Characterization of transition states in dichloro(1,4,7-triazacyclononane)copper(II)-catalyzed activated phosphate diester hydrolysis

Article abstract of DOI:10.1021/ja952306p

The reaction mechanism for Cu[9]aneN₃CI₂-catalyzed hydrolysis of ethyl 4-nitrophenyl phosphate was probed using kinetic isotope effects and isotope exchange experiments. The solvent deuterium isotope effect ((D)k = 1.14), combined with the absence of ¹⁸O incorporation into 4-nitrophenol, suggests that hydrolysis proceeds through intramolecular attack of the metal-coordinated hydroxide at the phosphorus center. The secondary ¹⁵N isotope effect (¹⁵k = 1.0013 ± 0.0002) implies that loss of the leaving group occurs at the rate-limiting step with approximately 50% bond cleavage in the transition state. This study is one of the first applications of the secondary ¹⁵N isotope effect to simple metal-promoted hydrolysis reactions, and the result is consistent with concerted bond formation and cleavage. A mechanism consistent with the isotope studies is presented.

Full text of DOI:10.1021/ja952306p

J. Am. Chem. Soc. 1996, 118, 1713-1718
1713
Characterization of Transition States in
Dichloro(1,4,7-triazacyclononane)copper(II)-Catalyzed Activated
Phosphate Diester Hydrolysis
Kim A. Deal,†,§ Alvan C. Hengge, and Judith N. Burstyn*,†
‡
Contribution from the Department of Chemistry, 1101 UniVersity AVenue,
UniVersity of Wisconsin, Madison, Wisconsin 53706, and Institute for Enzyme Research,
1
710 UniVersity AVenue, UniVersity of Wisconsin, Madison, Wisconsin 53705
X
ReceiVed July 13, 1995
Abstract: The reaction mechanism for Cu[9]aneN3Cl2-catalyzed hydrolysis of ethyl 4-nitrophenyl phosphate was
D
probed using kinetic isotope effects and isotope exchange experiments. The solvent deuterium isotope effect ( k )
1
.14), combined with the absence of 18O incorporation into 4-nitrophenol, suggests that hydrolysis proceeds through
15
intramolecular attack of the metal-coordinated hydroxide at the phosphorus center. The secondary N isotope effect
1
5
(
k ) 1.0013 ( 0.0002) implies that loss of the leaving group occurs at the rate-limiting step with approximately
1
5
5
0% bond cleavage in the transition state. This study is one of the first applications of the secondary N isotope
effect to simple metal-promoted hydrolysis reactions, and the result is consistent with concerted bond formation and
cleavage. A mechanism consistent with the isotope studies is presented.
Metal ions are essential cofactors in the reactions of many
nucleases, and yet the precise role of the metal ion in the
hydrolytic mechanism is unclear.1 Similarly, many other
biochemically important phosphoryl transfer reactions involve
the action of metalloenzymes, frequently containing labile metal
ions such as zinc(II) or magnesium(II).2 The mechanism of
metal complex-promoted phosphate diester hydrolysis has been
explored in detail for substitutionally inert metal complexes of
via a pentacovalent phosphorane intermediate or via a concerted
reaction with a phosphorane-like transition state structure. Metal
complex-promoted hydrolysis of phosphate monoester mono-
anions and all phosphate diesters has been proposed to occur
through the same SN2-type mechanism as the alkaline hydrolysis
7
of phosphate diesters. A metal-bound hydroxide nucleophile
has been proposed to attack the phosphorus center to form a
pentacovalent phosphorane intermediate followed by rate-
3
8
Ir(III), but similarly detailed work has not been reported for
limiting loss of the leaving group, but evidence for a phos-
labile metal systems involving Cu(II). There are several features
of the reaction mechanism for labile metal-promoted phosphate
diester hydrolysis which remain unclear, including the source
of the nucleophile and the type of intermediate or transition
state structure.
phorane intermediate has been limited. Indirect evidence for
such an intermediate has been obtained from labeling studies
involving cobalt(III)-promoted hydrolysis of a bound phosphate
9
monoester monoanion, and a phosphorane intermediate has
3,9
been invoked in the proposed mechanism for most inert and
4
,5
labile metal complexes. The generally proposed mechanism
for phosphate diester hydrolysis has been challenged recently
by studies of alkaline hydrolysis of activated phosphate diesters.
The method of heavy atom isotope effects suggests that a
The nucleophile has been clearly identified for Ir(III)-
promoted phosphate diester hydrolysis.3 Through labeling
studies, it was shown that phosphate diester hydrolysis occurred
via an intramolecular attack of a metal-coordinated hydroxide
at the phosphorus center. A similar mechanism has been
proposed for labile metal complex-promoted phosphate diester
10
concerted mechanism is operative. Hydrolysis of an activated
phosphodiester by the zinc(II)-dependent enzyme phosphodi-
esterase I proceeded with approximately 60% bond cleavage
in the transition state, clearly implicating a concerted reaction
hydrolysis.4
,5
Although intramolecular attack is the most
commonly hypothesized mechanism, it is also possible that the
metal-coordinated hydroxide could act as a general base to
activate a solvent water molecule for attack at the phosphorus
1
1
path.
Although the authors attribute the efficacy of this
enzyme to protonation of the phosphoryl group in the transition
state, the role of the zinc ion in the catalysis could scarcely be
assessed because of the paucity of relevant mechanistic data
for metal-promoted hydrolyses.
6
center.
At present, there are few experimental data available to
distinguish between metal-promoted phosphodiester hydrolysis
Heavy atom isotope effects have been used for many years
1
2
†
‡
§
to study enzymatic reaction mechanisms.
Recently, the
Department of Chemistry.
Institute for Enzyme Research.
In partial fulfillment of the Ph.D. degree in Chemistry, Ph.D. Disserta-
15
secondary N isotope effect has been shown to provide a
measure of the amount of bond cleavage in the transition state
tion, December 1993.
*
Author to whom correspondence should be addressed.
Abstract published in AdVance ACS Abstracts, January 15, 1996.
(7) Hendry, P.; Sargeson, A. M. In Progress in Inorganic Chemistry;
Lippard, S. J., Ed.; John Wiley & Sons: New York, 1990; Vol. 38, pp
201-258.
X
(1) Serpersu, E. H.; Shortle, D.; Mildvan, A. S. Biochemistry 1987, 26,
1
2
289-1300.
(8) Chin, J. Acc. Chem. Res. 1991, 24, 145-152.
(9) Jones, D. R.; Lindoy, L. F.; Sargeson, A. M. J. Am. Chem. Soc. 1983,
105, 7327-7336.
(
(
2) Mildvan, A. S.; Grisham, C. M. Struct. Bonding 1974, 20, 1-21.
3) Hendry, P.; Sargeson, A. M. J. Am. Chem. Soc. 1989, 111, 2521-
527.
(10) Hengge, A. C.; Cleland, W. W. J. Am. Chem. Soc. 1990, 112, 7421-
7422.
(
(
(
4) Morrow, J. R.; Trogler, W. C. Inorg. Chem. 1988, 27, 3387-3394.
5) Koike, T.; Kimura, E. J. Am. Chem. Soc. 1991, 113, 8935-8941.
6) AdVances in Physical Organic Chemistry; Gold, V., Ed.; Academic
(11) Hengge, A. C.; Cleland, W. W. J. Am. Chem. Soc. 1991, 113, 5835-
5841.
Press: New York, 1967.
(12) O’Leary, M. H. Ann. ReV. Biochem. 1989, 58, 377-401.
0
002-7863/96/1518-1713$12.00/0 © 1996 American Chemical Society

1
714 J. Am. Chem. Soc., Vol. 118, No. 7, 1996
Deal et al.
1
0,13
for phosphoryl transfer reactions.
Measurement of the
The solvent deuterium isotope effect for the hydrolysis of 2 by 1
1
5
was determined. Reactions were performed in H O or 99.99% D O at
secondary N isotope effect on the alkaline hydrolysis of 3,3-
dimethylbutyl 4-nitrophenyl phosphate revealed about 50% bond
cleavage to the 4-nitrophenol leaving group in the transition
2
2
pH > 9.0, corresponding to the pH-independent region of the pH vs
14
rate profile, in order to eliminate any rate effect due to differences
between pH and pD.16 As a control, the reaction was run in CHES
10
state, corresponding to a concerted mechanism. Application
of the secondary 15N isotope effect to metal complex-promoted
hydrolysis of activated phosphate diesters will allow for the
determination of whether the loss of the leaving group is rate
limiting and, if so, will provide an estimate of the percentage
of bond cleavage in the transition state. This information will
improve our understanding of metal-mediated phosphodiester
hydrolysis.
D
buffer at pH ( 0.5 from the calculated pD, and the k was determined
to be constant. In a typical experiment, freshly prepared 1 stock (11
2 2
mM) in either H O or D O was added to a reaction mixture containing
5
0 mM CHES buffer (pH or pD > 9.1) in the same solvent and heated
to 50 °C in the spectrophotometer. A reference cuvette, identical in
all respects except lacking 1, was similarly prepared. Freshly prepared
2 stock (51 mM) in the appropriate solvent was added to the reaction
mixtures, and the mixtures were allowed to equilibrate for 5 min. The
reactions in D
2 2
O and H O were carried out in parallel under identical
conditions. The total reaction volume was 3.00 mL, the final
concentration of 1 was 1.17 mM, the final concentration of 2 was 2.57
4
mM, and the ionic strength was adjusted to 0.1 M with 5 M NaClO .
We have conducted detailed mechanistic studies of labile
metal complex-promoted activated phosphate diester hydrolysis.
Previously, we described the hydrolytic activity of Cu[9]aneN3-
14
1
5
Cl2 (1), an effective catalyst for phosphate diester hydrolysis.
Isotope Incorporation Determination. The 18O isotope incorpora-
In order to probe the mechanism in greater detail, a series of
isotope effect studies have been performed. To obtain informa-
tion on the source of the nucleophile, the solvent deuterium
isotope effect on the rate of 1-catalyzed phosphate diester
hydrolysis has been determined. To assist in the determination
of the rate-limiting step as well as to distinguish between a
pentacovalent phosphorane intermediate and a concerted mech-
tion in the hydrolysis of 2 by the catalyst derived from 1 was measured.
18
2
The reaction mixture consisted of 50 mM 2, 10 mM 1, and 30% OH ,
and the pH was held constant at 7.2 with 0.5 M HEPES. The solution
was heated at 70 °C in a heat block for 3 weeks, at which time the
reaction had reached about 40% completion, as determined by the
amount of 4-nitrophenolate formed. The 4-nitrophenol was isolated
by acidification followed by ether extraction. The ether layer was dried
over magnesium sulfate, and the ether was removed by rotary
anism, the secondary 15N isotope effect for the hydrolysis of
ethyl 4-nitrophenyl phosphate (2) by 1 has been measured. These
studies have enabled us to develop a more complete understand-
ing of the mechanism of metal-promoted hydrolysis.
evaporation. The resulting yellow residue of 4-nitrophenol was
1
dissolved in 400 µL of D
The D
-nitrophenol was submitted for high-resolution mass spectrometry to
2
O, and its identity was confirmed by H NMR.
2
O was removed by rotary evaporation, and the sample of
4
18
measure O incorporation.
Secondary 15N Isotope Effect Determination. The secondary 15
N
isotope effect for the hydrolysis of 2 by 1 was determined with the
natural abundance of nitrogen in the substrate by the method of Hengge
and Cleland.10 For these experiments, reactions involved isolation and
measurement of the 15N/ N ratio for 4-nitrophenol as described.
14
17
Specifically, the reaction mixture contained 5 mM 2, 1.25 mM 1, and
0 mL of 0.1 M HEPES buffer, pH adjusted to 7.2. The reaction
2
Experimental Methods
mixtures were heated and stirred at 70 °C in an oil bath. After the
reaction reached 40-60% completion, the reaction was stopped and
the amount of 4-nitrophenol produced was determined. The reaction
mixture was acidified with 1 M HCl and the 4-nitrophenol extracted
three times with distilled diethyl ether. The ether extracts were dried
over magnesium sulfate and concentrated to dryness. The aqueous
layer, containing the unreacted substrate as well as 1, was subjected to
rotary evaporation to remove dissolved ether. To prevent interference
with the 15N measurement, 1 was removed by passage through a 5 mL
Materials. The biological buffers HEPES (N-(2-hydroxyethyl)-
piperazine-N′-ethanesulfonic acid) and CHES (2-(N-cyclohexylamino)-
ethanesulfonic acid) were purchased from Sigma Chemical. Chelex
resin (sodium form) was also purchased from Sigma Chemical. The
18
isotopically labeled compounds D
2
O (99.99+%) and OH (60-65%)
2
were purchased from Cambridge Isotope Laboratories. All chemicals
were used without further purification. Anhydrous analytical grade
diethyl ether (Mallinckrodt) was distilled at 34-35 °C to remove higher
boiling impurities shortly before use. Sodium ethyl 4-nitrophenyl
phosphate and dichloro(1,4,7-triazacyclononane)copper(II) were pre-
Chelex column, and the column was washed with 20-25 mL of H O
2
to ensure recovery of all the unreacted substrate. The eluent was
hydrolyzed completely by the addition of enough NaOH to make a 1
M solution, followed by heating at 95 °C. Complete hydrolysis was
ensured by heating the basic solution for at least 30 h, which is 10
times the half-life for base hydrolysis at this pH and temperature. The
4-nitrophenol produced was isolated as before.
1
4
pared as described. All aqueous solutions were prepared with water
purified by passage through a Millipore purification system. The
1
4
instrumentation was as described.
Kinetic Procedures. The initial rate of production of 4-nitropheno-
late was monitored spectrophotometrically at 400 nm as described.14
All experiments were run in triplicate, and the data reported represent
the average of these experiments with less than 5% deviation among
the measurements.
The effect of temperature on the hydrolysis of 2 by the catalyst
derived from 1 was determined over the temperature range 30-90 °C.
Reactions were performed in 50 mM CHES buffer, the pH of the buffer
was adjusted to 9.1 at each reaction temperature, and the reaction
mixture contained 1.00 mM 1 and 10.0 mM 2. The ionic strength was
The purification and preparation of the 4-nitrophenol samples for
17
combustion was performed by the method of Cleland. Briefly, the
4
9
-nitrophenol residues were purified by sublimation under vacuum at
5 °C, and the sublimed product was transferred to a quartz tube by
rinsing with distilled diethyl ether. The ether was removed from the
tubes under vacuum, and the tubes were prepared for combustion by
the addition of a layer of CuO, a layer of copper metal, and a thin
piece of silver foil. The samples were combusted at 850 °C for 2 h,
then at 550 °C for 8 h before cooling to room temperature.
0
.1 M and was controlled with 5 M NaClO
4
. Reactions were prepared
in parallel to insure identical conditions of pH, concentration, and ionic
strength and were pre-equilibrated at the desired temperatures. Four
different temperatures were used: in one experiment, the actual
temperatures used were 32.0, 50.2, 65.3, and 84.0 °C. The measured
temperatures are accurate to (0.5 °C. To maintain a constant
temperature at 32 °C, an external ice bath was used to cool the water
circulator.
After combustion, the nitrogen gas was separated from the water
vapor and carbon dioxide on a high-vacuum line by selective trapping
of the latter two gases with dry ice/alcohol and liquid nitrogen traps,
respectively. The nitrogen gas was deposited onto molecular sieves
cooled with liquid nitrogen and then released from the molecular sieves
by heating to 200 °C for 15 min before transfer to the isotope ratio
mass spectrometer where the isotopic composition of the N
2
gas was
(
(
(
13) Cleland, W. W. FASEB J. 1990, 4, 2899-2905.
14) Deal, K. A.; Burstyn, J. N. Inorg. Chem., in press.
15) Burstyn, J. N.; Deal, K. A. Inorg. Chem. 1993, 32, 3585-3586.
(16) Glasoe, P. K.; Long, F. A. J. Phys. Chem. 1960, 64, 188-190.
(17) Cleland, W. W. Methods Enzymol. 1995, 249, 341-373.

Metal Complex-Promoted Reactions of Phosphate Esters
J. Am. Chem. Soc., Vol. 118, No. 7, 1996 1715
∆Gq b
Table 1. Activation Parameters for Phosphate Ester Hydrolysisa
q
q
∆
H
∆S
-1
(kJ mol-1)
(J mol K-1
)
(kJ mol-1)
k
calcd × 10 s
c
6
-1
ref
nucleophile
9]Cu-OH+
substrate
[
(
(
2
2
85
69
71
80
-79
-93
108
97
90
0.58
70
900
d
3
9
2
+
en)
2
Ir-OH
Co-OH
2
+
en)
2
NPP
BDNPP
-65
-
OH
-107
111
0.15
24
a
Abbreviations: [9],1,4,7-triazacyclononane; en, ethylenediamine; NPP, 4-nitrophenyl phosphate; BDNPP, bis(2,4-dinitrophenyl) phosphate.
b
G ) ∆H - T∆S , calculated at 25 °C. Rate constants calculated at 25 °C for given activation parameters. d Activation parameters calculated
q
q
q
c
∆
from kobsd ) ν/[LCu], where [LCu] is defined as previously.14
determined. The hydrolysis of 2 by 1 was run in two triplicate sets,
and the resulting values were averaged.
Solvent Deuterium Isotope Effect Analysis. The solvent
D 20
kinetic isotope effect k was determined in order to distinguish
between a nucleophilic reaction path and a general base reaction
path. A general base mechanism is usually associated with a
solvent deuterium isotope effect of >2, while a nucleophilic
mechanism is usually associated with an isotope effect ranging
To determine the isotopic composition of the starting material, 50
µmol of 2 was placed in 1 M NaOH and heated at 95 °C for 30 h to
ensure complete hydrolysis. The 4-nitrophenol was extracted and
converted to nitrogen gas and the isotopic composition determined as
described above (Vide supra). The isotopic composition of the starting
material was also determined by direct combustion of 5 mg of solid 2.
The isotopic ratios were identical within experimental error, and the
average was used.
6
from 0.8 to 1.5. The solvent isotope effect observed for
D
1
-catalyzed hydrolysis of 2 is k ) 1.14, for reactions carried
out under otherwise identical conditions. This value is consis-
tent with the intramolecular attack of a coordinated nucleophile
where proton transfer does not occur in the rate-determining
step. Table 2 lists published deuterium solvent isotope effects
for various phosphate ester hydrolyses for comparison with this
measured value.
The isotope effects were calculated using the isotopic ratio obtained
from the isotope ratio mass spectrometer. The three sets of isotope
ratios obtained were the product at partial reaction (R
substrate (R ), and the starting material (R ). Measuring the isotopic
p
), the residual
s
o
ratio of both the product at partial completion and the residual substrate
allowed two independent calculations of the isotope effect for each
18O Isotope Incorporation Determination of Bond Cleav-
age Site. Nucleophilic attack at the phosphorus center is the
most common mode of phosphate diester cleavage, but nucleo-
philic attack at the aromatic carbon is also possible. For
example, the alkaline hydrolysis of methyl 2,4-dinitrophenyl
phosphate occurred by a mixed mechanism where both C-O
18
p o s o
experiment, one involving R and R and one using R and R .
Results
Computation of Activation Parameters. To determine the
effect of 1 on the magnitude of the activation barrier for the
hydrolysis of 2, the activation parameters were measured in the
2
1
and P-O bond cleavages were observed. For the hydrolysis
of 2 by 1, mass spectral analysis of the 4-nitrophenol indicated
1
8
18
no incorporation of O. The upper limit of O incorporated
into the sample was 2%, well within experimental error.
pH-independent region of the previously determined pH vs rate
profile.14 The initial rate data were analyzed by use of the
Secondary N Isotope Effect Analysis. The secondary 15N
isotope effect can provide a direct measure of the transition state
bond cleavage of phosphate esters with 4-nitrophenol leaving
groups. The effect arises from the resonance stabilization of
the partial negative charge resulting from partial cleavage of
the bond to the leaving group, through contribution from the
quinoid resonance structure, 4. Nitrogen bonds to oxygen are
15
modified Arrhenius equation. The plot of ln initial rate versus
q
-1
q
1
-
/T was linear, yielding ∆H ) 85 ( 1.7 kJ mol and ∆S )
-
1
-1
143 ( 5 J mol K . These values were determined over a
1
0
limited temperature range; the error in the experimental
determinations is small, but extrapolation beyond this temper-
ature range may not be valid.
In order to compare the measured activation parameters to
reported values, the parameters were calculated from the first-
and second-order rate constants. The rate constants were
calculated by applying the experimental reaction orders previ-
ously found for 1 (half-order) and 2 (first order) to the
experimentally measured initial rate.14 The magnitude of ∆H
q
q
is independent of reaction order, but ∆S is dependent on the
1
9
reaction order. If the initial rates of reaction are converted to
q
kobsd by removing the 1 dependence, then ∆H ) 85 ( 1 kJ
-
1
q
-1
-1
mol and ∆S ) -79 ( 1 J mol K . Since the reaction
was catalytic in 1 and there was a 10-fold excess of 2 to 1, the
concentration of 1 is negligible; therefore, this calculation gives
q
more stiffening than nitrogen bonds to carbon in terms of
a reasonable value for ∆S . If the initial rates of reaction are
15
vibrational frequencies, and the N label will enrich in the more
stiffly bonded neutral species. A normal isotope effect (value
greater than 1.0000) is therefore expected for the formation of
the 4-nitrophenolate anion.
converted to second-order rate constants by removing both the
q
-1
q
1
-
and 2 dependencies, then ∆H ) 85 kJ mol and ∆S )
-
1
-1
40 J mol K .
Relatively few workers have reported activation parameters
The isotope effect was calculated from the ratio of masses
obtained from the isotope ratio mass spectrometer by the
for metal complex-promoted phosphate diester hydrolysis. A
compilation of the reported values for metal complex-promoted
phosphate diester, metal complex-promoted phosphate mono-
ester, and alkaline phosphate diester hydrolyses is given in Table
1
8
literature method. A measure of the precision of the experi-
ment can be obtained by comparing the isotope effect calculated
from Rp and Ro to the isotope effect calculated from Rs and Ro.
1.
(
20) Northrop, D. B. In Isotope Effects on Enzyme-Catalyzed Reactions;
(
(
18) O’Leary, M. H. Methods Enzymol. 1980, 64, 83-104.
Cleland, W. W., O’Leary, M. H., Northrop, D. B., Eds.; University Park
Press: Baltimore, MD, 1977; p 122.
(21) Kirby, A. J.; Younas, M. J. J. Chem. Soc. B 1970, 1165-1172.
19) InVestigations of Rates and Mechanisms of Reaction, 4th ed.;
Bernasconi, C. F.; Ed.; John Wiley & Sons: New York, 1986; Part I.

1
716 J. Am. Chem. Soc., Vol. 118, No. 7, 1996
Deal et al.
Table 2. aSolvent Deuterium Isotope Effect for Phosphate Ester
equilibrium, the protonation constant for the coordinated water,
substrate exchange on the catalyst, and the hydrolysis reaction.
The forward and backward rates in the monomer-dimer
equilibrium, as well as the substrate exchange and the depro-
tonation steps, are expected to be fast, largely influenced by
rapid proton transfer and ligand exchange rates. Ligand
Hydrolysis
nucleophile
substrate
Dkb
mechanismc
ref
+
1.14d
1.55
1.20
1.0
[
9]Cu-OH
2
n
?
n
n
n
this work
24
9
21
21
-
OH
(
pyridine
BDNPP
NPP
MDNPP
MDNPP
en)
2
Co-OH2
+
8
exchange on Cu(II) typically occurs with rate constants of 10
acetate ion
1.1
-1 22
s
.
Exchange rates and activation parameters for solvent
a
2+
Abbreviations: [9],1,4,7-triazacyclononane; en, ethylenediamine;
NPP, 4-nitrophenyl phosphate; BDNPP, bis(2,4-dinitrophenyl) phos-
exchange on Cu(MeOH)6 have been reported; an exchange
7
-1
q
rate of kexch ) 3 × 10 s (at 25 °C) and a ∆H ) 17 kJ/mol
b
phate; MDNPP, methyl 2,4-dinitrophenyl phosphate. The isotope
23
were measured for this system.
For the purposes of the
2
0
effect notation used is that of Northrop, in which isotope effects are
H
D
following qualitative comparisons, we have assumed that since
loss of the leaving group is rate limiting in hydrolysis of 2 by
1, the measured temperature dependence is largely due to this
step in the reaction mechanism.
shown by a leading superscript. The kinetic isotope effect k /k is
D
c
written as k, where D corresponds to deuterium. Mechanism assigned
by criteria described in the text: n, nucleophilic; ?, not conclusive.
d
Ratio of ν
H
D
/ν measured under identical conditions of metal complex
concentration, substrate concentration, temperature, and pH.
It is of interest to determine the extent to which the presence
of 1 decreases the free energy barrier to hydrolysis of 2. To
directly compare the activation energy for 1 hydrolysis of 2 to
that for alkaline hydrolysis of 2, it is necessary to estimate the
energy barrier for alkaline hydrolysis of 2. The rate of this
reaction is sufficiently slow that temperature dependence studies
are not feasible; the data available in the literature is for the
Table 3. Comparison of the Secondary 15N Isotope Effect for
Hydrolysis of Alkyl 4-Nitrophenyl Phosphate by a Series of
a
Nucleophiles
approx %
hydrolytic systemb
15kc
bond cleavaged
+
[
9]Cu-OH
1.0013 ( 0.0002
1.0016 ( 0.0001
1.0009 ( 0.0002
1.0013 ( 0.0001
1.0017 ( 0.0002
46
57
32
47
61
-
24
OH
H
alkaline hydrolysis of bis(2,4-dinitrophenyl) phosphate. In
+
general, the rate constant for the hydrolysis of activated
â-cyclodextrin
phosphodiesterase I
phosphate diesters increases with a more acidic pKa of the
2
1,24
q
leaving group;
therefore, the ∆G for alkaline hydrolysis
a
Reported errors are standard errors. b Data for comparison with
11 c
of bis(2,4-dinitrophenyl) phosphate is expected to be smaller
than the activation barrier for alkaline hydrolysis of 2, since
2,4-dinitrophenol is more acidic than 4-nitrophenol. The rate
+
[
9]Cu-OH are from Hengge and Cleland. The isotope effect notation
used is that of Northrop, in which isotope effects are shown by a
leading superscript. The kinetic isotope effect k /k is written as k,
where 15 denotes N. The percentage of bond cleavage in the
transition state is calculated relative to a standard reference according
20
1
4
15
15
-
9 -1
1
5
d
constant for alkaline hydrolysis of 2 was estimated as 10
s
3
q
at 25 °C by Hendry and Sargeson, and therefore, ∆G is
expected to be 10 kJ mol higher than that for bis(2,4-
dinitrophenyl) phosphate. From this data we estimate ∆G to
18
to the method of O’Leary. The use of this standard and the calculation
-1
1
5
method assume a value of k ) 1.0028 to represent complete bond
cleavage in the transition state (see text) and assume a linear relation
between isotope effect and the fraction of bond cleavage.
q
-
1
be 120 kJ mol for the alkaline hydrolysis of 2.
The ∆Gq estimated for the alkaline hydrolysis of 2 is
compared to the hydrolysis of 2 by 1 or (en)2Ir-OH . The
presence of the catalyst, 1, causes an estimated decrease in ∆G
of 12 kJ mol relative to the estimated ∆G for alklaine
hydrolysis of 2. The corresponding increase in the rate constant
2+
The two values are expected to be the same, and the values
q
reported here are consistent with one another to within
-
1
q
experimental error. The overall isotope effect of 15k ) 1.0013
(
0.0002 was obtained by averaging the isotope effects
calculated for both Rs, and Rp for a total of six experiments. A
2
is calculated to be 10 , approximately the same magnitude as
1
4
comparison with secondary 15N isotope effects for phosphate
diester hydrolyses in several other systems is given in Table 3.
the rate increase that was observed previously for the
hydrolysis of bis(4-nitrophenyl) phosphate by 1 when compared
to the alkaline hydrolysis of the same substrate at comparable
concentrations of 1 and hydroxide. A decrease in ∆G of 22
kJ mol is noted for (en)2Ir-OH -promoted hydrolysis of 2
compared to the alkaline hydrolysis of 2. Although the decrease
q
Discussion
-
1
2+
Effect of Metal Complexes on the Activation Parameters.
There has been no attempt in the literature to interpret the
activation parameters for metal complex-promoted phosphate
diester hydrolysis, and the relatively few parameters that have
been reported are compiled in Table 1. The complexity of
hydrolysis reactions combined with the large number of species
in solution prohibit extensive interpretation; however, several
generalizations are noted. There is a considerable range of
q
in ∆G is consistent with the preorganization of the metal-bound
substrate, the entropic contribution is about the same as that
found in 1 hydrolysis of 2. Preorganization of the nucleophile
and substrate in the metal complex is expected to result in a
q
less negative ∆S ; however, no difference in the entropic effect
q
is observed. Interestingly, a substantial ∆H decrease is
2+
observed for (en)2Ir-OH -promoted hydrolysis of 2 compared
q
q
values observed for both ∆H and ∆S but no trend is evident.
to 1 hydrolysis of 2. The 3+ charge on the metal as well as
the preorganization of the complex in an orientation favorable
for nucleophilic attack presumably contributes to lower the
q
In all cases, ∆S is less than zero as expected for bringing
together multiple reactive species in solution. The free energy
q
barriers, ∆G for various phosphate diester hydrolyses have been
q
2+
overall activation enthalpy. The ∆G for (en)2Ir-OH
-
q
q
calculated from the reported ∆H and ∆S values and are listed
-1
q
promoted hydrolysis of 2 is 10 kJ mol less than the ∆G for
q
in Table 1. From the ratio of ∆G s for the hydrolysis of 2 by
1
-catalyzed hydrolysis of the same substrate, suggesting that
2
+
q
1
5
or (en)2Ir-OH , it is noted that an increase in ∆G of about
3+
the overall thermodynamic contribution of the Ir(en)2 complex
is substantial. The lower activation barrier corresponds to a 2
orders of magnitude increase in the rate constant of Ir(III)
-
1
kJ mol results in a 10-fold decrease in the rate constant.
q
The presence of a metal complex lowered ∆G for phosphate
diester hydrolysis, as would be predicted for any compound
which promotes a reaction.
(22) Wilkins, R. G. Kinetics and Mechanism of Reaction of Transition
Metal Complexes, 2nd ed.; VCH: Weinheim, Germany, 1991.
In the case of hydrolysis of 2 by 1, it should be noted that
the initial rates measured reflect the composite effect of
temperature on four distinct processes: the monomer-dimer
(
23) Helm, L.; Lincoln, S. F.; Merbach, A. E.; Zbinden, D. Inorg. Chem.
1
986, 25, 2550-2552.
(24) Kirby, A. J.; Younas, M. J. J. Chem. Soc. B 1970, 510-513.

Metal Complex-Promoted Reactions of Phosphate Esters
J. Am. Chem. Soc., Vol. 118, No. 7, 1996 1717
compared to the rate constant of 1. Assuming that the lowering
of the activation barrier can be assigned to the combination of
preorganization of the metal-substrate complex and more
effective charge neutralization, approximately 2 orders of
Position of Bond Cleavage. While the solvent deuterium
isotope effect indicates that the reaction occurs by nucleophilic
attack of the coordinated hydroxide, the solvent isotope effect
does not distinguish between attack at phosphorus and attack
at the aromatic carbon. Although attack at the aromatic carbon
appears unlikely, nucleophilic aromatic substitution at a ring
carbon has been observed in the hydrolysis of other aryl
-
7
-1
magnitude of the 6 × 10
-catalyzed hydrolysis of 2 can be assigned to the equilibrium
between the catalyst and the phosphate diester substrate. The
s
rate constant measured for
1
-
5
-1
21
remaining value of 6 × 10
s
can be attributed to the
phosphate esters. To gain information on the site of bond
hydrolysis of the phosphate diester.
cleavage, the hydrolysis of 2 by 1 was run in the presence of
1
8
Intramolecular versus Intermolecular Nucleophilic Attack.
In general, intramolecular nucleophilic attack has been associ-
ated with large rate accelerations due to the increased order of
OH2. If the coordinated hydroxide attacks at the phosphorus
1
8
center, then the O label will not be incorporated into the
4-nitrophenol product. On the other hand, coordinated hydrox-
ide attack at the aromatic carbon will result in O label
25
18
the system. Intramolecular attack of a coordinated hydroxide
at the phosphorus center has been conclusively demonstrated
for phosphate diester hydrolysis by substitutionally inert metal
complexes of Co(III) and Ir(III).3,9 A similar mechanism has
incorporation into 4-nitrophenol. The insignificant quantity of
1
8
O detected by mass spectral analysis of the 4-nitrophenol
produced by 1-catalyzed hydrolysis of 2 indicates that the
reaction proceeded essentially completely through attack at the
phosphorus with the corresponding P-O bond cleavage, with
negligible or no competing C-O bond cleavage.
2+
been proposed for Cu(2,2′-bipyridine) -catalyzed hydrolysis
of 2, but no direct evidence was reported.4
A classic method for probing the difference between a
nucleophilic mechanism and a general base mechanism is to
measure the solvent deuterium isotope effect.6 For the reaction
of 1 with 2, the two possible reaction pathways are either direct
nucleophilic attack by the coordinated hydroxide at the phos-
phorus or nucleophilic attack at the phosphorus by an activated
water molecule. Since no transferable proton will be involved
in the rate-limiting step for the former mechanism, a normal
solvent deuterium isotope effect will be observed only for the
1
5
Secondary N Isotope Effect. The generally postulated
mechanism for activated phosphate diester hydrolysis by a metal
complex involves the formation of a pentacovalent phosphorane
intermediate followed by the rate-limiting loss of the leaving
group. Although attempts have been made to detect the
3,9
phosphorane intermediate, all have been unsuccessful. Instead
of directly detecting the intermediate, the reaction mechanism
15
can be probed at the rate-limiting step with the secondary
isotope effect.
N
D
10,26
latter case. The measured k of 1.14 for 1-catalyzed hydrolysis
of 2 (Table 2) implies that no proton transfer is involved in the
rate-limiting step. This observation suggests that the mechanism
involves intramolecular nucleophilic attack by a copper(II)-
coordinated hydroxide. An alternative explanation, that an
inverse isotope effect on a different step in the reaction
mechanism is obscuring a large normal isotope effect, can be
ruled out by the following considerations. As mentioned
previously, exchange rates on Cu(II) are much more rapid than
hydrolysis; therefore, kinetic isotope effects on processes
dominated by ligand exchange, such as the monomer-dimer
equilibrium, will not be measured in this experiment. The same
argument holds true for proton transfers. Any equilibrium
solvent isotope effect on the protonation constant for the metal
complex is eliminated by carrying out the reactions well above
the pKa of the coordinated water. An equilibrium solvent
isotope effect on the monomer-dimer equilibrium is possible,
although the monomer-dimer interconversion most reasonably
occurs via ligand dissociation or aquation and not proton
transfer.
Since an isotope effect will only be observable to the extent
that the isotopically sensitive step is rate limiting, the presence
1
5
or absence of an observed N isotope effect will give
mechanistic information. For metal complex-promoted phos-
phate diester hydrolysis with a good leaving group like
4-nitrophenol, the mechanism can be associative or concerted.
In an associative mechanism, the formation of a pentacovalent
phosphorane intermediate is rate limiting, followed by the rapid
loss of the activated leaving group. Since the loss of 4-nitro-
phenol will not be involved in the rate-limiting step, no
1
5
secondary N isotope effect is expected. A concerted reaction
mechanism, where the bond between the phosphate and the
4-nitrophenol is breaking at the same time as the bond between
the nucleophile and the phosphate is forming, will yield a normal
1
5
secondary N isotope effect proportional to the extent of bond
cleavage in the transition state. A dissociative mechanism is
considered unlikely for phosphodiester hydrolysis, even in the
1
1
presence of an activated leaving group.
Previous work on the alkaline hydrolysis of 4-nitrophenyl
phosphate suggests that the reaction is concerted with a
The solvent deuterium isotope effect strongly implicates an
intramolecular reaction, with attack of the coordinated hydroxide
on the phosphorus atom. Such isotope effects have not been
measured previously for labile metal complex-promoted phos-
phate diester hydrolyses. Solvent deuterium isotope effects that
have been measured for organic nucleophiles used to promote
phosphate ester hydrolysis indicate an intramolecular nucleo-
2
7,28
15
dissociative transition state.
The secondary N isotope
effect for the alkaline hydrolysis of 4-nitrophenyl phosphate 15
k
) (1.0028) is the largest observed, supporting the earlier
conclusion that bond cleavage to the leaving group is far
10
15
advanced in the transition state. The maximum k, represent-
ing complete bond cleavage in a dissociative mechanism, is
unknown, but the large value observed for alkaline hydrolysis
2
1
philic mechanism. The solvent isotope effect for hydrolysis
of the coordinated 4-nitrophenyl phosphate in the metal complex
of 4-nitrophenyl phosphate has been used as a standard,
+
10
Co(en)2(4-nitrophenyl phosphate)(OH) has been measured and
representing virtual total bond cleavage.
This reference
9
was also assigned to an intramolecular nucleophilic mechanism.
reaction is a phosphate monoester hydrolysis, and while it is
not ideal for comparison with phosphate diester hydrolyses, it
is the best that is available. Calculating the percentage of bond
cleavage at the transition state relative to the alkaline hydrolysis
of 4-nitrophenyl phosphate therefore provides a lower limit on
Interestingly, the alkaline hydrolysis of bis(2,4-dinitrophenyl)
phosphate gave a solvent deuterium isotope effect of 1.55, a
24
value which the authors assigned to a general base mechanism,
although this value is lower than that usually found for general
base reactions.6 A solvent deuterium isotope effect in the range
(
26) Hengge, A. C.; Tobin, A. E.; Cleland, W. W. J. Am. Chem. Soc.
1
.5-2.0 is generally considered inconclusive for either type of
1
995, 117, 5919-5926.
mechanism.
(27) Kirby, A. J.; Jencks, W. P. J. Am. Chem. Soc. 1965, 87, 3209-
3
216.
(25) Menger, F. M. Acc. Chem. Res. 1985, 18, 128-134.
(28) Westheimer, F. H. Chem. ReV. 1981, 81, 313-326.

1
718 J. Am. Chem. Soc., Vol. 118, No. 7, 1996
Deal et al.
Scheme 1
Scheme 2
the extent of bond cleavage.10,11 The calculation assumes a
linear relationship between isotope effect and the fraction of
bond cleavage; this model provides the best current method to
evaluate the data, albeit in the absence of experimental valida-
tion. For 1-catalyzed hydrolysis of 2, the secondary 15N isotope
effect was measured as 1.0013 ( 0.0002. The presence of a
normal isotope effect confirms that the loss of the leaving group
occurs at the rate-limiting step. In the transition state for the
hydrolysis of 2 by 1, the leaving group bond cleavage is about
D would be observed, and therefore, no 15N isotope effect (15k
) 1) in the leaving group is expected since the bond to the
leaving group has not broken. The experimentally observed
1
5
magnitude of k thus rules out this associative mechanism. An
alternative direct route from A to D via concerted deprotonation
D
and nucleophilic attack predicts a significant value for k and
15
50%, and this value corresponds best to a concerted mechanism,
a k of unity and is therefore also ruled out by the experimental
as discussed in detail below.
data.
Mechanistic Analysis. The isotope effect studies discussed
1
4
above, combined with the kinetic data reported previously,
Conclusion
enable us to evaluate the detailed mechanism for 1-catalyzed
hydrolysis of activated phosphate diesters. The mechanistic
possibilities, illustrated in Schemes 1 and 2, are now discussed
in detail. The catalytically active copper(II) hydroxide complex
binds directly to the phosphate diester to form A, positioning
the coordinated hydroxide for intramolecular nucleophilic attack
at the phosphorus center. One possibility is that species A may
react via nucleophilic attack simultaneous with leaving group
departure in a concerted mechanism, followed by rapid depro-
tonation to give E. This mechanism, which we believe to be
operative, is shown in Scheme 1. There is no proton in flight
in the rate-determining step, consistent with the near unity value
Isotope effect studies have provided detailed mechanistic
information for the hydrolysis of 2 by 1 that could not be
obtained from basic kinetic studies. This work demonstrates
1
5
the utility of secondary N isotope effects for studying metal
complex-promoted reactions of phosphate esters with a 4-nitro-
phenol leaving group. From the data reported above, 1-cata-
lyzed hydrolysis of activated phosphate diesters is shown to
proceed through a concerted reaction mechanism. Further data
for labile and inert metal complex-promoted hydrolysis of
activated phosphate diesters will need to be collected to
determine if the proposed mechanism holds true for other metal-
promoted hydrolyses.
D
of k observed. Partial cleavage of the bond to the leaving group
1
5
is in agreement with the observed moderate value for k.
An alternative possibility is an associative mechanism via a
phosphorane intermediate. This mechanism can be ruled out
by careful consideration of the relative rates of competing
Acknowledgment. The authors wish to acknowledge I.
Treichel for her expert technical assistance and W. W. Cleland
for his invaluable support. K.A.D. acknowledges a predoctoral
fellowship from the Department of Education. The support of
the Wisconsin Alumni Research Foundation is gratefully
acknowledged. Financial support for the isotope ratio mass
spectroscopy facility was provided to W. W. Cleland by NIH
D
15
processes and the observed kinetic isotope effects, k and k.
The species initially formed upon intramolecular nucleophilic
attack of the metal-bound hydroxide is shown as B in Scheme
2
. The hypothetical phosphorane B could partition either of
two ways depending on the relative rates of proton loss from
the bridging oxygen atom and competing loss of p-nitro-
phenolate. Given the very low pKa of the proton in B and the
alkaline conditions of the reaction, proton loss from B will be
extremely rapid. In contrast, loss of p-nitrophenolate is known
from the isotope effect data to be the rate-limiting step in the
overall mechanism. From these relative rate considerations,
B-D-E is clearly the most likely reaction path. The formation
of intermediate D will be rapid and essentially irreversible under
the alkaline reaction conditions, and D will therefore partition
completely forward by loss of p-nitrophenolate to give E. Under
these circumstances, only isotope effects on the formation of
(
GM18938).
Supporting Information Available: Tables of initial rates
for the temperature-dependent hydrolysis of 2 by 1 and of
calculated isotope effect values and an Arrhenius plot (4 pages).
This material is contained in many libraries on microfiche,
immediately follows this article in the microfilm version of the
journal, can be ordered from the ACS, and can be downloaded
from the Internet; see any current masthead page for ordering
information and Internet access instructions.
JA952306P