The dinuclear Zn(II) complex catalyzed cyclization of a series of 2-hydroxypropyl aryl phosphate RNA models: Progressive change in mechanism from rate-limiting P-O bond cleavage to substrate binding

Article abstract of DOI:10.1021/ja076847d

A methoxide-bridged dinuclear Zn(II) complex of 1,3-[N,N′-bis(1,5,9- triazacyclododecane)]propane (1-Zn(II)₂:(^-OCH ₃)) was prepared, and its catalysis of the cyclization of a series of 2-hydroxypropyl aryl phosphates (4a-g) was investigated in methanol at _s^spH 9.8, T = 25°C by stopped-flow spectrophotometry. An X-ray diffraction structure of the hydroxide analogue of 1-Zn(II) ₂:(^-OCH₃), namely 1-Zn(II)₂:( ^-OH), reveals that each of the Zn(II) ions is coordinated by the three N's of the triazacyclododecane units and a bridging hydroxide. The cyclizations of substrates 4a-g reveal a progressive change in the observed kinetics from Michaelis-Menten saturation kinetics for the poorer substrates (4-OCH₃ (4g); 4-H (4f); 3-OCH₃ (4e); 4-Cl (4d); 3-NO ₂, (4c)) to second-order kinetics (linear in 1-Zn(II) ₂;(^-OCH₃)) for the better substrates (4-NO ₂,3-CH₃ (4b); 4-NO₂, (4a)). The data are analyzed in terms of a multistep process whereby a first formed complex rearranges to a reactive complex with a doubly activated phosphate coordinated to both metal ions. The kinetic behavior of the series is analyzed in terms of change in rate-limiting step for the catalyzed reaction whereby the rate-limiting step for the poorer substrates (4g-c) is the chemical step of cyclization of the substrate, while for the better substrates (4b,a) the rate-limiting step is binding. The catalysis of the cyclization of these substrates is extremely efficient. The k_cat/K_M values for the catalyzed reactions range from 2.75 × 10⁵ to 2.3 × 10⁴ M^-1 s^-1, providing an acceleration of 1 × 10⁸ to 4 × 10⁹ relative to the methoxide reaction (k₂^OCH3, which ranges from 2.6 × 10 ^-3 to 5.9 × 10^-6 M^-1 s^-1 for 4a-g). At a _s^spH of 9.8 where the catalyst is maximally active, the acceleration for the substrates ranges from (1-4) × 10 ¹² relative to the background reaction at the same _s ^spH, Detailed energetics calculations show that the transition state for the catalyzed reaction comprising 1-Zn(II)₂, methoxide, and 4 is stabilized by about -21 to -23 kcal/mol relative to the transition state for the methoxide reaction. The pronounced catalytic activity is attributed to a synergism between a positively charged catalyst that has high affinity for the substrate and for the transition state for cyclization, and a medium effect involving a reduced polarity/dielectric constant that complements a reaction where an oppositely charged reactant and catalyst experience charge dispersal in the transition state.

Full text of DOI:10.1021/ja076847d

Published on Web 11/30/2007
The Dinuclear Zn(II) Complex Catalyzed Cyclization of a
Series of 2-Hydroxypropyl Aryl Phosphate RNA Models:
Progressive Change in Mechanism from Rate-Limiting P-O
Bond Cleavage to Substrate Binding
Shannon E. Bunn, C. Tony Liu, Zhong-Lin Lu,† Alexei A. Neverov, and
R. Stan Brown*
Contribution from the Department of Chemistry, Queen’s UniVersity,
Kingston, Ontario, Canada K7L 3N6
Received September 10, 2007; E-mail: rsbrown@chem.queensu.ca
Abstract: A methoxide-bridged dinuclear Zn(II) complex of 1,3-[N,N′-bis(1,5,9-triazacyclododecane)]propane
(1-Zn(II)2:(-OCH3)) was prepared, and its catalysis of the cyclization of a series of 2-hydroxypropyl aryl
phosphates (4a-g) was investigated in methanol at sspH 9.8, T ) 25 °C by stopped-flow spectrophotom-
etry. An X-ray diffraction structure of the hydroxide analogue of 1-Zn(II)2:(-OCH3), namely 1-Zn(II)2:(-OH),
reveals that each of the Zn(II) ions is coordinated by the three N’s of the triazacyclododecane units and a
bridging hydroxide. The cyclizations of substrates 4a-g reveal a progressive change in the observed kinetics
from Michaelis-Menten saturation kinetics for the poorer substrates (4-OCH3 (4g); 4-H (4f); 3-OCH3 (4e);
4-Cl (4d); 3-NO2, (4c)) to second-order kinetics (linear in 1-Zn(II)2:(-OCH3)) for the better substrates (4-
NO2,3-CH3 (4b); 4-NO2, (4a)). The data are analyzed in terms of a multistep process whereby a first formed
complex rearranges to a reactive complex with a doubly activated phosphate coordinated to both metal
ions. The kinetic behavior of the series is analyzed in terms of change in rate-limiting step for the catalyzed
reaction whereby the rate-limiting step for the poorer substrates (4g-c) is the chemical step of cyclization
of the substrate, while for the better substrates (4b,a) the rate-limiting step is binding. The catalysis of the
cyclization of these substrates is extremely efficient. The kcat/KM values for the catalyzed reactions range
from 2.75 × 105 to 2.3 × 104 M-1 s-1, providing an acceleration of 1 × 108 to 4 × 109 relative to the
methoxide reaction (k2OCH3, which ranges from 2.6 × 10-3 to 5.9 × 10-6 M-1 s-1 for 4a-g). At a sspH of 9.8
where the catalyst is maximally active, the acceleration for the substrates ranges from (1 - 4) × 1012
relative to the background reaction at the same sspH. Detailed energetics calculations show that the
transition state for the catalyzed reaction comprising 1-Zn(II)2, methoxide, and 4 is stabilized by about -21
to -23 kcal/mol relative to the transition state for the methoxide reaction. The pronounced catalytic activity
is attributed to a synergism between a positively charged catalyst that has high affinity for the substrate
and for the transition state for cyclization, and a medium effect involving a reduced polarity/dielectric constant
that complements a reaction where an oppositely charged reactant and catalyst experience charge dispersal
in the transition state.
1. Introduction
systems comprising mono- and dinuclear Zn(II)- and Cu(II)-
containing complexes of 1,5,9-triazacyclododecane and 1,3-bis-
N1-(1,5,9-triazacyclododecyl)propane (1) in methanol3 and
Zn(II) in ethanol4 where the reduced polarity media help to
promote the cleavage of a well-studied RNA model, namely
2-hydroxypropyl p-nitrophenyl phosphate (2). The most notable
aspect of this work relates to the very effective catalysis achieved
for the cleavage of 2 promoted by 1-Zn(II)2 in the presence of
one equivalent of methoxide in methanol (which produces the
Intense effort has been devoted to the study of model systems
for the cleavage of phosphodiesters mediated by simple mono-
and dinuclear metal ion complexes.1 Much of this work has
been driven by the fact that phosphodiesters are extremely stable
entities well suited for the storage of genetic information in RNA
and DNA, and yet Nature has provided Zn2+-containing
enzymes that accelerate their cleavage by impressive factors of
up to 1016-fold.2 Our own work involved investigation of
† Present address: Beijing Normal University, College of Chemistry,
Beijing, China 100875. E-mail: luzl@bnu.edu.cn.
(1) (a) Mancin, F.; Tecillia, P. New J. Chem., 2007, 31, 800. (b) Weston, J.
Chem. ReV. 2005, 105, 2151. (c) Molenveld, P.; Engbertsen, J. F. J.;
Reinhoudt, D. N. Chem. Soc. ReV. 2000, 29, 75. (d) Williams, N. H.;
Takasaki, B.; Wall, M.; Chin, J. Acc. Chem. Res. 1999, 32, 485. (e) Mancin,
F.; Scrimin, P.; Tecilla, P.; Tonellato, U. Chem. Commun. 2005, 2540. (f)
Morrow, J. R.; Iranzo, O. Curr. Opin. Chem. Biol. 2004, 8, 192.
(2) (a) Cowan, J. A. Chem. ReV. 1998, 98, 1067. (b) Wilcox D. E. Chem. ReV.
1996, 96, 2435. (c) Stra¨ter, N.; Lipscomb, W. N.; Klabunde, T.; Krebs, B.
Angew. Chem., Int. Ed. Engl. 1996, 35, 2024.
(3) (a) Neverov, A. A.; Lu, Z.-L.; Maxwell, C. I.; Mohamed, M. F.; White, C.
J.; Tsang. J. S. W.; Brown, R. S. J. Am. Chem. Soc. 2006, 128, 16398. (b)
Lu, Z.-L.; Liu, C. T.; Neverov, A. A.; Brown, R. S. J. Am. Chem. Soc.
2007, 129, 11642.
(4) Liu, C. T.; Neverov, A. A.; Brown, R. S. Inorg. Chem. 2007, 46, 1778.
9
16238
J. AM. CHEM. SOC. 2007, 129, 16238-16248
10.1021/ja076847d CCC: $37.00 © 2007 American Chemical Society

Dinuclear Zn(II) Complex-Catalyzed Cyclization
A R T I C L E S
Scheme 1 a
we report a detailed analysis that shows the energies associated
with the various steps of methoxide binding to the catalyst,
subsequent substrate binding, and activation energies for the
catalytic steps with substrates 4a-g which are compared with
the methoxide promoted reactions.
a S ) phosphodiester substrate, charges omitted for simplicity.
s
complex 1-Zn(II)2:(-OCH3) and sets the spH5 of the medium
at 9.4-9.8) where the second-order rate constant (k2obs) of 275,-
000 M-1 s-1 is several orders of magnitude larger than anything
reported to date for related catalysts in water.3a,6-8 This result
demands further investigation when viewed in the light of the
1-Zn(II)2:(-OCH3)-promoted catalysis of cleavage of a DNA
model, methyl p-nitrophenyl phosphate (3), where the plot of
2. Experimental
2.1. Materials. Methanol (99.8% anhydrous), sodium methoxide
(0.50 M solution in methanol, titrated against N/50 certified standard
aqueous HCl solution and found to be 0.49 M), tetrabutylammonium
hydroxide in methanol (1 M, titrated against N/50 certified standard
aqueous HCl solution and found to be 1.087 M), Zn(CF3SO3)2,
p-nitrophenol (98%), 4-chlorophenol (99+%), 3-nitrophenol (99%),
3-methyl-4-nitrophenol (98%), phenol (99%), 4-methoxyphenol (99%),
3-methoxyphenol (96%), triethylamine (g99.5%), propylene oxide
(ReagentPlus, 99%), phosphorus oxychloride (99%), and Ba(OH)2 were
purchased from Aldrich and used without further purification. HClO4
(70% aqueous solution, titrated to be 11.40 M) was purchased from
Acros Organics and used as supplied. The sodium salts of all the
2-hydroxylpropyl aryl phosphates were prepared by a modification9 of
a prior procedure.10 The corresponding disodium salts of the aryl
monophosphates required for the preparation of 4a-g were synthesized
kobs vs [1-Zn(II)2:(-OCH3)]free exhibited Michaelis-Menten
behavior with KM and kmax values of 0.37 ( 0.07 mM and (4.1
( 0.3) × 10-2 s-1, respectively.3a That two closely related
substrates exhibited such different kinetic behavior led us to
propose the mechanism given in Scheme 1 where the rate-
limiting step with 2 was its binding (k1 or k2), whereas in the
case of the slower-reacting 3 it was the chemical step of cleavage
of the bound substrate (kcat) to release p-nitrophenol.
We recently reported that the cleavages of 2 and 3 promoted
by the corresponding Cu(II)2 complex, 1-Cu(II)2:(-OCH3),3b
follow a pathway similar to the mechanism given in Scheme 1.
However, there are sufficient differences in the chemical,
physical, and coordination properties of the two metal ions that
one might wonder about the general similarity of their catalytic
processes. A better test for the proposed scheme requires study
of the 1-Zn(II)2:(-OCH3)-catalyzed cleavage of a series of
hydroxypropyl aryl phosphates where the rate of the chemical
cleavage step (kcat) can be altered by changing the nature of the
departing aryloxy group. In the following we report a kinetic
study of the cleavage of such a series of substrates (4a-g)
promoted by 1-Zn(II)2:(-OCH3) where there is a gradual change
according to a literature procedure.11 Each of 4a-g had 1H NMR, 31
P
NMR, and exact MS spectra consistent with those of the structure (see
the Supporting Information). 1,3-Bis-N1-(1,5,9-triazacyclododecyl)-
propane (1) was prepared as described.12 The dinuclear 1-Zn(II)2:(--
OCH3) complex was prepared as a 2.5 mM stock solution in anhydrous
methanol at 25 °C by sequential addition of aliquots of stock solutions
of sodium methoxide, 1,3-bis-N1-(1,5,9-triazacyclododecyl)propane, and
Zn(CF3SO3)2 such that the relative ratios were 1:1:2. This order of
addition is essential for the formation of the catalyst complex which
takes ∼40 min in methanol (as monitored by the change in catalytic
activity over time).
1
2.2. Methods. H NMR and 31P NMR spectra were determined at
+
400 and 162.04 MHz. The CH3OH2 concentrations were determined
potentiometrically using a combination glass electrode (Radiometer
model no. XC100-111-120-161) calibrated with certified standard
aqueous buffers (pH ) 4.00 and 10.00) as described in previous
papers.13 The sspH values in methanol were determined by subtracting
the correction constant of -2.2413a from the readings obtained from
the electrode, and the autoprotolysis constant for methanol was taken
to be 10-16.77 M2. The sspH values for the kinetic experiments were
simply measured from solutions containing the complex and generally
found to be in the range of 9.8 ( 0.2 at the concentrations of catalyst
employed. We have found that the addition of buffers to control the
from Michaelis-Menten kinetics to second-order kinetics as
s
s
the pKa of the leaving phenol group is decreased. In addition,
s
spH inhibits the catalytic reaction, probably due to the associated
counterions that bind to the catalyst, and thus, all the kinetic studies
were done under buffer-free conditions. The first and second macro-
(5) For the designation of pH in nonaqueous solvents we use the forms
recommended by the IUPAC, Compendium of Analytical Nomenclature:
DefinitiVe Rules 1997, 3rd ed.; Blackwell: Oxford, U.K. 1998. If one
calibrates the measuring electrode with aqueous buffers and then measures
-
s
scopic pKa values for 1:Zn2:(HOCH3) and 1:Zn2:(HOCH3)( OCH3)
s
(0.4 mM) were determined to be 9.41 ( 0.013b and 10.11 ( 0.01 from
duplicate measurements of the sspH at half neutralization, whereby the
[1-Zn(II)2:(-OCH3)]/[1-Zn(II)2:(HOCH3)] or [1-Zn(II)2:(-OCH3)2]/
the pH of an aqueous buffer solution, the term wwpH is used; if the
electrode is calibrated in water and the pH of the neat buffered methanol
solution is then measured, the term swpH is used; and if the electrode is
calibrated in the same solvent and the pH reading is made, then the term
sspH is used. Since the autoprotolysis constant of methanol is 10-16.77
,
(9) Tsang, J. S.; Neverov, A. A.; Brown, R. S. J. Am. Chem. Soc. 2003, 125,
1559.
neutral sspH is 8.4.
(6) Feng, G.; Mareque-Rivas, J. C.; Williams, N. H. Chem. Commun. 2006,
1845.
(10) Brown, D. M.; Usher, D. A. J. Chem. Soc. 1965, 6558.
(11) Williams, A.; Naylor, R. A. J. Chem. Soc. (B) 1971, 10, 1973.
(12) Kim, J.; Lim, H. Bull. Korean Chem. Soc. 1999, 20, 491.
(13) (a) Gibson, G.; Neverov, A. A.; Brown, R. S. Can. J. Chem. 2003, 81,
495. (b) Gibson, G. T. T.; Mohamed, M. F.; Neverov, A. A.; Brown, R. S.
Inorg. Chem. 2006, 45, 7891.
(7) Feng, G.; Natale, D.; Prabaharan, R.; Mareque-Rivas, J. C.; Williams, N.
H. Angew. Chem., Int. Ed. 2006, 45, 7056.
(8) Iranzo, O.; Kovalevsky, A. Y.; Morrow, J. R.; Richard, J. P. J. Am. Chem.
Soc. 2003, 125, 1988.
9
J. AM. CHEM. SOC. VOL. 129, NO. 51, 2007 16239

A R T I C L E S
Bunn et al.
[1-Zn(II)2:(-OCH3)] ratios were 1.0. The half neutralization involved
treating a 0.4 mM methanol solution of 1-Zn(II)2:(-OCH3), prepared
as described in the Materials section above, with 0.5 equiv of HClO4
or NaOCH3 (diluted to a 0.05 M stock solution in anhydrous methanol)
and measuring the sspH.
Table 1. Crystal Data and Structural Refinement Details for
1-Zn(II)2:(-OH)(CF3SO3-)3(HOCH3)
empirical formula
formula weight
temperature
C25H47F9N6O11S3Zn2
1005.61
180(2) K
wavelength
0.71073 Å
2.3. Stopped-Flow Kinetics in Methanol. A 2.5 mM stock solution
of 1-Zn(II)2:(-OCH3) in methanol was prepared in an oven-dried vial,
stoppered with a needle-permitting lid, and then sealed with Parafilm
under nitrogen gas. This solution was used to prepare stock catalyst
solutions of concentrations ranging from 0.4 mM < [1-Zn(II)2:(-OCH3)]
< 2.5 mM, which were loaded into one syringe of the stopped-flow
reaction analyzer. Substrate stock solutions at concentrations of 8 ×
10-5 M (for 4b-e) and 1.6 × 10-4 M (for 4f,g) were prepared and
loaded into the second syringe. The final respective substrate concentra-
tions were 4 × 10-5 M (4b-e) and 8 × 10-5 M (4f,g). At each
concentration five kinetic runs were recorded, and the average values
were used for the kobs for appearance of phenol product vs [catalyst]
plots. The production of the phenol was monitored at the following
wavelengths: 4b, 323 nm; 4c, 340 nm; 4d, 284 nm; 4e, 282 nm; 4f,
280 nm; 4g, 292 nm.
crystal system
space group
unit cell dimensions
monoclinic
P2(1)/c
a ) 10.9062(10) Å
b ) 22.998(2) Å
c ) 16.7888(14) Å
R ) 90°
â ) 102.656(2)°
γ ) 90°
volume
Z
density (calculated)
absorption coefficient
F(000)
crystal size
θ range for data collection
index ranges
4108.7(6) Å3
4
1.626 Mg/m3
1.419 mm-1
2064
0.40 × 0.10 × 0.08 mm3
1.53 to 25.00°.
-12 e h e 12,
-27 e k e 27,
-19 e l e 17
23896
The spKa values in methanol of the corresponding phenol leaving
s
group for 4a, 4c, 4d, 4f, and 4g were obtained from previous work14
reflections collected
and were plotted against the known spKa values in water and fit by
independent reflections
completeness to θ ) 25.00°
absorption correction
max. and min. transmission
refinement method
7225 [R(int) ) 0.0303]
99.9%
empirical (Bruker SADABS)
1.0000 and 0.8649
full-matrix least-squares on F2
7225/7/529
s
s
linear regression to the relationship, pKaMeOH ) (1.09 ( 0.06) pKaHOH
s
+ (3.40 ( 0.50). This relationship was used to interpolate the spKa
s
values in methanol for the 4b and 4e phenols.
data/restraints/parameters
The base-dependence of the 1:Zn(II)2-catalyzed cyclization of 4f was
studied through “pH jump” experiments. The concentration of the
initially prepared catalyst (1-Zn(II)2:(-OCH3)) was held constant at 0.8
mM, and the initial concentration of 4f was 8 × 10-5 M. To the solution
containing 4f, a required amount of perchloric acid or sodium methoxide
was added such that when mixed in the stopped-flow reaction analyzer
for rapid monitoring of the change in absorbance at 280 nm, the
[CH3O-]/[1:Zn2] ratio was varied from 0 through 2.
goodness-of-fit on f 2
1.000
final R indices [I > 2σ(I)]
R indices (all data)
largest diff. peak and hole
R1 ) 0.0373, wR2 ) 0.0940
R1 ) 0.0573, wR2 ) 0.1005
0.648 and -0.509 e‚Å-3
Table 2. Second-Order Rate Constants for the Methoxide
Promoted Cleavage of 4a-g, and pKa Values for the
Corresponding Phenols in Methanol, T ) 25 °C
s
s
Base-catalyzed reactions were also conducted for each of 4b-g.
Tetrabutylammonium hydroxide (1.087 M in methanol) was used as
the base, and it was assumed to be completely converted to tetrabu-
tylammonium methoxide in methanol. For each substrate, the reaction
media were prepared in duplicate with at least three concentrations in
the range of 0.02-0.3 M (tetrabutylammonium hydroxide) along with
0.4 mM of substrate in a methanol solvent. The reactions were
monitored using a Cary 100 UV-visible spectrophotometer with the
cell compartment thermostatted at 25.0 ( 0.1 °C. Second-order
methoxide rate constants were evaluated from the gradients of the
observed first-order rate constant (kobs) v s[methoxide] plots. (The
concentration of water at 0.02 M of tetrabutylammonium hydroxide in
methanol was titrated (Karl Fisher) to be 46 ( 2 mM, and at 0.3 M it
was found to be 480 ( 4 mM. Anhydrous methanol as received was
found to contain 16 ( 1 mM of water.)
-
1
k2 OMe (M-1 s-
)
s
substrate
phenol pKa
s
4a (2)
4b
4c
4d
4e
11.30
11.50
12.49
13.59
13.93
14.33
14.77
(2.6 ( 0.2) × 10-3a
(5.3 ( 0.2) × 10-4
(6.7 ( 0.2) × 10-4
(5.2 ( 0.2) × 10-5
(1.68 ( 0.06) × 10-5
(1.21 ( 0.03) × 10-5
(5.9 ( 0.2) × 10-6
4f
4g
a Rate constant from ref 9.
determined at 25 °C under pseudo-first-order conditions of
excess [CH3O-] (0.02-0.3 M in methanol). In Table 2 are the
second-order rate constants (k2-OMe) for each substrate along
with the spKa values for the corresponding phenols in metha-
s
2.4. X-ray Diffraction. A crystal of 1-Zn(II)2:(-OH)(CF3SO3-)3-
(HOCH3) suitable for X-ray diffraction was grown from a methanol
solution containing ligand 1, Zn(CF3SO3)2, and NaOCH3 in a 1:2:1
ratio. The solution was allowed to evaporate exposed to the atmosphere
at ambient temperature over 4 days. Structural details are given in Table
1 below and the ORTEP diagram for 1:Zn2:(-OH) at the 50%
probability level is given in Figure 6. Full structural parameters and
the CIF file are given in the Supporting Information.
s
s
nol. A Brønsted plot of the log k2-OMe vs pKa (not shown) fits
-OMe
a linear regression of log k2
(5.36 ( 1.08).
) (-0.72 ( 0.08) spKa +
s
3.2. Cyclization Reaction of 4a-g Promoted by 1-Zn(II)2:
(-OCH3) in Methanol. The plots of the kobs for cyclization of
4b, 4c, and 4f vs [1-Zn(II)2:(-OCH3)]free shown in Figures 1,
2, and 3 demonstrate progressive differences in passing from
good aryloxy leaving groups to poorer ones with higher spKa
3. Results
s
values. Previously we have established that triflate ion is an
inhibitor of the catalysis afforded by 1:Zn(II)2:(-OCH3) with
an inhibition constant of Ki ) 14.9 mM3a from which one can
determine the free catalyst concentration at any [triflate]. All
the catalyst concentrations reported in the figures and used for
computation of the kinetic parameters refer to the free concen-
trations corrected for triflate inhibition ([1-Zn(II)2:(-OCH3)]free).
3.1. Methoxide-Promoted Reactions. In order to compare
the catalysis afforded by 1-Zn(II)2:(-OCH3) for cleavage of the
substrates (4a-g) to the background reactions, the second-order
rate constants for their methoxide-promoted cleavages were
(14) Liu, T.; Neverov, A. A.; Tsang, J. S. W.; Brown, R. S. Org. Biomol. Chem.
2005, 3, 1525.
9
16240 J. AM. CHEM. SOC. VOL. 129, NO. 51, 2007

Dinuclear Zn(II) Complex-Catalyzed Cyclization
A R T I C L E S
3 for the less reactive 4f, and fitting of these data to eq 1 gives
max
KM ) (1.00 ( 0.08) × 10-4 M and kcat ) 3.10 ( 0.06 s-1
.
Original data for all substrates and plots of kobs vs [1-Zn(II)2:
(-OCH3)]free are given as Supporting Information.
There are several points of note required to understand the
fitted data. First, inspection of all the plots reveals that there is
a significant X-intercept which results from incomplete forma-
tion of the catalyst at low concentrations. Equation 1 is a
universal binding equation that is applicable to both strong and
weak binding situations15
kobs ) kcat(1 + KB[S] + [Cat]KB - X)/(2KB)/[S] (1)
Figure 1. Plot of kobs vs [1-Zn(II)2:(-OCH3)]free for the catalyzed
methanolysis of 4b (4 × 10-5 M) determined from the rate of appearance
of product phenol at 323 nm, sspH 9.8 ( 0.1, and 25.0 ( 0.1 °C. The linear
regression gives k2obs ) (1.94 ( 0.03) × 105 M-1 s-1 with an intercept of
A ) 0.190 ( 0.013 mM.
where:
2
X ) (1 + 2KB[S] + 2[Cat]KB + KB [S]2 -
2KB [Cat][S] + [Cat]2KB )
2
2 0.5
and KB refers to the Cat + Sub binding constant (units of M-1),
the reciprocal of which is defined here as the Michaelis-Menten
constant KM with units of M. The [Cat] term in eq 1 used for
the fitting corresponds to the actual concentration of Viable
catalyst, free from triflate inhibition, which is derived according
to the expression [Cat] ) ([1:Zn(II)2:(-OCH3)]free - A), where
A is an independently fitted parameter generally having the value
around 0.2 mM, corresponding to the observable intercept.
While we acknowledge this approach is not rigorous, it is the
same as what we have used previously to model the La3+
-
catalyzed methanolysis of carboxylate esters and neutral phos-
phate triesters16a-c where only the concentration of a catalytically
active La3+-containing dimer was taken into account. We have
also demonstrated the validity of this approach in the case of
Figure 2. Plot of kobs vs [1-Zn(II)2:(-OCH3)]free for the catalyzed
methanolysis of 4c (4 × 10-5 M) determined from the rate of appearance
of product phenol at 340 nm, sspH 9.8 ( 0.2, and 25.0 ( 0.1 °C. Fitting of
the data to the expression given in eq 1 gives KM ) (6.3 ( 0.9) × 10-4 M,
kcat
) 190 ( 16 s-1 and an intercept A ) 0.128 ( 0.005 mM.
max
the Zn2+-catalyzed cyclization of HPNPP in ethanol.4 Given in
max
Table 3 are the best-fit KM and kcat
constants for substrates
4c-g, along with the experimentally observed second-order rate
constants for substrates 4a (or 23a) and 4b. The kcat
values
max
for substrates 4a and 4b were computed under the assumption
that values for all the substrates adhere to a Brønsted relationship
s
s
max
lg
a
kcat
) 10(â ‚ pK +C). Nonlinear least-squares fitting of the
s
max
available kcat vs pKa data gives âlg ) -0.97 ( 0.05 and C
s
) 14.36 ( 0.57 from which values of 2300 ( 800 and 1470 (
450 s-1 can be extrapolated for substrates 4a and 4b.17 These
are used to provide the KM values from the expression KM
)
kcatmax/k2obs. Given in Figure 4 is a plot of the KM vs phenol
spKa values that shows an upward curvature commencing at
s
the more acidic phenols corresponding to substrates 4c, 4b, and
4a. Also included on the figure are two complex constants, the
components of which refer to various rate and equilibrium
binding constants given in Scheme 1 which will be discussed
later.
Figure 3. Plot of kobs vs [11-Zn(II)2:(-OCH3)]free for the catalyzed
methanolysis of 4f (4 × 10-5) M determined from the rate of appearance
of phenol at 280 nm, sspH 9.8 ( 0.1, and 25.0 ( 0.1 °C. Fitting of the data
max
to the expression given in eq 1 gives KM ) (1.00 ( 0.08) × 10-4 M, kcat
) 3.10 ( 0.06 s-1 and intercept A ) 0.171 ( 0.001 mM.
(15) Equation 1 was obtained from the equations for equilibrium binding and
for conservation of mass by using the commercially available MAPLE
software, Maple 9.00, June 13, 2003, Build ID 13164, Maplesoft, a division
of Waterloo Maple Inc. 1981-2003, Waterloo, Ontario, Canada.
(16) (a) Brown, R. S.; Neverov, A. A.; Tsang, J. S. W.; Gibson, G. T. T.;
Montoya-Pela´ez, P. J. Can. J. Chem. 2004, 82, 1791. (b) Brown, R. S.;
Neverov, A. J. Chem. Soc., Perkin Trans. 2 2002, 1039. (c) Tsang, J. S.;
Neverov, A. A.; Brown, R. S. J. Am. Chem. Soc. 2003, 125, 7602.
max
Substrates 4a and 4b do not show saturation kinetics at the
catalyst concentrations used, and the linear regressions of the
respective plots give second-order rate constants for the ap-
pearance of phenol product of 275,0003a and 194,000 M-1 s-1
.
The plot for substrate 4c shows apparent saturation over the
same range of [1-Zn(II)2:(-OCH3)]free from which nonlinear
least-squares (NLLSQ) fitting of the data to the expression given
(17) The linear regression of log kcatmax vs spKa fits the expression log kcat
)
s
(-0.89 ( 0.13) spKa + (13.20 ( 1.78). Although the nonlinear and linear
s
fittings give Brønsted âlg values that are within experimental error, the
max
in eq 1, vide infra, gives KM ) (6.3 ( 0.9) × 10-4 M and kcat
nonlinear fit gives non-log-weighted error limits which reduce the
) 190 ( 16 s-1. A more pronounced saturation is seen in Figure
uncertainty in the extrapolated kcat terms for 4a and 4b.
max
9
J. AM. CHEM. SOC. VOL. 129, NO. 51, 2007 16241

A R T I C L E S
Bunn et al.
obs
Table 3. Kinetic Constants (Maximum Rate Constant, kcatmax, Michaelis-Menten Constants (KM), and Second-Order Rate Constants (k2
or kcatmax/KM) for the Cleavage of 4a-g Mediated by 1-Zn(II)2:(-OCH3) at sspH 9.8 and 25.0 ( 0.1 °C
phenol s pKa in
kcat
KM
k2
(M-1 s-
)
acceleration by
max
obs
s
catalyst at sspH 9.8c
1
substrate
methanol
(s-
1
)
(M)
4a (2)a
4ba
4c
4d
4e
11.30
11.50
12.49
13.59
13.93
14.33
14.77
2300 ( 800
1470 ( 450
190 ( 16
15.0 ( 0.4
3.32 ( 0.09
3.10 ( 0.07
2.37 ( 0.07
(8.3 ( 2.9) × 10-3
(7.2 ( 2.3) × 10-3
(6.3 ( 0.9) × 10-4
(8.0 ( 0.8) × 10-5
(1.0 ( 0.1) × 10-4
(7.9 ( 0.8) × 10-5
(1.0 ( 0.1) × 10-4
(2.75 ( 0.04) × 105
(1.94 ( 0.03) × 105
(3.0 ( 0.5) × 105b
(1.9 ( 0.2) × 105b
(2.6 ( 0.3) × 104b
(3.9 ( 0.4) × 104b
(2.3 ( 0.3) × 104b
1.0 × 1012d
3.4 × 1012d
2.6 × 1012
2.7 × 1012
1.8 × 1012
2.4 × 1012
3.8 × 1012
4f
4g
obs
a KM values in italics computed as described in text. The k2 values for 4a3a and 4b are the experimentally observed values from the slopes of the kobs
s
s
obs
max
vs [1-Zn(II)2:(-OCH3)]free plots. b k2 values computed from kcatmax/KM. c Acceleration in terms of kcat
relative to the methoxide reaction at pH 9.8
where [-OCH3] )1.07 × 10-7 M. d Acceleration in terms of 1 mM catalyst reacting with observed second-order rate constant at sspH 9.8.
Figure 4. Plot of KM vs spKa values for the phenols of substrates 4a-g.
s
KM values for the substrates with the two lowest spKa’s (4a,b) computed
s
as described in text; KM values for 4c-g are experimentally determined
from fits of the saturation kinetic plots. Line through the data computed on
·sspKa+const.)
the basis of NLLSQ fit to KM ) C + 10(â
where âlg has a
Figure 6. Molecular structure of 1-Zn(II)2:(-OH) (CF3SO3-)3(HOCH3)
shown as an ORTEP diagram at the 50% probability level, counterions
and methanol of solvation omitted for clarity. Structural details given in
Supporting Information.
lg
computed value of -1.02 ( 0.13, constant is 12.5 ( 1.5 and C ) K-1k-2
/
k2 is set as a constant of 9 × 10-5 M (the average KM value for substrates
4d-g).
for the product appearance contains one methoxide, one 1-Zn-
(II)2 complex, and one substrate, or their kinetic equivalents.
A crystal of 1-Zn(II)2:(-OH) (CF3SO3-)3(HOCH3) suitable
for X-ray diffraction was grown from a methanol solution where
the components 1, Zn(CF3SO3-)2, and NaOCH3 were introduced
in a 1:2:1 ratio. Surprisingly, the structure of the core unit shown
in Figure 6 (devoid of the three triflate counterions and the
methanol of solvation) shows a hydroxide ion bridging between
the two Zn(II) ions, the source of which must be adventitious
water that was present in the methanol solution or in the
atmosphere to which the solution was exposed during its
evaporation.
Figure 5. Plot of the observed first-order rate constants for the methanolysis
of 0.04 mM 4f catalyzed by 0.4 mM 1-Zn(II)2 as a function of the [CH3O-]/
[1:Zn(II)2] ratio at 25.0 ( 0.1 °C.
4. Discussion
4.1. X-ray Diffraction Structure. Despite having been grown
from a methanol solution with added methoxide, the only
crystallized complex we obtained contains a bridging hydroxide
as 1-Zn(II)2:(-OH). Its molecular structure, shown in Figure 6
devoid of counterions and methanol of solvation, shows that
each metal ion is coordinated to a single bridging HO- and the
three N’s of the opposing triazacyclododecane units that are
joined by a 1,3-propano linker in a fashion that is reminiscent
of “earmuffs”. The Zn(II)2-coordinated hydroxide is hydrogen
bonded to a triflate ion in the adjacent unit cell which may be
a stabilizing force for HO- in the solid state, although it is
possible that there is some other subtle preference for hydroxide
that is retained in solution, but this cannot be ascertained from
Shown in Figure 5 is a plot of the first-order rate constant
for the appearance of phenol product from 4 × 10-5 M 4f
promoted by 0.4 mM 1-Zn(II)2 containing catalyst where the
[methoxide]/[1-Zn(II)2] ratio varies from 0 to 2. A 0.8 mM stock
solution of the 1-Zn(II)2:(-OCH3) catalyst was prepared as
described in the Experimental Section and then mixed in a
stopped-flow reaction analyzer with equal volumes of a second
solution containing 0.08 mM of 4f and varying amounts of
added HClO4 or NaOCH3 after which the rate of appearance of
product phenol was determined. As the figure shows, the
maximum rate constant is observed when the [methoxide]/[1-
Zn(II)2] ratio is unity, consistent with a process where the TS
9
16242 J. AM. CHEM. SOC. VOL. 129, NO. 51, 2007

Dinuclear Zn(II) Complex-Catalyzed Cyclization
A R T I C L E S
the studies we have at present. Each of the Zn(II) ions
experiences a distorted tetrahedral coordination with the HO--
Zn-N angles varying between 105° and 117°. The Zn--OH
bond lengths are 1.912(2) and 1.915(2) Å, while the Zn--O-
(H)-Zn angle is 147.19(14)°. The Zn-Zn separation is 3.670
Å which can be compared with Cu(II)-Cu(II) separations of
3.26 and 3.68 Å in two related 1-Cu(II)2:(-OH)(OH2) and 1-Cu-
(II)2:(-OH)((ArCH2O)2PO2-) complexes previously reported.3b
We have repeatedly tried, without success, to obtain a dibenzyl
phosphate complex of 1-Zn(II)2:(-OH), but the only crystalline
methanol) as in eq 2. The methanol solvent actually reduces
the base-catalyzed reaction rates relative to the situation in water.
For example with 4a, the k2-OMe is 2.6 × 10-3 M-1 s-1, 9 while
that for the hydroxide-promoted cyclization is reported as 9.9
× 10-2 M-1 s-1 23 or 6.5 × 10-2 M-1 s-1 24
.
-
products are those containing polymeric Zn2+2((ArCH2O)2PO2 )2
with no ligand present.
4.2. Methoxide-Promoted Cyclization of 4a-g. The Brøn-
s
sted plot of the log k2-OMe vs phenol pKa data (Table 2) has a
s
âlg value of -0.72 ( 0.08. This value can be compared with
âlg values of -0.62 and -0.58 for the hydroxide-promoted
cleavage of 2-hydroxypropyl aryl phosphates and for the
hydroxide-promoted hydrolysis of methyl aryl phosphate di-
esters18 and values of -0.54 for the hydroxide reaction of aryl
uridine-3′-phosphates,19a and -0.64 for the attack of aryloxy
anions on aryl methyl phosphate diesters.19b Williams19,20 has
argued that nucleophilic displacement of good leaving groups
from neutral or monoanionic phosphates by oxyanions is
concerted or involves an intermediate in a very shallow energy
well at the top of a barrier. The more recent study21 indicated
the base-catalyzed ring closure of a series of uridine 3′-phosphate
esters with leaving groups spanning a range of 5-17 pKa units
proceeded through an intermediate, the formation and break-
down of which was rate limiting for good and poor leaving
groups respectively (âlg ) -0.52 and -1.34, respectively). In
the latter report21 it was argued that the original Brown and
Usher data for the base-promoted cyclization of 2-hydroxypropyl
aryloxy and alkoxy phosphates10 could be reasonably reinter-
preted as being consistent with a stepwise process proceeding
through a five-membered phosphorane intermediate.
This rate reduction is probably a consequence of the reduced
polarity/dielectric constant of methanol vs water (ꢀ ) 31.5 vs
7825) which opposes the required pre-equilibrium formation of
the dianionic form (4-). Nevertheless, as will be shown later,
this same reduced dielectric constant/polarity effect considerably
enhances the reactions between the substrates 4 and the net
positively charged 1-Zn(II)2:(-OCH3) catalyst.
4.3. Cyclization Reactions of 4a-g Catalyzed by 1-Zn-
(II)2:(-OCH3). 4.3.1. Stepwise or Concerted kcat Process?
Inspection of the kobs vs [1-Zn(II)2:(-OCH3)]free plots provided
in Figures 1-3 for 4b,c,f, as well as those figures in the
Supporting Information for 4d,e,g, indicates that all substrates
but 4a3a and 4b exhibit Michaelis-Menten behavior over the
range of 0.2 mM < [1-Zn(II)2:(-OCH3)]free < 1 mM. The KM
and kcatmax constants in Table 3 for 4c-g are derived from fitting
the kobs vs [1-Zn(II)2:(-OCH3)]free data to the expression in eq
1. Since the kobs plots for 4a,b are linear over the concentration
range investigated, only second-order rate constants for the
appearance of the corresponding phenol products are experi-
mentally accessible. However, under the assumption that the
The data we have with the methoxide reaction of 4a-g are
consistent with a single rate-limiting step for these substrates
max
kcat for 4a and 4b should lie on the same Brønsted plot that
having a limited spKa range for the various aryloxy-leaving
is defined by the other members of series 4, respective values
of 2300 and 1470 s-1 can be determined for these. This
assumption might be questioned since it requires that the
mechanism for the kcatmax term involving cleavage of the bound
phosphate should be a common one throughout this series of
closely related aryloxy derivatives. In our opinion it is likely
that the present 1-Zn(II)2:(-OCH3)-catalyzed cyclization reac-
tions of 4a-g all comprise two steps since the strong electro-
static interaction of the dinuclear core of the catalyst should
stabilize greatly the dianionic phosphorane intermediate and the
transition state leading to it, thereby deflecting the reaction away
from a concerted pathway. Since we are dealing with good
aryloxy-leaving groups, the probable rate-limiting step is the
formation of the intermediate, but the slightly more negative
Brønsted âlg value for the catalyzed kcat reaction (-0.97 ( 0.05)
relative to that of the second-order methoxide-promoted reaction
(-0.72 ( 0.08) weakly suggests that there is more P-OAr bond
cleavage at the TS in the former.
s
groups. This includes a two-step process where formation of
the phosphorane intermediate is rate limiting with fast loss of
the good leaving groups. At the transition state, the change in
bonding of ArO-P bond has progressed some 40-45% from
starting material to product phenolate.22 The observed reaction
with substrates 4 in both water and methanol does not
incorporate lyoxide, but rather involves specific base-promoted
cyclization with the expulsion of the substituted phenoxide
(which immediately produces phenol at these sspH values in
(18) Williams, N. H.; Takasaki, B.; Wall, M.; Chin, J. Acc. Chem. Res. 1999,
32, 485.
(19) (a) Davis, A. M.; Hall, A. D.; Williams, A. J. Am. Chem. Soc. 1998, 110,
5105. (b) Ba-Saif, S. A.; Davis, A. M.; Williams, A. J. Org. Chem. 1989,
54, 5483.
(20) (a) Bourne, N.; Williams, A. J. Am. Chem. Soc. 1984, 106, 7591. (b) Bourne,
N.; Chrystiuk, E.; Davis, A. M.; Williams, A. J. Am. Chem. Soc. 1988,
110, 1890.
(21) Lo¨nnberg, H.; Stro¨mberg, R.; Williams, A. Org. Biomol. Chem. 2004, 2,
2165.
(22) The extent of breaking of the P-OAr bond in the TS can be measured by
the Leffler parameter, a, which measures the change in the Brønsted âlg
for the TS relative to the beq for equilibrium transfers of the phosphoryl
group between oxyanion nucleophiles. In the case of the transfer of the
(RO)P(dO)O- group20a, the âeq value is -1.74 with the O-Ar oxygen in
the starting material having a net effective charge of +0.74. When
methoxide is the nucleophile, the Leffler parameter of âlg/âeq ) 0.43
suggests that the P-OAr cleavage is 43% of the way from starting material
to product.
(23) O’Donoghue, A. M.; Pyun, S. Y; Yang, M.-Y.; Morrow, J. R.; Richard, J.
P. J. Am. Chem. Soc. 2006, 128, 1615.
(24) Bonfa´, L.; Gatos, M.; Mancin, F.; Tecilla, P.; Tonellato, U. Inorg. Chem.
2003, 42, 3943.
(25) Harned, H. S.; Owen, B. B. The Physical Chemistry of Electrolytic Solutions,
3rd ed.; ACS Monograph Series 137; Reinhold Publishing: New York,
1957; p 161.
9
J. AM. CHEM. SOC. VOL. 129, NO. 51, 2007 16243

A R T I C L E S
Bunn et al.
s
s
a first-formed complex that evolves into reactive form where
we suggest that the phosphate is doubly coordinated by the Cu-
(II)2 moiety.3b
Among all the terms present in Scheme 1, the kcatmax constant
should be most sensitive to the nature of the aryloxy group,
and the binding terms K-1, k2, and k-2 are expected to be much
less so. For the analysis, these will be assumed to be invariant
constants for the series 4 substrates. The form of eq 4 breaks
4.3.2. Break in the KM vs pKa Plot. From the extrapolated
kcatmax values for 4a and 4b can be derived KM constants of 8.3
× 10-3 and 7.2 × 10-3 M respectively, with errors of about
(30%. All the KM values are plotted in Figure 4 against the
s
s
phenol pKa values and NLLSQ fitting of the data to KM ) C
‚
sspKa+ const.)
+10(â
provides a âlg value of -1.02 ( 0.13. This
lg
value is largely defined by the kcatmax term which starts to come
into play for the substrates where the leaving phenols have
spKa values of less than ∼13.
the KM term into two components, the first of which (K-1k-2
/
s
k2) should be insensitive to the nature of the leaving group, and
thus largely defines the KM values for the poorer substrates.
Accordingly, for substrates 4d-g, all the experimental KM values
are very close, lying in the range of 8 × 10-5 M to 1 × 10-4
M, and this is shown graphically in Figure 4 as a dotted
horizontal line. For the substrates with the better leaving groups,
the second term in eq 4, K-1kcatmax/k2, comes into play, causing
increases in the KM to 6.3 × 10-4 M for 4c and calculated values
of 7.2 × 10-3 and 8.3 × 10-3 M for 4b and 4a as a result of an
increasing kcatmax (190 to 1470 and then 2300 s-1 respectively).
The upward curvature seen in Figure 4 largely depends on how
the electronic changes in the substrate affects the ratio of (k-2
+ kcatmax)/k2 and provides evidence for a change in the rate-
limiting step from chemical cleavage for poor substrates
The Michaelis-Menten expression given in eq 3 can be
derived for the process given in Scheme 1,
kcatmax [1 - Zn(II)2:(-OCH3)]free[4]
dP
rate )
KM )
)
(3)
dt
KM + [4]
max
cat
max
cat
k-1 k-2 + k
k
k
•
) K-1 -2 + K-1
k1
k2
k2
k2
(4)
with the KM constant being defined in eq 4. In our model, the
k-1/k1 term refers to an equilibrium dissociation constant of a
first-formed complex where the substrate is bound to one of
the Zn(II) ions (herein termed K-1 for simplicity). The k2 term
refers to a subsequent step proposed to be an intramolecular
rearrangement where the substrate becomes bound to both of
the Zn(II) ions, experiencing a double activation as shown in
Scheme 1, and the kcatmax term corresponds to the chemical step
of intramolecular cyclization which produces the observed
phenol product. We note that it is not mathematically required
that we include the first equilibrium constant k-1/k1 or K-1 in
the analysis but suggest it is necessary for a more complete
description based on two considerations. First, it is difficult to
envision that a productive substrate association involving
phosphate coordination to both metal ions can occur in a single
step with the highly congested 1-Zn(II)2:(-OCH3) as it would
involve formation of two P-O-Zn(II) bonds along with the
cleavage of at least one Zn(II)--OCH3 linkage in a very
sterically restricted complex. This explains the rather small,
second-order rate constants of 275,000 and 194,000 M-1 s-1
observed for catalyzed reactions of 4a and 4b26 that seem low
for a binding event involving a facile ligand exchange on Zn-
(II) which, in favorable cases, can be as high as 107-108 M-1
(leaving-group spKa values are >13) to binding for good
s
substrates (leaving-group spKa values are <12.5) for which
s
max
kcat
> k-2.
Some information about the values of the complex constants
comes from analysis of the relationship given in Figure 4 where
the average KM for substrates 4d-g is 9 × 10-5 M and is equal
to K-1k-2/k2 in Scheme 1. This number is very close to the
observed inhibition constant of 1.6 × 10-4 M for the essentially
unreactive, but structurally analogous, complex of dibenzyl
phosphate and 1-Zn(II)2:(-OCH3).3a Using the experimentally
determined KM and kcatmax values of 6.3 × 10-4 M and 190 s-1
for 4c (Table 3), one calculates from eq 4 values of 32 s-1 for
k-2 and 2.84 × 10-6 M‚s for the ratio K-1/k2 or its reciprocal
of 352,000 M-1 s-1 for k1k2/k-1. This latter number is the
computed second-order rate constant for the binding of substrate
to form the Michaelis complex, and can be compared with the
experimental values of 275,000 and 194,000 M-1 s-1 found for
substrates 4a and 4b26. Given the assumptions, the closeness of
the computed and experimental numbers for a substrate, where
saturation kinetics are observed, and two others, where only
second-order kinetics are observed, gives support for the
qualitative correctness of the proposed change in the rate-
limiting step from chemical cleavage to substrate binding as
we progress from poor to good substrates.
s-1 27
The rates of binding to M(II) complexes in general are
.
known28 to be highly sensitive to crowding effects such as is
expected with 1-Zn(II)2:(-OCH3) (see structure in Figure 6).
Second, for the analogous 1-Cu(II)2:(-OCH3) complex, stopped-
flow spectrophotometry reveals that the kinetics of changes of
the Cu(II) absorbance bands in the presence of some phosphate
substrates including 4a and 3 are consistent with two events; a
bimolecular binding of substrate and 1-Cu(II)2:(-OCH3) fol-
lowed by a unimolecular rearrangement of what appears to be
4.4. Catalysis and Mechanistic Questions. Presented in
Scheme 2 is a more detailed version of Scheme 1 that gives
the possible candidates involved in the 1-Zn(II)2:(-OCH3)-
catalyzed cleavage of substrates 4. Inspection of Figure 5
indicates that for substrate 4f where Michaelis-Menten kinetics
are observed and the concentration of catalyst is great enough
that the Michaelis complex is essentially completely formed,
the kcat rate constant for production of product maximizes at 1
equiv of methoxide per 1:Zn(II)2 unit. This stopped-flow
experiment is conducted starting with the fully formed 1-Zn-
(II)2:(-OCH3) catalyst which is mixed rapidly with 4f and an
appropriate amount of HClO4 or Bu4N+(-OCH3). Under the
assumptions that the substrate binding and acid/base reactions
(26) According to our proposed mechanism, the chemical step for catalyzed
cleavage of 4a and 4b is not rate limiting, but the binding steps are. Steady-
state analysis of the binding with k2 being rate limiting gives kobs ) k1k2/
(k-1 + k2).
(27) Another possibility for the rather slow binding of the substrate to the catalyst
could be an unfavourable pre-equilibrium opening of one of the Zn(II)-
OCH3 linkages, followed by a rapid attack of the substrate, which then
undergoes double activation and subsequent reaction. This process is
kinetically equivalent to what we have proposed.
(28) (a) Dukes, G. R.; Margerum, D. W. Inorg. Chem. 1972, 11, 2952. (b) Pitteri,
B.; Marangoni, G.; Viseutin, F. V.; Cattalini, L.; Bobbo, T. Polyhedron
1998, 17, 475.
9
16244 J. AM. CHEM. SOC. VOL. 129, NO. 51, 2007

Dinuclear Zn(II) Complex-Catalyzed Cyclization
A R T I C L E S
Scheme 2. Simplified Proposed Scheme for Binding and Catalyzed Cyclization of 2-Hydroxypropyl Aryl Phosphates Promoted by
1:Zn2:(-OCH3)
Scheme 3
are fast and the metal ion does not dissociate from the complex
within this time scale, the results suggest that the TS for the
catalytic formation of the phenol product contains one each of
methoxide, substrate, and 1-Zn(II)2, or the kinetic equivalent.
Substrate 4a exhibits a linear dependence on [catalyst] for the
production of p-nitrophenol product and shows a quite different
dependence for the plot of kobsvs the [-OCH3]/[1:Zn(II)2] ratio.
This plot maximizes at ∼0.1-0.2 at which point the kobs is about
1.6-fold larger than when the[-OCH3]/[1:Zn2] ratio is unity, and
falls dramatically at lower ratios. This is consistent with a rate-
limiting step where there is preferential 4a binding to a 1-Zn-
(II)2 complex without an associated methoxide. The drop in rate
at [-OCH3]/[1-Zn(II)2] ratio of ∼0.1-0.2 is then consistent with
a reduction in the free [-OCH3] to the point that the production
of the product, which requires a deprotonation of the bound
substrate, now becomes rate limiting.
actual substrate-binding step may involve the deprotonated 2-OH
form of the substrate (4-) and the protonated form of the catalyst
(5-Zn(II)2:(OH2)), rather than the alternative 2-OH form of the
substrate with the deprotonated form of the catalyst (see Scheme
3). While this mechanism cannot be ruled out in general for
the processes in water, it can be excluded for 1-Zn(II)2:(-OCH3)
in methanol. The pKa for deprotonation of the 2-HO group of
anionic 4a in methanol should be at least as large or greater
s
than that of methanol (Kw/[MeOH] ) (10-16.77/30)M) or pKa
s
g 18.2. At a spH of 9.8 where the 1-Zn(II)2 catalyst is
s
operative, the dianionic [4-] would be one part in 108.4. Since
the second-order rate constants for binding of substrates 4a and
4b, in whatever form, by the 1-Zn(II)2 catalyst are observed to
be 275,000 and 194,000 M-1 s-1 respectively, these reactions
would have to exceed the diffusion limit of 1010 M-1 s-1 by at
least 104 if the dianionic form of the substrate was the requisite
form for binding. This factor must be even larger if one
acknowledges that at sspH 9.8 essentially all of the catalyst is
present as 1-Zn(II)2:(-OCH3) and only a small proportion exists
in the form required to bind the anionic substrate, namely 1-Zn-
(II)2:(HOCH3). Note this does not preclude a kinetically
equivalent situation where there is transient binding of 1:Zn-
(II)2:(-OCH3) and 4 with a subsequent fast formation of a bound
1:Zn(II)2:4- form along with simultaneous or subsequent double
activation and cyclization.
There are two other aspects for which information is now
available to rule out some of the mechanistic alternatives. The
first is the state of ionization of the catalyst and substrate for
the binding steps. It has been suggested,8,23 on the basis of
studies of the cleavages of some RNA models including uridine
3′-p-nitrophenyl phosphate and 4a by the dinuclear Zn(II)catalyst
5 and some other metal ion-containing catalysts,29,30 that the
(29) Farquhar, E. R.; Richard, J. P.; Morrow, J. R. Inorg. Chem. 2007, 46, 7169.
(30) Mathews, R. A.; Rossiter, C. S.; Morrow, J. R.; Richard, J. R. Dalton Trans.
2007, 3804.
9
J. AM. CHEM. SOC. VOL. 129, NO. 51, 2007 16245

A R T I C L E S
Bunn et al.
Scheme 4
Our favored mechanism shown in Scheme 2 thus involves
the binding of 1-Zn(II)2:(-OCH3) to the aryloxy hydroxypropyl
phosphate (4) first through one of the metal ions, followed by
a rearrangement to doubly activate the phosphate31 via binding
to the second Zn(II) ion. In accordance with the X-ray structure
shown in Figure 6, we believe the resting state of 1:Zn2:(-OCH3)
in solution has the methoxy group bridging the two Zn(II) ions.
In Scheme 2 the first formed complex (Comp 1) is formulated
as having the phosphate bound to one of the Zn(II) ions so that
its binding to the bridging methoxy group is loosened or lost
completely. The process by which the double activation occurs
is not clear from the present data, but could involve: (1)
phosphate closure on the second Zn(II) to yield Comp 1′ which
retains the methoxy coordinated to the second Zn(II); (2)
complete loss of the methoxide during the phosphate closure
to give Comp 0 plus liberated methoxide; or (3) formation of
Comp 2 where the departing methoxide has removed the 2-OH
proton of the bound substrate during the process of double
activation. If this process is reversible, it is equivalent to specific
base removal of the 2-OH proton by an external methoxide
followed by subsequent ring closure. The available kinetic
evidence indicates the specific base role for external methoxide
is possible, but only if deprotonation occurs at rates approaching
the diffusion limit. For the fastest substrate where Michaelis-
Menten behavior is observed in the concentration range
investigated here (4c), the experimental kcat value for the
(II)2:(-OCH3):4 complex or its kinetic equivalent, and (3) the
free energy of activation for the reaction of methoxide with 4.
∆GM is the free energy for association of 4 and 1-Zn(II)2:
(-OCH3) computed from the Michaelis constant KM for each
substrate. In eq 6 the free energy calculation is specific for the
reactions with 4a,b that show only a first-order dependence on
[1-Zn(II)2:(-OCH3)]free throughout the concentration range
investigated.
∆∆Gqstab ) (∆GBind - ∆GM + ∆Gqcat ) -
(kcat/KM)(Ka/Kw)
∆GqNon ) - RT ln
(5)
k-2 OMe
[
]
∆∆Gqstab ) (∆GBind + ∆G 1:Zn(II)2:(OCH3)) -
cat
(ko2bs)(Ka/Kw)
∆GqNon ) -RT ln
(6)
k-2 OMe
[
]
completely bound substrate is 190 s-1. Assuming that external
s
s
methoxide is the base, its concentration at pH 9.8 would be 1
The various terms in eqs 5,6 are: (1) the kcat/KM values given
in Table 3. The observed second-order rate constant (k2obs) for
the reaction of 1-Zn(II)2:(-OCH3) with 4a,b is equivalent to this
and is also found in Table 3; (2) Ka/Kw which can be readily
shown to be formally equivalent to the binding constant of
methoxide to the dinuclear complex 1-Zn(II)2 to form 1-Zn-
(II)2:(-OCH3). Ka refers to the first acid dissociation constant
of 1-Zn(II)2:(HOCH3), determined from half-neutralization3b as
× 10-7 M. Thus, if proton removal occurs at the diffusion limit
of ∼1010 M-1 s-1, the kcat would have an upper limit ∼1000
s-1 which encompasses the observed value for 4c, and those
values predicted for 4a and 4b.
4.5. Catalytic Acceleration of the Cyclization of 4 and
Energetic Considerations. Presented in Table 3 are accelera-
tions provided by 1-Zn(II)2:(-OCH3) for the cyclization reaction
of 4a-g relative to the rates contributed by the methoxide
reactions (Table 2) at sspH 9.8. Interestingly, all accelerations
lie in the range of 1 - 4 × 1012-fold. An alternative assessment
compares the second-order rate constants for reaction of 1-Zn-
10-9.41, and Kw is the autoprotolysis constant for methanol
(10-16.77); and (3) k2
-OMe
is the second-order rate constant for
the methoxide reaction of each substrate given in Table 2.
Listed in Table 4 are the (kcat/KM)/k2-OMeand (kcat/KM)(Ka/
Kw) constants and computed ∆∆Gqstab values which are -21 to
-23 kcal/mol. The first ratio provides the acceleration in second-
order rate constant for the complex-catalyzed reaction relative
to the methoxide reaction, while the second, when inserted into
eqs 5,6, provides the free energy of stabilization (∆∆Gqstab) of
the complex-catalyzed reaction relative to the methoxide reac-
tion. Of these constants, the Ka/Kw term contributes -10.0 kcal/
mol to the ∆∆Gqstab for the series, and the (kcat/KM)/k2-OMe term
contributes -11.0 to -13.2 kcal/mol with the slowest-reacting
substrates having the more negative values.
(II)2:(-OCH3) with 4a-g (defined as kcat/KM) with the k2
-OMe
values for the methoxide reactions (Table 2). These fit into a
domain of (1-4.5) × 108-fold acceleration for 4a-d, and a
second domain of (1.5-4) × 109-fold for the less reactive
substrates 4e-g.
A convenient way23 to quantify the catalytic efficiency is to
calculate the difference in energy of a transition state containing
only [CH3O-:4], and one where the latter is bound to the
catalyst32 (1-Zn(II)2) to form a hypothetical transition state
[1-Zn(II)2:(-OCH3):4]q. Presented in Scheme 4 is a thermody-
namic cycle involving both the methoxide reaction and the 1-Zn-
(II)2:(-OCH3)-promoted (or kinetically equivalent) reactions of
4c-g, for which eq 5 provides the free energy difference
(∆∆Gqstab.) between [1:Zn(II)2:(-OCH3):4]q relative to [CH3O-:
4]q. In Scheme 4, ∆Gbind, ∆Gqcat and ∆GqNon are the respective
free energies for: (1) methoxide binding to 1-Zn(II)2, (2) the
free energy of activation for unimolecular reaction of the 1-Zn-
It instructive to look at the graphical representation of the
activation energy diagram shown in Figure 7 for the reaction
of one example of the substrates, 4c, with methoxide and the
catalyst (represented as 1:Zn(II)2:(-OCH3)),33 all species at a
standard state of 1 M, where the ground state comprises CH3O-,
1-Zn(II)2, and substrate 4c partners. The ∆Gq
for the
non
methoxide reaction in the absence of catalyst is 21.7 kcal/mol
above the ground state. Figure 7 shows that binding of the three
components to form the Michaelis complex, formulated as 1:Zn-
(II)2:(-OCH3):4c, is exothermic by -14.4 kcal/mol, which is
almost exactly offset by the activation energy of 14.3 kcal/mol
(31) Double Lewis, activation of phosphates is a well-known motif of the binding
of a phosphate mono or dianion to a dinuclear core. See refs 1, 7, 8, 18,
and Kinoshita, E.; Takahashi, M.; Takada, H.; Shiro, M. Koike, T. Dalton
Trans. 2004, 8, 1189.
(32) Wolfenden, R. Nature 1969, 223, 704.
9
16246 J. AM. CHEM. SOC. VOL. 129, NO. 51, 2007

Dinuclear Zn(II) Complex-Catalyzed Cyclization
A R T I C L E S
-OMe
Table 4. (kcat/KM)(Ka/Kw) and (kcat/KM)/k2
Constants and Computed Free Energies of Formation of Michaelis Complex (∆GBind - ∆GM),
the Free Energies of Activation for kcat (∆Gqcat), and the Free Energies of Stabilization of the Methoxide Transition State through Binding to
1-Zn(II)2 (∆∆Gq
)
stab
for the Reaction of 1-Zn(II)2:(-OCH3) with Substrates 4a-g at 25 °C in Methanol
a
(kcat/KM)(Ka/Kw)
∆
G
Bind − ∆GM
∆
Gq
∆∆Gq
stab
(kcal/mol)
cat
-OMe
1 b,c
substrate
(kcat/KM)/k2
(M-2 s-
)
(kcal/mol)d
(kcal/mol)e
4a
4b
4c
4d
4e
4f
1.15 × 108
3.7 × 108
4.5 × 108
3.6 × 109
1.5 × 109
3.2 × 109
3.9 × 109
6.3 × 1012
4.5 × 1012
6.9 × 1012
4.4 × 1012
6.0 × 1011
9.0 × 1011
5.3 × 1011
-12.9
-13.0
-14.4
-15.6
-15.5
-15.6
-15.5
12.8
13.1
14.3
15.8
16.7
16.8
16.9
-21.0
-21.7
-21.8
-23.1
-22.6
-23.0
-23.1
4g
a ∆Gqstab computed from application of kinetic and equilibrium constants to eqs 5, 6; k2-OMe constants from Table 2. kcat/KM from Table 3 c Ka determined
b
from half-neutralization3b to be 10-9.41; Kw ) 10-16.77; Ka/Kw ) 2.3 × 107. d Computed as (∆GBind - ∆GM) ) -RT ln((Ka/Kw)/KM). e Computed from ∆Gq
cat
) -RT ln(kcat/(kT/h)) from the Eyring equation where (kT/h) ) 6 × 1012 s-1 at 298 K.
components (1-Zn(II)2, CH3O-, and 4)33 into a ternary complex
that is stabilized by -13 to -15.5 kcal/mol relative to the free
energies of the noncomplexed partners. Placing the ternary
complex (or its kinetic equivalent, for example 1-Zn(II)2:4-)
into a deep thermodynamic well does little for promoting the
cyclization of 4 unless the transition state for the catalyzed
reaction is stabilized to a greater extent. As shown in column 5
of the Table, the ∆Gqcat values for the cylcization of the bound
substrates are endothermic by ∼13-17 kcal/mol such that the
transition-state energy for the catalyzed reaction is 21-23 kcal/
mol lower than the transition-state energy for the methoxide-
catalyzed cyclization. Put another way, the stabilization energy
associated with the binding of 1:Zn(II)2 to the [CH3O-:4]q
transition state is -21 to -23 kcal/mol. This brings the free
energy of the transition state for the catalyzed reactions for all
the substrates to within 0 to 1.4 kcal/mol of the free energy of
the ground-state components.
5. Conclusions
The remarkable catalysis of the cleavage of the RNA models
Figure 7. Activation energy diagram for the reaction of CH3O- and 1:Zn-
4a-g that is exhibited by a simple system comprising 1:Zn-
(II)2:(-OCH3) and a medium effect imbued by the methanol
solvent stands out among all known catalytic systems for the
cleavage of 2-hydroxypropyl aryl phosphates. The available data
permit a detailed analysis of the change in mechanism from
(II):(-OCH3) with substrate 4c at standard state of 1 M and T ) 25 °C
showing the calculated energies of binding the methoxide by 1:Zn(II)2, of
binding of the substrate to 1:Zn(II)2:(-OCH3), and the calculated activation
-OMe
energies associated with kcat and k2
.
max
associated with the kcat
term.34 Even for the substrates that
substrate binding to chemical cleavage of the catalyst-bound
do not show saturation kinetics (4a,b), the diagram shows that
the computed energy of activation for the second-order rate
constant for reaction of 4a,b with 1:Zn(II)2:(-OCH3) is 10.0 and
10.2 kcal/mol, respectively, again essentially offsetting exactly
the energy of binding of methoxide to 1:Zn(II)2. It seems an
important consequence that the catalyzed process of cyclization
of the good substrates is energy neutral in passing from the
ground state to the transition state of the activated complex,
and that the overall rate of the reaction should be controlled by
the rate of association of the components.
s
s
substrate as the pKa values of the substrates’ leaving groups
increase. It has been pointed out35 that enzyme active sites
comprise a domain of effectively reduced dielectric constants,
but at the same time are decorated with specifically oriented
functional groups to create a specific medium tailored for a given
reaction. Since 1-Zn(II)2 is a poor catalyst for the cleavage of
phosphate diesters in water36 but is an exceptional one in
methanol, we view the net positively charged catalyst as one
important component that further requires an intimately associ-
ated low dielectric constant/polarity environment in order to
exert its maximum catalytic efficiency. The dinuclear complex
1:Zn(II)2 concentrates considerable positive net charge for
binding, orienting, and promoting the cleavage of anionic
As presented in column 4 of Table 4, the catalyzed cyclization
of all substrates 4 involves an assembly of the essential
(33) There is a kinetic ambiguity as to whether the active form of the catalyst
is 1-Zn(II)2 which binds substrate that subsequently undergoes deprotonation
by external methoxide, or whether the active form is 1-Zn(II):(-OCH3)
which binds substrate that undergoes a spontaneous decomposition to
cyclized product mediated by the “internal” methoxide. Although these
routes are kinetically equivalent and thus thermodynamically equivalent,
we prefer the 1-Zn(II)2:(-OCH3) formulation due to its simplicity in
recruiting the required catalytic components into one entity and the fact
that the reaction conditions where there is optimal reactivity require a
1:Zn2+:-OCH3 ratio of 1:2:1.
(35) Richard, J. P.; Ames, T. L. Bioorg. Chem. 2004, 32, 354.
(36) Kim and Lim (ref 12) have reported that the cleavage of bis-(2,4-
dinitrophenyl)phosphate promoted by 1:Zn(II)2 in water as a function of
pH is not markedly different from the catalysis afforded by the mono-Zn-
(II) complex of 1,5,9-triazacyclododecane, the rate constant for the former
being 7 × 10-3 M-1 s-1 at pH 7. In contrast, we have found that the second-
order rate constant for methanolysis of the far less reactive substrate bis-
(34) Energies of activation calculated from the Eyring equation as ∆Gq ) -RT
ln(kcat/(kT/h)).
(p-nitrophenyl)phosphate promoted by 1:Zn(II)2:(-OCH3) is 20 M-1 s-1
.
Brown, R. S.; Lu, Z.-L. Unpublished results.
9
J. AM. CHEM. SOC. VOL. 129, NO. 51, 2007 16247

A R T I C L E S
Bunn et al.
phosphate substrates. A synergistic medium effect in methanol
obviously enhances the catalysis in a number of possible ways
including desolvation of the ionic components, increasing the
energy of interaction of the oppositely charged reaction partners,
and stabilizing charged dispersed transition states relative to
charge localized reactants, but more work is required to delineate
these and other modes of action.
efficient enzymes, bearing in mind that the two catalysts operate
in different media. The present model system suggests that a
considerable enhancement in efficiency can be realized by a
synergistic tailoring of the catalyst and medium in which the
reaction occurs, and that this may be a general and exploitable
phenomenon in the creation of man-made catalysts for the
cleavage of phosphoesters and related materials.
It has been pointed out8 that enzymes that promote the
cleavage of phosphodiesters37 have kcat/KM values typically in
the range of 106-107 M-1 s-1 and so provide ∼20 kcal/mol of
stabilization energy to the transition state of the catalyzed
reaction relative to that of the background reaction. Energetics
calculations of the catalysis of the cyclization of 4a promoted
by complex 5 in water originally presented by Morrow and
Richard8 indicated that the ∆Gstabq for 5 and hydroxide-promoted
cyclizations was -9.3 kcal/mol or ∼50% of what could be
expected of the highly evolved and efficient protein catalysts.38
Acknowledgment. We gratefully acknowledge the financial
assistance of the Natural Sciences and Engineering Research
Council of Canada (NSERC), The Canada Council of the Arts
(CCA), and the Canada Foundation for Innovation (CFI). We
are indebted to Dr. Ruiyao Wang for his expertise in solving
the X-ray diffraction structure of 1-Zn(II)2:(-OH) and Dr. Dave
Edwards for helpful discussions. In addition, S.E.B. thanks
NSERC for an Undergraduate Summer Research Award, C.T.L.
thanks NSERC for a PGS-2 postgraduate scholarship, and R.S.B.
thanks the CCA for a Killam Research Fellowship, 2007-2009.
q
In the present case the ∆Gstab of -21 kcal/mol for 1-Zn(II)2:
Supporting Information Available: Tables of pseudo-first-
order rate constants for reactions of 4b-g with 1-Zn(II)2:
(-OCH3); NMR and mass spectroscopic data for the identity
of substrates 4a-g; structural report and CIF file for 1-Zn
(II)2(-OH)(CF3SO3-)3(CH3OH). This material is available free
of charge via the Internet at http://pubs.acs.org.
(-OCH3)-promoted cyclization of 4a in methanol would appear
to have approached that which is maximally expected for such
(37) Raines, R. T. Chem. ReV. 1998, 98, 1045.
(38) The reader will appreciate that any catalyst offering 50% of the 20 kcal/
mol of energy of stabilization of the protein system will still be 10 million
times less efficient than the enzyme in terms of rate constant. See Tang,
M.-Y.; Iranzo, O.; Richard, J. P.; Morrow, J. R. J. Am. Chem. Soc. 2005,
127, 1064.
JA076847D
9
16248 J. AM. CHEM. SOC. VOL. 129, NO. 51, 2007