A study of the kinetics of La3+-promoted methanolysis of S-aryl methylphosphonothioates: Possible methodology for decontamination of EA 2192, the toxic byproduct of VX hydrolysis

Article abstract of DOI:10.1021/ic200007v

The kinetics of the La³⁺-catalyzed methanolysis of a series of S-aryl methylphosphonothioates (4a-e, phenyl substituents = 3,5-dichloro, 4-chloro, 4-fluoro, 4-H, 4-methoxy) were studied at 25 °C with ss pH control. The reaction involves saturation binding of the anionic substrates to dimeric La³⁺/methoxide catalysts formulated as La₂ ³⁺(^-OCH₃)_x, where x = 2-5 depending on the solution ss pH. Cleavage of the La³⁺-bound methylphosphonothioates is fast, ranging from 5 × 10^-3 s ^-1 to 5.5 × 10^-5 s^-1 for substrates 4a-e at a ss pH of 8.4 and 1.6 × 10^-1 s^-1 to 4 × 10^-3s^-1 at a ss pH of 11.7. The rate accelerations for the methanolysis of substrates 4a-e, relative to their background methoxide-promoted reactions, average 7 × 10¹⁰ and 1.5 × 10⁹, respectively, at sspH's of 8.4 and 11.7. The catalytic system is predicted to cleave EA 2192 (S-2(N,N-di-iso-propylaminoethyl) methylphosphonothioate), a toxic byproduct of the hydrolysis of VX, with a t_1/2 between 4 and 8 min at a ss pH of 8.4, and 27 min at a ss pH of 11.7.

Full text of DOI:10.1021/ic200007v

ARTICLE
pubs.acs.org/IC
A Study of the Kinetics of La3þ-Promoted Methanolysis of S-Aryl
Methylphosphonothioates: Possible Methodology for
Decontamination of EA 2192, the Toxic Byproduct of VX Hydrolysis
Basab B. Dhar, David R. Edwards, and R. Stan Brown*
Department of Chemistry, Queen’s University Kingston, Ontario, Canada, K7L 3N6
S Supporting Information
b
ABSTRACT: The kinetics of the La3þ-catalyzed methanolysis
of a series of S-aryl methylphosphonothioates (4a-e, phenyl
substituents = 3,5-dichloro, 4-chloro, 4-fluoro, 4-H, 4-methoxy)
were studied at 25 °C with sspH control. The reaction involves
saturation binding of the anionic substrates to dimeric La3þ
/
methoxide catalysts formulated as La23þ(-OCH3)x, where x =
2-5 depending on the solution sspH. Cleavage of the La3þ-bound methylphosphonothioates is fast, ranging from 5 ꢀ 10-3 s-1 to
5.5 ꢀ 10-5 s-1 for substrates 4a-e at a sspH of 8.4 and 1.6 ꢀ 10-1 s-1 to 4 ꢀ 10-3 -1
s
at a sspH of 11.7. The rate accelerations for the
methanolysis of substrates 4a-e, relative to their background methoxide-promoted reactions, average 7 ꢀ 1010 and 1.5 ꢀ 109,
s
s
respectively, at pH’s of 8.4 and 11.7. The catalytic system is predicted to cleave EA 2192 (S-2(N,N-di-iso-propylaminoethyl)
methylphosphonothioate), a toxic byproduct of the hydrolysis of VX, with a t1/2 between 4 and 8 min at a sspH of 8.4, and 27 min at a
sspH of 11.7.
1. INTRODUCTION
The phosphonofluoridate G agents (1) and phospho-
nothioate V agents (2) are potent acetylcholine esterase
inhibitors that have been widely employed as chemical
warfare materials. Such G agents readily undergo hydrolysis
of the P-F bond to yield a nontoxic O-alkyl methylpho-
sphonic acid product. For example, Sarin (1a) is hydrolyzed
with a halftime (t1/2) of <6 min in water at a pH of 10 and
As part of an ongoing research program to develop methods to
at 25 °C to produce O-iso-propyl methylphosphonate.1 By
deactivate organophosphate CW agents and pesticides, we have
contrast, the base-promoted hydrolysis of V agents such as
investigated the potential of various metal ions to promote the
VX (2a) and Russian-VX (2b) is slower and more proble-
alcoholysis of a variety of neutral carbon esters and organopho-
matic, owing to their propensity to undergo a P-OR
sphorous (OP) esters. These metal-ion catalyzed alcoholysis
cleavage in competition with the desired P-SR cleavage
(MICA) reactions, using lanthanides (Ln3þ) and complexes of
reaction. For example, hydrolysis of VX with 0.1 M NaOH
certain transition metal ions such as Zn2þ and Cu2þ, are very
produces an 87% yield of the desired cleavage product 3a
effective for cleaving neutral phosphate and phosphorothioate
(O-ethyl methylphosphonic acid, EMPA) and 13% 3b S-(N,
triesters,8 O-ethyl-O-aryl methylphosphonate esters,9 PdS-con-
N-(diethylaminoethyl) methylphosphonic acid, commonly
taining phosphorothioate triesters,10 and carboxylate esters.11
known as EA 2192) arising from P-OEt cleavage (eq 1).2
Changing from water to the light alcohols leads to increased
Being some 3700 times less reactive to hydroxide nucleo-
solubility of the neutral OP substrates and metal ions (as their
philic attack than VX, EA 2192 is highly resistant to further
kinetically active alkoxide forms), and in general, metal-ion
degradation and yet retains a high level of oral and i.v.
catalyzed alcoholysis in methanol and ethanol is far more rapid
toxicity.3 Alternative simple strategies for the decomposi-
than the corresponding hydrolysis in water.12 Importantly, the
tion of VX that avoid formation of EA 2192 are much sought
overall process is one of transesterification and not true hydro-
after, and highly activated stoichiometric oxidants and
lysis. This is significant because the hydrolysis of neutral sub-
nucleophiles such as H2O2, O-iodosylcarboxylates, oxone,
strates leads to the production of acidic OP products that can
and NaOCl have been investigated for their potential to
bind, as their anions, to the metal ion and inhibit catalysis.
degrade VX.4 Other hydrolytic methodologies including the
use of latex dispersions,5 glycerophosphodiesterases,6 and
metallocene complexes7 [e.g., Cp2Mo(OH)(OH2)] have
also been disclosed.
Received: January 3, 2011
Published: March 09, 2011
r
2011 American Chemical Society
3071
dx.doi.org/10.1021/ic200007v Inorg. Chem. 2011, 50, 3071–3077
|

Inorganic Chemistry
ARTICLE
We recently reported13 that the methanolysis of a series of VX
simulants (O-ethyl S-aryl methylphosphonates) was strongly
catalyzed by [La3þ(-OCH3)]2 and [Zn(II)(-OCH3)(1,5,9-tri-
azadodecane)], which allowed us to predict a 0.3 and 18 s half-
time for the destruction of VX in the presence of 1 mM of each
respective catalyst at near neutral pH. A major advantage of using
a light alcohol such as methanol or ethanol for the decomposition
of VX is that any competitive P-OEt cleavage that accompanies
the desired P-SR displacement from 2a or 2b results in a neutral
transesterification intermediate (MeOP(dO)(CH3)(SR)) that
is subject to further reaction to displace the P-SR group. The
predicted rapid La3þ-promoted transformation of VX into its
methanolysis product (O-ethyl O-methyl methylphosphonate,
EMMP) was subsequently verified by tests with a live agent
performed at the Edgewood Chemical and Biological Center
(ECBC).14
b. Methods. The sspKa values in methanol for the thiophenol leaving
groups of 4a-4e were obtained from previous work.13 The sspH values in
methanol reported here were determined by subtracting the correction
constant of -2.24 from the electrode readings as described.16 The sspH
values (defined as in ref 16) in the kinetic experiments were measured at
the end of the reactions to avoid the possibility of chloride leaching from
the electrode (Cl- is an inhibitor of La3þ catalysis in methanol). For the
titrations of La(OTf)3 and 4e, the [CH3OH2þ] was determined
potentiometrically using a combination glass electrode (Fisher Scientific
Accumet electrode model #13-620-183A) calibrated with certified
standard aqueous buffers (pH = 4.00 and 10.00). The values of the
species formation constants in methanol were calculated using the
computer program Hyperquad 2000 (version 2.1 NT),17 with the
autoprotolysis constant of pure methanol taken to be 10-16.77 at 25.0 (
0.1 °C .18 The formation constants for the species (La3þ)2:4e:(CH3O-)n
(n = 1-5) were determined from the analysis of the potentiometric
titration of a methanol solution containing 2 mM La(OTf)3 and 1 mM 4e.
The results for these experiments can be found in the Supporting
Information.
Herein, we describe an extension of these studies13,14 demon-
strating that La(OTf)3 in methanol cleaves anionic phospho-
nothioate models, namely, S-aryl methylphosphonothioates,
4a-4e. The linear free energy study of the methanolysis of these
allows us to predict a t1/2 for the decomposition of EA 2192 of
∼4-8 min at 25 °C and essentially neutral pH in methanol.
c. Kinetics. La3þ-catalyzed reactions were followed by UV/vis
spectrophotometry under buffered conditions. Reaction progress at a
sspH of 8.4 was monitored at 242 nm (4a), 246 nm (4b), 235 nm (4c),
s
s
242 (4d), and 240 nm (4e), while at a pH of 11.7, reactions were
monitored at 282 nm (4a), 282 (4b), 270 (4c), 280 (4d), and 262 nm
(4e). Amine buffers, partially acidified with triflic acid, were employed to
s
control pH: N-methylimidazole (sspKa = 7.60), N-ethylmorpholine
s
(sspKa = 8.28), and triethylamine (sspKa = 10.78). Total [buffer] varied
from (1-4) ꢀ 10-2 M.
d. Identification of Reaction Products in the La3þ-Cata-
lyzed Cleavage of 4b. Substrate 4b (1 ꢀ 10-4 M), La(OTf)3 (1.5 ꢀ
10-3 M), and a NEt3 buffer (40 mM, pH 11.7) were combined in a
s
s
2. EXPERIMENTAL SECTION
1-cm-path-length cell and the UV-vis spectrum obtained following
complete conversion. The spectrum was identical to that of a methanol
solution containing 4-chlorothiophenol (1 ꢀ 10-4 M), La(OTf)3 (1.5
ꢀ 10-3 M), and NEt3 (40 mM, sspH 11.7).
a. Materials. Kinetic reactions were carried out in commercially
available EMD Drisolv Methanol (99.8%, anhydrous). All other solvents
used in the study were anhydrous and subsequently treated with an
Innovative Technology, Inc. Pure Solv solvent purification system
comprising alumina packed columns. Reagents, sodium methoxide
(0.50 M solution in methanol, titrated against N/2 certified standard
aqueous HCl solution and found to be 0.50 M), La(CF3SO3)3, CF3SO3H
(titrated to be 167 mM in ethanol), thiophenol (97%), and triethylamine
(99%) were purchased from Aldrich and used without further purifica-
tion. The compounds 4-fluorothiophenol (98%) and 3,5-dichlorothio-
phenol (97%) were obtained from TCI America. 4-Chlorothiophenol
(98%) was obtained from Acros Organics. 4-Methoxythiophenol (97%)
was obtained from Alfa Aesar, and methylphosphonyl dichloride
(97%þ) was purchased from Fluka.
e. Determination of Second Order Rate Constants for
Base-Promoted Methanolysis of 4a-4e. An aliquot of a 1 M
solution of tetrabutylammonium hydroxide in methanol was added
individually to substrates 4a-4e in a standard 5 mm NMR tube. It is
assumed that all hydroxide is converted to methoxide under these
reaction conditions.19 Substrate concentrations ranged from 7 to
16 mM, and in all cases, the base consumed during the course of the
reaction was less than 5%, so standard pseudo-first-order reaction
conditions are assumed. The NMR tubes were thermostatted at
25 °C, and the reaction progress was determined by 31P NMR. Fits
of the integrated intensities of 31P signals corresponding to substrates
4a-e and the product O-methyl methylphosphonate to a standard
single exponential equation provided the observed pseudo-first-order
rate constants (kobs-OMe) at a given [base]. The second order rate
constant (k2-OMe) was assigned as the gradient of the plot of
kobs-OMe vs [NBu4OMe].
The phosphonothioates (4) were prepared from methylphosphonyl
dichloride following a general route.15 Thiophenol (5 mmol) was added
to a solution of methylphosphonyl dichloride (2.5 mmol) in 10 mL of
dry CH2Cl2, after which triethylamine (5 mmol) was slowly added to the
mixture. The reaction mixture was stirred at 40 °C for 4 h, after which
the crude reaction mixture was washed with water (6 ꢀ 15 mL) and the
combined aqueous layers were extracted with CH2Cl2 (3 ꢀ 15 mL). The
combined organic layers were washed with brine and then dried over
anhydrous Na2SO4. The crude product, obtained after filtering away the
drying agent and removing the solvent, was added to a well-stirred
mixture of 3 mL of 5 M KOH in 16 mL of 1,4-dioxane; stirring was
continued at 50 °C for 8 h. The solution was concentrated under
reduced pressure to remove 1,4-dioxane, and the aqueous layer was
acidified with 20 mL of 3 M aqueous HCl and subsequently extracted
with CH2Cl2 (3 ꢀ 15 mL). The combined extracts were washed with
brine and dried over Na2SO4. CH2Cl2 was removed under reduced
pressure, and 4a-4e were obtained in nonoptimized yields of 40-80%.
The physical data for 4a-4e are given in the Supporting Information.
3. RESULTS
a. Identification of Reaction Products. Substrate 4b was
subjected to methanolysis at 25 °C in the presence of 1.5 ꢀ
10-3 M La(OTf)3 at a sspH of 11.7, maintained with a 40 mM
triethylamine buffer partially acidified with triflic acid. Com-
parison of the UV-vis spectra obtained after complete
reaction to that of an authentic sample of 4-chlorothiophenol
(ε275 nm = 1.38 ꢀ 104 M-1 cm-1) under experimental condi-
tions confirmed methanolytic cleavage of the P-SAr bond as
depicted in eq 2.
3072
dx.doi.org/10.1021/ic200007v |Inorg. Chem. 2011, 50, 3071–3077

Inorganic Chemistry
ARTICLE
Scheme 1. Simplified Scheme for the La3þ-Catalyzed
Methanolysis of 4a from which eq 3 is Obtained
Figure 1. Plot of kobs versus [La(OTf)3] for the methanolysis of 4a
s
(1 ꢀ 10-4 M) determined at 25.0 ( 0.1 °C; pH = 8.4 (0.04 M N-
s
ethylmorpholine buffer). NLLSQ fit of the data to eq 3 provides k1 =
some phosphate diesters were strikingly similar to those presented
here.20
(28.0 ( 3.0) M-1 s-1 and kcat8.4 = (5.9 ( 0.2) ꢀ 10-3 s-1
.
0:5
kobs ¼ ð2k1Kðð1 þ 8½La3þꢁt=KdÞ - 1Þ þ kcatð4½La3þꢁt=Kd
0:5
- ð1 þ 8½La3þꢁt=KdÞ þ 1ÞÞ=ð8K=Kd þ 4½La3þꢁt=Kd
0:5
- ð1 þ 8½La3þꢁt=KdÞ þ 1Þ
ð3Þ
pH
c. Brønsted Correlation of kcat Values. Phosphonothio-
ates 4a-4e (1 ꢀ 10-4 M) were subjected to La3þ-catalyzed
methanolysis at 25 °C under buffered conditions at sspH’s of 8.4
(N-ethylmorpholine) and 11.7 (triethylamine). The [La(OTf)3]
was varied between 1 and 2.5 mM, and the observed pseudo-first-
order rate constants (kobs) were determined to be independent of
the [La3þ
]total. This indicates that the substrate is fully bound to
the La3þ species at these concentrations, and so kobs is the first-
order rate constant for cleavage of the S-aryl group from a
complex comprising (La3þ)2, (methoxide)x, and a bound 4.
Figure 2. Plot of kobs versus [La(OTf)3] for the methanolysis of 4a (1 ꢀ
10-4 M) determined at 25.0 ( 0.1 °C; sspH = 11.7 (0.04 M triethylamine
buffer). NLLSQ fit of the data to eq 3 gives a k1 indistinguishable from
The averages of the various kobs values are given as the kcat8.4 and
11.7
kcat
values reported in Table 1. The data are presented
0 M-1 s-1 and a kcat11.7 = (1.64 ( 0.01) ꢀ 10-1 s-1
.
graphically in Figure 3 as the Brønsted correlations that fit linear
regressions of log(kcat8.4) = (-0.67 ( 0.03)sspKHa SAr þ (3.89 (
0.43) and log(kcat11.7) = (-0.55 ( 0.04)sspKHa SAr þ (4.15 (
b. La3þ-Catalyzed Methanolysis of Phosphonothioate 4a.
Shown in Figures 1 and 2 are plots of kobs versus [La3þ] obtained
for the decomposition of 4a at a sspH of 8.4 and 11.7, respectively.
The rate data obtained as a function of [La3þ] were fit to eq 3,
derived for the simplified reaction mechanism presented in
Scheme 1. In this model, k1 refers to a bimolecular reaction
between the anionic substrate and metal ion, kcat represents
the unimolecular rate constant for a substrate fully bound to a
dimeric La3þ -complex and the constants K and Kd are defined as
0.45) at
a
sspH of 8.4 and 11.7. Substrates 4a-4e
(1 ꢀ 10-4 M) were also subjected to reaction at 25 °C employing
a commercially available La3þ-containing decontamination solu-
tion known as System 421 operating at a sspH of 9.2. The observed
first order rate constants, kSyst4, for this series of experiments
are presented in Table 1, and the experimental protocols are given
in the Supporting Information. The Brønsted correlation for the
System 4 reactions shown in Figure 3 is log(kSyst4) = (-0.72 (
0.03)sspKHa SAr þ (4.22 ( 0.33).
indicated in Scheme 1 (the kcatpH referred to later is defined as
d. Potentiometric Titration of La(OTf)3 in the Presence of
Phosphonothioate 4e. In Figure 18S (Supporting Information)
are two titration curves (sspH vs added [-OMe]) acquired from a
2 mM solution of La(OTf)3 in methanol with and without 1 mM
4e, while in Table 18S (Supporting Information) are compiled
the computed stability constants and the associated microscopic
sspK2a:1x data. The calculated speciation diagram illustrating the
s
s
the kcat at specific pH values). The errors associated with the
individual constants from the multiparameter nonlinear least-
squares fits are quite large, owing to the fact that the four fitted
constants are heavily correlated. A simpler method for determin-
ing the experimental kcatpH values takes an average rate constant
in the plateau regions seen in Figures 1 and 2 at high [La3þ]. The
so-determined kcatpH values for reaction of the metal bound4a are
(5.9 ( 0.2) ꢀ 10-3 s-1 and (1.64 ( 0.01) ꢀ 10-1 s-1 at a sspH of
8.4 and 11.7, respectively. This treatment of the rate
data for the La3þ-catalyzed methanolysis of phosphonothioate 4a
is the same as was presented earlier for the La3þ-catalyzed cleavage
of phosphate diesters containing aryloxy leaving groups where
the rate constant vs [La3þ} profiles for catalyzed methanolysis of
concentrations of the various [(La3þ)2:4e:(-OCH3)x] complexes
s
s
as a function of the pH determined at 2 mM La(OTf)3 and
1 mM 4e is shown in Figure 19S (Supporting Information).
The dependence of the rate of La3þ-catalyzed methanolysis of
substrate 4b on [-OMe] was determined at 7.6 e sspH e 11.7 at
pH
25 °C under buffered conditions, and the kcat values for
cleavage of 4b when bound to La3þ2(CH3O-)x, determined
3073
dx.doi.org/10.1021/ic200007v |Inorg. Chem. 2011, 50, 3071–3077

Inorganic Chemistry
ARTICLE
Table 1. Rate Constants k2-OMe for Methoxide Reaction of Substrates 4, kcat8.4 and kcat11.7 (at a sspH of 8.4 and 11.7), and kSyst4
Determined for the La3þ-Catalyzed Methanolysis of Substrates 4, T = 25 ( 0.1 °C
-OMe
8.4
11.7
k2
(M-1 s-1
kcat ꢀ 103
kcat ꢀ 102
rate accel.d
sspH 11.7
b
sub
sspKHa SAra
)
(s-1
)
c sspH 8.4
rate accel.d sspH 8.4
(s-1
)
e sspH 11.7
)
kSyst4 ꢀ 103 (s-1 f sspH 9.2
4a
4b
4c
4d
4e
9.08
10.47
11.07
11.28
11.98
(4.7 ( 0.1) ꢀ 10-5
(2.7 ( 0.1) ꢀ 10-6
(9.8 ( 0.2) ꢀ 10-7
(6.1 ( 0.4) ꢀ 10-7
(1.5 ( 0.4) ꢀ 10-7
5.9 ( 0.2
0.58 ( 0.05
0.33 ( 0.03
0.230 ( 0.001
0.055 ( 0.005
2.9 ꢀ 1010
5.1 ꢀ 1010
8.0 ꢀ 1010
8.9 ꢀ 1010
8.6 ꢀ 1010
16.4 ( 0.1
1.79 ( 0.08
1.22 ( 0.01
0.8 ( 0.1
4.0 ꢀ 108
7.9 ꢀ 108
1.2 ꢀ 109
1.6 ꢀ 109
3.1 ꢀ 109
5.6 ( 0.1
0.5 ( 0.1
0.40 ( 0.09
0.047 ( 0.003
a Data as reported in ref 13. b Second order rate constants, k2-OMe, determined from slopes of plots of kobs vs [NBu4OMe], 25 oC. c kcat8.4 determined
from the plateau region in the plots of kobs vs [La3þ] at a high metal ion concentration, a sspH of 8.4 (0.04 M N-ethylmorpholine buffer), and T = 25 °C.
d The rate accelerations are computed from the kcatpH/k-OMe ratio, where k-OMe has units of s-1 and is determined as k2
ꢀ [-OMe] at either sspH
-OMe
8.4 or 11.7 using the autoprotolysis constant Kauto = 10-16.77 for methanol; [-OMe] = Kauto/[Hþ]. e kcat11.7 determined from the plateau region in the
plots of kobs versus [La3þ] at a high metal ion concentration at a sspH of 11.7 (0.04 M NEt3 buffer) and 25 °C. f kSyst4 determined from duplicate runs
using System 4 decontamination system.21
Table 2. Table of Catalytic Rate Constants, kcatpH, for the
La3þ-Catalyzed Methanolysis of 4b Determined under Buf-
fered Conditions at 25.0 ( 0.1 °C from a sspH of 7.6-11.7a
sspH
kcat ꢀ 103 (s-1
pH
)
7.6
8.4
0.32 ( 0.05
0.58 ( 0.01
3.5 ( 0.1
8.9
9.9
7.3 ( 0.2
10.9
11.7
12.4 ( 0.6
17.9 ( 0.8
a Catalytic rate constants computed as described in the text as the
average of the values in the plateau region seen at high [La3þ] in the
plots of kobs vs [La3þ].
Figure 3. Brønsted plots for (a) La3þ-catalyzed methanolysis of 4a-4e
Table 3. Kinetic Constants (kcat2:1:x) for the Methanolysis of
4b Catalyzed by the Various (La3þ)2:4b:(CH3O-)x Species
Present at [La3þ] = 2.0 mM and 25.0 ( 0.1 °C
s
determined at a pH of 11.7, T = 25.0 ( 0.1 °C, linear regression
s
log(kcat11.7) = (-0.55 ( 0.04)sspKHa SAr þ (4.15 ( 0.45), r2 = 0.98 (solid
line, b); (b) La3þ-catalyzed methanolysis of 4a-4e determined at a sspH
of 8.4, T = 25.0 ( 0.1 °C, linear regression log(kcat8.4) = (-0.67 (
0.04)sspKHa SAr þ (3.89 ( 0.43), r2 = 0.99 (dashed line, /); (c) reaction of
4a, 4b, and 4e with the System 4 decontamination solution at T = 25.0 (
0.1 °C, linear regression log(kSyst4) = (-0.72 ( 0.03)sspKHa SAr þ (4.22 (
0.33), r2 = 0.99 (solid line, Δ); and (d) methoxide promoted reaction of
4a-4e linear regression log(k2-OMe) = (-0.85 ( 0.02)sspKHa SAr þ (3.40
( 0.17), r2 = 0.9989 (solid line, 0).
(La3þ)2:4b:(CH3O-)x
x = 1
x = 2
x = 3
x = 4
x = 5
kcat2:1:x a (s-1
)
0b
(2.5 ( 1.0) (19 ( 5) (11 ( 3) (27 ( 4)
a kcat2:1:x determined using a fit of the data to eq 4. b kcat2:1:1 constrained
to a value of zero for fitting of the sspH/rate data to eq 4. All attempts to
carry out an unconstrained fit resulted in excessively large errors.
pH
as described above, are presented in Table 2. A plot of log(kcat
)
= (0.28 ( 0.07) sspKa2:1:x - (1.9 ( 0.8), r2 = 0.88.
versus sspH (not shown) provides a poor linear correlation with a
slope of 0.44 ( 0.07 and r2 = 0.8997. The kcatpH values are shown
overlaid on the speciation diagram of Figure 19S to give an
indication of the pH dependence for the decomposition of the
(La3þ)2:4b:(-OCH3)x complexes. The kcatpH vs sspH data were
then analyzed as a linear combination of the individual rate
constants by fitting to eq 4 to determine the first order rate
constants (kcat2:1:1, kcat2:1:2, ...kcat2:1:x) for spontaneous de-
composition of the specific metal ion complexes
[(La3þ)2:4b:(-OCH3)x]; where x = 1-5. This analysis rests
on a reasonable assumption that the complex stability constants
k2 ¼ ∑ðkcat ½ðLa3þÞ2 : Sub
: ðCH3O-ÞxꢁÞ=½La2ꢁtotal
cat
2:1:x
ð4Þ
e. Methoxide Promoted Cleavage of 4a-4e. The metha-
nolysis of substrates 4a-4e was carried out at 25 °C under
pseudo-first-order conditions of excess NBu4OMe, and the rates
were determined by integration of the 31P NMR peaks corre-
sponding to unreacted starting material and product. The
intensity vs time data were fit to a standard first order exponential
equation to obtain the kobs values, and the second order rate
constants, k2-OMe, were obtained as the slopes of the linear plots
of kobs vs [NBu4OMe], the values of these being given in Table 1.
with 4b can be accurately modeled by those of the slower
2:1:x
reacting material, 4e. The first order rate constants, kcat
,
appear in Table 3, and a plot of log(kcat2:1:x) versus the
s
2:1:x
microscopic pKa2:1:x values from Table 2 is fit to log(kcat
)
s
3074
dx.doi.org/10.1021/ic200007v |Inorg. Chem. 2011, 50, 3071–3077

Inorganic Chemistry
ARTICLE
-OMe
Included in Figure 3 is a Brønsted correlation of the k2
Scheme 2. Proposed Mechanism for the La3þ-Catalyzed
Methanolysis of 4a-4e Where the Charges and Associated
Counter Ions Have Been Omitted for Visual Clarity
values vs the corresponding sspKaHSAr values of the HSAr leaving
groups: log(k2-OMe) = (-0.85 ( 0.02)sspKHa SAr þ (3.40 ( 0.17).
When viewed as pseudo-first-order rate constants obtained at a
standard state of 1 M, the methoxide data presented in Figure 3
can be easily compared with the metal-catalyzed reactions. The
raw rate data and 31P NMR spectra are included in the Support-
ing Information.
4. DISCUSSION
2,4-dinitrophenyl phosphate,24 the Ln(OH)2þ-promoted hy-
drolysis of bis-4-nitrophenyl phosphate,25 and the metal-
hydroxo attack on CdO-containing compounds.26 In the
present case, the βNu of þ0.28 for the La3þ-catalyzed metha-
nolysis of 4b is in the range expected for a bifunctional process
involving delivery of a metal-coordinated nucleophile and
Lewis acid activation of the coordinated phosphorothioate.
Plots of kobs versus [La3þ] for the catalyzed methanolysis of 4a
at sspH’s of 8.4 and 11.7 are complex, albeit consistent with the
simplified model shown in Scheme 1 derived previously for the
related La3þ-catalyzed methanolysis of O-methyl O-aryl phos-
phodiesters.20 Acceptable fits of the rate data to eq 3, derived for
the simplified process given in Scheme 1, appear in Figures 1 and
2. However, the kinetic constants are heavily correlated, so a
simpler treatment of the rate data is applied as described in the
Results section, whereby the kobs constants for the methanolysis
of La3þ2-bound 4 determined between 1 and 2.5 mM were
In Scheme 2 is a mechanism originally proposed for the La3þ
-
catalyzed cleavage of monoanionic phosphate diesters,20 which is
consistent with the present results for the catalyzed methanolysis
of monoanionic phosphonothioate substrates. The common
feature is the bis-coordination of the two phosphate or phospho-
nothioate oxygens to the dimeric La3þ-complex to provide double
Lewis acid activation toward nucleophilic attack by a metal-
coordinated methoxide. A bridging methoxide bound by two
La3þ ions likely does not possess sufficient nucleophilicity to
displace the -SAr leaving group, and accordingly we propose
that the nucleophilic CH3O- is coordinated to a single metal ion.
The Brønsted plots for the La3þ-catalyzed methanolysis of
4a-4e at a sspH of 11.7 and 8.4 have very similar gradients (βLG
coefficients) of -0.55 ( 0.04 and -0.67 ( 0.04, respectively,
but given the bifunctional nature of the catalysis and lack of
pH
obtained, and averaged, to give kcat (the rate constant for
cleavage of 4 by the La3þdimer at a desired sspH), which was used
for all further analyses.20
The methanolytic cleavage of anionic phosphonothioates
promoted by La3þ is relatively insensitive to lyoxide concentra-
tion, as exemplified by the reaction of 4b, which displays less than
a 60-fold rate difference (3.2 ꢀ 10-4 to 1.8 ꢀ 10-2 s-1) over a
four sspH unit increase. The plot of log(kcat) vs sspH has a nonunit
slope of 0.44. This result is consistent with the spontaneous
decompositions of a set of pH dependent complexes formulated
s
s
simply as (La3þ)2:4:(CH3O-)x; where x = 1-5. The computed
kinetic constants given in Table 2 pertaining to the decomposi-
tion of each of the individual complexes, kcat2:1:x indicate that the
(La3þ)2:4:(CH3O-)x complexes with 3, 4, and 5 associated
methoxides (or their kinetic equivalents where an external
methoxide strikes the La3þ-complex having 2, 3, or 4 associated
methoxides) are the most active, although modest activity is also
attributed to the x = 2 species. From the speciation diagram of
Figure 19S (Supporting Information) and the catalytic constants
kcat2:1:x, we compute that 96% of the reaction occurs by the
decomposition of (La3þ)2:4b:(CH3O-)2 with about 4% coming
from (La3þ)2:4b:(CH3O-)3 at a sspH of 8.4. At a sspH of 11.7, the
complexes (La3þ)2:4b:(CH3O-)4 and (La3þ)2:4b:(CH3O-)5
account for 41 and 54% of the reaction with <5% coming from
the decomposition of (La3þ)2:4b:(CH3O-)3.
β
Equilibrium values for phosphonyl transfer between oxyanion and
thiolate nucleophiles, a detailed analysis of the transition state
charges or extent of cleavage of the P-SAr bond cannot be made.
Phosphonothioates 4a, 4b, and 4e are also readily cleaved by the
System 4 decontamination solution,21 and the limited plot of
log(kSyst4) vs sspKHa SAr gives a βLG of -0.72 that can essentially be
overlaid on the plot constructed for reactions carried out at a sspH
of 8.4 (see Figure 3). The Brønsted coefficients determined here
are similar to that observed for the La3þ-catalyzed methanolysis
of the neutral OP V-agent simulants EtO(CH3)P(dO)SAr (sspH
9.1, βLG = -0.75)13 and suggest a similar level of charge
development on the leaving groups is occurring in the TS for
all of these reactions. The βLG of -0.85 determined for metho-
The Brønsted plot of log(k22:1:x) vs sspK2a:1:x for the cleavage of
substrate 4b is linear with a small positive slope of βNu = þ0.28. A
comparison can be made to the methanolytic cleavage of
phosphate triesters catalyzed by the series of metal alkoxides
(La3þ)2:(CH3O-)x where the βNu was determined to be þ0.02.22
The low value of the βNu found in that study was proposed to
result from a competing combination of increased nucleophi-
licity for the CH3O- attached to the species with the highest
sspKa values, coupled with a concomitant decreased Lewis
acidity/electrophilic activation of the coordinated substrate
due to the reduced net positive charge on the La3þ ions in
the complex. Slightly positive, or even negative, values of βNu
are reported for reactions promoted by metal coordinated
lyoxides, including the dinuclear Zn(II)-complex-promoted
hydrolysis of bis-4-nitrophenyl phosphate,23 the mononucle-
ar Zn(II) triamine-complex-catalyzed hydrolysis of diethyl
xide-promoted cleavage of 4a-4e indicates a slightly greater
s
dependence of the reaction rate on the pKHa SAr of the leaving
s
groups as compared to the metal-catalyzed reactions. Compar-
ison of the latter value and those obtained for the La3þ-promoted
processes suggests that the metal-containing complexes exhibit a
catalytic preference for less activated substrates with poorer
leaving groups.
The methoxide promoted reactions of 4a-4e have computed
second-order rate constants falling in the range of 1.5 ꢀ 10-7 to
4.7 ꢀ 10-5 M-1 s-1. A visual comparison of the rates of meth-
oxide reactions to those of the La3þ-catalyzed processes is seen in
Figure 3, if the methoxide second order rate constants are
brought to first order ones at the standard state of 1 M, T =
25 °C. The vertical separation of the various plots is indicative
of the level of catalysis achieved by coordinating the methoxide
and substrate to a (La3þ)2:(-OCH3)x-1 metal ion complex in
3075
dx.doi.org/10.1021/ic200007v |Inorg. Chem. 2011, 50, 3071–3077

Inorganic Chemistry
ARTICLE
the transition state to form [(La3þ)2:4:(-OCH3)x]‡. An alterna-
available System 4 decontamination solution21 operating at a sspH
-OMe
tive assessment of catalytic efficiency comes from the kcat/(k2
of 9.2 performed comparably to the simple La3þ-contaning
[-OMe]) value at the same pH, where [-OMe] is given as
solution at a pH of 8.4. This represents a significant advan-
s
s
s
s
K
Auto/[Hþ] or alternatively 10-16.77/10-pH. The computed rate
tage of metal ion catalyzed alcoholysis for the decomposition of
the CW agent VX,13 particularly munitions grade material or
that which has been exposed to adventitious hydrolysis during
storage.14
accelerations range from (2.9-8.6) ꢀ 1010 at a sspH of 8.4 to (0.4-
3.1) ꢀ 109 at a sspH of 11.7. Somewhat higher catalytic efficiencies
are calculated for the less activated substrates with higher sspKaHSAr
values as a direct result of the observed slopes in the Brønsted plots
of Figure 3.
’ ASSOCIATED CONTENT
a. Predicted Rate of Decomposition of EA 2192. A pre-
dicted rate acceleration for decomposition of this substrate can
be calculated if one assumes that the Brønsted correlations
described for cleavage of thioaryl leaving groups from 4 are appli-
cable to the reaction of EA 2192 and that the conditions for
cleavage have the La3þ-containing catalyst in excess of the EA
2192 as it is in this study. The leaving group of EA 2192 at a sspH
of 8.4 is the N-protonated zwitterion -SCH2CH2NþH(iPr)2,
and the sspKa of HSCH2CH2NþH(iPr)2 dissociating to form the
zwitterion in methanol is estimated to be 9.54.13 Inserting this
sspKa value into the Brønsted relationship, log(kcat8.4) = (-0.67 (
0.04)sspKHa SAr þ (3.89 ( 0.43) gives a predicted rate constant
(t1/2) of 3.15 ꢀ 10-3 s-1 (corresponding to a t1/2 of 4 min) for
S
Supporting Information. Descriptions of syntheses and
b
physical data of 4a-4e; 31P NMR spectra for determining the
rate constants for methoxide catalyzed cleavage of 4a-4e; proce-
dures for UV/vis kinetics and potentiometric titration of La3þ in
the presence of 4e; and speciation diagrams and the computed
stability constants for the La3þ2:4e:(-OCH3)x complexes (24
pages). This material is available free of charge via the Internet at
http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
*E-mail: rsbrown@chem.queensu.ca.
the cleavage of EA 2192 at a sspH of 8.4. Unfortunately, since we
s
s
do not know the pKa value for acid dissociation of the thiol
group of HSCH2CH2N(iPr)2, we cannot calculate the antici-
pated rate constant for the cleavage of EA 2192 at a sspH of 11.7.27
The large structural disparity between the thioaryl and (di-iso-
propylaminoethyl)thio- leaving groups makes it unclear as to
whether the same Brønsted relationship can be used for aryl
thiols and aliphatic thiols, so we consider an alternative estimate.
The methoxide-promoted cleavage reaction of EA 2192 is
reported4a to have a t1/2 of 140 h in 0.3 M methoxide, corres-
’ ACKNOWLEDGMENT
The authors gratefully acknowledge the generous financial
support of this work by the United States Defense Threat Reduc-
tion Agency—Joint Science and Technology Office, Basic and
Supporting Sciences Division through the award of grant
HDTRA-08-1-0046. In addition, they acknowledge stimulating
discussions and ongoing technical support from Dr. H. Dupont
Durst, Head, Chemical Methodology Team, Edgewood Chemi-
cal and Biological Center.
ponding to a second order rate constant of 4.6 ꢀ 10-6 M-1 -1
s
from which one can calculate kobs values of 2 ꢀ 10-14 and 4 ꢀ
10-11 s-1 at sspH’s of 8.4 and 11.7. The average rate accelerations
for cleavage of 4 by saturating La3þ catalysis relative to the
s
s
background methoxide promoted reactions provided at pH
’ REFERENCES
values of 8.4 and 11.7 are 7 ꢀ 1010 and 1.5 ꢀ 109, respectively.
Assuming that a similar level of catalysis is obtained with EA
2192, we predict that at saturating [La3þ] the t1/2 for cleavage of
EA 2192 at these sspH values is 8 and 27 min. The predicted t1/2
for the reaction of EA 2192 is likely an underestimate at the lower
sspH since it is expected to be N-protonated at a sspH of 8.4 and so
may result in a substantially more activated phosphonothioate
owing to a combination of nucleophilic attack on a formally
neutral substrate.
(1) Gustafson, R. L.; Martell, A. E. J. Am. Chem. Soc. 1962, 84, 2309.
(2) Yang, Y. C.; Berg, F. J.; SzaFraniec, L. L.; Beaudry, W. T.; Bunton,
C. A.; Kumar, A. J. Chem. Soc., Perkins Trans. 2 1997, 607.
(3) Munro, N. B.; Talmage, S. S.; Griffin, G. D.; Waters, L. C.;
Watson, A. P.; King, J. F.; Hauschild, V. Environ. Health Perspect. 1999,
107, 933.
(4) For reviews, see: (a) Yang, Y.-C. Acc. Chem. Res. 1999, 32, 109.
(b) Morales-Rojas, H.; Moss, R. A. Chem. Rev. 2002, 102, 2497.
(5) Seabolt, E. E.; Ford, W. T. Langmuir 2003, 19, 5378.
(6) Ghanem, E.; Li, Y.; Xu, C.; Raushel, F. M. Biochemistry 2007,
46, 9032.
(7) Kuo, L. Y.; Adint, T. T.; Akagi, A. E.; Zakharov, L. Organometallics
2008, 27, 2560.
(8) Liu, C. T.; Neverov, A. A.; Tsang, J. S. W.; Brown, R. S. Org.
Biomol. Chem. 2005, 3, 1525.
(9) Lewis, R. E.; Neverov, A. A.; Brown, R. S. Org. Biomol. Chem.
5. CONCLUSION
La3þ ions in methanol are shown to be highly effective for
cleaving anionic S-aryl methylphosphonates (4). The reaction
proceeds via a pre-equilibrium binding of the substrate to a
dimeric La3þ complex formulated as La3þ2:(-OCH3)x to form
2005, 3, 4082.
La3þ2:4:(-OCH3)x, where x can assume values of 2-5 depend-
(10) Lu, Z. L.; Neverov, A. A.; Brown, R. S. Org. Biomol. Chem. 2005,
3, 3379.
(11) Neverov, A. A.; Sunderland, N. E.; Brown, R. S. Org. Biomol.
Chem. 2005, 3, 65.
(12) Brown, R. S.; Neverov, A. A. Adv. Phys. Org. Chem. 2008, 42, 271.
(13) Melnychuk, S. A.; Neverov, A. A.; Brown, R. S. Angew. Chem.,
Int. Ed. Engl. 2006, 45, 1767.
(14) Durst, H. D.; Sumpter, K.; Brown, R. S.; Neverov, A. A.;
Andrea, T. Abstracts of Papers, 238th ACS National Meeting,
Washington, DC, United States, August 16-20, 2009; American
s
s
ing on the pH of the solution. Methanolytic cleavages of the
phosphorothioate bound to the catalyst relative to cleavage by
the background methoxide reactions at sspH values of 8.4 and 11.7
are 7 ꢀ 1010 and 1.5 ꢀ 109. The La3þ2:(-OCH3)x system is also
predicted to be active for the decontamination of EA 2192. The
t1/2 values for the methanolysis cleavage reaction of EA 2192 in
the presence of saturating La3þ are estimated to be 4-8 and
27 min at a sspH of 8.4 and 11.7, respectively. The commercially
3076
dx.doi.org/10.1021/ic200007v |Inorg. Chem. 2011, 50, 3071–3077

Inorganic Chemistry
ARTICLE
Chemical Society: Washington, DC, 2009; INOR-396. CODEN:
69LVCL AN 2009:983730.
(15) Kobayashi, Y.; Morisawa, F.; Saigo, K. J. Org. Chem. 2005, 71, 606.
(16) (a) Gibson, G.; Neverov, A. A.; Brown, R. S. Can. J. Chem. 2003,
81, 495.(b) For the designation of pH in non-aqueous solvents, we use
the nomenclature recommended by IUPAC: Compendium of Analytical
Nomenclature. Definitive Rules 1997, 3rd ed.; Blackwell, Oxford, U.K.,
1998. The pH meter reading for an aqueous solution using an electrode
calibrated with aqueous buffers is designated as wwpH. 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. In methanol, swpH-(-2.24) = sspH, and since the autoprotolysis
constant of methanol is 10-16.77, neutral sspH is 8.4.
(17) Hyperquad 2000 is a computer program that applies nonlinear
curve fitting to a titration by adjusting a series of formation constants
(log β) for species input to a model, published in: Gans, P.; Sabatini, A.;
Vacca, A. Talanta 1996, 43, 1739.
(18) (a) Canals, I.; Portal, J. A.; Bosch, E.; Rosꢀes, M. Anal. Chem.
2000, 72, 1802. (b) Bosch, E.; Rived, F.; Rosꢀes, M.; Sales, J. J. Chem. Soc.
Perkin Trans. 2 1999, 1953. (c) Rived, F.; Rosꢀes, M.; Bosch, E. Anal.
Chim. Acta 1998, 374, 309. (d) Bosch, E.; Bou, P.; Allemann, H.; Rosꢀes,
M. Anal. Chem. 1996, 68, 3651.
(19) The maximum concentration of -OH in ∼30 M methanol
solution for these NMR experiments is 16 mM, or 5000 times less than
the methanol. The autoprotolysis constant of water in methanol,
reported as pKH O(HS), while not known with precision, is reported
2
(Fonrodona, G.; Rꢁafols, C.; Bosch, E.; Marti Rosꢀes, M. Anal. Chem. Acta
1996, 335, 291) as 20 ( 2 in methanol, while that of methanol in
methanol is 16.77, such that only about 0.001% or less of the total
lyoxide is present as -OH, the rest being -OCH3.
(20) (a) Gibson, T. T. G.; Neverov, A. A.; Teng, A. C.-T.; Brown,
R. S. Can. J. Chem. 2005, 83, 1268. (b) Neverov, A. A.; Brown, R. S. Inorg.
Chem. 2001, 40, 3588.
(21) Solution 4 is a patent-protected commercially available MICA
solution for the decomposition of neutral organophosphate, phospho-
nate, phosphorothioate, and phosphorothionate esters as well as all OP-
based chemical weapons of the G- and V-agent classes. Information
regarding this solution can be obtained from Mr. Jason Hendry,
Manager, Commercial Development, Parteq Innovations, Queen’s
University, Kingston, Ontario, Canada, K7L 3N6, e-mail: J.Hendry@
parteqinnovations.com.
(22) Edwards, D. R.; Liu, C. T.; Garrett, G. E.; Neverov, A. A.;
Brown, R. S. J. Am. Chem. Soc. 2009, 131, 13738.
(23) Bencicni, A.; Berni, E.; Bianchi, A.; Fedi, V.; Giorgi, C.; Paoletti,
P.; Valtancoli, B. Inorg. Chem. 1999, 38, 6323.
(24) Itoh, T.; Tada, T.; Yoshikawa, Y.; Hisada, H. Bull. Chem. Soc.
Jpn. 1996, 1265.
(25) Bencicni Gomez-Tagle, P.; Yatsimirsky, A. Inorg. Chem. 2001,
40, 3786.
(26) Martin, R. B. J. Inorg. Nucl. Chem. 1976, 38, 511.
(27) The predicted rate constant for cleavage of EA 2192 by Solution
4 can be estimated knowing the operating sspH of 9.2, the speciation of
the La3þ dimeric catalyst present at that sspH, and considering that the
amino group of the EA 2192 should be protonated so the relevant sspKa
of the N,N-di-iso-propylaminoethanothiol is 9.54.
3077
dx.doi.org/10.1021/ic200007v |Inorg. Chem. 2011, 50, 3071–3077