Entropy and Enthalpy Contributions to Solvent Effects on Phosphate Monoester Solvolysis. The Importance of Entropy Effects in the Dissociative Transition State

Article abstract of DOI:10.1021/jo981160k

The solvolysis reactions of a series of aryl phosphates in tert-butyl alcohol and in tert-amyl alcohol have been examined. The dianion of p-nitrophenyl phosphate reacts 7500- and 8750-fold faster in these solvents, respectively, than the corresponding aqu

Full text of DOI:10.1021/jo981160k

6680
J . Org. Chem. 1998, 63, 6680-6688
En tr op y a n d En th a lp y Con tr ibu tion s to Solven t Effects on
P h osp h a te Mon oester Solvolysis. Th e Im p or ta n ce of En tr op y
Effects in th e Dissocia tive Tr a n sition Sta te
Richard H. Hoff and Alvan C. Hengge*
Department of Chemistry and Biochemistry, Utah State University, Logan, Utah 84322-0300
Received J une 16, 1998
The solvolysis reactions of a series of aryl phosphates in tert-butyl alcohol and in tert-amyl alcohol
have been examined. The dianion of p-nitrophenyl phosphate reacts 7500- and 8750-fold faster in
these solvents, respectively, than the corresponding aqueous reactions. The monoanion reacts 14-
and 16-fold slower respectively in tert-butyl alcohol and in tert-amyl alcohol. Analysis of the
activation parameters shows that the rate enhancement for the dianion is due solely to entropic
factors, while the slower reaction of the monoanion is due to increased enthalpy of activation. The
significantly more positive entropy of activation for the solvolysis of p-nitrophenyl phosphate dianion
in tert-butyl alcohol supports the original proposal that racemization at phosphorus in this reaction
is caused by a switch to a DN + AN mechanism, rather than subsequently proposed mechanisms
which avoid the formation of metaphosphate. Rate enhancements of similar magnitudes are seen
for the dianion reactions of all of the aryl phosphates examined; the slope of a plot of the rate
constants for solvolysis versus the aqueous pKa of the leaving phenols has a slope of -1.1, within
experimental error of the value for the aqueous reaction. However, in the reactions in tert-amyl
alcohol, para-substituted and meta-substituted aryl phosphates fall on separate but parallel lines
with para-substituted compounds reacting faster than meta-substituted reactants with leaving
groups of similar pKa. The pKa values for a series of para- and meta-substituted phenols in tert-
butyl alcohol and in tert-amyl alcohol were determined and were found to have a linear relationship
with the aqueous pKa values, with no distinction between para and meta substitution. Thus the
different Brønsted behavior of para- and meta-substituted aryl phosphates in these solvents is not
due to differential solvent-induced perturbations of the pKa values of the leaving groups. The
mechanistic implications of these results and their relevance to enzymatic phosphoryl transfer are
discussed.
In tr od u ction
This intermediate may or may not pseudorotate before
expelling a leaving group in a second step. Mechanistic
possibility C is a concerted ANDN reaction in which the
nucleophile enters and the leaving group departs in a
single step, with no intermediate.
The chemistry of phosphoryl transfer from phosphate
monoesters and diesters is essential and ubiquitous in
biological systems. A considerable amount of work over
the past decades has been devoted to gaining an under-
standing of the mechanisms of these reactions in solution
and to understanding the mechanisms of enzyme-
catalyzed phosphoryl transfer. Of particular interest is
whether the mechanistic pathways and the transition
states involved in the enzymatic reactions differ from
uncatalyzed phosphoryl transfer reactions in solution.
Three distinct reaction mechanisms, have been ob-
served for various types of phosphate esters under
different conditions, which are illustrated diagrammati-
cally in Figure 1.1,2 Two are two-step mechanisms; A is
an SN1 (DN + AN in the IUPAC nomenclature3) type of
mechanism in which a metaphosphate intermediate is
formed in the rate-determining step; this highly reactive
species is then attacked by a nucleophile in a subsequent
rapid step. Mechanism B is addition-elimination (AN
+ DN) and involves nucleophilic attack in the first step
to form a pentacoordinate phosphorane intermediate.
The primary factor that influences which mechanism
is followed is the alkylation state of the phosphate.
Monoesters are generally observed to follow, depending
on conditions, either a concerted mechanism with a
highly dissociative transition state (in nucleophilic sol-
vents) or, in very nonnucleophilic solvents such as tert-
butyl alcohol, a DN + AN mechanism with a metaphos-
phate intermediate. Diesters and triesters follow
successively more associative mechanisms, either con-
certed ones if the leaving group is good (i.e., an aryloxy
group), with more nucleophilic participation in the tran-
sition state, or fully associative mechanisms via phos-
phorane intermediates.
Mechanism C for the aqueous hydrolysis of phosphate
monoester dianions is supported by a very small entropy
of activation,4 a large (-1.2) âlg and a small ânuc,4 and
5
the occurrence of inversion of configuration when the
phosphoryl group is made chiral.6 The reaction of the
(1) Hengge, A. C. In Comprehensive Biological Catalysis; Sinnott,
M., Ed.; Academic Press: San Diego, CA, 1998; Vol. 1, pp 517-542.
(2) Thatcher, G. R. J .; Kluger, R. Adv. Phys. Org. Chem. 1989, 25,
99-265.
(3) Guthrie, R. D.; J encks, W. P. Acc. Chem. Res. 1989, 22, 343-
349.
(4) Kirby, A. J .; J encks, W. P. J . Am. Chem. Soc. 1965, 87, 3209-
3216.
(5) Kirby, A. J .; Varvoglis, A. G. J . Am. Chem. Soc. 1967, 89, 415-
423.
(6) Friedman, J . M.; Freeman, S.; Knowles, J . R. J . Am. Chem. Soc.
1988, 110, 1268-1275.
S0022-3263(98)01160-8 CCC: $15.00 © 1998 American Chemical Society
Published on Web 08/28/1998

Solvent Effects on Phosphate Monoester Solvolysis
J . Org. Chem., Vol. 63, No. 19, 1998 6681
F igu r e 1. The dissociative (A) and the associative (B) mechanistic extremes, and the concerted (C) pathway for phosphoryl
transfer. The concerted pathway is drawn to indicate a dissociative transition state, which is typical for phosphate monoester
reactions. At the bottom is the proposed mechanism for the first step in phosphoryl transfers of monoanions; for less basic -OR
groups such as phenols, the proton transfer to the leaving group may be concerted with P-O bond cleavage.4
monoanion is also believed to proceed by a dissociative
mechanism, with the proton transferred to the leaving
group either in a preequilibrium step as shown in Figure
1 or, for less basic OR groups, simultaneously with P-O
bond cleavage.5 This aqueous reaction also shows a very
small entropy of activation,5 and the reaction of pNPP
in methanol proceeds with inversion of configuration at
phosphorus.7
amyl alcohol and compared these reactions to the aque-
ous ones. This report presents the results of those
studies.
Exp er im en ta l Section
Gen er a l. Solvents were obtained from commercial sources,
and all distillations were done under anhydrous nitrogen. tert-
Butyl alcohol was distilled from NaH, tert-amyl alcohol
(t-AOH) was distilled from Mg, pyridine was distilled from
CaH2, and acetone was dried over molecular sieves and
distilled. Other solvents were used as received. Tetrapheny-
larsonium chloride and sodium tetraphenylborate were ob-
tained from commercial sources. Tetraphenylarsonium chlo-
ride was recrystallized from acetone, while tetraphenylborate
was recrystallized from acetone/toluene.9
Very large rate accelerations are observed in enzyme-
catalyzed reactions of phosphate monoesters. Apart from
the direct contribution from catalytic groups, the mi-
croenvironment in the active site experienced by the
substrate is one possible source of the catalytic efficiency.
One technique which has been used to address the effects
of microenvironment on phosphoryl transfer reactions is
solvent effects on the rate and mechanism of the reaction.
Several effects have been noted to arise in the aqueous
reactions of phosphate monoesters from the addition of
organic cosolvents, or of carrying out reaction in anhy-
drous solutions of alternative phosphoryl acceptors. Abell
and Kirby found that added DMSO or HMPA accelerated
the aqueous hydrolysis of pNPP dianion by up to 106- to
107-fold, by effects that were attributed to decreased
hydrogen bonding of the solvent to the anionic nonbridge
oxygen atoms.8 Knowles et al. found that the stereo-
chemical outcome of phosphoryl transfer from pNPP,
made chiral by the use of oxygen isotopes, changes from
inversion in protic solvents such as methanol to racem-
ization in tert-butyl alcohol.6 No kinetic studies were
reported, but the reported approximate half-life of the
reaction in tert-butyl alcohol was significantly faster than
for the aqueous reaction at the same temperature.
To better understand the origins of these rate accelera-
tions we have conducted kinetic and thermodynamic
studies of the reactions of a series of aryl phosphate
monoesters in anhydrous tert-butyl alcohol and in tert-
Syn th esis of Ar yl P h osp h a tes. Substituted phenyl phos-
phates were synthesized by the method of Bourne and Wil-
1
liams.10 The identities of the products were confirmed by H
and 31P NMR analysis. The phosphate monoesters were
converted to the free acid forms by cation exchange with
Dowex-50X8-100 cation-exchange resin in the proton form
followed by drying in vacuo. However, with para-substituted
halogenated phenyl phosphates, cation exchange was ac-
companied by significant hydrolysis. With these monoesters
the bis-cyclohexylammonium salts in water were made 1 N in
HCl and then extracted three times with diethyl ether. The
ether layers were dried over MgSO4 and filtered, and solvent
was removed under vacuum. This yielded the pure acid form
of the phosphate monoester.
Syn th esis a n d P u r ifica tion of Tetr a bu tyla m m on iu m
a n d Tetr a p h en yla r son iu m Sa lts. The synthesis of tetra-
butylammonium tetrabutylborate is described elsewhere in
detail.11 The product was confirmed by NMR analysis, el-
emental analysis, and melting point.
Tetraphenylarsonium tetraphenylborate was synthesized by
combining aqueous solutions of tetraphenylarsonium chloride
and sodium tetraphenylborate. An aqueous solution was made
of each of these compounds, which when combined produced
a gelatinous suspension of tetraphenylarsonium tetraphenyl-
(7) Buchwald, S. L.; Friedman, J . M.; Knowles, J . R. J . Am. Chem.
Soc. 1984, 106, 4911-4916.
(8) Abell, K. W. Y.; Kirby, A. J . Tetrahedron Lett. 1986, 27, 1085-
1088.
(9) Cox, B. G.; Hedwig, G. R.; Parker, A. J .; Watts, D. W. Aust. J .
Chem. 1974, 27, 477-501.
(10) Bourne, N.; Williams, A. J . Org. Chem. 1984, 49, 1200-1204.
(11) Hoff, R. H.; Hengge, A. C. J . Org. Chem. 1998, 63, 195.

6682 J . Org. Chem., Vol. 63, No. 19, 1998
Hoff and Hengge
borate. This product was filtered, dried in vacuo, and recrys-
tallized from nitromethane. 1H NMR analysis confirmed the
desired product.
sealed into this assembly under nitrogen. Dried t-AOH was
added to the other side of the cell under a blanket of nitrogen,
and the glass electrode and titration buret were introduced,
with the tip of the titration buret immersed in the solution.
Phenol was added to the cell, with constant stirring, and the
cell was allowed to stabilize for at least 10 min. Once the cell
voltage was stable, titration was started. Titrations were
carried at least to the neutralization point, the data were
plotted, and the pKa was calculated by graphic interpolation
of the half-neutralization point. Experiments in t-AOH were
performed at least in duplicate. No effort was made to
evaluate junction potentials in this cell, as these measure-
ments are intended only for comparison to each other. The
stability and reproducibility of the junction potentials was
indicated by the reproducibility of the titration results.
Tetraphenylarsonium chloride is noted in the literature for
problems resulting from the presence of water of hydration.9
It was necessary to obtain a water-free tetraphenylarsonium
halide for evaluation of its solubility in nonaqueous solvent.
Therefore, the anion was exchanged by metathesis in water
with sodium iodide, the iodide salt of the tetraphenylarsonium
ion being much less soluble in water than the chloride salt.
The tetraphenylarsonium iodide was then recrystallized from
dry acetone.
Kin etics Meth od s. The acid form of the substituted phenyl
phosphate monoester was dissolved in dried, freshly distilled
t-AOH in
a dried, screw-top test tube. Typical solution
concentrations were 6-20 mM. This solution was brought to
the appropriate temperature, and a stoichiometric amount of
tetrabutylammonium hydroxide (Aldrich 1.0 M solution in
methanol, stored under nitrogen) was added. To study the
dianion reactions, a slight excess of 2 equiv of base was added,
while to study the monoanion, only 1 equiv was added.
Over time, aliquots were removed and added to measured
portions of a solution of 0.1 N NaOH. The reaction rates for
phosphate monoesters in basic aqueous solution are many
times slower than the rates in t-AOH, so this transfer
effectively stopped the reaction. This solution was analyzed
spectrophotometrically to determine the concentration of free
phenol. The plot of absorbance versus time over the first 1-5%
of reaction was analyzed by assuming first-order kinetics using
the initial rates method. To determine initial substrate
concentration, an aliquot of the reaction mixture was subjected
to complete hydrolysis by alkaline phosphatase in pH 9.0 Tris
buffer, 100 mM, containing 1 mM ZnCl2 and MgCl2, and the
final absorbance of this solution was determined.
One manifestation of the method in this experiment is noted
at the beginning of each titration. When the titration is in its
early stages, the phenol is predominantly present in the
protonated form. Since this is the only electrolyte in the
solution where the indicating electrode is located and the
titrant is not in significant concentration yet, this solution has
almost no charge carriers and the impedance is very high. The
very low current flow produces a regular oscillation in the
voltage reading of about (2 mV, most likely due to the stirring
of the solution. As the titration proceeds, phenol dissociates
and electrolyte concentration rises, corresponding to the
disappearance of this oscillation.
Solu bility Exp er im en ts. Experiments were performed to
evaluate the solubilities of tetrabutylammonium tetrabutyl-
borate, tetraphenylarsonium tetraphenylborate, tetraphenyl-
arsonium iodide, sodium tetraphenylborate, and sodium iodide
in t-AOH and water. We were unable to reliably quantify the
very slight solubilities of the solutes tetrabutylammonium
tetrabutylborate and tetraphenylarsonium tetraphenylborate,
and others have reported similar problems with direct mea-
surement of these solubilities.13 For the more soluble salts, a
series of tubes were prepared with 5.0 mL of freshly dried
t-AOH and sufficient mass of dried solute to exceed saturation
concentration. Saturation concentrations were estimated on
the basis of preliminary experiments of a similar nature. One
set of solutions was stirred at the experimental temperature
representing solutions approaching saturation from under-
saturation conditions, while another set, representing solutions
approaching saturation from super-saturation, was stirred at
elevated temperature for a period of time and then allowed to
cool to the experimental temperature. Samples were analyzed
over time until apparent equilibrium concentrations were
reached. Analysis was spectrophotometric (Ph4AsI at 265 nm,
NaBPh4 at 225 nm) for all but NaI, which was quantified by
an Ag/Cr colorimetric titration.14 Some inconsistencies in data
were observed, and others have reported the decomposition
of the tetraphenyl solutes in alcohols and in water.9,13,15,16 Due
to these problems, the solubility of tetraphenylarsonium
tetraphenylborate was not determined directly in either water
or t-AOH.
For each different substituted phenol, a spectroscopic study
was performed comparing free phenolate to the corresponding
phenyl phosphate. For each, a wavelength was selected for
kinetic analysis such that no correction for the presence of the
aryl phosphate was required.
p Ka Deter m in a tion s. The pKas in tert-butyl alcohol and
t-AOH of the phenols listed in this study were evaluated using
the method of Marple and Fritz12 with an H-cell custom-made
for this work. A drawing of this cell is included in the
Supporting Information. The description below refers to the
procedure using t-AOH; the procedure used for tert-butyl
alcohol was identical, with tert-butyl alcohol replacing t-AOH
in both reference and titration cells. The reference was an
Accumet saturated calomel electrode with a cracked bead
junction. This junction extended into a saturated aqueous KCl
solution. Above this solution was t-AOH, also saturated in
KCl, which was in contact with a frit. On the other side of
this frit was a solution of t-AOH, saturated in tetrabutyl-
ammonium bromide. This solution was in contact with a frit
which joined the reference assembly to the titration cell. In
the titration cell, a glass pH indicating electrode connected to
a pH meter was used to monitor pH and electrode potential
as titrations were performed at 25 °C under nitrogen. Before
use of the cell for any data collection the response of the cell
was tested by titration with p-toluenesulfonic acid monohy-
drate. Over a range of 5 pH units, the response of the cell
was found to be 57 ( 1 mV/pH. Phenols were dissolved in
tert-amyl or tert-butyl alcohol and titrated with tetrabutyl-
ammonium hydroxide (Acros Organics, 0.1 N solution in
toluene/methanol) that was normalized against a standard HCl
solution. The mass of phenol used in each titration experiment
was equivalent to about 0.2 mmol, to allow a titration volume
of about 2 mL.
P a r tition in g Exp er im en ts. Phenyl phosphate was pre-
pared in the bis-cyclohexylammonium salt form as described
above. Conversion of this compound to the tetrabutylammo-
nium form using cation-exchange resin resulted in incomplete
exchange. Therefore the potassium salt was prepared by
treating a 10 mM solution of the free acid form with KOH to
pH 9, followed by cation exchange using Dowex 50X8-100 resin
(13) Fuchs, R.; Bear, J . L.; Rodewald, R. F. J . Am. Chem. Soc. 1969,
91, 5797-5800.
(14) Greenberg, A. E.; Trussel, R. R.; Clesceri, L. S. Selected Physical
and Chemical Standard Methods for Students, Based on Standard
Methods for the Examination of Water and Wastewater, 16th ed.;
Greenberg, A. E., Trussel, R. R., Clesceri, L. S., Eds.; American Public
Health Association: Washington, DC, 1986.
(15) Popovych, O.; Friedman, R. M. J . Phys. Chem. 1966, 70, 1671-
1673.
Between experiments, the H-cell and reference electrode
annular container were cleaned with solvent and distilled
water and oven dried. The dried glassware was assembled
and filled, beginning with the reference cell solutions. Once
solutions were at the proper level, the reference electrode was
(16) Alexander, R.; Parker, A. J .; Sharp, J . H.; Waghorne, W. E. J .
Am. Chem. Soc. 1972, 94, 1148-1158.
(12) Marple, L. W.; Fritz, J . S. Anal. Chem. 1962, 34, 796-800.

Solvent Effects on Phosphate Monoester Solvolysis
J . Org. Chem., Vol. 63, No. 19, 1998 6683
Ta ble 1. Ra tes a n d Activa tion P a r a m eter s for
Rea ction s of p-Nitr op h en ylp h osp h a te Dia n ion a n d
Mon oa n ion
tert-butyl
tert-amyl
aqueous
alchol
alchol
A. Dianion
k at 39 C, s-1
ksolv/kaqueous
∆Hq, kcal/mol
∆Sq, eu
1.6 × 10-8
1.2 × 10-4
1.4 × 10-4
8750
30.9 ( 0.3
+23.0 ( 0.3
23.7 ( 0.3
this work
1
7500
30.6
+3.5
29.5
31.6 ( 1.0
+24.5 ( 0.8
24.0 ( 1
∆Gq, kcal/mol
source
ref 3
this work
B. Monoanion
k at 39 C, s-1
ksolv/kaqueous
∆Hq, kcal/mol
∆Sq, eu
1.1 × 10-6
7.7 × 10-8
6.81 × 10-8
0.06
32.6 ( 1.0
+13.4 ( 0.8
28.4 ( 1.0
this work
1
0.07
25.4
-4.5
26.8
37.1 ( 0.9
+27.7 ( 0.7
28.5 ( 0.9
this work
∆Gq, kcal/mol
source
ref 4
in the tetrabutylammonium form. This yielded a clean solu-
tion of the bis-tetrabutylammonium salt of phenyl phosphate
(8 mM).
Partitioning experiments were performed with the 8 mM
aqueous solution described above at 4 °C in a coldroom. This
solution and the alcohol were brought to the experimental
temperature, and 5.00 mL of the aqueous solution and 5.00
mL of t-AOH were introduced into screw-top tubes with PTFE-
lined caps. The tubes were placed on a shaker for a measured
period of time. To sample the solutions, tubes were removed
from the mixers and centrifuged to speed separation of the
layers. All equipment was maintained at the experimental
temperature, and all handling of samples was performed in a
coldroom. Aliquots were removed by pipet from both the
aqueous and organic layers and introduced into tubes contain-
ing premeasured mixtures for spectrophotometric quantifica-
tion. These mixtures were prepared such that, after addition
of the experimental sample, each tube contained 2.0 mL of 50%
ethanol 0.1 N in NaOH, 0.5 mL of water, and 0.5 mL of t-AOH.
The concentration of phenyl phosphate was determined at 267
nm. In addition, samples were scanned for detection of
phenolate ion at 289 nm, which would indicate hydrolysis of
the phenyl phosphate.
F igu r e 2. Brønsted plot for solvolysis of substituted aryl
phosphates in tert-amyl alcohol: circles represent para-
substituted aryl phosphates, diamonds represent meta-
substituted aryl phosphates. The slope of the best-fit line for
para-substituted compounds is -1.14 ( 0.12 and that for meta-
substituted compounds is -1.11 ( 0.02. The phenyl phosphate
rate constant was used for the slope calculations for both
groups.
Ta ble 2. p Ka Va lu es for Su bstitu ted P h en ols in Aqu eou s
Solu tion , in ter t-Bu tyl Alcoh ol a n d in ter t-Am yl Alcoh ol
pKa
substituent aqueousa tert-amyl alcoholb tert-butyl alcoholb
p-nitro
7.14
7.95
10.0
10.57 (10.4824
12.5
)
)
p-cyano
p-amido
p-bromo
p-phenyl
H
3,5-dinitro
m-nitro
m-chloro
11.83
13.95
15.00
15.75
16.20
9.37
8.5630
9.34
15.35
9.51
10.00
6.7331
8.39
16.82 (16.6424
9.98
13.84
Resu lts
12.97
14.06
Kin etics of Rea ction s in ter t-Bu tyl Alcoh ol a n d
ter t-Am yl Alcoh ol. Values of the first-order rate con-
stants for the reactions of aryl phosphate dianions and
monoanions were determined by spectrophotometrically
following the liberation of the free phenols or phenolate
anions. The rate constants for the reactions of the
monoanion and the dianion of p-nitrophenyl phosphate
in both tert-butyl alcohol and in tert-amyl alcohol gave
good linear Eyring plots (Supporting Information) which
were used to calculate the enthalpies and entropies of
activation, ∆Hq and ∆Sq. The dianion of pNPP was found
to undergo phosphoryl transfer to solvent approximately
9000-fold faster in tert-amyl alcohol than in water, while
the monoanion reacted about 16-fold more slowly in tert-
amyl alcohol. The rate and activation parameter data
together with literature values for the aqueous reactions
are shown in Tables 1a and 1b.
The kinetic data for the reactions of the dianions of a
series of aryl phosphates in tert-amyl alcohol were used
to construct the Brønsted plot in Figure 2. The rates for
meta-substituted and for para-substituted compounds fall
on separate but parallel lines with a slope of -1.1 ( 0.1.
P h en olic p Ka Va lu es in ter t-Am yl Alcoh ol a n d
ter t-Bu tyl Alcoh ol. The pKa values for the substituted
phenols were measured in the anhydrous alcohols as
described in the Experimental Section. The values are
9.02
a
Aqueous values are from the CRC Handbook of Biochemistry
and Molecular Biology, Physical Chemical Data, Part 1 unless
b
otherwise specified. Values in tert-butyl alcohol and tert-amyl
alcohol are from this work unless otherwise specified.
tabulated in Table 2 and are plotted against their
literature aqueous pKa values in Figure 3. The wider
range of pKa values for the phenols in the nonaqueous
solvent as compared to the range observed in water is
consistent with previous observations.17
Solven t P a r tition in g Stu d ies. Partitioning of the
phosphate dianion between water and t-AOH was studied
with the bis-tetrabutylammonium salt of phenyl phos-
phate. Phenyl phosphate was chosen due to its relatively
slow rate of hydrolysis for the dianion in t-AOH. The
experiments were performed at 4 °C in order to further
minimize hydrolysis. The free energy of transfer from
water to t-AOH is calculated by the equation
∆Gtr(water/t-AOH)(bis-tetrabutylammonium
18
phenyl phosphate) ) RT ln(P)
(17) Marple, L.; Fritz, J . S. Anal. Chem. 1963, 35, 1223-1227.
(18) Leo, A.; Hansch, C.; Elkins, D. Chem. Rev. 1971, 71, 525.

6684 J . Org. Chem., Vol. 63, No. 19, 1998
Hoff and Hengge
F igu r e 4. The thermodynamic cycle used to determine the
free energy of solvation of Ph4AsBPh4 in tert-amyl alcohol, with
the concentrations at saturation at 25 °C.
tetrabutylammonium tetrabutylborate or to solve for its
solubility through a thermodynamic cycle. These efforts
were hindered by the extremely low solubility of tetra-
butylammonium tetrabutylborate and the high solubility
of some of the salts in t-AOH which precluded using a
thermodynamic cycle to obtain its free energy of solvation
indirectly. Due to these difficulties, we decided to use a
similar approach with tetraphenylarsonium tetraphen-
ylborate, a more frequently used and widely accepted
subject of studies involving the extrathermodynamic
assumption.
Since the direct determination of the concentration of
Ph4AsBPh4 in t-AOH was not reliable, saturation con-
centrations of Ph4AsI, NaBPh4, and NaI in t-AOH were
measured at 25 °C for use in the thermodynamic cycle
shown in Figure 4. Solubilities and free energies of
solvation in t-AOH are given in Figure 4. From this
result, applying the extrathermodynamic assumption the
free energy of solvation of the Ph4As+ ion in t-AOH )
9.4 kcal/mol.
The next step in this process is to compare the ∆Gs(t-
AOH) to the ∆Gs(water) to calculate the free energy of
transfer of the species from water to t-AOH. The solubil-
ity of Ph4AsBPh4 in water is very low, and in our hands
direct attempts to determine the equilibrium concentra-
tion again proved to be unreliable. The high solubilities
of NaI and NaBPh4 in water mean that activities cannot
be assumed to be approximately equal to concentration,
and so the thermodynamic cycle approach cannot be used.
A literature value for ∆Gs(water) of Ph4AsBPh4 which is
derived from indirect methods is 23.5 kcal/mol9 which
leads to ∆Gs(water)(Ph4As+) ) 11.8 kcal/mol.
F igu r e 3. Comparison of aqueous and nonaqueous pKas of
the substituted phenols from Table 2: substituted phenolic
pKa in tert-butyl alcohol are represented by diamonds, pKas
in tert-amyl alcohol are represented by circles; filled shapes
represent meta-substituted phenols, open shapes represent
para-substituted phenols.
where P is the ratio of solubilities in water and t-AOH.
The value for P was found to be 9.4 ( 0.1. The free
energy of transfer is calculated to be 1.24 ( 0.01 kcal/
mol. This represents a destabilization of the ground state
of the bis-tetrabutylammonium phenyl phosphate in
changing solvent from water to t-AOH.
Sin gle Ion Solva tion Stu d ies of th e Ar yl P h os-
p h a te Dia n ion . To isolate the thermodynamics of the
solvation of single ions, extrathermodynamic assump-
tions are used. These assumptions are based on the
observation that when both the cation and anion of an
ion pair are large, structurally similar, and radially
symmetric, the charge is essentially shielded from the
solvent, and solvent interactions are with a dispersed
charge density on the outer surface of the ion. Therefore,
solvent interactions with both ions are essentially equal,
and any thermodynamic parameters of solvation for the
ion pair can be divided equally between the ions. In some
applications of this extrathermodynamic assumption,
investigators have used compounds with somewhat dis-
similar structures, for example the use of triisoamyl-
butylammonium tetraphenylborate.19 If the extrather-
modynamic assumption can be used to equate the thermo-
dynamic contributions of these two ions, then it is
reasonable for us to assume that the free energy of
transfer from water to t-AOH for the tetraphenylarso-
nium ion is a good approximation for that of the tetra-
butylammonium ion.
The activities of the solutes used in this study were
assumed to be close enough to unity to be disregarded,
given the dilute concentrations at saturation and the
qualitative nature of the conclusions that are drawn. The
free energies of solvation can then be approximated from
the solubility data using the equation ∆Gs ) -RT
ln(Ksp).
The solubilities of tetrabutylammonium tetrabutyl-
borate and salts of its ions were investigated in an
attempt to either directly measure the solubility of
This allows calculation of the free energy of transfer
for the Ph4As+ ion from water to t-AOH according to the
equation
∆Gtr(water/t-AOH)(Ph4As+) )
∆Gs(t-AOH)(Ph4As+) - ∆Gs(water)(Ph4As+)
Using our experimental results and the literature
value for ∆Gs(water)(Ph4AsBPh4), we obtain a ∆Gtr(water/
t-AOH)(Ph4As+) of -2.4 kcal/mol. To a first approxima-
tion this result is reasonable, given the free energies of
transfer from water to various alcohols for the Ph4As+
ion as reported by Marcus:19 -5.8 kcal/mol for methanol,
-5.1 kcal/mol for ethanol, and -6.0 kcal/mol for n-
-
propanol and -3.5 kcal/mol for the BPh4 ion in 2-pro-
(19) Marcus, Y. Pure Appl. Chem. 1983, 55, 977-1021.
panol.

Solvent Effects on Phosphate Monoester Solvolysis
J . Org. Chem., Vol. 63, No. 19, 1998 6685
The experimental cycle to derive the ∆G for transfer
of the phenyl phosphate dianion can be completed by
measuring the partitioning of a bis-tetraphenylarsonium-
substituted phenyl phosphate between water and in
t-AOH. This will allow calculation of the free energy of
transfer for the compound, from which the contributions
of the tetraphenylarsonium ions can be subtracted.
Unfortunately, we were unable to isolate the pure bis-
tetraphenylarsonium salt of phenyl phosphate. As an
approximation, using the extrathermodynamic assump-
tion that the free energies of solvation for tetraphenyl-
arsonium and for tetrabutylammonium are similar, we
can estimate this quantity as follows:
F igu r e 5. Diagram showing the effect of changing the solvent
from water to tert-amyl alcohol on the free energy of activation
and the contributions of ground state versus transition state
effects.
∆Gtr(water/t-AOH)(phenyl phosphate) )
∆Gtr(water/t-AOH)((n-Bu4N+)2 phenyl phosphate) -
2∆Gtr(water/t-AOH)(Ph4As+)
entropic and enthalpic effects in the rate enhancements
observed in tert-butyl alcohol and in t-AOH.
∆Gtr(water/tAOH)(phenyl phosphate) )
Th e Dia n ion Rea ction . Our work reported in this
study was motivated by an interest in analyzing how the
microenvironment can affect the kinetics and mechanism
of phosphoryl transfer reactions. An essential difference
in the phosphoryl transfer from the dianion of pNPP to
anhydrous tert-butyl alcohol, where this acceptor also is
the reaction solvent, is the observation of racemization
at phosphorus when the reactant p-nitrophenyl phos-
phate was made chiral by the use of oxygen isotopes.6
This outcome was interpreted to indicate the formation
of a free metaphosphate intermediate. This outcome was
in contrast to the observation in more nucleophilic
solvents such as methanol, where the product had
inverted configuration consistent with nucleophilic par-
ticipation in the transition state of a concerted mecha-
nism.7 Because we were interested in performing solvent
partitioning studies in order to determine differences in
solvation free energies, we also examined the reaction
in tert-amyl alcohol, which is structurally very similar
but is much less miscible with water. Since this alcohol
is, if anything, even more hindered as a nucleophile than
tert-butyl alcohol, we assume that the stereochemical
course of the phosphoryl transfer reaction in these two
alcohols is the same. The very similar kinetics and
activation parameters measured for the two reactions
supports this assumption.
The reduced activation barrier in the tertiary alcohols
can occur by raising the free energy of (destabilizing) the
reactant, by lowering the free energy of the transition
state, or a combination of both. The portion of the more
favorable entropy of activation from reduced solvent
reorganization could arise from less effective solvation
(increased free energy) of the substrate or from stabiliza-
tion (reduced free energy) of the transition state, relative
to the aqueous reaction. To determine the contribution
of differences in the free energies of solvation (∆Gs) of
the reactant, we compared the free energies of the
tetrabutylammonium salt of p-nitrophenyl phosphate in
water and in tert-amyl alcohol. The difference in the
solvation free energies in the two solvents, shown as ∆Gtr-
(water/t-AOH) in Figure 5, represents the free energy of
transfer of the substrate from water to the alcohol. The
solvent partitioning experiments thus indicate that the
overall reduction in ∆Gq of 5.8 kcal/mol in the dianion
reaction can be attributed to approximately 1 kcal/mol
of ground state destabilization, which presumably reflects
1.24 kcal/mol - 2(-2.4 kcal/mol) ) 6 kcal/mol
This demonstrates a significant destabilization of the
phenyl phosphate dianion in t-AOH relative to water,
presumably arising from less efficient solvation by
t-AOH.
Discu ssion
Phosphate monoesters exhibit bell-shaped pH-rate
profiles with an optimum pH of around 4, with a lower,
pH-independent rate at higher pH values, indicating that
the monoanionic species is more reactive than the dian-
ion. Nevertheless, the less reactive dianion is typically
the substrate for phosphatases, and these enzymes
exhibit catalytic efficiencies (defined as the ratio of kcat
to the uncatalyzed rate constant) of 1010 or more. The
large rate accelerations provided by phosphatases have
led to proposals that, while dissociative transition states
occur in uncatalyzed reactions, enzymatic phosphoryl
transfer from monoesters may proceed by a more as-
sociative mechanism. Since all known phosphatases bear
positively charged residues and/or metal ions at their
active sites, it is postulated that this charge may elec-
trostatically function in the same way as alkylation does
in causing a mechanistic shift; that is, by neutralizing
the negative charge on the substrate and enhancing the
electrophilicity of the phosphorus atom toward attack.
This notion is intuitively attractive since phosphate
triesters are considerably more reactive than their mo-
noester counterparts. Another motivating factor has
been a perceived difficulty in rationalizing how an
enzyme can stabilize a dissociative transition state.20
Methods by which an enzyme could stabilize a dissocia-
tive phosphoryl transfer have been considered.21
This descriptive term can be misleading however; the
aqueous hydrolysis reference reaction for phosphatases,
while having a dissociative transition state, is nonethe-
less ANDN. Entropy effects are therefore still important,
and overcoming entropic barriers is a potential source of
catalytic efficiency. Despite this, entropic contributions
have been omitted from many discussions of the catalytic
efficiencies of phosphatases. A discussion follows of
(20) Hasset, A.; Blattler, W.; Knowles, J . R. Biochemistry 1982, 21,
6335-6340.
(21) Admiraal, S. J .; Herschlag, D. Chem. Biol. 1995, 2, 729-739.

6686 J . Org. Chem., Vol. 63, No. 19, 1998
Hoff and Hengge
less effective solvation of the substrate in tert-amyl
alcohol, with the balance being a transition state effect.
The change to a two-step mechanism in the tertiary
alcohols offers one explanation for the faster rate, since
the entropy of activation should be more favorable for a
reaction involving rate-limiting unimolecular dissociation
than for one in which nucleophilic participation by a
second molecule occurs. This is revealed in the more
favorable ∆Sq for the reactions in tert-butyl alcohol and
tert-amyl alcohol compared to that of the aqueous reac-
tion. Another possible energetic source for the enhanced
rate is the weaker hydrogen-bonding ability of these
solvents compared to that of water. The driving force
for the dissociative reaction (whether in a fully two-step
dissociative mechanism or in a concerted mechanism with
a dissociative transition state) can be rationalized as
coming from the internal electron donation of the nega-
tive charges on the nonbridge oxygen atoms. Indeed, the
loss of these charges by alkylation deactivates this
pathway and leads to mechanisms with greater nucleo-
philic participation. If these charges are less involved
in stabilizing interactions such as hydrogen bonds with
the solvent, they ought to be more capable of acting as
internal nucleophiles in expelling the leaving group. This
effect has been postulated to explain the acceleration of
the hydrolysis of p-nitrophenyl phosphate dianion in
aqueous DMSO or HMPA.8 In the reactions examined
in this study in tert-butyl alcohol and tert-amyl alcohol,
the tetra-N-butylammonium counterion is also unable to
interact effectively with these nonbridge oxygen atoms,
due to the steric shielding of the positive charge by the
bulky n-butyl groups. The single ion ∆Gtr of 6 kcal/mol
measured for transfer of the phenyl phosphate dianion
from water to t-AOH confirms that this ion is consider-
ably less solvated in the latter solvent. An increased
hydrolysis rate due to the loss of stabilizing hydrogen-
bonding interactions with the solvent should be analo-
gous to the rate accelerations observed in the alkaline
hydrolysis of esters caused by added aprotic cosolvents;
the enthalpy of activation for the saponification of ethyl
acetate and ethyl benzoate are reduced by 4 and 6 kcal/
mol, respectively, in aqueous DMSO versus aqueous
ethanol.22 The increased rates and lower ∆Hq were
attributed to decreased solvation of hydroxide ion. While
these are bimolecular reactions and the phosphoryl
transfers in the tertiary alcohols are unimolecular, the
rate-limiting steps share the features of an anionic
species providing the driving force for the reaction, with
negative charge becoming more dispersed in the transi-
tion state. In the ester saponification reactions a larger
destabilization of the reactants relative to the transition
state resulted in a significant net reduction in ∆Hq.22
Thus the expectation is reasonable that rate enhance-
ment caused by desolvation of the dianionic phosphate
reactant should show up as a decrease in ∆Hq.
oxygens may not be a major factor in the rate accelera-
tions in these alcohols and that any such destabilization
is evidently felt equally in both the ground state and
transition state. A possible alternative explanation for
the similar enthalpies of activation is a coincidental
compensation for the loss of a stabilizing bonding inter-
action in the transition state with the nucleophile
(present in the aqueous concerted reaction but absent in
the stepwise reaction in t-AOH) by a more favorable ∆Hq
for the unimolecular dissociation in t-AOH compared with
the same hypothetical process in water. Given a similar
degree of bond cleavage to the leaving group in the
transition state, the unimolecular dissociation should
have a higher ∆Hq than the concerted reaction since in
the latter case the enthalpic cost of bond cleavage is
partly offset by formation of the new bond. Though bond
formation to the nucleophile in the transition state is
small in the aqueous reaction, there should be some
enthalpic effect in lowering ∆Hq which will not be present
in the reaction in t-AOH. It is possible that solvation
differences lower the enthalpic barrier for unimolecular
dissociation in t-AOH by an amount similar to the lost
bonding interactions, leaving an unchanged overall ∆Hq.
Though perhaps unlikely, this possibility cannot be ruled
out.
There are two sources for the more favorable entropy
of activation. One is the aforementioned change from a
bimolecular concerted reaction in water to a mechanism
with rate-limiting unimolecular dissociation. Even though
bonding between the phosphorus atom and the nucleo-
phile is thought to be small in the transition state of the
aqueous reaction, a solvent molecule must still be re-
cruited to assume an apical position in the trigonal
bipyramidal transition state, with an associated loss of
entropy. In addition, water is a more highly structured
solvent than the tertiary alcohols. Solvent reorganization
must occur in the transition state, driven by the change
in geometry and in charge distribution. The overall
charge of the transition state is the same as that of the
ground state; however, since the size of the structure
increases and charge is dispersed, a larger solvation shell
is required in the transition state with an accompanying
loss in entropy. This solvent effect should be more
pronounced in the more highly structured solvent water
than in the tertiary alcohols. The latter solvents are not
without structure, however, and so the potential entropic
acceleration obtainable from eliminating solvent reorga-
nization is probably not fully realized in these solvents.
Although a nucleophilic role is eliminated in the rate-
determining step in the reactions in tert-butyl alcohol and
tert-amyl alcohol, isotope effect data23 indicate that this
occurs with no change in the transition state structure
of the pNPP reactant. The 18O isotope effects in the
nonbridging oxygen atoms and in the bridging oxygen
atom and the 15N isotope effects for the reaction of
p-nitrophenyl phosphate in tert-butyl alcohol are very
similar to those for the aqueous reaction.23 That the
transition state structure within the reactant is un-
changed in the DN + AN reaction compared with the ANDN
one is not surprising in view of the very small degree of
bond order to the nucleophile in the transition state of
the aqueous reaction. Evidently the loss of this interac-
tion is not accompanied by significant changes in the
Surprisingly, the activation parameters (Table 1a) for
the reaction of the dianion of pNPP reveal that the rate
enhancements for the reaction in tert-butyl alcohol and
in tert-amyl alcohol are due entirely to a more favorable
entropy of activation. The activation enthalpy is un-
changed within experimental error from the value in the
aqueous reaction. This indicates that the loss of stabiliz-
ing hydrogen-bonding interactions with the nonbridge
(22) Haberfield, P.; Friedman, J .; Pinkston, M. F. J . Am. Chem. Soc.
1972, 94, 71-74.
(23) Hengge, A. C.; Edens, W. A.; Elsing, H. J . Am. Chem. Soc. 1994,
116, 5045-5049.

Solvent Effects on Phosphate Monoester Solvolysis
J . Org. Chem., Vol. 63, No. 19, 1998 6687
degree of bond cleavage to the leaving group or in the
transition state structure of the phosphoryl group. Thus
the entropic accelerations observed do not result from
making the reaction mechanism more dissociative with
respect to the degree of transition state bond cleavage to
the leaving group, but arise solely from eliminating the
need for nucleophilic participation and from possible
contributions from solvent reorganization.
A fair question is whether the rate enhancement
observed with pNPP also occurs with less activated
phosphate esters. The rates were therefore measured for
the phosphoryl transfer reaction with a range of substi-
tuted phenyl phosphates, and the results were plotted
versus the aqueous pKa values for the leaving groups in
Figure 2. Although the rates for meta- and para-
substituted aryl phosphates fall on separate but parallel
lines, the Brønsted slopes of -1.1 are very similar to the
Brønsted slope of -1.25 reported for the aqueous reaction.
Thus similar rate enhancements are seen with all of the
aryl phosphates tested.
The occurrence of separate lines for meta- and para-
substituted aryl phosphates contrasts with the aqueous
reactions where the hydrolysis rates of meta- and para-
substituted phenyl phosphates fall on the same line in
the analogous Brønsted plot, both for the reaction of the
dianion and for that of the monoanion.5 We initially
hypothesized that the reason might be differential changes
in the pKa values of meta-substituted phenols compared
to para-substituted ones in tert-amyl alcohol. If this were
the case then if the rate constants were plotted against
the true pKa values in the reaction solvent, the meta- and
para-substituted compounds should fall on the same line
as they do in the aqueous reaction.
We measured the pKa values of the phenols in anhy-
drous tert-butyl alcohol and in tert-amyl alcohol using an
electrochemical method24 described in the Experimental
Section. The results are given in Table 2. The measured
pKa values in tert-amyl alcohol are higher than those in
water and cover a larger span compared to the aqueous
values. Both of these effects are expected from the
results of previous work on nonaqueous pKa values of
phenols.17 While there is always some uncertainty in the
assignment of absolute pKa values in nonaqueous sol-
vents, insofar as they can be directly compared with the
aqueous values, the method can be counted upon to give
precise measurements of the relative nonaqueous pKa
values of the phenols. The pKa values determined in tert-
amyl alcohol are plotted against the aqueous values in
Figure 3, which reveals a linear correlation between the
two values. Meta- and para-substituted phenols fall on
the same correlation line, indicating that there is no
anomalous perturbation of pKa values in the alcohol by
meta versus para substituents. Thus the hypothesis that
meta versus para substitution causes a differential
perturbation in the leaving group pKa values is incorrect.
We are presently uncertain as to the source of the
difference in rates between these two classes of com-
pounds in tert-amyl alcohol.
The monoanions react more slowly in the alcohols than
in aqueous solution, in contrast to the behavior of the
dianion. In both alcohols, a more favorable entropy of
activation is more than offset by an increased enthalpy
of activation.
This increase in ∆Hq most likely arises from greater
difficulty in transferring the proton from a nonbridge
oxygen atom of the reactant to the bridging oxygen atom.
In the alcohol solvents the pKa value of the phosphate
group will be higher, making proton removal more
energetically difficult. In addition, in water this proton
transfer could well occur via an intervening water
molecule; this would allow a six-membered transition
state for the proton transfer process, which is sterically
preferable to the four-membered one which would be
operative in its absence. The steric bulk of the tertiary
alcohols makes them less able to similarly assist in the
proton transfer.
Stereochemical studies of the monoanion reaction in
tert-butyl alcohol have not been carried out. Therefore
the more favorable entropies of activation cannot with
assurance be ascribed to a change in the molecularity of
these reactions, though it is likely. In addition, the
reason for the less favorable entropy of activation in the
reaction of the monoanion in tert-amyl alcohol compared
to tert-butyl alcohol is uncertain. In both reactions the
predominant reason for the slower rates is a higher
enthalpic barrier relative to the aqueous reaction.
Sign ifican ce for En tr opic Con sider ation s in P h os-
p h a ta se Rea ction s. The results presented here have
significance for the enzymatic reactions of phosphatases.
Enzymes confer significant entropic advantage relative
to uncatalyzed reactions in solution. It has been esti-
mated that 108 M is the approximate maximum value
for the advantage of an enzymatic reaction compared
with a corresponding bimolecular reaction from entropic
advantages, in reactions in which the transition state is
relatively tight so that a large loss of entropy is required
for its formation.25 The catalytic benefit of induced
intramolecularity in a dissociative phosphoryl transfer
reaction has previously been pointed out, and a catalytic
enhancement of approximately 102-fold was attributed
to the positioning of a metal-bound hydroxide ion and
phosphoryl group by Mg2+ ions in a phosphorylated
4-morpholinopyridine/Mg2+ complex.26 In general, how-
ever, entropy considerations are often omitted in assess-
ments of the sources for the catalytic efficiencies of
phosphatases and in considerations of how these enzymes
might catalyze a reaction having a dissociative transition
state. This may be because dissociative transition states
such as those operative in phosphate monoester reactions
may seem like unlikely candidates for potential accelera-
tion from entropic effects.
The term “dissociative”, while generally an accurate
description of the aqueous reaction, may convey the
impression that there is no nucleophilic involvement in
the transition state. However, the aqueous reference
reactions are concerted, with nucleophilic participation
in the transition state; even the weak bond formation to
the nucleophile in this reaction nonetheless requires its
proper positioning in the trigonal bipyramidal transition
state. The removal of the associated entropic barrier
must be considered as one source of the enzymatic
Th e Mon oa n ion Rea ction . The thermodynamic and
kinetic data for the monoanion reaction are collected in
Table 1b. In contrast to the situation in aqueous solution
where the monoanions of phosphate monoesters are the
more reactive species, in the tertiary alcohols the monoan-
ion reaction is 2000-fold slower than that of the dianion.
(25) J encks, W. P. Adv. Enzymol. 1975, 43, 219-410.
(26) Herschlag, D.; J encks, W. P. Biochemistry 1990, 29, 5172-5179.
(24) Fritz, J . S.; Marple, L. W. Anal. Chem. 1962, 34, 921-924.

6688 J . Org. Chem., Vol. 63, No. 19, 1998
Hoff and Hengge
efficiencies of phosphatases. In the ground states of a
phosphatase-substrate complex, a nucleophile is already
preassociated and the hydration shell has been removed,
eliminating the need for solvent reorganization. The rate
enhancement observed in these studies of nearly 104 is
probably a lower limit for the rate enhancement from
entropic considerations that are enjoyed by phosphatases.
This is not to say that these enzymatic reactions proceed
by a metaphosphate intermediate as in the tertiary
alcohols. Quite the contrary, with a preassociated nu-
cleophile in place a free metaphosphate intermediate is
unlikely. However the entropic advantage enjoyed in the
enzymatic reaction is the same as that gained from
eliminating the need to recruit a disordered solvent
molecule to serve the nucleophilic role in the tertiary
alcohol reactions. The factor of 104 is a significant
fraction of the overall catalytic rate enhancement achieved
by some phosphatases.
tive mechanism were operative, since nucleophilic par-
ticipation is still important in the transition state, no
significant change in the entropy of activation should be
seen. If the phosphorane mechanism were operative,
then the entropy of activation should be large and
negative, as it is for reactions of phosphate triesters
which have such associative transition states. Typical
reactions of this type exhibit entropies of activation of
-35 eu,29 dramatically different from the experimental
value reported here. The reported results are in best
agreement with the originally reported explanation for
racemization, namely, a change to a dissociative SN1-type
of mechanism.
Ack n ow led gm en t. The authors wish to thank Pro-
fessor G. M. Swain for advice in conducting the elec-
trochemical pKa determinations and Professor V. D.
Parker for discussions regarding the free energy of
solvation experiments. Financial support of this work
came from NIH Grant GM47297 to A.C.H. and from the
U.S. Army Advanced Civil Schooling Program for sup-
port of R.H.H.
Mech a n istic Con clu sion s
After the report of racemization in the reaction of
p-nitrophenyl phosphate in tert-butyl alcohol,6 alternative
explanations have been suggested that avoid the forma-
tion of metaphosphate. One is the possibility that a
bridge-protonated tert-butyl phosphate is initially formed
which undergoes reversible and rapid phosphoryl trans-
fer with other solvent molecules before loss of the proton
which is necessary to form a stable product.27 It has also
been postulated that an associative mechanism occurs,
forming a pentacoordinate intermediate which may
undergo pseudorotations, leading to racemized product.28
Both of these alternative explanations are unlikely in
view of the significantly more positive entropy of activa-
tion in the tert-butyl alcohol reaction. If the first alterna-
Su p p or tin g In for m a tion Ava ila ble: Supporting Infor-
mation Available: Eyring plots for the reactions of the
monoanion and the dianion of p-nitrophenyl phosphate in tert-
butyl alcohol and in tert-amyl alcohol and a drawing of the
electrochemical cell used in the determinations of the phenolic
pKa values in tert-butyl alcohol and in t-AOH (3 pages). This
material is contained in libraries on microfiche, immediately
follows this article in the microfilm version of the journal, and
can be ordered from the ACS; see any current masthead page
for ordering information.
J O981160K
(29) Khan, S. A.; Kirby, A. J . J . Chem. Soc. B 1970, 1172-1182.
(30) Caldwell, S. R.; Newcomb, J . R.; Schlecht, K. A.; Raushel, F.
M. Biochemistry 1991, 30, 7438.
(27) Herschlag, D.; J encks, W. P. J . Am. Chem. Soc. 1989, 111,
7579-7586.
(28) Florian, J .; Warshel, A. J . Phys. Chem. B 1998, 102.
(31) Dean, J . A. Lange’s Handbook of Chemistry, 1962.