Proton activating factors and keto-enol-zwitterion tautomerism of 2-, 3- And 4-phenylacetylpyridines

Article abstract of DOI:10.1039/a701546a

Equilibrium constants for keto-enol tautomerism of 2-, 3- and 4-phenylacetylpyridines in aqueous solution at 25°C have been measured as pK_T^E = 3.35, 4.2 and 3.1 respectively (K_T^E = [enol]/[ketone], PK_T^E = -log K_T^E). Corresponding values for the N-protonated ketones are 1.64, 2.80 and 1.54. These enol contents are consistently higher than those of the isomeric phenacylpyridines, except in the case of the (unprotonated) 2-isomer where the greater enol content of the latter (pK_T^E = 2.0) can be attributed to more effective stabilization by hydrogen-bonding to the pyridyl nitrogen in a six- than five-membered ring. The tautomeric constants were obtained by combining rate constants for enolisation, measured by halogen trapping, with rate constants for relaxation of the enol tautomer (generated by quenching the enolate anion into acid or acidic buffers) to its more stable keto isomer. Approximate tautomeric constants for zwitterion formation (pK_T^Z = 4.6, 7.4 and 6.1 for 2-, 3- and 4-isomers respectively) are inferred from measurements of ionisation constants and keto-enol tautomeric constants for N-methylated ketones by taking the N-methylated enolate anions as models for the zwitterions and correcting for the substituent effect of the methyl group. The tautomerism is discussed in terms of the relationship pK_T = ΔpK_ab + log PAF which dissects tautomeric constants into contributions from (a) a difference in pK_as of non-interacting acidic and basic tautomeric sites (ΔpK_ab) and (b) a mutual stabilisation of these sites from conjugative, inductive or hydrogen-bonded interactions between them. This stabilisation is described by a proton activating factor (PAF) measuring the effect of protonation at one tautomeric site upon the ionisation constant at the other.

Full text of DOI:10.1039/a701546a

View Article Online / Journal Homepage / Table of Contents for this issue
Proton activating factors and keto–enol-zwitterion tautomerism of 2-,
3- and 4-phenylacetylpyridines
Geraldine M. McCann, Rory A. More O’Ferrall* and Sinead M. Walsh
Department of Chemistry, University College, Belfield, Dublin 4, Ireland
Equilibrium constants for keto–enol tautomerism of 2-, 3- and 4-phenylacetylpyridines in aqueous
solution at 25 ЊC have been measured as pKTE = 3.35, 4.2 and 3.1 respectively (KTE = [enol]/[ketone],
pKTE = Ϫlog KTE). Corresponding values for the N-protonated ketones are 1.64, 2.80 and 1.54. These enol
contents are consistently higher than those of the isomeric phenacylpyridines, except in the case of the
(unprotonated) 2-isomer where the greater enol content of the latter (pKTE = 2.0) can be attributed to more
effective stabilization by hydrogen-bonding to the pyridyl nitrogen in a six- than five-membered ring. The
tautomeric constants were obtained by combining rate constants for enolisation, measured by halogen
trapping, with rate constants for relaxation of the enol tautomer (generated by quenching the enolate
anion into acid or acidic buffers) to its more stable keto isomer. Approximate tautomeric constants for
zwitterion formation (pKTZ = 4.6, 7.4 and 6.1 for 2-, 3- and 4-isomers respectively) are inferred from
measurements of ionisation constants and keto–enol tautomeric constants for N-methylated ketones by
taking the N-methylated enolate anions as models for the zwitterions and correcting for the substituent
effect of the methyl group. The tautomerism is discussed in terms of the relationship pKT = ∆pKab ϩ log
PAF which dissects tautomeric constants into contributions from (a) a difference in pKas of non-
interacting acidic and basic tautomeric sites (∆pKab) and (b) a mutual stabilisation of these sites from
conjugative, inductive or hydrogen-bonded interactions between them. This stabilisation is described by a
proton activating factor (PAF) measuring the effect of protonation at one tautomeric site upon the
ionisation constant at the other.
There have been many investigations of the tautomerism of
aldehydes and ketones in the past fifteen years,1–8 including a
number of studies of α-heterocyclic ketones.3–6 Less attention
has been given to ketones in which a heterocyclic substituent is
bound directly to the carbonyl group,7,8 however, and, as
examples of these, in this paper we report measurements of
keto–enol tautomeric constants for 2-, 3- and 4-phenylacetyl-
pyridines (PhCH2COPy). The results are of interest for com-
parison with previous measurements for the isomeric phen-
acylpyridines (PhCOCH2Py)3,4 and as a preliminary to studies
of the more complex tautomerism of pyridacyl pyridines
(PyCOCH2Py).9
they are zwitterions.3 The distinction between these structures
rests on the presence or absence of direct resonance, leading to
charge neutralisation, between the ionic centres created by the
proton shift. This resonance is strongly stabilising for the 2- and
4-phenacylpyridines, and these enaminones are more stable
than their respective enols, whereas for 3-phenacylpyridine and
the phenylacetylpyridines the enols are more stable than the
zwitterions. Resonance and, indeed, electrostatic or inductive
interactions between their charge centres are an important
influence upon the stabilisation of the tautomers, and in this
paper we describe the use of ‘proton activating factors’ to
measure this stabilisation.10
Unlike simple ketones N-heterocyclic ketones can tautomer-
ise through migration of hydrogen from a carbon to a nitrogen
atom as well as from carbon to oxygen. Shown below are all
three tautomers of 2-phenacylpyridine (1–3) and 2-phenyl-
acetylpyridine (4–6) together with the N-protonated tautomers
(only) of the corresponding 3- and 4-phenacyl- and phenyl-
acetyl-pyridines (7–10). For the 2- and 4-phenacylpyridines the
H
+
H
N
O–
N
O
Ph
Ph
8
7
H
N
+
N
H
+
N
N
H
N
H
Ph
O
Ph
O
Ph
O
Ph
Ph
O–
O–
10
1
2
3
9
A further influence on the stabilities of enol and enaminone
tautomers is hydrogen-bonding. The hydrogen-bonded enol of
2-phenacylpyridine (3) amounts to 1% of an equilibrium mix-
ture of tautomers in aqueous solution at 25 ЊC compared with
less than one part in 104 for 4-phenacylpyridine. A point of
interest in comparing phenacyl- and phenylacetyl-pyridines
therefore is to determine whether the enol of 2-phenylacetyl-
pyridine (5), in which the hydrogen bond occurs in a five- rather
N
Ph
N
Ph
N
Ph
O
O
H+
O–
H
4
5
6
N-protonated tautomers are enaminones (2 and 7) and for the
3-phenacylpyridine (8) and the phenylacetylpyridines (6, 9, 10)
J. Chem. Soc., Perkin Trans. 2, 1997
2761

View Article Online
than a six-membered ring, is comparably stabilised. It is
assumed that in these structures (and 3) the hydrogen-bonding
maintains the configuration of the exocyclic double bond in a
Z-configuration.11 Further consideration of E–Z isomerism of
enol or enaminone tautomers is postponed to a later paper.
The 3- and 4-phenylacetylpyridines were studied previously
by Bunting and Stefanidis,12–14 who measured rate and equi-
librium constants for ionisation of the ketones to their enolate
anions. The ketones are quite acidic (pKas 13.1 and 12.3 respect-
ively) and are appreciably ionised in dilute aqueous sodium
hydroxide. This suggested to us that the enol tautomers might
be generated by quenching solutions of their enolate ions in
excess acid or acidic buffer. Tautomeric constants could then be
measured by combining rate constants for relaxation of the
enol to their stable keto tautomers with the corresponding value
for enolisation of the ketone, measured in the usual way by
trapping the enol as it is formed from the ketone with iodine or
bromine.6 This indeed was the starting point for the present
investigation.
Fig. 1 Log k–pH profile for enolisation of 2-phenylacetylpyridine
(lower line) and ketonisation of its enol (upper line)
Results
respectively. Values of ko from measurements in solutions of
sodium hydroxide or hydrochloric acid are given in Table 3.
Measurements for borate buffers, which showed practically no
buffer catalysis at low buffer concentrations, are also included
in Table 3.
The equilibrium constant for keto–enol tautomerism KT may
be obtained from a ratio of rate constants for enolisation and
ketonisation of 4 measured under the same conditions at a pH
where both enol and ketone (rather than their conjugate acids
or bases) are reactants. The most complete set of such meas-
urements relates to lutidine and acetate buffers. For the lutidine
buffers, in which catalysis by buffer base only is observed, KT
may be evaluated from eqn. (4) in which the superscripts E and
2-Phenylacetylpyridine
This compound (4) was not studied by Bunting and Stefanidis12
and so was investigated in more detail than its isomers. Spec-
trophotometric measurements of ionisation constants in acidic
and basic media gave pKa = 2.30 0.09 and pKa = 12.02 0.09
for N-protonation of the ketone and C-protonation of its eno-
late anion respectively. Stopped flow measurements of rates of
enolate ion formation in the range 0.01–0.12  hydroxide ion
gave pKa = 11.9 from the ratio of slope to intercept of a plot of
first-order rate constants against base concentration, in good
agreement with the directly measured value. Rate constants for
ketonisation were measured by quenching a basic solution of
the enolate anion into HCl or acidic buffers and monitoring
relaxation of the initially formed enol (5) to the thermo-
dynamically more stable keto tautomer by stopped-flow
spectrophotometry.
The measured first-order rate constants for tautomerisation
in buffer solutions (kobs) reflect contributions from second order
rate constants for general acid and/or general base catalysis
(kGA and kGB) by acidic (AH) and basic (AϪ) components
of a buffer and a first-order buffer independent rate constant
(ko) for reaction with H2O, Hϩ or OHϪ as shown in eqn. (1).9
Values of kGA and kGB were obtained from slopes (k) of
K
KT = kGBE/kGB
(4)
K refer to enolisation and ketonisation respectively. Substitut-
E
K
ing values of kGB and kGB from Tables 1 and 2 gives
KT = 5.3 × 10Ϫ4. This compares with KT = 4.0 × 10Ϫ4 based on
E
K
kGA and kGB for acetate buffers. Taking the average of these
values gives KT = 4.7 × 10Ϫ4 and pKT = 3.33.
Strictly speaking kGBK in the denominator of eqn. (4) should
be corrected for a contribution from the rate constant for
enolisation, since kGBK is measured for an equilibrium relaxation
and represents a sum of rate constants for ketonisation and
kobs = ko ϩ kGA[AH] ϩ kGB[AϪ]
(1)
enolisation. In practice kGB is too small for this correction to
E
be required. However, a further correction arises because KT is
really an ‘effective’ equilibrium constant reflecting the com-
bined formation of enol (EH) and zwitterion (ZH) tautomers
from the ketone (KH). This means that KT = ([EH] ϩ [ZH])/
[KH] = KTE ϩ KTZ, where KTE = [EH]/[KH] and KTZ = [ZH]/
[KH] are the tautomeric constants for formation of enol and
zwitterion alone. Evaluation of KTZ, which allows correction of
KT to KTE is described below.
The buffer independent rate constants for both enolisation
and ketonisation are conveniently displayed as pH-profiles by
plotting log ko against pH as in Fig. 1. In this Figure, the lower
plot represents enolisation and the upper plot (the faster) keton-
isation. Both reactions show acid- and base-catalysed pathways
at low and high pHs respectively. At a sufficiently low pH the
acid catalysis becomes saturated, as expected of N-protonation
of ketone and enol reactants. For enolisation the inflection in
the pH-profile is consistent with the measured pKa for N-
protonation of the ketone. At high pH base-catalysis of keton-
isation is also saturated, reflecting ionisation of the enol to its
enolate anion. The two profiles intersect at the pKa for the basic
ionisation of the keto tautomer (pH = 12.0).
plots of kobs against the concentration of the base compon-
ent of the buffer at different (constant) buffer ratios (R = [AH]/
[AϪ]) which, as shown in eqn. (2), have ko as intercept. Plotting
kobs = ko ϩ k[AϪ]
(2)
values of k against the buffer ratio then gave kGA and kGB as a
further slope and intercept (eqn. 3) based on measurements at
three or four buffer ratios.
k = kGB ϩ RkGA
(3)
Rate constants for the reverse enolisation reaction were
measured from reaction of the ketone with bromine and iod-
ine.6,15 These may also be expressed as in eqn. (1) and separated
into values of ko, kGA and kGB using eqns. (2) and (3). Occasion-
ally, for either ketonisation or enolisation, equations had to be
modified to take account of partial protonation of the reactant
or reversibility of the tautomerisation reaction.3,6
Values of k, ko, kGA and kGB for ketonisation and halogen-
ation of 2-phenylacetylpyridines in lutidine (2,6-dimethylpyr-
idine) and acetic acid buffers are listed in Table 1 and Table 2
The limiting pH-independent rate constants in acidic solu-
tions from the two pH-profiles represent enolisation (kH OE) and
2
2762
J. Chem. Soc., Perkin Trans. 2, 1997

View Article Online
Table 1 Slopes and intercepts of buffer plots and rate constants for general acid- and general base-catalysis for relaxation (ketonisation) of the
enols of 2-, 3- and 4-phenylacetylpyridines and the N-methyl-2-phenylacetylpyridinium ion in aqueous solution at 25 ЊC and ionic strength 0.1 
Substrate/buffer
pH
Ra
ko/sϪ1
kb/dm3 molϪ1 sϪ1
kGB/dm3 molϪ1 sϪ1 kGA/dm3 molϪ1 sϪ1
2-Phenylacetylpyridine
Acetate
3.58
4.04
4.54
5.18
6.66
7.27
12
0.038
0.030
—
—
0.12
—
9.4
9.0
6.8
5.8
19.1
18.6
5.0c
18.8
4.5c
4.2
1.3
0.33
1
Lutidine
0.25
N-Methyl-2-phenylacetylpyridinium ion
Acetate
4.60
5.08
5.51
6.37
6.85
7.33
7.83
9.07
1
0.33
0.125
3
1
0.33
1
—
0.05
—
0.016
—
—
7.32
7.13
5.74
7.10
4.03
3.45
1.77
4.32
7.8
Lutidine
9.96
N-Ethylmorpholinec
0.005
0.022
14.0d
34.3d
Boratec
1
3-Phenylacetylpyridine
Acetate
3.92
4.41
4.85
5.26
6.50
6.85
7.18
4.6
1.5
0.59
0.22
1.5
0.66
0.25
—
—
—
(0.07)
—
1.0
6.0
18.8
15.5
13.8
13.4
184
176
186
13.0c
2.44c
Lutidine
182
4-Phenylacetylpyridine
Cyanoacetate
1.87
2.36
2.70
2.97
3.43
3.82
3.93
4.41
4.84
5.27
6.50
6.85
7.17
2.5
0.0090
0.0074
0.0076
0.0065
—
—
—
—
—
—
—
0.068
0.068
0.071
0.062
0.060
0.60
4.0
4.0
3.75
4.0
4.34
0.83
0.40
4.9
1.7
0.77
4.6
Glycolate
Acetate
0.63c
3.65c
1.56c
1.17c
1.5
0.56
0.22
1.50
0.67
0.25
Lutidine
55.1
49.8
50.8
49.0
—
—
a Ratio of buffer acid to buffer base concentration. b Slope of plot of measured first order rate constants against concentration of buffer base.
c Separation of kGA and kGB required correction for partial protonation of enol reactant using eqns. (12)–(14) and assumes that kGA/kGB is the same
as in Table 2. d Rate constant for reaction of enol with base: the reverse of the measured reaction of ketone with base represented by k in preceding
column.
ketonisation (kH OK) of the N-protonated keto and enol taut-
enol tautomeric constant for the neutral (KTE = KaKH/KaEH) and
2
KH2؉
EH2؉
EH2؉
omers (KH2ϩ and EH2ϩ respectively). The ratio of these values
protonated (KTE = KaKHKa
/KaEHKa
could also have been obtained
ketones respectively.
K
EH
kH OE/kH O thus corresponds to the keto–enol tautomeric
In principle Ka and Ka
2
2
constant KTE = [EH2ϩ]/[KH2ϩ] for the N-protonated ketone.
from a best fit of calculated to experimental data, but the preci-
sion of values of ko extrapolated from the buffer plots [eqn. (2)]
was not very high.
Rate measurements for solutions of HCl (<pH 2) yield
kH OE = 7.4 × 10Ϫ4 sϪ1 and kH OK = 3.2 × 10Ϫ2 Ϫ1, from which
s
2
2
KTE = 0.023 and pKTE = 1.64. For economy of notation the
symbol KTE is used for both neutral and protonated ketones.
The lines through the experimental points for ketonisation in
Fig. 1 are based on eqn. (5), in which the rate constants kHK and
The buffer-independent rate constants for enolisation (ko )
E
E
KH2؉
may be fitted to eqn. (6) where kHE, kOH and Ka
are rate
KH2؉
kHEKa
E
(6)
ko
=
ϩ kOHE[OHϪ]
KH2؉
{1 ϩ Ka
/[Hϩ]}
kHK[Hϩ]
(1 ϩ [Hϩ]/Ka
kOHK[OHϪ]
(1 ϩ KaEH/[Hϩ])
K
(5)
ko
=
ϩ
EH2؉
constants and ionisation constants for the keto reactant analo-
gous to those for the enol in eqn. (5), and the superscript E for
the rate constants denotes enolisation. The line drawn through
)
kOHK refer to the acid and base-catalysed reactions respectively
(again the superscript K refers to the ketonisation reaction),
and KaEH (= [EϪ][Hϩ]/[EH], where EϪ denotes the enolate anion)
E
E
the points corresponds to a best fit of values of kH and kOH
and the independently measured values of ionisation constants
to the data. In the figure the vertical separation of the plot for
enolisation from that for ketonisation in the pH range 4–9,
where the reaction involves interconversion of neutral enol and
ketone tautomers, corresponds to the effective tautomeric con-
stant pKT = 3.35. Below pH 3, where the enol and ketone are
protonated, the separation reduces to pKTE = 1.64, the tauto-
meric constant for the protonated species.
EH2؉
ϩ
and Ka
(= [EH][Hϩ]/[EH2ϩ], where EH2 is the N-
protonated enol) are ionisation constants for the enol and its
conjugate acid. The rate constants are chosen to give a best fit
of calculated to experimental data, and the ionisation constants
EH
EH2؉
Ka and Ka
are based on combination of the correspond-
ing ionisation constants for the keto tautomer (KaKH = [EϪ][Hϩ]/
[KH] and Ka
KH2؉
= [KH][Hϩ]/[KH2ϩ]) with values of the keto–
J. Chem. Soc., Perkin Trans. 2, 1997
2763

View Article Online
Table 2 Slopes and intercepts of buffer plots and rate constants for general acid- and general base-catalysis of halogenation of 2-, 3- and 4-
phenylacetylpyridines and N-methyl-2-phenylacetylpyridinium ion in aqueous solution at 25 ЊC and ionic strength 0.2 
Substrate/buffer pH
2-Phenylacetylpyridine
Ra
ko/10Ϫ4 sϪ1
kb/10Ϫ2 dm3 molϪ1 sϪ1
kGB/10Ϫ2 dm3 molϪ1 sϪ1
kGA/10Ϫ2 dm3 molϪ1 sϪ1
Acetate (I2)
3.68
4.04
5.29
5.66
9
4.2
0.25
0.11
1.0
(0.26)
(0.12)
—
—
—
1.85
1.05
0.258
0.166
1.01
0.202
0.18
Lutidine (I2) 6.67
1.0
7.30
7.66
0.25
0.11
(0.53)
0.99
1.00
0.987
N-Methyl-2-phenylacetylpyridinium ion
Acetate (Br2) 4.02
4.64
4
1
0.25
0.0035
0.0035
—
90.0
109.0
60.5c
100.0
(I2)
5.28
3-Phenylacetylpyridine
Acetate (I2)
3.44
3.86
4.54
5.17
5.57
14
5.1
1.2
0.28
0.13
1.5
0.53
0.19
0.10
0.10
—
0.632
0.256
0.227
0.156
0.183
0.603
0.586
0.581
0.165
0.577
0.031
—
0.06
0.12
0.18
0.74
Lutidine (I2) 6.46
6.94
7.35
4-Phenylacetylpyridine
Acetate (I2)
3.43
3.86
4.54
5.17
5.57
14
5.2
1.2
0.28
0.13
1.5
0.53
0.19
0.54
—
0.18
—
0.22
0.47
1.6
2.20
1.28
0.519
0.476
0.441
1.92
0.428
1.87
0.133
Lutidine (I2) 6.48
6.94
7.35
1.89
1.88
0.31
a Ratio of buffer acid to buffer base concentration. b Slope of plot of measured first order rate constants against concentration of buffer base. c Not
used in evaluating kGB (see text).
Table 3 First order rate constants for ionisationa and bromination of 2-, 3- and 4-phenylacetylpyridines and the N-methyl-2-phenylacetyl-
pyridinium ion (2-NMeϩ) and relaxation of their enols in aqueous HCl, NaOH and borate buffersb at different pHs at 25 ЊC
2-Isomer
pH
2-NMeϩ c
pH
3-Isomer
pH
4-Isomer
pH
ko/sϪ1
ko/sϪ1
ko/sϪ1
ko/sϪ1
Relaxation
Ionisation
0.70
0.30
2.20
8.50
9.41
10.05
0.0326
0.0329
0.0314
2.94
3.06
3.58
0.62
1.4
1.82
2.41
2.72
2.93
0.04
0.92
1.69
1.98
0.0478
0.0491
0.0490
0.0475
0.92
1.69
1.91
0.008 15
0.007 99
0.008 70
12.0
12.1
12.5
12.65
12.8
13.0
3.14
4.51
15.4
21.7
25.2
35.3
11.2
11.4
11.7
12.0
8.46
15.2
30.1
58.1
k/10Ϫ4 sϪ1
k/10Ϫ4 sϪ1
k/10Ϫ4 sϪ1
k/10Ϫ4 sϪ1
Bromination
0.90
1.20
1.60
1.88
2.20
2.60
7.42
8.02
6.90
7.42
5.69
4.28
0.70
1.47
2.17
20.2
20.4
25.8
0.78
1.17
1.78
2.47
2.57
2.88
0.89
1.06
0.98
0.86
0.61
0.41
0.78
1.78
2.17
2.69
2.77
2.87
2.38
2.44
2.32
2.06
1.93
1.64
a Reaction of ketone with OHϪ to form enolate anion. b The ionic strength is 0.1  in the borate buffers, and was not maintained constant in
solutions of HCl or NaOH. c N-Methyl-2-phenylacetylpyridinium ion.
The rate constants for base catalysis of the tautomerisation,
kOH and kGB, can be converted to microscopic (or ‘molecular’)6
rate constants for the individual steps of C-protonation of the
enolate anion kBH and proton abstraction from the α-carbon of
the ketone kB based on the usual mechanism for base-catalysed
interconversion of ketone and enol tautomers shown in Scheme
2764
J. Chem. Soc., Perkin Trans. 2, 1997

View Article Online
1, in which again KH, EH and EϪ represent ketone, enol and
ZE
pKT
EH
ZH
EH
kBϪ[AϪ
]
[Hϩ]/Ka
KH
EϪ
BH[AH]
EH
(6.98)
3.35
k
Scheme 1
+
KH2
KH
2.30
EH
enolate anion respectively and Ka the ionisation constant of
Scheme 4
the enol.
E
E
Ϫ
Values of kB correspond to kGB and kOH for the enolis-
ation reaction and kBH to kGBK(Ka/KaEH) for ketonisation, where
Ka is the ionisation constant of the buffer acid AH. If KT is
known these relationships may be combined with eqn. (4) to
KH2؉
constant and pKa
= 2.30 for the pKa of the N-protonated
ZE
ketone allows pKT to be evaluated as 6.98 Ϫ 3.35 Ϫ 2.30 =
1.33, which implies that the zwitterion is more than 20-fold less
stable than the enol.
Ϫ
obtain values of both kBH and kB for each acid base pair. Using
A deficiency in this evaluation of pKTZE is the lack of correc-
tion for the substituent effect of replacing H by methyl in the
ionisation reaction of Scheme 3. However, an indication that
this correction should be small comes from measurements of
the keto–enol tautomeric constant for the N-methylphenyl-
acetylpyridium ion 11 described below. It should also perhaps
be noted that the value of pKTE = 3.35 used in Scheme 4 is a
the same equation kBH for H2O (kBHH O) may be obtained from
2
kOH (= kOHE/KT) as kBHH O = kOHKKw/KaEH, where Kw is the
K
2
Ϫ
autoprotolysis constant of water. Values of kB and kBH for
lutidine, acetate and hydroxide bases are listed in Table 4.
The interpretation of the acid-catalysed tautomerisation
reaction is less straightforward than that for the base reaction,
as was previously found for phenacylpyrazine.5 This is a con-
sequence of the expectation that enolisation proceeds not by a
two-step mechanism with protonation occurring on the oxygen
atom of the ketone, as is the case for simple ketones, but by the
three-step mechanism shown in Scheme 2, in which protonation
corrected value based on the relationship KT = KTE ϩ KT
Z
where KT is the experimentally measured constant (pKT = 3.33).
Since KT cannot be evaluated without knowing KTZ, it is
E
obtained in principle by iterative substitution in Scheme 4. In
practice, the correction is very small and iteration is unnecessary.
ϩ
KH2
EZ
[Hϩ]/Ka
kAϪ[AϪ
]
1/KT
KH
KH
ZH
EH
ϩ
2
1-Methyl-2-phenylacetylpyridinium toluene-p-sulfonate
kAH[AH]
A keto–enol tautomeric constant for this compound (11) was
determined in the same way as for its unmethylated counter-
part, 2-phenylacetylpyridine (1), by combining kinetic meas-
urements of rate constants for ketonisation and enolisation. As
before rates of ketonisation were measured by quenching the
(zwitterionic) conjugate base of the ketone (12) into solutions
of a strong acid or acidic buffers at pHs below the pKa of the
ketone (6.98) and observing relaxation of the initially formed
N-methylated enol (13) to its keto tautomer. Rates of enolis-
ation were measured from reaction of the ketone with bromine
or iodine. The observed rate constants may be expressed as in
eqns. (1)–(3) and derived rate constants k, ko and kGB from
measurements in acetate and lutidine buffers as well as hydro-
chloric acid are listed in Tables 1–3. Values of kGB were
Scheme 2
yields the N-protonated ketone, KH2ϩ, and is followed by depro-
tonation from carbon to form initially the zwitterion tautomer
(ZH) which then rapidly tautomerises to the enol (EH). Evi-
dence that the reaction proceeds in this manner comes from
measurements of rate constants that are considerably larger
than expected for enolisation of a non-heterocyclic ketone
of comparable reactivity: thus kH = ~0.15 Ϫ1 sϪ1 for 2-phenyl-
acetylpyridine compared with an expected value of ~10Ϫ5 Ϫ1
sϪ1 5
.
Ϫ
The rate constant kA for proton abstraction from the proto-
E
nated ketone presents no difficulty. This is related to kGA for
KH2؉
KH2؉
enolisation as kA = kGAEKa
/Ka where Ka
and Ka are
Ϫ
Ϫ
converted to microscopic rate constants kA and kAH based
ionisation constants for the N-protonated ketone and buffer
acid respectively. However, to obtain kAH knowledge of the
equilibrium constant for tautomerisation of enol (EH) to zwit-
E
Ϫ
on Scheme 5, with kA = kGB for enolisation and kAH
=
+
terion (ZH, 6), KTEZ = [ZH]/[EH] is required. Thus from
kAH[BH]
KaEHMe/[H+]
Scheme 2, kAH = kGAK/KT
.
EZ
kA-[B–]
EZ
N
Ph
+
N
Ph
N
Ph
+
+
In the absence of a direct measurement of KT a value may
be inferred by taking the N-methyl zwitterion (12) formed from
ionisation of the N-methyl-2-phenylacetylpyridinium cation
(11) as a model for the N–H zwitterion. A pKa for this ionis-
ation (Scheme 3) was measured spectrophotometrically in
Me
OH
Me
O
Me
O–
ZMe
12
EHMe+
KHMe+
13
11
Scheme 5
؉
EHMe؉
kGBK(Ka/KaEHMe )/(1 ϩ KTE) for ketonisation, where Ka
is
+
H+
N
Ph
(7)
+
the dissociation constant for the enolic hydrogen of 13 and
KTE = [EHMeϩ]/[KHMeϩ] is the tautomeric constant for enolis-
ation of the methylated ketone (11). As expected only base cata-
lysis of the tautomerisation was observed but the microscopic
+ N
Ph
Me
ZMe
O–
Me
O
KMeH+
11
12
Ϫ
Ϫ
rate constants are denoted kA and kAH (rather than kB and
kBH) because of the analogy between N-protonated and N-
methylated substrates. The term (1 ϩ KTE) appears in the
expression for kAH because the tautomeric equilibrium does not
lie strongly in favour of the ketone. Rates of ketonisation, there-
fore, comprise a sum of rate constants kK ϩ kE = kK(1 ϩ KTE).
In practice interpretation of the buffer measurements for the
ketonisation reaction also depended upon the pH. In acetic acid
buffers the measured rate constants for reaction of acetate ion
correspond to kGBK ϩ kGBE, with only a minor correction for
Scheme 3
borate, lutidine and acetate buffers as 6.98, and this value was
KH2؉(C)
substituted for pKa
, the pKa for deprotonation of the N-
protonated phenylacetylpyridine (KH2ϩ) at its methylene car-
bon atom, in the cycle shown in Scheme 4, in which, again, ZH,
EH and KH represent the zwitterion, enol and ketone tauto-
mers and the equilibria are shown as single rather than double
arrows to indicate the directions of reaction to which the pKs
refer. Substituting pKT = 3.35 for the keto–enol equilibrium
deprotonation of the reactant (EMeHϩ
ZMe ϩ Hϩ). For
J. Chem. Soc., Perkin Trans. 2, 1997
2765

Table 4 Rate constantsa for reaction of 2-, 3- and 4-phenylacetylpyridines, their enol and zwitterion tautomers and N-methyl-2-phenylacetylpyridinium ion with oxygen and nitrogen acids and bases in aqueous
solution at 25 ЊC
N-Methyl-2-
phenylacetylpyridinium ion
2-Phenylacetylpyridine
3-Phenylacetylpyridine
4-Phenylacetylpyridine
c
d
e
f
c
d
c
d
e
f
a
d
e
f
؊
Acid (AH or BH)b
H3Oϩ
pKa
kAH
kA
kBH
kB
kAH
kA
kAH
kA
kBH
kB
kAH
kA
kBH
kB
؊
؊
؊
؊
؊
؊
Ϫ1.76
7.44 × 103 7.44 × 10Ϫ4
55.5
/
2.24 × 103/ 2.24 × 10Ϫ3/ 7.8 × 105 9.8 × 10Ϫ5
55.5
/
1.19 × 105 2.38 × 104/
55.5
55.5
55.5
CNCH2COOH
HOCH2COOH
CH3COOH
2,6-Lutidinium ion
N-Ethylmorpholinium ion
Boric acid
2.43
3.83
4.76
6.69
7.70
4.34 × 103 2.33 × 10Ϫ3
1.56 × 103 2.10 × 10Ϫ2 1.34 × 105 4.98 × 10Ϫ?
90.0
0.52
3.68 × 104 2.02 × 10Ϫ3 170.0
2.14 × 103 1.0 × 10Ϫ2 2.5
0.30
0.98
2.4
1.77
6.3 × 103 4.8 × 10Ϫ2 4.34 × 105 1.65 × 10Ϫ3 1.17 × 103 0.134
1.87 × 104 5.77 × 10Ϫ3
9.17 × 104 2.91 × 103
6.95 × 103 1.87 × 10Ϫ2
9.24
2.49 × 10Ϫ2 4.32
5.80 × 10Ϫ4/ 5.80 × 103
55.5
H2O
Ϫ15.76
3.63/55.5 347
6.02/5.55g 38g
1.10/55.5g 60.3g
a Units molϪ1 sϪ1 b AH and BH (and their conjugate bases AϪ or BϪ) are the acids (or bases) indicated in subscripts of rate constants. c Rate constants for protonation of zwitterion tautomer of substrate by acid AH.
.
d Rate constant for reaction of N-protonated (or N-methylated) ketone with base. e Rate constant for protonation of enolate anion. f Rate constant for reaction of ketone to enolate anion. g Rate constants calculated from
data in ref. 12.

View Article Online
more basic buffers, however, the enol reactant becomes largely
deprotonated and for lutidine the observed reaction is the
reversible protonation of the zwitterion to form the ketone
(EMe ϩ AH
KMeHϩ ϩ AϪ), from which the microscopic
Ϫ
rate constants kAH and kA are directly determinable (with now
a minor correction for protonation of the zwitterion). In N-
ethylmorphine and borate buffers (and aqueous sodium hydrox-
ide) the only reaction observable is conversion of the ketone
reactant to enolate zwitterion, with rate constant kA . Values of
kGB for these buffers, therefore, are calculated from kA . The
؊
Ϫ
appropriate rate constants are included in Tables 1–4. For the
Ϫ
lutidine buffers the values of kAH and kA may be combined to
evaluate the pKa of the ketone (KMeϩ) as 6.7, in fairly satisfac-
tory agreement with the spectrophotometrically determined
value of 7.0.
E
Values of the tautomeric constant KT for the methylated
ketone were obtained by combining corresponding rate con-
stants for enolisation and ketonisation measured in HCl solu-
tions and acetic acid buffers. These gave a mean value of
KTE = 0.13 and pKTE = 0.9.
Fig. 2 Log k–pH profiles for enolisation (or ionisation) of N-methyl-
2-phenylacetylpyridine 11 (᭹) and relaxation of its enol tautomer
13 (᭺). The dashed line represents ketonisation of the enol and its
conjugate base 12.
N
H
Ph
+
O
ation shows an inflection at pH 6 corresponding to the pKa of
the enol. At pHs below the pKa of the ketone (6.9) ketonisation
is no longer separately observable from enolisation, but the
dashed line traces the pH-dependence of the reaction of the
enolate anion with H3Oϩ and its pH-dependent (and thermo-
dynamically unfavourable) protonation by water.
14
E
The value of pKT for 11 may be compared with the corres-
ponding value of 1.64 for the N-protonated ketone (14). The
larger value for the protonated ketone is at first surprising in the
E
light of previous measurements showing that pKT for N-
The line drawn through the points of the pH-profile for eno-
lization (or enolate anion formation) is based on eqn. (7) in
protonated 3-phenacylpyridine is smaller (3.89 compared with
4.23) than for its N-methylated counterpart.3 The most likely
explanation of this would seem to be that the ketone (14) is
stabilised by hydrogen-bonding. It seems likely that the sub-
stituent effect of replacing H by methyl upon the dissociation
of protonated and methylated ketones to their respective zwit-
terions (Scheme 3) will be small, because replacement of methyl
by H in the zwitterion 12 should lead to a similar stabilising
influence from hydrogen-bonding to that in the ketone. No cor-
rection is applied therefore to the value of KTZE based on use of
the ionisation constant of Scheme 3 in Scheme 4. An ionisation
koE = kH OE ϩ kOHE[OHϪ]
(7)
2
E
E
which kH O and kOH represent rate constants for the pH-
2
independent and hydroxide ion-catalysed reactions respectively.
The dashed line for ketonisation is based on eqn. (8) in which
kH OK ϩ kOHK[OHϪ]
2
K
(8)
ko
=
EMeH
1 ϩ Ka
ϩ/[Hϩ]
constant for the enol of the N-methyl-2-phenylacetylpyrid-
EHMeϩ
H2O
k
K and kOHK are the corresponding rate constants for keton-
inium ions (pKa
= 6.1) was obtained by combining the
EH
isation (with unionised enol as reactant) and Ka is the acid
dissociation constant of the enol. The full line drawn through
the experimental points for ‘ketonisation’ represents the sum of
these two expressions. The rate constants were chosen to give a
best fit to the experimental points and the ionisation constant
of the enol is taken as pKa
measured value of the pKa for the ketone and the tautomeric
constant pKTE = 0.9. The pH dependence of buffer catalysis of
ketonisation measurements in acetic acid buffers provided a
qualitative conformation of this value.
EMeHϩ
Values of ko from the buffer plots and direct measurements in
solutions of HCl and NaOH were used to construct the pH-
profiles of log ko plotted against pH shown in Fig. 2 for both
enolisation (filled circles) and ketonisation (open circles). The
ionisation measurements above pH 7–8 are included with the
enolisation rate constants as filled circles. As already noted, the
‘ketonisation’ rate constants are really relaxation rate constants
corresponding to sums of rate constants for ketonisation and
enolisation. At high pHs, when enolisation (strictly ionisation
to the enolate anion) becomes faster than ketonisation the two
plots merge. A similar pH-profile has been reported by Bunting
and Kanter for the tautomerisation of ketoesters,7 which also
have relatively large enol contents. The pH-dependence of the
‘true’ ketonisation rate constants are shown by the dashed line.
The lack of acid catalysis of the tautomerisation reactions is
evident in Fig. 2 from the pH-independence of the reactions in
acidic solutions. The reactions correspond in acid to rate-
determining deprotonation of the ketone by a water molecule
in the enolisation direction and to protonation of the enolate
anion by H3Oϩ in the reverse ketonisation. Above pH 7–8 the
hydroxide ion replaces water as the base leading to base-
catalysed enolisation (ionisation). The pH-profile for ketonis-
= 6.1, as determined above.
The hydroxide and water rate constants for enolisation cor-
H2O
HO
Ϫ
Ϫ
respond to the microscopic rate constants kA
and kA
.
/
K
Those for ketonisation yield rate constants kAHH O = kH O
3
2
ϩ
EHMeϩ
2
Ka
and kAHH O = kOHK/kwKaEHMe . Values for these are
listed in Table 4.
3- and 4-Phenylacetylpyridines
Studies of the 3- and 4-phenylacetylpyridines (15 and 16) closely
N
N
Ph
Ph
O
16
O
15
followed the pattern for the 2-isomer (1) save that the N-methyl
compounds, which had been previously studied by Bunting and
Stefanidis,12 were not prepared. Measurements of pKas for N-
protonation gave 2.59 and 2.66 for the 3- and 4-isomers respect-
ively, and the pKas of 13.1 and 12.26 for ionisation to their
J. Chem. Soc., Perkin Trans. 2, 1997
2767

View Article Online
enolate anions were measured or estimated by Bunting and
Stefanidis.12 Rate constants for ketonisation and enolisation
were measured in the same way as for the 2-compound, and
slopes of buffer plots for lutidine and acetate buffers are given
in Table 1, together with corresponding values of ko from inter-
cepts of buffer plots; rate measurements in aqueous HCl are
shown in Table 3.
N
N
Ph
Ph
O
O
pKET
pKa
3.1
13.26
4.2
12.1
Rate constants for the reactions in aqueous sodium hydrox-
ide had been measured by Bunting and Stefanidis and were not
duplicated.12 For 4-phenylacetylpyridine, the buffer measure-
ments were extended to cyanoacetic and glycolic acids. Contri-
butions from both acid- and base-catalysed pathways were
observed for acetate and glycolate buffers but only an acid con-
tribution for the stronger cyanoacetic acid.
N
Ph
Ph
O
O
pKET
pKa
3.35
5.15
12.02
Values of kGA and kGB were converted to microscopic rate
constants and these are included in Table 4. Again it was pre-
sumed that acid-catalysed tautomerisation occurs via a zwitter-
ion tautomer and corrected values of pKTZE, which were
required to obtain values of kAH for ketonisation, were derived
N
O
O
O
KHMeϩ
from Bunting’s and Stefanidis’s measurements12 of pKa
N
Ph
N
Ph
Ph
for the N-methyl-3- and N-methyl-4-phenylacetylpyridinium
KHMeϩ
ions. For the 3-compound pKa
to 10.00 for pKa
= 10.30 and this is corrected
KH2؉(C)
on the basis of a previously estimated
pKET
pKa
4.4
12.74
4.86
2.0
ϩ
difference between KHMeϩ and KH2 ionising to zwitterions
13.65
13.27
for N-protonated and N-methylated 3-phenacylpyridine.3
A
ϩ = 9.02 for the 4-
KMeH
Fig. 3 Comparison of keto–enol tautomeric constants and ionisation
constants of the keto tautomers of phenyl- and phenylacetyl-pyridines
and deoxybenzoin
similar correction was applied to pKa
isomer12 to give pKa = 8.7 for deprotonation from carbon of
KH2ϩ. These corrections contrast with the lack of correction to
KHMeϩ
pKa
for the 2-isomer, where it is supposed that hydrogen
bonding in KH2 stabilises the ketone (above). It is also note-
E–
ϩ
worthy that for the N-methyl-4-phenacylquinolinium ion,
8.67
(1.3)
12.02
where deprotonation yields an enaminone rather than zwitter-
KHMeϩ
(7.4)
ion, pKa
= 7.02 is less than for the corresponding proto-
(4.6)
nated ketone (7.54).11
3.35
ZH±
KH
EH
KH
Combining rate constants for ketonisation and enolisation
gives values of pKT = 4.2 and 3.1 for keto–enol tautomerism of
the 3- and 4-phenylacetylpyridine isomers respectively. Rate con-
stants from Table 2 may also be combined with Bunting’s and
Stefanidis’s measurements in NaOH to construct pH profiles
for ketonisation and enolisation. These are similar to those for
the 2-isomer in Fig. 1 and are not shown, but rate constants kH
for ketonisation and enolisation reactions could be derived
based on eqns. (5) and (6), and these were converted to the
appropriate microscopic rate constants, as well as tautomeric
constants pKTE = 2.80 and 1.54 for the N-protonated 3- and 4-
isomers respectively, in the same manner as for the 2-isomer;
values of kOH were taken from the measurements of Bunting
and Stefanidis.12 These values are also included in Table 4.
N
Ph
(5.3)
1.64
(6.9)
4.0
O
2.30
+
+
EH2
KH2
E–
8.9
(3.0)
13.2
(5.9)
(7.3)
4.2
ZH±
KH
EH
KH
N
Ph
(7.1)
2.80
9.9
O
4.1
EH2
2.59
+
+
KH2
Discussion
Tautomeric and ionisation constants
For all three phenylacetylpyridine isomers the order of tauto-
mer stability in aqueous solution at 25 ЊC is confirmed to be
keto > enol > zwitterion. Enol contents are greater than for
many simple ketones15 but in no case exceed one part in a thou-
sand. Quantitative measurements of the tautomeric and ionis-
ation constants are summarised in Scheme 6 as networks of pKs
for equilibria linking keto (KH), enol (EH) and zwitterionic
forms (ZH ) with their common N-protonated conjugate acids
(KH2ϩ and EH2ϩ) and single conjugate base, the enolate anion
(EϪ).3 As before, the normal double arrows depicting equilibria
are replaced by single arrows indicating directions of reactions.3
If opposite signs are assigned to pKs for clockwise and anti-
clockwise arrows the pKs in each reaction cycle sum to zero.
The pKs of N᎐H compounds estimated from measurements for
their N-methyl analogues are shown in brackets.
E–
9.16
3.0
(7.2)
1.54
12.26
(6.2)
6.1
N
3.1
ZH±
KH
EH
KH
Ph
8.7
O
4.22
EH2
2.66
+
+
KH2
Scheme 6
also shown and it can be seen that replacement of either of the
two phenyl rings by pyridyl groups increases the enol content.
Interestingly, for 3- and 4-pyridyl groups the enol content is
increased to a greater extent when substitution occurs adjacent
to the carbonyl group as in the phenylacetylpyridines (PhCH2-
COPy) than at the α-carbon atom in the phenacylpyridines
In Fig. 3 keto–enol tautomeric constants and ionisation con-
stants of the keto tautomers of all the phenacyl- and phenyl-
acetyl-pyridines are compared. Values for deoxybenzoin3,16 are
2768
J. Chem. Soc., Perkin Trans. 2, 1997

View Article Online
(PyCH2COPh). As discussed by Stefanidis and Bunting,13 it is
likely that in addition to stabilisation of the enols by conju-
gation between the pyridine and phenyl rings (17 and 18)
replacement of phenyl by the electron-withdrawing pyridyl
group decreases the stability of the ketone. The effect is greater
for the phenylacetyl- than phenacyl-pyridines because of the
greater proximity of the pyridyl and carbonyl groups. Appar-
ently, this outweighs any stabilising conjugation between the
pyridyl ring and hydroxy group in the enols of the latter (e.g.
18).
The difference in enol contents between the phenylacetyl-
and phenacyl-pyridines is even more marked for the N-
protonated ketones, and indeed now extends to the 2-isomer
(Scheme 6). For all three (neutral) isomers also ionisation to an
enolate anion occurs more readily for the phenylacetyl pyridyl
ketones.
Proton activating factors
The measurement of rate and equilibrium constants for the
phenylacetylpyridines allows evaluation of ‘proton activating
factors’10 for these compounds. Proton activating factors meas-
ure the effect of protonation at one reaction site upon suscepti-
bility to attack by a base or nucleophile at another. Originally
they were conceived by Stewart and Srinivasan for reaction
rates,10 but there seems no reason not to apply them also to
equilibria.3,5
The connection between proton activating factors and
tautomeric equilibria may be seen by dissecting these equilibria
into acid and base ionisation reactions for proton addition and
proton loss to and from the two tautomers. This is illustrated
for the interconversion of enol and zwitterion tautomers of a
phenylacetylpyridine in Scheme 7, which shows that the proton
EH + H+
OH
a
b
N
N
E– + 2H+
+
pKT
EH2
HO
17
18
c
d
ZH± + H+
Only for the 2-isomers is there a greater enol content for a
phenacyl- than phenylacetyl-pyridine. For 2-phenacylpyridine
there is a large increase in enol content (pKTE = 2.0) relative to
its 4-isomer (pTE = 4.20), which can be attributed to stabilis-
ation of the enol by hydrogen-bonding (3).1 For the phenyl-
acetylpyridines, in contrast, the enol content is less for the 2-
isomer (pKTE = 3.35) than for the 4-isomer (pKTE = 3.1).
Inspection of the pKTs and pKas in Scheme 6 offers no evi-
dence of hydrogen-bonding in 2-phenylacetylpyridine (19).
Thus the pKa for ionisation of the enol (8.67) is normal while
that for phenacylpyridine enol is high (11.27), consistent with
an additional stabilisation of the latter. On the other hand,
Bunting and Kanter have proposed a small (ten-fold) stabilis-
ation of the enol of pyridacyl esters and amides (e.g. 20) on the
basis of a correlation of pKas of enols and ketones for a series
of keto esters and keto amides.7 Moreover, the difference in
pKas of N-methyl and N᎐H phenacylpyridinium ions is consist-
ent with a similar degree of stabilisation for the structurally
related ketone 16. It seems likely therefore that there is some
hydrogen-bonding in 19 but that its effect is substantially less
than in the phenacylpyridine enol 3. The difference presumably
reflects more favourable hydrogen-bonding in the six- than five-
membered ring.
Scheme 7
transfer and tautomeric equilibria may be written as a network
of thermodynamic cycles. In Scheme 7, the usual symbols for
the tautomers and their ionised forms are used.
Reaction schemes similar to Scheme 7 have been widely used
to represent tautomerisation and ionisation of amino acids.17–19
They show that the position of a tautomeric equilibrium can be
expressed in terms of acid dissociation constants relating the
tautomers to a common conjugate acid or conjugate base. Thus
in Scheme 7 if the pKas of the reactant tautomer acting as acid
and base respectively are denoted a and b and those of the
product tautomer c and d, the tautomeric constant pKT is given
by a Ϫ c or d Ϫ b.
For such equilibria normally one tautomer is considerably
more stable than the other. The information readily available
then is limited to the pKas for the stable tautomer (say a and b).
If the acid and base sites of the tautomers are well separated, as
in a long chain amino acid [e.g. ϩH3N(CH2)nCOOϪ where n is
large], extents of ionisation at the two sites are nearly independ-
ent of each other and, to a good approximation, pKT corres-
ponds to the difference in pKas of acidic and basic sites a Ϫb,
which may be denoted ∆pKab. More commonly, however, ioni-
sations at the two sites are not independent and ∆pKab has to be
corrected for their interaction. The appropriate correction fac-
tor is provided by the (log of the) proton activation factor, PAF,
which describes the effect of protonation at one tautomeric site
upon proton loss at the other. Thus in Scheme 7 log PAF is
given by a Ϫ d or c Ϫ b and its relationship to pKT and ∆pKab is
expressed by eqns. (9) and (10).
X
N
Ph
N
O
O
O
H
H
X = OMe or NH2
19
20
The proportion of zwitterion tautomer in the phenylacetyl-
pyridines is consistently less than that of the enol. This con-
trasts with the phenylacylpyridines where for the 2- and 4-
isomers the enaminone is more stable than the enol (e.g.
pKTM = 0.88 {where KTM = [enaminone]/[ketone]} compared
with pKTE = 2.0 for 2-phenacylpyridine). The largest proportion
of zwitterion relative to enol is observed for the 2-
∆pKab = a Ϫ b
log PAF = c Ϫ b
(9)
(10)
(11)
pKT = ∆pKab Ϫ log PAF = a Ϫ c
Eqn. (11) summarises the relationship between pKT, ∆pKab
and log PAF. Normally it is not used for evaluating pKT because
this is better done by estimating one of the unknown pKas in
Scheme 7 from a free energy relationship. For amino acids, for
example, use of a corrected pKa for a methyl or ethyl ester,
taken as a model for the neutral acid, provides the most effect-
ive method for evaluating pKT.
phenylacetylpyridine (1) for which pKTZE = 1.3 compared with
ZE
3.1 and 3.0 for the 3- and 4-isomers respectively (KT
=
[zwitterion]/[enol]). This presumably reflects the juxtaposition
of positive and negative charges in the 2-isomer. The stability of
the 2-phenylacetyl zwitterion (6) is also apparent from the
enhanced acidity of its N-methyl derivative, with pKa = 7.0
compared with pKa = 10.30 and 9.02 for the corresponding
N-methylated 3- and 4-phenylacetylpyridines.
Occasionally, an unknown pKT or pKa may be estimated from
eqn. (11). For the phenylacetylpyridines, for example, it is not
J. Chem. Soc., Perkin Trans. 2, 1997
2769

View Article Online
Table 5 Contributions of intrinsic acidity and basicity (∆pKab)a and proton activation (log PAF) to enol–zwitterion (or enaminoneb) and
keto–zwitterion tautomeric constants (pKT) for phenacyl- and phenylacetyl-pyridines
EH → ZH
KH → ZH
∆pKab
EZ d
KZ f
∆pKab
log PAFc
pKT
log PAFe
pKT
Phenylacetyl
2-
3-
4-
4.8
4.8
4.95
3.55
1.7
2.0
1.15
3.1
2.95
9.7
10.4
9.6
5.1
3.1
3.6
4.6
7.3
6.0
Phenacyl
2-
3-
4-
6.75
2.95
2.10
7.70
1.95
4.1
Ϫ0.95
1.0
Ϫ2.0
8.25
8.80
8.1
7.2
2.9
5.7
1.05
5.9
2.4
a Difference in pKas of basic and acidic ionisation sites of reactant tautomers. b For 2- and 4-phenacylpyridines the zwitterion structure is replaced by
c
an enamine. PAF measures the effect of N-protonation upon deprotonation from the oxygen of the enol (or O-protonation upon deprotonation
from the nitrogen of the zwitterion or enaminone). d KTEZ = [enol]/[zwitterion]. e PAF measures the effect N-protonation upon C-deprotonation of
the ketone (or of C-protonation upon N-protonation of the zwitterion). f KTKZ = [zwitterion]/[ketone].
possible to write a cycle corresponding to Scheme 7 for keto–
enol tautomerism because, although pKTE is known, ionisation
constants for O-protonation of the ketone (or C-protonation of
the enol) are not known. However, we note that the difference in
pKas of O-protonated ketone and (unprotonated) enol corres-
ponds to a proton activating factor. If for 4-phenylacetyl-
pyridine (e.g.) we suppose that log PAF is the same as that for
acetophenone, which has been evaluated as 14.34,3 then from
eqn. (11) or Scheme 7, the pKa of the O-protonated ketone is
Closer inspection of Table 5 shows that ∆pKab for enol–
zwitterion and keto–zwitterion tautomerism varies little
between 2-, 3- and 4-pyridyl substituents, except where there is
strong stabilisation from hydrogen bonding, as in the case of
the enol of 2-phenacylpyridine. On the other hand log PAF
varies considerably. Relative magnitudes are consistently in the
order 2- > 3- > 4-, although the difference between 3- and 4- is
much greater (1.95 compared with 4.1) for the phenacylpyrid-
ines, where the 4-pyridinium substituent can interact strongly
with a negative charge centre by resonance, than for the pheny-
lacetylpyridines, where it cannot. The unusually large values for
2-phenyacylpyridine (log PAF = 7.7 and 7.2) again reflect the
influence of hydrogen-bonding.
EH
given by pKaEH Ϫ 14.34. Since pKa for the enol is 9.16, the
desired pKa is estimated as Ϫ5.2.
This estimate may be compared with pKa = Ϫ4.0 for
acetophenone.20 Considering the electron-withdrawing effect
of the nitrogen atom the value appears to be reasonable and
suggests the possibility of deriving pKas for other phenylacetyl
or phenacyl pyridines. Only the 2-derivatives, where the enol
is stabilised by intramolecular hydrogen-bonding, might be
expected to give unrealistic values.
In conclusion, this analysis of the tautomeric measurements
illustrates the use of proton activating factors to express the
contributions of inductive, resonance and hydrogen-bonding
interactions between ionising sites to magnitudes of tautomeric
equilibrium constants.
However, probably the main usefulness of eqn. (11) lies in its
factorisation of tautomeric constants into contributions from
an ‘intrinsic’ term representing ionisation constants at the two
protonation sites and an ‘interaction’ term expressing inductive,
resonance or hydrogen-bonding interactions between them. For
glycine for example pKT = 5.34 for conversion of the zwit-
terionic to neutral form can be broken down into an intrinsic
contribution ∆pKab = 3.38 and interaction term log PAF = 1.94.
The much smaller interaction term for glycine than acetophe-
none reflects a purely electrostatic interaction between charge
centres compared with the strong resonance interaction for
acetophenone (21 and 22).
Experimental
Preparations
The 2-, 3- and 4-phenylacetylpyridines were prepared by Grig-
nard reactions following the procedure of Chu and Teague,
and LaForge.21 Details are given only for 2-phenylacetyl-
pyridine. The 3- and 4-isomers have also been prepared recently
by Stefanidis and Bunting.12
2-Phenylacetylpyridine. Benzyl chloride (11.4 g) was added
slowly to magnesium turnings (2.2 g). The solution was stirred
and 2-cyanopyridine was added at a rate to maintain gentle
reflux. After standing overnight 20% aqueous ammonium
chloride (100 ml) and concentrated hydrochloric acid (25 ml)
were added. The mixture was filtered and the precipitate stirred
into a solution of excess sodium hydrogen carbonate. Extrac-
tion with chloroform followed by drying (Na2SO4) and evapor-
ation of solvent gave a brown solid. This was purified by silica
flash chromatography using diethyl ether–light petroleum (40–
60 ЊC) as eluent. On refrigeration, the product was obtained as
a pale yellow low-melting solid (3.5 g). (Found: C, 79.3; H, 5.6;
N, 7.1. C16H11NO requires C, 79.1; H, 5.6; N, 7.0%); δH(CDCl3)
4.55 (2H, s, CH2), 7.2–7.35 (5H, m, Ar), 7.45–8.1 (3H, m, H-3,
H-4 and H-5), 8.7–8.75 (1H, d, H-6). Satisfactory analytical
data and expected NMR spectra were also obtained for the 3-
and 4-isomers.
+OH
C
OH
C
+H3NCH2COO–
–
Ph
CH2
Ph
CH2
22
21
log PAF = 1.94
log PAF = 14.34
With respect to the phenylacetylpyridines, Table 5 shows
tautomeric constants for converting the enol to zwitterion
tautomer (pKTEZ) and ketone to zwitterion (pKTKZ) factored
into contributions from ∆pKab and log PAF. The same analysis
is shown for 2- and 4-phenacylpyridines, for which the zwitter-
ion tautomers are replaced by enaminones. As already
explained the analysis cannot be applied to keto–enol tautomer-
ism because of lack of ionisation constants for the O-
protonated ketone.
It is evident from the table that when tautomeric constants
are dissected in this way the interaction term makes a substan-
tial contribution to pKT. Indeed pKT becomes quite a small
difference between larger values of ∆pKab and log PAF.
1-Methyl-2-phenylacetylpyridinium toluene-p-sulfonate.
Methylation of 2-phenylacetylpyridine was achieved by reac-
tion of 1.8 g with the same quantity of methyl toluene-p-
sulfonate at 150 ЊC for 4 h. After cooling overnight the product
was obtained as a brown oil that was crystallised from
methanol–diethyl ether (yield 0.71 g) (Found: C, 65.3; H, 5.5; N,
2770
J. Chem. Soc., Perkin Trans. 2, 1997

View Article Online
Table 6 Spectra and conditions of measurement of pKas of phenylacetylpyridines (PhAcPy) in aqueous solution at 25 ЊC
Substrate
2-PhAcPy
Acid–base–buffer
λ/nma
ε/10Ϫ4 dm3 molϪ1 cmϪ1 b Species
Ic/
pKa
NaOH
345
340
410
265
260
1.43
0.030
0.14
anion
cation
zwitterion
cation
—
12.02
2.30
7/0d
2.59
2.66
HCl–acetate
lutidine–borate–acetate
HCl
0.2
0.2
—
N-Me-2-PhAcPyϩ
3-PhAcPy
4-PhAcPy
0.55
HCl–acetate
0.050
cation
—
a λmax unless otherwise indicated. b From limiting absorbance at high or low pH. c More intense peaks at 232 and 270 nm gave small changes in
absorbance with pH and less reproducible values of pKa. c The ionic strength was not kept constant for measurements in HCl and NaOH solutions.
d Corrected for presence of 10% enol so that Ka refers to the ketone tautomer.
3.55. C21H21NO4S requires C, 65.8; H, 5.5; N, 3.7%); δH(CF3-
COOH) 2.50 (3H, s, CH3), 4.48 (2H, s, CH2), 4.66 (3H, s,
NCH3), 7.3–7.8 (13H, m, Ar).
NMR spectra were recorded on a Jeol JNM-GX 270FT spec-
trometer operating at 270.05 MHz.
In acetic acid buffers enolisation and ketonisation were sub-
ject to catalysis by both buffer acid and buffer base. In order to
separate second order rate constants kGA and kGB, however, the
measured first order rate constants kobs had to be corrected for
N-protonation of the reactant. For enolisation this presented
little difficulty as the extent of protonation was small in the pH-
range studied and the pKa for protonation was independently
known.
Ionisation constants
Ionisation constants for N-protonation of 2-, 3- and 4-
phenylacetylpyridines, for formation of an enolate anion from
2-phenylacetylpyridine and for formation of an N-
methylenolate zwitterion from the N-methyl-2-phenylacetyl-
pyridinium ion, were measured spectrophotometrically in
aqueous solution at 25 ЊC. The pH-dependence of absorption
maxima for cationic, anionic or zwitterionic species were moni-
tored using a Pye-unicam SP8-400 or Perkin-Elmer Hitachi 124
spectrophotometer, as described earlier.3,6 Measured pKas, ionic
strengths, wavelengths and extinction coefficients of absorption
maxima used in the measurements are given in Table 6; further
details of the measurements are provided elsewhere.22 Spectral
changes associated with the most intense absorption maxima
(ε у 104) were too small for satisfactory determination of pKas
for N-protonation, but consistent results were obtained from
measurements with weaker, longer wavelength absorptions. The
pKas obtained from these equilibrium measurements provided a
satisfactory fit of calculated-to-observed measurements for the
kinetic dependence of rate constants for enolisation upon pH
in the appropriate pH-regions and, in the case of N-methyl-
2-phenylacetylpyridinium ion, were corroborated by kinetic
measurements in lutidine buffers. The pKas in Table 6 are dir-
ectly measured values at the ionic strength indicated. Normally
no correction for ionic strength was required because under the
conditions of the measurements positive and negative ions were
conserved between reactants and products of the equilibrium.
In the case of N-methyl-2-phenylacetylpyridinium ion the pKa
is corrected for the presence of a small amount of enol so that
it refers specifically to the keto form.
For ketonisation, on the other hand, the enol reactant was
significantly protonated in acetic acid buffers. The dependence
of observed rate constants upon buffer acid and buffer base
concentration is then given by eqn. (12), in which R is the ratio
ko ϩ kGB[AcOϪ] ϩ kGA[AcOH]
(12)
kobs
=
EH2؉
1 ϩ RKaAcOH/Ka
of buffer acid to buffer base concentrations and RKaAcOH in the
denominator of the equation has been substituted for [Hϩ].
Values of kGB and kGA may be extracted from this expression
EH2؉
if the ionisation constant of the enol Ka
is known. In prin-
ciple this may be obtained by combining the pKa of the ketone
with keto–enol tautomeric constants for the neutral and N-
protonated ketones evaluated from measurements of rates of
enolisation in lutidine buffers and HCl respectively. In practice,
an alternative procedure was followed, in which the ratio kGA
/
kGB was assigned the value independently determined from the
EH2؉
enolisation reaction. This has the advantage that pKa
also be evaluated. Values of kGA, kGB and Ka
can
EH2؉
were then
derived from the slopes of plots of kobs against [AcOϪ] which
are given by eqn. (13), in which x (= 0.9 for 2-phenylacetyl
k
obs Ϫ ko kGB(1 ϩ xR)
(13)
=
[AcOϪ]
(1 ϩ yR)
AcOH
pyridine) is the ratio of kGB/kGA for enolisation and y = Ka
/
EH2؉
Ka
Cross multiplying by (1 ϩ xR) and taking reciprocals gives
eqn. (14), from which it can be seen that a plot of the left hand
.
Kinetic measurements
Kinetic measurements of ketonisation and enolisation were
carried out in the same manner as described for the tautomeris-
ation of the phenacylpyridines.5 Normally rates of ketonis-
ation were fast enough to require stopped flow measurements
and a Durrum 110 spectrometer was used for these. A freshly
prepared solution of enolate anion in dilute sodium hydroxide
in one syringe of the instrument was quenched with excess HCl
or buffer acid in the other, and reaction of the enol so generated
monitored at a wavelength close to its λmax, which was usually at
a slightly shorter wavelength than that of the enolate anion.
Reactions at pHs above the pKa for enolate anion formation
were measured by quenching ketone reactant into aqueous
sodium hydroxide or buffer base, and were monitored from the
increase in absorption arising from formation of enolate anion.
Rate constants for enolisation in acetic acid or lutidine buf-
(1 ϩ xR)[AcOϪ]
1
yR
(14)
=
ϩ
k
obs Ϫ ko
kGB kGB
side against R yields 1/kGB as intercept and y/kGB as slope. This
analysis gave kGA = 5.0 Ϫ1 sϪ1 and kGB = 4.5 Ϫ1 sϪ1 and
EH2؉
pKa
= 4.5 (corrected for ionic strength)19 for 2-phenylacet-
EH2؉
ylpyridine. Although the value of pKa
agrees only moder-
ately well with the more reliable value of 4.0 from combining
the keto–enol tautomeric constants and pKa of the ketone as
described above, the agreement is probably within the limits of
uncertainty of the analysis. Similar treatments of ketonisation
in acetic acid buffers for 3- and 4-phenylacetylpyridines gave
enol pKas of 3.6 and 4.2 respectively, which gave similar and
better agreement with the independently determined values of
4.1 and 4.22.
Ϫ
fers were measured spectrophotometrically from uptake of I3
ion accompanying conversion of the ketone reactant to its enol
tautomer. At low pHs, however, iodine was replaced by bromine
as scavanger of the enol because of reversibility of the iodin-
ation reaction.
Some minor discrepancies were encountered in the measure-
ments. For the N-methyl-2-phenylaceylpyridinium ion meas-
urements of rates of enolisation in acetic acid buffers using
J. Chem. Soc., Perkin Trans. 2, 1997
2771

View Article Online
6 A. R. E. Carey, G. Fukata and R. A. More O’Ferrall, J. Chem.
Soc., Perkin Trans. 2, 1985, 1711.
iodine rather than bromine as a trapping agent gave a reduction
of 50% in kGB. The iodine value was disregarded on the grounds
that some interference from the reversibility of iodination
might be expected; however, this rate constant gave better
agreement with KE determined in HCl solutions than the bro-
mination constant. When the same substrate was reacted in
lutidine buffers a similar discrepancy was observed in kGB
values when the reaction was carried out in forward and reverse
directions. It is possible that this was due to the presence of a
non-equilibrium mixture of E and Z enol structures formed
from neutralisation of the enolate anion. The possible influence
of E and Z structures of enols or enaminones upon rates and
equilibria of tautomerisation will be the subject of a later
paper.
7 J. W. Bunting and J. P. Kanter, J. Am. Chem. Soc., 1993, 115, 11 705.
8 A. Fontana and R. A. More O’Ferrall, J. Chem. Soc., Perkin Trans.
2, 1994, 2453; A. Fontana and P. De Maria, J. Chem. Soc., Perkin
Trans. 2 in the press.
9 A. R. E. Carey, Ph.D. Thesis, National University of Ireland, 1991.
10 R. Stewart and R. Srinivasan, Acc. Chem. Res., 1978, 11, 271.
11 G. Fukata, C. O’Brien and R. A. More O’Ferrall, J. Chem. Soc.,
Perkin Trans. 2, 1979, 792.
12 J. W. Bunting and D. Stefanidis, J. Am. Chem. Soc., 1988, 110, 4008.
13 D. Stefanidis and J. W. Bunting, J. Am. Chem. Soc., 1990, 112, 3163.
14 D. Stefanidis and J. W. Bunting, J. Am. Chem. Soc., 1991, 113, 991.
15 A. J. Kresge and J. R. Keefe, in ref. 1.
16 A. J. Kresge, personal communication.
17 R. Stewart, The Proton: Applications to Organic Chemistry,
Academic Press, Toronto, 1985.
18 A. Albert and E. P. Sergeant, The Determination of Ionization
Constants, Chapman and Hall, London, 2nd edn., 1971.
19 A. J. Hoefnagel, M. A. Hoefnagel and B. M. Wepster, J. Org. Chem.,
1978, 43, 4720.
Acknowledgements
We thank a referee for a number of corrections to the paper.
20 A. Bagno, V. Lucchini and G. Scorrano, J. Phys. Chem., 1991, 95,
345.
References
21 C. C. Chu and P. C. Teague, J. Org. Chem., 1958, 23, 1578;
F. B. LaForge, J. Am. Chem. Soc., 1928, 50, 2477; H. Schwarz,
R. Herrman and R. Wolfschütz, J. Heterocycl. Chem., 1975, 12, 633.
22 G. McCann, Ph.D. Thesis, National University of Ireland, 1991.
23 D. D. Perrin and B. Dempsey, Buffers for pH and Metal Ion Control,
Chapman and Hall, London, 1974.
1 The Chemistry of Enols, ed. Z. Rappoport, Wiley, New York, 1990.
2 E. A. Jefferson, J. R. Keefe and A. J. Kresge, J. Chem. Soc., Perkin
Trans. 2, 1995, 2041 and references therein.
3 A. R. E. Carey, S. Eustace, R. A. More O’Ferrall and B. A. Murray,
J. Chem. Soc., Perkin Trans. 2, 1994, 2285.
4 A. R. Katritzky, H. Z. Kucharska and J. D. Rowe, J. Chem. Soc.,
1965, 3093.
5 A. R. E. Carey, R. A. More O’Ferrall, M. G. Murphy and B. A.
Murray, J. Chem. Soc., Perkin Trans. 2, 1994, 2471.
Paper 7/01546A
Received 4th March 1997
Accepted 26th August 1997
2772
J. Chem. Soc., Perkin Trans. 2, 1997