identical values were obtained
at lower temperatures.[20] New data (Table 1) confirm trends
observed previously for solvolyses of benzyl chlorides (1, Z ¼Cl,
H, Me and OMe) in 50–95% EW.[21]
c Multiple injections (5 Â 20 ml) of a 2% solution of the chloride
in acetonitrile were made, so the solvent also contained 2%
acetonitrile.
d Similar values were observed at 50 8C (20%, S ¼ 4.5; 10%, 4.5;
5%, 4.6), and for 20% ethanol at 25 8C (S ¼ 4.4, Reference [20]).
Experimental difficulties were reduced by studying
4-nitrobenzyl tosylate (2, R ¼ 4-tolyl, Table 2). For highly aqueous
mixtures, some data were obtained for the more soluble mesylate
(2, R ¼ Me) to confirm that products were formed from reactions
in homogeneous solutions. Much of the required kinetic data has
already been published,[22] and new kinetic data including kinetic
solvent isotope effects (KSIE) are in Table 3.
DISCUSSION
Choice of substrate
The nature of intermediates or transition states from solvolyses of
benzyl substrates depends on the initial substrate and on the
solvent.[12,13,21–27] For solvolyses of 1, Z ¼OMe, the increase in S,
as water is added to alcohol can be explained by SN1 reactions
with product formation from a carbocation intermediate.[28–30]
Although S values for 1, Z ¼ NO2 are similar to those for 1,
Z ¼OMe, the mechanism for the former is SN2.[22,31] S values help
www.interscience.wiley.com/journal/poc
Copyright ß 2008 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2008, 21 251–256

A KINETIC MODEL FOR WATER REACTIVITY
usually an increase in S as water is added to alcohol, but the small
decrease in S in highly aqueous methanol may be due to a
change in mechanism. For 1, Z ¼ NO2 there is a monotonic increase
in S as water is added to alcohol, and hence no indication of a
mechanistic change as the solvent composition is varied.
Table 2. Product selectivities (S, Eqn (2)) for solvolyses of
4-nitrobenzyl tosylate (2, R ¼ 4-tolyl) in alcohol–watera
Ethanol
Methanol
45 8C
S values for 4-nitrobenzyl sulphonates (2) in MW depend only
slightly on temperature (at 25 and 45 8C, Table 2), and are similar
those for 4-nitrobenzyl chlorides (1, Z ¼ NO2) at 75 8C (Table 1). S
values for mesylates support the reliability of data for tosylates in
highly aqueous media, when product ratios could be influenced
by formation of non-homogeneous solutions. In previous work,
we have found solvolyses of sparingly soluble substrates where
rate constants and/or product selectivities may be anomalous,
but the solutions appear to be homogeneous.[32] One expla-
nation is that aggregates of a few molecules may dissociate
during the time-scale of the experiment. Data for 4-nitrobenzyl
tosylate (2, R ¼ 4-tolyl) were selected for further detailed
interpretation (see below).
Alcohol (% v/v)
45 8C
25 8C
97
95
93
90
85
80
70
60
50
40
30
20
10
5
0.97 (0.89b)
0.85 (0.95b)
1.23
1.74
1.93
2.12
2.3
2.6
2.8
3.1
3.1
3.7
3.8
3.8
3.8
3.8
3.7
1.51
1.65
1.88
2.1
2.3
2.5
2.8
3.1
1.12
1.40
2.1
3.3
2.7 (2.7b)
2.8 (2.7b)
2.9 (2.7b)
2.8 (2.7b)
3.1 (2.8b)
3.5 (3.6c)
3.6 (3.6c)
3.7 (3.7c)
3.7 (3.7c)
3.7 (3.7c)
Rate-product correlations[33]
For second order reactions such as SN2 processes, the increase in
S from alcohol to water has been explained by an increase in the
nucleophilicity of alcohol relative to water.[21] New equations are
given below to investigate the solvent dependence of the
nucleophilicity of both alcohol and water in alcohol–water
mixtures.
Equation (2) may be derived on the assumption that S is the
ratio of two second order rate constants:[9] S ¼ ka/kw, where ka
refers to the second order rate constant for formation of
ArCH2OR and kw refers to the second order rate constant for
formation of ArCH2OH. For any alcohol–water mixture, we
assume that the pseudo-first order rate constants can be
calculated using Eqn (3) (so there are independent competing
second order hydrolysis and alcoholysis reactions). Values of ka in
alcohol, and kw in water are then derived from Eqns (4) and (5),
respectively, using experimentally observed rate constants (kobs).
a Determined by HPLC from the ratio of product areas, and a
molar response factor [ether]/[alcohol] of 1.00; typical errors
Æ5%.
b Selectivity values for 4-nitrobenzyl mesylate (2, R ¼ Me) at
45 8C in ethanol–water.
c Selectivity values for 4-nitrobenzyl mesylate (2, R ¼ Me) at
45 8C in methanol–water.
to identify mechanistic changes due to changes in product-
determining steps.[21] Differences from solvolyses of 1, Z ¼ NO2
and OMe are apparent for solvolyses of 1, Z ¼ Me, which shows
almost no change in S as the solvent composition is varied and is
difficult to classify mechanistically.[26] Variations in S are also
relatively small for solvolyses of 1, Z ¼ H. For 1, Z ¼Cl, there is
kcalc ¼ kw½waterŠ þ ka½alcoholŠ
kobs
In pure alcohol: ka ¼
½alcoholŠ
(3)
(4)
Table 3. Additional kinetic data for solvolyses of 4-nitrobenzyl
tosylate (2, R ¼ 4-tolyl) and mesylate (2, R ¼ Me) in methanol–
water at 45 8Ca
kobs
In pure water: kw
¼
(5)
½waterŠ
Rate constant (k/10À4sÀ1
)
Solvolyses in highly aqueous media reach a plateau (Smax
)
between 30% and 5% alcohol water, and it is assumed that the
value of ka in water containing only a trace of alcohol can be
obtained from Eqn (6); in other words, as S remains relatively
constant from 30% to 5% alcohol, we assume that there will be no
sudden changes in this trend between 5% alcohol and water.
Similarly a value for kw in alcohol can be obtained from Eqn (7),
where Smin is an estimate of the value of S in alcohol containing
only a trace of water; in this case a short extrapolation from data
in 95% and 97% alcohol is required. Equations (4–7) each provide
one piece of experimental data (a total of two for water and two
for each alcohol), from which the rate constants and product
compositions in binary alcohol–water mixtures will be calculated
as described below.
Solvent (% v/v)
% Methanol
Tosylate
Mesylate
kOTs/kOMs
100
40
30
20
10
5
Water
0.409 Æ 0.006b 0.204 Æ 0.011c
1.99
2.06
2.37
2.25
2.13
2.09
5.18
2.51 Æ 0.01
3.12 Æ 0.02
4.30 Æ 0.02
5.53 Æ 0.09
6.07 Æ 0.02
6.59 Æ 0.04e
7.38 Æ 0.14
9.68 Æ 0.24
11.8 Æ 0.6
12.7 Æ 0.2
13.8d
a Determined conductimetrically in duplicate; errors shown are
average deviations.
b In MeOD, 104k ¼ 0.321 Æ 0.007 so the KSIE is 1.27 Æ 0.05.
c In MeOD, 104k ¼ 0.169 Æ 0.003 so the KSIE is 1.21 Æ 0.08.
d Estimated assuming a OTs/OMs ratio of 2.1 (Reference [34]).
e In D2O, 104k ¼ 5.61 Æ 0.02, so the KSIE is 1.17 Æ 0.02.
In water: ka ¼ kwðSmax
Þ
(6)
(7)
ka
In alcohol: kw
¼
Smin
J. Phys. Org. Chem. 2008, 21 251–256
Copyright ß 2008 John Wiley & Sons, Ltd.
www.interscience.wiley.com/journal/poc

T. W. BENTLEY ET AL.
The calculations based on Eqns (4–7) show that values of ka
and kw are solvent-dependent (e.g. values of ka in water are over
60-times greater than the value in ethanol (Table 4)). The
following procedure for interpolating values of ka and kw in pure
solvents to values in mixed solvents was devised. Logarithms of
observed first order rate constants for solvolyses of 4-nitrobenzyl
tosylate (2, R ¼ 4-tolyl) correlate well with YOTs, a measure of
solvent ionising power based on solvolyses of 1- and 2-adamantyl
tosylates[34] (Eqn (8), Fig. 1).
Figure 1. Correlation of logarithms of first order rate constants for
solvolyses in methanol–water of 4-nitrobenzyl tosylate (2, R ¼ 4-tolyl)
at 45 8C (slope: 0.304 Æ 0.008, intercept: À4.05 Æ 0.02, r ¼ 0.998, n ¼ 9),
contrasting with those for 4-nitrobenzenesulfonyl chloride (3) at 25 8C and
4-nitrobenzoyl chloride (4) at 25 8C) with YOTs; kinetic data from Table 3
and References [16,17,22]; YOTs values from Reference [34] (for this range
of solvents YOTs and YCl behave similarly – refer to Reference [36]).
Assuming from the results for first order rate constants (Fig. 1)
that the second order rate constants for solvolyses of (2,
R ¼ 4-tolyl) also correlate with YOTs, values of ka and kw in any
solvent can be calculated by interpolation by adapting Eqn (8).
Calculated slopes (ma and mw) are shown in Table 4.
log k ¼ mYOTs þ c
(8)
(9)
ꢀ
ꢁ
in Y and N. One parameter may be adequate for these two binary
mixtures, because changes in N correlate with changes in Y.[35]
The variation in m from 0.210 to 0.345 (Table 4) may be explained
as follows. When water is added to alcohol: (i) the increase in
solvent polarity causes an increase in rates (a well-established
effect[10,11]), (ii) although selectivities (S, Eqn (2)) increase, the
nucleophilicities of the incoming nucleophiles (water and
alcohol) decrease, (iii) the nucleophilicity of water decreases
ka½alcoholŠ
ðmol % etherÞcalc
¼
 100
kcalc
Values of ka and kw, calculated from Eqn (8) can be substituted
into Eqn (3) to obtain calculated values (kcalc) of first order rate
constants, and the mol% ether can be calculated from Eqn (9).
Small but systematic deviations are seen; calculated rate
constants are all slightly too high and the calculated mol%
ether are all slightly too low (Table 5). Calculated rate constants
for MW mixtures (Table 5) are in good agreement with
experimental data, whereas calculated rate constants for EW
mixtures are up to 1.5 fold too high. Considering the simplicity of
the theory, and the absence of adjustable parameters (only two
fixed values of k and of S are required), the results are satisfactory.
According to the above theory, one slope parameter (m)
indicates the responses of competing SN2 solvolyses to changes
Table 5. Calculated and observed rate constants (k) and
mol% ether for solvolyses of 4-nitrobenzyl tosylate (2, R ¼
4-tolyl) in alcohol–water at 45 8C
k/(10À4sÀ1
)
Mol% ether
Calcb Obsdc Calcd Obsde
a
% Alcohol (v/v)
YOTs
Methanol–water
80
60
50
40
30
Table 4. Calculations of second order rate constants (MÀ1sÀ1),
and slope parameter (m) for rate-product correlations for
solvolyses of 4-nitrobenzyl tosylate (2, R ¼ 4-tolyl) in alco-
hol–water at 45 8C
0.47
1.52
2.00
2.43
2.97
3.39
1.32
2.92
4.10
5.43
7.72
9.82
1.26
2.72
3.91
5.18
7.38
9.68
77
60
52
44
36
26
82
67
60
51
41
29
20
Ethanol–water
Alcohol
parameter
Ethanol
Methanol
Equation
90
80
60
40
30
À0.77
0.00
0.92
1.97
2.84
0.50
0.94
1.98
4.04
6.72
0.48
0.73
1.32
2.90
4.81
74
60
40
27
22
76
63
50
36
27
ka (in alcohol)a
kw (in water)b
kw (in alcohol)
ka (in water)
1.06 Â 10À6
2.49 Â 10À5
1.32 Â 10À6
6.97 Â 10À5
0.210
1.70 Â 10À6
2.49 Â 10À5
1.21 Â 10À6
9.20 Â 10À5
0.262
4
5
6
7
8
8
mw (slope for kw)c
ma (slope for ka)c
a From Reference [34].
b Equation (3).
0.300
0.345
a Observed first order rate constant from Reference [22].
b Observed first order rate constant from Table 3.
c YOTs data from Reference [34]; refer to Table 5.
c Kinetic data from Table 3 and Reference [22].
d Equation (9).
e From the observed ether/alcohol product ratios.
www.interscience.wiley.com/journal/poc
Copyright ß 2008 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2008, 21 251–256

A KINETIC MODEL FOR WATER REACTIVITY
more sharply than that of alcohol, possibly because additional
hydrogen bonds reduce the availability of free lone pairs to act as
nucleophiles.[6] In addition, the positive charge in the attacking
solvent nucleophile may be more dispersed by greater hydrogen
bonding when water attacks than when alcohol attacks, so values
of mw may be lower than for ma due to different responses to YOTs
(Table 4).
Figure 1 also shows a clear difference between the linear plot
for the SN2 (second order) solvolyses of 2, R ¼ 4-tolyl with a low
KSIE of 1.2 (Table 3, footnotes b, c and e), and shallow curved plots
for solvolyses of 3 and 4, for which the KSIE is relatively large
(>2.0) and rate-product correlations can be achieved assuming
that competing solvolyses are third order.[16,17] It is preferable to
choose a Y value corresponding to the leaving group to account
for electrophilic solvation effects.[34] However, differences in
shapes between the three plots in Fig. 1 are not due to the
selection of YOTs rather than YCl, because these two scales
correlate linearly in EW mixtures (solvents of similar electro-
philicities).[36]
The assignment of molecularity to pseudo-first order solvolytic
reactions has been controversial for decades, and was initially
concerned with the distinction between uni- and bimolecular
reactions.[37–39] As a large dynamic pool of solvent molecules
contributes to the observed solvent effect, molecularity is difficult
to assign. Nevertheless, third order solvolyses of acid chlorides
are well established from the rate equations for alcoholyses in a
large excess of relatively inert solvents such as ether and
acetonitrile.[40]
Consequently, we propose a simple picture for second and
third order hydrolyses (Scheme 1). Key differences are: (i) the
timing of the deprotonation of the solvent nucleophile, and the
high KSIE for third order reactions is explained by partial cleavage
of an O—H bond in the attacking nucleophile, (ii) the greater
extent of charge delocalisation of both positive and negative
charges of the transition states in the third order reactions – in
several cases,[15–18] rate-product correlations can be achieved by
ignoring the solvent dependence (m, Eqn (8)) of the third order
rate constants, (iii) for second order reactions, the charge is more
localised and an intermediate (ROHþ2 ) may be formed, which is
rapidly deprotonated.[41]
Table 6. Activity coefficients (g) for alcohol–water mixtures
Solvent
g (water)
g (alcohol)
T/8C
Water (MeOH)a
Water (MeOH)a
23.4% v/v MeOH
23.4% v/v MeOH
Water (EtOH)a
1.000b
1.000b
1.007
1.003
1.000b
1.10
1.77
1.63
1.60
1.53
3.9
35c
25c
35c
25c
25d
25
50.5% v/v EtOH
2.0
a Effect of alcohol solute in very dilute solution.
b By definition.
c Reference [48].
d Reference [49].
graphical data,[2] giving values between 0.95 and 1.1. Literature
data for water in alcohol–water mixtures also show relatively
small activity coefficients (g, Table 6). Consequently, replacement
of water concentration by water activity (e.g. as in Eqn (3)) would
have only a small effect, and the main effects of added organic
cosolvents are due to changes in water concentration, solvent
[10]
polarity
and solvent nucleophilicity.[11–13,42–44]
Activity coefficients for alcohols in water show much larger
variations (Table 6), and as alcohol is added to water S decreases
and galcohol decreases (g ¼ 1.00 in pure alcohol). However, the
order of S values (SMeOH > SEtOH > Swater) is not the order of g (or N
values[42–44]).
For SN1 solvolyses of 1-adamantyl substrates (1AdX) in 50–95%
v/v EW at 100 8C, S (¼ kw/ka, the inverse of our definition)
correlates linearly with the ratio of activity coefficients (gwater
/
gethanol) with slopes varying from 0.1 for X ¼ Br and 0.6 for
X ¼OTs.[45] Over the full range of compositions, the gwater/gethanol
ratio varies by over 10-fold at 50 8C,[46] whereas S values for 2 vary
only 3-fold (Table 2); a plot (not shown) of S values (Table 2)
versus gethanol/gwater is sharply curved, although the region
between 50% and 95% EW is approximately linear with a slope of
about 0.8. Values of gmethanol/gwater vary only 3.2-fold at 50 8C,[47]
and show a shallow curve when plotted against the 2.5-fold
range of S values for 2 at 45 8C. In earlier work,[21,45] product ratios
were obtained by GC for a restricted range of solvent
compositions. A wider and more informative solvent range can
nowadays be investigated using reverse phase HPLC, and the
more extensive data show that ratios of activity coefficients do
not adequately explain trends in S values.
CONCLUSIONS
Scheme 1. A simplified picture of some differences between second and
third order hydrolyses for RX (second order) and ArSO2Cl (third order)
Product selectivities (Table 1) support independent evidence that
solvolyses of benzyl chlorides are susceptible to mechanistic
changes over the range of reaction conditions from alcohol to
water. Solvolyses of 4-nitrobenzyl substrates in alcohol–water
mixtures are an exception, and react by classical SN2 reactions
throughout the solvent range.
Rate constants and products of solvolyses of 4-nitrobenzyl
tosylate in a full range of alcohol–water mixtures can be
explained quantitatively by competing second order reactions
based on Eqn (3). The new theory (Eqns (3–9)) leads to
rate-product correlations, which explain why S values (defined
Role of activities
The influence of solvent activity on the kinetics of a neutral ester
hydrolysis was considered in detail recently.[2] Using a water
activity meter, it was shown that added acetonitrile or THF
increases water activity, whereas PEG 400 decreases it. The
authors’ commented: ‘water activity hardly contributes to the
observed decrease in rate constant’.[2] Although activity
coefficients were not quoted, these can be calculated from the
J. Phys. Org. Chem. 2008, 21 251–256
Copyright ß 2008 John Wiley & Sons, Ltd.
www.interscience.wiley.com/journal/poc

T. W. BENTLEY ET AL.
by a ratio of second order rate constants, S ¼ ka/kw)[9] may be
solvent-dependent. Addition of alcohol to water decreases the
rate of reaction due to a decrease in solvent ionising power
(Ingold rules:[10] the SN2 transition state is more polar than
the initial state of two neutral species). S decreases because the
nucleophilicity of water increases relative to alcohol (possible due
to a decrease in hydrogen bonding). Hence, the decrease in S as
alcohol is added to water is explained using kinetic rather than
thermodynamic concepts (such as activities, which do not
account for selectivities when the full range of alcohol–water
mixtures is examined).
[4] M. H. Abraham, M. J. Blandamer, J. Am. Chem. Soc. 2002, 124,
7853–7856.
[5] G. W. Neilson, P. E. Mason, S. Ramos, D. Sullivan, Phil. Trans. R. Soc.
Lond. (A) 2001, 359, 1575–1591.
[6] M. C. R. Symons, Phil. Trans. R. Soc. Lond. (A) 2001, 359, 1631–
1646.
[7] F. Franks (Ed.), Water, a Comprehensive Treatise, Plenum Press, New
York, 1979.
[8] M. C. R. Symons, Acc. Chem. Res. 1981, 14, 179–187.
[9] R. Ta-Shma, Z. Rappoport, Adv. Phys. Org. Chem. 1992, 27, 239–291.
[10] C. K. Ingold, Structure and Mechanism in Organic Chemistry, Cornell
University Press, Ithaca, New York, 1953, 346–350.
[11] F. L. Schadt, T. W. Bentley, Pv. R. Schleyer, J. Am. Chem. Soc. 1976, 98,
7667–7674.
[12] D. N. Kevill, H. Ren, J. Org. Chem. 1989, 54, 5654–5655.
[13] D. N. Kevill, M. J. D’Souza, H. Ren, Can. J. Chem. 1998, 76, 751–757.
[14] T. W. Bentley, K. Roberts, J. Chem. Soc. Perkin Trans. 2 1987, 293–299.
[15] T. W. Bentley, H. C. Harris, Z. H. Ryu, G. T. Lim, D. D. Sung, S. R. Szajda,
J. Org. Chem. 2005, 70, 8963–8970.
For second order solvolyses (e.g. of 2), initial attack by the
a rapid
solvent nucleophile (Scheme 1) is followed by
deprotonation step, and the kinetic isotope effect (KSIE) is low
(1.2).[41] For third order solvolyses (e.g. of 3 and 4), a second
molecule of solvent could act as a general base at the same time
as the nucleophilic attack, so the KSIE is larger (2.0).[15–18]
Logarithms of first order rate constants for solvolyses of 2 in MW
correlate linearly with YOTs, but the plots (Fig. 1) for 3 and 4 are
shallow curves.
[16] T. W. Bentley, R. O. Jones, J. Chem. Soc. Perkin Trans. 2 1993,
2351–2357.
[17] T. W. Bentley, R. O. Jones, I. S. Koo, J. Chem. Soc. Perkin Trans. 2 1994,
753–759.
[18] I. S. Koo, K. Yang, K. Kang, I. Lee, T. W. Bentley, J. Chem. Soc. Perkin
Trans. 2 1998, 1179–1183.
[19] F. A. Cotton, G. Wilkinson, Advanced Inorganic Chemistry, 4th edn.
Wiley, New York, 1980, 545.
[20] T. W. Bentley, I. S. Koo, S. J. Norman, J. Org. Chem. 1991, 56,
1604–1609.
EXPERIMENTAL
Materials
[21] H. Aronovitch, A. Pross, J. Chem. Soc. Perkin Trans. 2 1978, 540–545.
[22] M. Fujio, T. Susuki, M. Goto, Y. Tsuji, K. Yatsugi, S. H. Kim, G. A.-W.
Ahmed, Y. Tsuno, Bull. Chem. Soc. Jpn. 1995, 68, 673–682.
[23] I. Lee, H. J. Koh, Y. S. Park, H. W. Lee, J. Chem. Soc. Perkin Trans. 2 1993,
1575–1582.
[24] I. M. Gordon, H. Maskill, J. Chem. Soc. Perkin Trans. 2 2001, 2059–2062.
[25] M. Fujio, T. Susuki, M. Goto, Y. Tsuji, K. Yatsugi, Y. Saeki, S. H. Kim, Y.
Tsuno, Bull. Chem. Soc. Jpn. 1994, 67, 2233–2243.
[26] A. Thibblin, J. Org. Chem. 1993, 58, 7427–7433.
[27] J. B. Kyong, B.-C. Park, C.-B. Kim, D. N. Kevill, J. Org. Chem. 2000, 65,
8051–8058.
Benzyl chlorides were obtained from Aldrich; and purity was
checked by HPLC analysis of methanolysis products; 1, X ¼Cl,
Z ¼ H and Cl were distilled under reduced pressure prior to use.
4-Nitrobenzyl sulphonates were prepared from 4-nitrobenzyl
alcohol (Aldrich) by standard methods.[31] Solvents for solvolyses
and chromatography were as described previously.[15–18]
Solvolyses
[28] A. Queen, Can. J. Chem. 1979, 57, 2646–2651.
[29] T. L. Amyes, J. P. Richard, J. Am. Chem. Soc. 1990, 112, 9507–9512.
[30] T. W. Bentley, Z. H. Ryu, J. Chem. Soc. Perkin Trans. 2 1994, 761–767.
[31] P. Dietze, W. P. Jencks, J. Am. Chem. Soc. 1989, 111, 5880–5886.
[32] T. W. Bentley, H. C. Harris, I. S. Koo, J. Chem. Soc. Perkin Trans. 2 1988,
783–789.
Data for chlorides were obtained from solvolyses at 75 8C in 5 ml
sealed ampoules. Rate and product data for sulphonates were
obtained by injecting a 1% stock solution (10 ml) into 5 ml of
rapidly stirred, thermostated solvent.
[33] For a recent discussion of rate-product correlations see T. W. Bentley,
In The Investigation of Organic Reactions and Their Mechanisms (Ed.: H.
Maskill ) Blackwell, Oxford, 2006, Chapter 2.
[34] T. W. Bentley, G. Llewellyn, Prog. Phys. Org. Chem. 1990, 17, 121–158.
[35] J. Kaspi, Z. Rappoport, Tetrahedron Lett. 1977, 2035–2038.
[36] T. W. Bentley, G. E. Carter, K. Roberts, J. Org. Chem. 1984, 49,
5183–5189.
[37] A. R. Olson, R. S. Halford, J. Am. Chem. Soc. 1937, 59, 2644–2647.
[38] P. D. Bartlett, J. Am. Chem. Soc. 1939, 61, 1630–1635.
[39] S. Winstein, J. Am. Chem. Soc. 1939, 61, 1635–1640.
[40] D. N. Kevill, C.-B. Kim, Bull. Chem. Soc. France 1988, 383–390.
[41] J. M. W. Scott, Can. J. Chem. 2005, 83, 1667–1719.
[42] D. N. Kevill, In Advances in Quantitative Structure-Property Relation-
ships (Ed.: M. Charton ) Jai Press, Greenwich, CT, 1996, 1, 81–115.
[43] S. Minegishi, S. Kobayashi, H. Mayr, J. Am. Chem. Soc. 2004, 126,
5174–5181.
[44] T. W. Bentley, M. S. Garley, J. Phys. Org. Chem. 2006, 19, 341–349.
[45] P. R. Luton, M. C. Whiting, J. Chem. Soc. Perkin Trans. 2 1979, 646–647.
[46] R. C. Pemberton, C. J. Mash, J. Chem. Thermodynamics 1978, 10,
867–888.
Chromatography
Products of solvolyses of chlorides were analysed using a 5 mm
Spherisorb ODS2 chromatography column (15 cm  1/400); typical
conditions were: eluent (70% MW), flow rate (1 ml minÀ1), UV
detection (l ¼ 266 nm, A ¼ 0.5). A more polar eluent (55–60%
methanol) was required to separate the products for the tosylate.
Acknowledgements
We are grateful to L. M. Howle (now Pollaud) for preliminary
investigations of the benzyl chlorides, N. Asaad, J. R. Conder and
D. N. Kevill for helpful discussions, and EPSRC for equipment
grants.
REFERENCES
[47] K. Kurihara, T. Minoura, K. Takeda, K. Kojima, J. Chem. Eng. Data 1995,
40, 679–684.
[48] S. D. Christian, E. H. Lane, E. E. Tucker, J. Soln. Chem. 1981, 10,
181–188.
[49] L. Nord, E. E. Tucker, S. D. Christian, J. Soln. Chem. 1984, 13, 849–867.
[1] M. J. Blandamer, J. B. F. N. Engberts, P. T. Gleeson, J. C. R. Reis, Chem.
Soc. Rev. 2005, 34, 440–458.
[2] T. Rispens, C. Cabaleiro-Lago, J. B. F. N. Engberts, Org. Biomol. Chem.
2005, 3, 597–602.
[3] W. P. Jencks, Catalysis in Chemistry and Enzymology, McGraw Hill,
New York, 1969, 599–605.
www.interscience.wiley.com/journal/poc
Copyright ß 2008 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2008, 21 251–256