Mechanistic studies of La3+ and Zn2+-catalyzed methanolysis of O-ethyl O-aryl methylphosphonate esters. An effective solvolytic method for the catalytic destruction of phosphonate CW simulants

Article abstract of DOI:10.1039/b511550g

The kinetics of methanolysis of six O-ethyl O-aryl methylphosphonates (6a-f) promoted by methoxide, La³⁺ and 1,5,9-triazacyclododecane complex of Zn²⁺(^-OCH₃) (5:Zn²⁺( ^-OCH₃)) were studied as simulants for chemical warfare (CW) agents, and analyzed through the use of Bronsted plots. The β_1g, values are, respectively, -0.76, -1.26 and -1.06, pointing to significant weakening of the P-OAr bond in the transition state. For the metal-catalyzed reactions the data are consistent with a concerted process where the P-OAr bond rupture has progressed to the extent of 84% in the La ³⁺ reaction and ca. 70% in the Zn²⁺ catalyzed reaction. The catalysis afforded by the metal ions is remarkable, being about 10 ⁶-fold and 10⁸-fold for poor and good leaving groups, respectively, relative to the background reactions at _s^spH 9.1. Solvent deuterium kinetic isotope studies for two of the substrates promoted by 5:Zn²⁺(^-OCH.,) give k_H/k _D = 1.0 ± 0.1, consistent with a nucleophilic mechanism. A unified mechanism for the metal-catalyzed reactions is presented which involves pre-equilibrium coordination of the substrate to the metal ion followed by intramolecular delivery of a coordinated methoxide. The Royal Society of Chemistry 2005.

Full text of DOI:10.1039/b511550g

View Article Online / Journal Homepage / Table of Contents for this issue
A R T I C L E
Mechanistic studies of La3+ and Zn2+-catalyzed methanolysis of
O-ethyl O-aryl methylphosphonate esters. An effective solvolytic
method for the catalytic destruction of phosphonate CW simulants†
Roxanne E. Lewis, Alexei A. Neverov and R. Stan Brown*
Department of Chemistry, Queen’s University, Kingston, Ontario, Canada K7L 3N6.
E-mail: rsbrown@chem.queensu.ca; Fax: 613-533-6669; Tel: 613-533-2400
Received 15th August 2005, Accepted 31st August 2005
First published as an Advance Article on the web 28th September 2005
The kinetics of methanolysis of six O-ethyl O-aryl methylphosphonates (6a–f) promoted by methoxide, La3+ and
1,5,9-triazacyclododecane complex of Zn2+(−OCH3) (5:Zn2+(−OCH3)) were studied as simulants for chemical warfare
(CW) agents, and analyzed through the use of Brønsted plots. The blg values are, respectively, −0.76, −1.26 and
−1.06, pointing to significant weakening of the P–OAr bond in the transition state. For the metal-catalyzed reactions
the data are consistent with a concerted process where the P–OAr bond rupture has progressed to the extent of 84%
in the La3+ reaction and ca. 70% in the Zn2+ catalyzed reaction. The catalysis afforded by the metal ions is
remarkable, being about 106-fold and 108-fold for poor and good leaving groups, respectively, relative to the
background reactions at sspH 9.1. Solvent deuterium kinetic isotope studies for two of the substrates promoted by
5:Zn2+(−OCH3) give kH/kD = 1.0 0.1, consistent with a nucleophilic mechanism. A unified mechanism for the
metal-catalyzed reactions is presented which involves pre-equilibrium coordination of the substrate to the metal ion
followed by intramolecular delivery of a coordinated methoxide.
ous work from our laboratories showed that metal-catalyzed
alcoholysis is very effective for the catalytic decomposition of
Introduction
Phosphonoflouridate esters such as soman and sarin (1 (GD),
2 (GB)) as well as phosphonothioate esters such as VX and
Russian VX (3, 4) are extremely effective acetylcholinesterase
inhibitors and have nefarious notoriety as chemical warfare
phosphate and phosphorothioate triesters, converting these into
non-toxic methoxy esters.7 The most effective metal ions are La3+
and the Zn2+ and Cu2+ complexes of 1,5,9-triazacyclododecane
(5), all of which are active when bound with one eq. of methoxide
(CW) agents.1 Due to their toxicity, and the 1992 Chemical
per metal.8 More recently we have shown that a dimethyl-
Weapons Convention Treaty2 that requires destruction of CW
stockpiles by the signatory nations by 2007, considerable effort
benzyl amine palladacycle is very effective in promoting the
=
methanolytic decomposition of some P S-containing pesticides
has been directed at developing methods for destruction of these
although the palladium based catalysts are not effective for
and related organophosphorus (OP) materials.1,3 In addition
9
=
promoting the methanolysis of P O based OP materials.
to the required large-scale destruction of these (estimated
The alcoholysis process offers some advantages over hydrol-
to be several tens of thousands of tons), recent geopolitical
ysis of these particularly noxious materials. We have found that
events have hastened military and civilian initiatives to develop
the metal-catalyzed reactions are very much faster in methanol
effective decontamination methods for the removal of smaller
than they are in water, presumably due to a medium effect which
amounts of materials from surfaces and equipment.4 Although
we speculate to involve better pre-equilibrium binding of the
some effective methods are available, none is applicable to all
substrate and metal ion and an enhanced rate of intramolecular
situations or classes of compounds thus spurring research into
delivery of the methoxide nucleophile.7 Second, the OP materials
alternative methods for their destruction.
are far more soluble in the lower polarity alcohol medium than in
water which obviates the use of cosolvents or micellar conditions
as is often required for OP decomposition in aqueous media.
Third, the reactions occur in mild conditions of essentially
neutral sspH10,11 and ambient temperature. Fourth, the products
of the reactions are relatively innocuous phosphate triesters
which, if necessary for large scale, can be safely disposed of
by conventional burning along with the solvent. Finally, the
P-containing products of alcoholysis are neutral phosphorus
esters which are non-inhibitory to the reaction,7,9 while those
from hydrolysis are phosphoric acids which can dissociate to
Numerous reports have appeared concerning the metal-
form anionic materials which bind to the metal ion and thereby
catalyzed hydrolyses of neutral phosphate triesters5 and a lesser
inhibit the catalysis.12
number concerning phosphonate diesters.6 While it may be
In this report we extend the methanolysis study to La3+ and
considered by some that hydrolytic media are convenient for
5:Zn2+(−OCH3) promoted methanolysis of a series of O-ethyl O-
the reactions of noxious substrates, any acceptable process
aryl methylphosphonates 6a–f, these being of the same general
considered for large scale application requires cleaning of the
class as the phosphonofluoridate CW agents. Both catalysts are
hydrolysate before returning it to the biosphere, a non-trivial
shown to be very effective for this type of OP material, such that
task that spurs research into alternative methodologies. Previ-
a solution containing 1 mmol dm−3 of the La3+-catalyst provides
about 108-fold acceleration of the methanolysis of 6a relative to
the background reaction at essentially neutral sspH and ambient
† Electronic supplementary information (ESI) available: physical char-
temperature.
acterisation data. See http://dx.doi.org/10.1039/b511550g
4 0 8 2
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 4 0 8 2 – 4 0 8 8
T h i s j o u r n a l i s
T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 5
©

View Article Online
(b) General UV-vis kinetics
The following materials were used to make stock solutions:
sodium methoxide (0.5 mol dm−3), 1, 5, 9-triazacyclododecane
(97%), tetrabutylammonium hydroxide (1.0 mol dm−3),
La(OTf)3 (99.999%), anhydrous methanol (99.8%) and N-
ethylmorpholine (99%), all being obtained from Aldrich and
used as received. Zn(OTf)2 (98%) was purchased from Strem
Chemical and Cu(OTf)2 (98%) was obtained from Acros Organ-
ics. Methanol-d1 was purchased from Aldrich (99 atom%) and
used as received.
Stock solutions of phosphonates 6 were formulated between 5
and 10 mmol dm−3 in anhydrous methanol. La(OTf)3, tetrabuty-
lammonium hydroxide, Zn(OTf)2, 1,5,9-triazacyclododecane
and Cu(OTf)2 stock solutions were made to a concentration
of 50 mmol dm−3 in anhydrous methanol.
Kinetic determinations were done by UV-vis spectropho-
s
tometry as reported in our earlier publications, as were pH
measurements.7 Each kinetic run using Zn2+ consisted of sfor-
mulating the 5:Zn2+(−OCH3) catalyst in situ by adding mea-
sured amounts of the Zn(OTf)2, 1,5,9-triazacyclododecane and
tetrabutylammonium hydroxide stock solutions to anhydrous
methanol to form a final volume of 2.5 mL. The ratios of the
components of the catalyst were 1 : 1 : 0.5, respectively, to self-
Experimental
(a) Materials
3-Nitrophenol (99%), 4-chlorophenol (99+%), 4-chloro-2-nitro-
phenol (98%), 4-methoxyphenol (99%), sodium 4-nitrophen-
oxide (∼90%), phenol (99+%), methylphosphonic dichloride
(98%) and triethylamine (99.5%) were obtained from Aldrich
and used without further purification.
Caution: phosphonates 6 are all acetylcholinesterase inhibitors.
These, and the methyl phosphonic dichloride precursor, should be
synthesized and used with due attention to safety protocols.
s
buffer the mixtures at pH 9.1. Kinetic runs using Cu2+ were
s
s
prepared in the same manner and self-buffered at pH 8.75.
s
Final 5:M2+ concentrations in the kinetic runs ranged from 0.2
to 1.2 mmol dm−3, while the O-ethyl O-aryl methylphosphonate
concentrations ranged from 0.5 to 2 × 10−4 mol dm−3. The water
content in the methanol resulting from the stock solutions was
not more than 0.1%.
For the solvent deuterium kinetic isotope effect studies, the
solutions were made exactly as for the protiated methanol
kinetics, except that the medium was 99% d1-methanol and to
preserve consistency the stock solutions were the same ones used
for the protiated methanoldeterminations. For these kinetics, the
maximum amount of protium is 8% depending on the [5:M2+].
Ethyl methylphosphonochloridoate (C3H8ClO2P)13
To a 500 mL three-necked round-bottom flask equipped with
a magnetic stirring bar and a condenser was added 5.62 g of
methylphosphonic dichloride. With use of an addition funnel,
100 mL of toluene (dried over Na) was added under argon,
followed by a solution comprising 2.5 mL of absolute ethanol,
7.0 mL of triethyl amine and 30 mL of dry toluene while the
contents of the flask were stirred in an ice bath. This reaction was
left to stir for 1 h under the described conditions after which the
temperature was raised to 40 ◦C and the contents were allowed
to stir for 2 h. After 2 h, the salt (HNEt3Cl) was filtered away and
the filtrate was evaporated to give a crude yield of 43.7%. This
material was immediately used for the subsequent preparations.
s
Although we did not measure the pH of the solutions, the
s
medium is self-buffered at the sspKa in methanol-d1 through the
addition of 0.5 eq. of tetrabutylammonium hydroxide.
The kinetics of disappearance of 6a–f or appearance of the
phenol/phenoxide products were monitored at 25 ◦C in dupli-
cate under pseudo-first order conditions of excess catalyst at 428,
312, 340, 290, 277 and 300 nm. Pseudo-first-order rate constants,
k
obs, were obtained by fitting the absorbance vs. time data to a
standard first-order exponential model and the second order rate
constants, k2, were obtained as the gradients of the kobs vs. [active
catalyst] plots. Reactions were also carried out using buffered
La3+ solutions at sspH 9.14 and monitored by both UV-vis
spectrometry and stopped-flow spectrometry using an Applied
Photophysics SX-17MV Sequential Stopped-Flow ASVD Spec-
trophotometer. The buffer was prepared using 17 mmol dm−3
N-ethylmorpholine and perchloric acid in a 4 : 1 ratio.
General route for the preparation of O-ethyl O-aryl
methylphosphonates (6)13
To 100 mg of the sodium salt of the desired phenol suspended
in 5 mL of toluene (dried over Na) was added 1 eq. of ethyl
methylphosphonochloridate (from above) dissolved in 2 mL
of dry toluene. The mixture was heated to reflux under an
argon atmosphere for 1 h, cooled to room temperature and then
filtered through glass wool packed into a Pasteur pipette. The
filtrate was stripped of solvent via rotary evaporation and the
residue was dissolved in 1 mL of dichloromethane and mixed
with silica for dry loading for flash column chromatography.
The dichloromethane was evaporated and the silica coated with
product was loaded onto a Biotage flash silica column (12 +
Background reactions with methoxide as the active species
were carried out in duplicate using three methoxide concentra-
tions between 0.02 and 0.1 mol dm−3.
(c) Turnover experiment
To an NMR tube charged with 600 ll of d4-methanol containing
5 mmol dm−3 of 4-chloro-2-nitrophenyl methylphosphonate (6a)
as well as 17 mmol dm−3 N-ethylmorpholine and perchloric
acid in a 4 : 1 ratio was added 6 ll of a 50 mmol dm−3 stock
solution of La(OTf)3 such that the final concentration of La3+
was 0.5 mmol dm−3. The 31P NMR spectrum was recorded
periodically at 25 ◦C and the disappearance of the starting
material signal at 28.20 ppm monitored as a function of time.
After 20 min the starting material signal was completely replaced
by a peak at 30.69 ppm corresponding to the product O-ethyl
O-methyl methylphosphonate.
˚
M, KP–SIL, 12 × 150 mm, 40–63 lm, 60 A) and purified
with a Biotage SP1 HPFC Purification System with an elution
profile based on a TLC separation using 50 : 50 hexanes–ethyl
acetate.
The solvent from the final product was evaporated using
rotary evaporation and then the compound was placed under
vacuum pump aspiration for several hours prior to use. All ma-
terials were characterized by 1H and 31P NMR and exact mass.
The analytical data are given in the electronic supplementary
information.
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 4 0 8 2 – 4 0 8 8
4 0 8 3

View Article Online
Table 1 Acid dissociation constants and second order rate constants for the various methanolysis reactions of phosphonates 6a–e promoted by
methoxide, La3+, 5:Zn2+(−OCH3) and 5:Cu2+(−OCH3)
Phosphonate
pKa
6.32
7.14
8.39
9.38
10.00
10.20
sspKa
kO2 Me/mol dm−3 s−1
kL2 a/mol dm−3 s−1
k52:Zn(OMe)/mol dm−3 s−1
k52:Cu(OMe)/mol dm−3 s−1
6a
6b
6c
6d
6e
6f
10.64
11.30
12.41
13.59
14.33
14.70
14.3 0.1 (21 1)a
2.70 0.02
1.44 0.02
0.118 0.001
0.0217 0.0003
0.008 4 0.0001
(2.60 0.20) × 104
(1.45 0.03) × 103
(2.60 0.20) × 102
3.97 0.03
517 3 (510 20)a
46.8 0.5
2100 300
243 10
20.8 0.1
N.a.
N.a.
N.a.
22
1
0.45 0.01
0.072 0.002
0.014 0.002
(4.0 0.1) × 10−1
(1.4 0 0.02) × 10−1
a In DOCH3, determined at a 5 : Zn2+
:
−OCH3 ratio of 1 : 1 : 0.5.
In Fig. 2 are alternative presentations of the data where the La3+
or 5:M2+(−OCH3) rate constants for methanolysis of 6a–f are
plotted vs. the methoxide rate constants. The linear regressions
for these are given in eqn (5), (6) and (7) with the latter line
having rather large error limits due to it being defined only by
three points.
Results
Given in Table 1 are the second order rate constants for the
methanolysis of 6a–f promoted by −OCH3, La3+, 5:Zn2+(−OCH3)
and, for three examples, 5:Cu2+(−OCH3). The limited number
of data with the Cu2+ complex was a consequence of its
being highly absorbing such that only those substrates with
nitrophenoxy leaving groups (6a–c) could be monitored by UV-
vis kinetics. The second order kinetic terms were determined
from the gradients of the plots of kobs vs. [methoxide] or
[metalx+]total. For 5:Zn2+ and 5:Cu2+ the solutions were self
buffered at sspH 9.1 and 8.75 through half neutralization of the
catalyst ([5:M2+(−OCH3)]/[5:M2+(HOCH3)] = 1). Since earlier
work showed that the catalytic activity was entirely due to the
methoxide form, 5:Zn2+(−OCH3),7b,d or 5:Cu2+(−OCH3),7e the
k25:Zn(OMe) and k52:Cu(OMe) values reported in Table 1 were calculated
based on half the added metal ion being in the active form.
La3+catalysis for all substrates was determined by observing
the kobs vs. [La3+]total in a 17 mmol dm−3 N-ethylmorpholine
log kL2 a = (1.65 0.07) log kO2 Me + (2.36 0.06);
r2 = 0.9923, 6 data
(5)
log k25:Zn(OMe) = (1.40 0.03) log kO2 Me + (1.08 0.04);
r2 = 0.9981, 6 data
(6)
(7)
log k25:Cu(OMe) = (1.86 0.56) log kO2 Me + (1.26 0.40)
s
s
buffer at an pH value of 9.14 which is the value where the
rate maximum for La3+ catalysis of phosphate esters occurs.7a,b
Also given in Table 1 are the sspKa values for the phenols which
are experimentally determined11b,14 or can be calculated from
their water values by the linear regression given in reference 11b
(pKMa eOH = 1.12pKHa 2O + 3.56).
In Fig. 1 are presented Brønsted plots of the second order
rate constants for methoxide, La3+ and 5:M2+(−OCH3) catalyzed
methanolysis of the phosphonates with the linear regressions
presented in eqn (1) to (4).
log kO2 Me = (9.24 0.81) − (0.76 0.06) sspKHa OAr
;
Fig.
2
Plots of log kM2 x+ vs. log kO2 Me
,
(La3+(−OCH3))2 (ꢁ),
5:Zn2+(−OCH3) (᭺), 5:Cu2+(−OCH3) (ꢂ).
r2 = 0.9896, 6 data
(1)
(2)
In order to confirm that the system was truly catalytic, a 31
P
log kL2 a = (17.78 0.84) − (1.26 0.06) sspKHa OAr
r2 = 0.9716, 6 data
;
NMR experiment in d4-methanol containing 5 mmol dm−3 of
6a along with 0.5 mmol dm−3 La3+ buffered with 17 mmol dm−3
NEM was performed. After the first five minutes of data
acquisition (128 scans), the reaction had progressed to the
extent of at least 10 turnovers per catalyst based on the 31P
signals for starting material at 28.20 ppm and O-ethyl-O-
methylmethylphosphonate product at 30.69 ppm.15
log k52:Zn(OMe) = (14.04 1.17) − (1.06 0.09) sspKHa OAr
r2 = 0.9734, 6 data
;
(3)
(4)
log k52:Cu(OMe) = (15.10 1.39) − (1.12 0.12) sspKHa OAr
Discussion
In constructing the various Brønsted plots for the La3+-catalyzed
processes, we used the second order rate constants obtained as
the gradients of plots of Dkobs/D[La3+] under buffered conditions
s
with 17 mmol dm−3 NEM buffered at pH 9.14, where the
maximal activity occurs, so the kL2 a vaslues in Table 1 are
presented per La3+ ion even though we have determined that
the actual catalytic species are dimers (La3+)2(−OCH3)2,3,4 the
relative catalytic importance of which depend on the sspH.16,17 At
this sspH > 90% of the catalysis occurs through La3+2(−OCH3)2.
The phosphonate esters encompass a range of 104 in leaving
group acidity and about 105 in k2La but only about 103 in k2OMe
.
Fig. 1 Brønsted plots of the second order rate constants for methoxide
and metal ion catalyzed methanolysis of 6a–e at 25 ◦C. (La3+(−OCH3))2
(ꢀ), 5:Cu2+(−OCH3) (♦), 5:Zn2+(−OCH3) (ꢁ), −OCH3 (ꢂ).
For the same series of phosphonates, the k52:Zn(OMe) at sspH 9.1 span
a range of 3 × 104 fold.
4 0 8 4
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 4 0 8 2 – 4 0 8 8

View Article Online
(b) La3+ and 5:Zn2+(−OCH3) catalyzed methanolysis
(a) Reaction of methoxide
The Brønsted plot for methoxide shown in Fig. 1 provides a
linear correlation with a blg value of −0.76 0.06 (n = 6, r2 =
0.9734). This blg value can be compared with the −0.70 found
for the attack of methoxide on O,O-diethyl O-aryl phosphate
triesters,7b values of −0.43 and −0.44 for hydroxide attack
on the same species,18,19 as well as −0.66 and −0.79 for the
attack of phenoxide and p-nitrophenoxide respectively on O-
aryl diphenylphosphinates.20 In general, low blg values observed
with strong nucleophiles on phosphates imply relatively little
cleavage of the aryloxy bond in the transition state and suggest
a two step reaction, where the rate limiting step is attack.
However, as the nucleophile becomes weaker, the blg values
become increasingly negative, signalling a shift in the reaction
mechanism so, if it is two steps, the second (involving the
departure of the leaving group) is rate limiting, or alternatively
the reaction is concerted.19,21,22,23 Buncel and co-workers have
presented evidence that the mechanism can change from con-
certed to stepwise depending on the solvent, citing ethoxide
promoted ethanolysis of (CH3)2P(O)–OAr and Ph2P(O)–OAr24
as stepwise in pure ethanol but respectively concerted25 or
stepwise26 in 90% water–dioxane. Apparently, the increased
basicity of the nucleophile in the less polar solvent has the effect
of enhancing the bond formation relative to the leaving group
departure.
Also given in Fig. 1 are Brønsted plots for the metal-catalyzed
methanolysis of the phosphonates, the blg values for La3+ and
5:Zn2+(−OCH3) being −1.26 0.06 and −1.06 0.09. The value
for 5:Cu2+(−OCH3) is −1.12 0.12, but is less well-defined since
it is based only on the three NO2-containing derivatives. These
large negative blg values suggest a metal catalyzed process where
there is more cleavage of the leaving group in the transition
state than is the case for the methoxide reaction. For La3+ and
5:Zn2+(−OCH3) promoted methanolysis of O,O-diethyl O-aryl
phosphates7b the blg values for the La3+ and Zn2+ catalysts are also
large and negative at −1.43 0.08 and −1.12 0.13 which we
interpreted as indicative of an associative process with concerted
displacement of the leaving group: the phosphonate Brønsted
data are also consistent with a concerted displacement. The sol-
vent deuterium kinetic isotope effect (dkie) for the methanolysis
of 6a promoted by methoxide and 5:Zn2+(−OCH3) in methanol
or DOCH3 are, respectively, kD/kH = 1.47 0.08 and 0.99
0.05, both cases supporting a nucleophilic mechanism rather
than a general base one. Although a complete analysis of the dkie
values for methanolysis of phosphate, phosphorothioate, phos-
phonate, phosphonothioate and carboxylate esters promoted
by 5:Zn2+(−OCH3) will be presented elsewhere,31 our available
phosphonate data are consistent with either of two kinetically
equivalent nucleophilic processes, termed internal methoxide
(IM) or external methoxide (EM)32 which are presented in eqn
(8), (9), (10) and (11). However, the observed second order
rate constants for the metal-catalyzed reactions, along with a
reasonable value for the equilibrium binding constant, rules out
the EM mechanism since the computed rate constants (kꢀ1) for
methoxide attack on the La3+ and 5:Cu2+-bound substrates for
at least one phosphonate (6a) exceeds the diffusion limit of 5 ×
109 mol−1 dm3 s−1.33,34
Following the ‘effective charge treatment’ described by
Jencks27 and Williams,28 the Brønsted blg value of −0.76 for
nucleophilic attack of methoxide on the aryl phosphonates 6
implies a process where the rate-limiting transition state has
appreciable changes in the P–OAr bond. This could be viewed
as resulting from either a two step process with the attack step
largely rate limiting because methoxide is a poorer leaving group
than any of the aryloxy anions29 or with a concerted process,
both of which are shown in Scheme 1. The progress of P–OAr
bond breaking in the TS is measured by the Leffler parameter (a)
that relates the Brønsted blg to the beq for equilibrium transfers
of acyl or phosphoryl groups between oxyanion nucleophiles.
K
b
2+
−
−−−−ꢀ
5 : M ( OCH3) + EtO(CH3)P(= O)-XAr
ꢁ−−−−
5 : M2+(−OCH3) : EtO(CH3)P(= O)XAr
(8)
(9)
For transfer of the (EtO)2P O group the beq value is 1.8719 with
=
the O–Ar oxygen in the starting material having a net effective
charge of +0.87. For transfer of the diphenylphosphinoyl group
k
1
5 : M2+(−OCH3) : EtO(CH3)P(= O)XAr −−−−−−→ P
=
((C6H5)2P O), between oxyanion nucleophilies the beq value
is 1.2520 with the O–Ar oxygen in the starting phosphinate
having a net charge of +0.25. The net charge difference on
the O–Ar of phosphates relative to phosphinates is due to the
ꢀ
K
b
−−−−ꢀ
ꢁ−−−−
5 : M2+ + EtO(CH3)P(= O)-XAr
=
high net electronegativity of the (EtO)2P O group relative to
5 : M2+ : EtO(CH3)P(= O)XAr
(10)
=
the Ph2P O group. Although the blg is not known for the
=
EtO(CH3)P O, we predict that its value would be roughly
midway between that of the phosphoryl and phosphinoyl groups,
ca. 1.5.30 The Leffler parameter of −blg/beq = 0.50 suggests
that for the methoxide reaction on 6, the P–OAr cleavage is
ca.50% of the way from starting material to product with the
departing aryloxy group having a net charge of −0.26. The other
mechanism for the strongly nucleophilic methoxide shown in
Scheme 1 involves rate-limiting formation of a highly unstable
five-coordinate intermediate and cannot be distinguished from
the concerted process.20
ꢀ
k
1
5 : M2+ : EtO(CH3)P(= O)XAr + −OCH3 −−−−−−→ P (11)
The favoured mechanism is thus the IM process of eqn (8)
and (9), for which the derived kinetic expressions are given in
eqn (12) and (13) where bb and b1 refer to the Brønsted blg
s
s
values for the binding and kinetic steps, and pKa refers to the
acid dissociation constant for the conjugate acid of the leaving
groups. The experimental Brønsted blg values
s
+ b1) pK
ko2bs = Kbk1 = CbC110(b
(12)
(13)
a
s
b
log ko2bs = {log Cb + log C1} + (bb + b1)sspKa
comprise the influence of the leaving group on the pre-
equilibrium binding step and on the intramolecular kinetic
step (bb + b1). The available literature data35 suggest that bb
should be slightly positive, so the observed negative blg should
be dominated by b1. The proposed transition states for the
(La3+(−OCH3))2 and 5:Zn2+(OCH3) catalyzed methanolysis of
the phosphonates are shown in 7 and 8. The Leffler parameter
for the La3+-catalyzed process is −blg/beq = 1.26/1.530 = 0.84,
Scheme 1
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 4 0 8 2 – 4 0 8 8
4 0 8 5

View Article Online
so transition state 7 has extensive cleavage of the P–OAr bond.
In the case of the 5:Zn2+(−OCH3) catalyzed reaction, a similar
mechanism is envisioned but this time transition structure 8 will
involve five-coordinate Zn2+ with a Leffler a of 0.7.
Fig. 3 A reaction coordinate diagram for the metal-methoxide cat-
alyzed methanolysis of phosphonates 6.
Under the assumption that both the reactions are concerted and
thus have the same rate limiting step in the forward and reverse
direction, we can locate the respective transition states for the
methoxide reaction as beq = bnuc − blg = − 1.5; bnuc = 0.75, blg = −
0.76, and for the metal-catalyzed ones as bnuc = 0.24–0.44 and
blg = − 1.26 to −1.06 for the La3+ and 5:Zn2+(−OCH3) reactions.
Making the leaving group better depresses the right side of the
diagram, forcing the TS in both cases toward starting materials
along the reaction coordinate, and toward the lower right side
perpendicular to the reaction coordinate. The result that the new
TS in both cases is more charge separated with the methoxide–
P distance being less and the P–OAr distance little changed.
The current data do not allow a detailed charge analysis,
but the emerging picture is that the metal ion must stabilize
the exploded, charge separated transition structures, perhaps
through an electrostatic stabilization as has been suggested in
related cases involving metal ion catalysis of the hydrolyses of
anionic phosphorus materials.38
(c) Comparision of kO2 Me and k25:Zn(OMe) or k2La
As in the case of phosphate and phosphorothioate ester
methanolysis,7b we assume here that the mechanism of CH3O−-
promoted methanolysis is similar enough for all the phos-
phonate esters with nucleophilic addition being either rate
limiting or concerted with leaving group departure so that the
rate constant, kO2 Me, can be used as an empirical measure of
the composite effects of structural changes that incorporate
both electronic and steric effects. Comparison of the La3+ and
5:M2+(−OCH3) kinetic data in Table 1 with the corresponding
kO2 Me kinetic data indicates that the metal ion systems are more
effective than methoxide for promoting the methanolysis of
this series of phosphonates, a phenomenon observed before
for the methanolysis of phosphates and phosphorothioates.7b,d
The kL2 a/kO2 Me ratio varies from 1800-fold for the most reactive
phosphonate 6a to16-fold the least reactive derivative 6f. Over
the same series, the k52:Zn(OMe)/k2OMe ratio varies only by 36-fold to
ca. 1.6-fold. Shown in Fig. 2 is a graphical representation the
Conclusion
The above study indicates that La3+ and 5:M2+(−OCH3) cat-
alyzed methanolysis is an effective strategy for the solvolytic
destruction of phosphonates. As far as we are aware, the data
obtained in this study provides the most complete mechanistic
picture of the metal-catalyzed solvolysis reactions of phospho-
nates, which may in turn provide mechanistic information of
relevance to the metal promoted solvolyses of CW G-agents
and spur development of effective methods for destroying these
materials.12 The phosphonates are about 100-fold more reactive
than the corresponding phosphate triesters with the same leaving
groups7b toward methanolysis promoted by methoxide, La3+
and the 5:Zn2+(−OCH3) systems. Nevertheless, the catalysis we
observe is quite impressive considering that the catalysts are
data plotting log kLa, log k25:Zn(OMe), and log k25:Cu(OMe) vs. log k2OMe
,
the respective gradi2ents of the lines being 1.65 0.07, 1.40 0.03
and a less defined 1.86 0.56 for the copper complex. The fact
that all of the plots adhere to respectable linearity suggests that
the changes in the leaving group influence both the methoxide
and metal-catalyzed reactions in much the same way, supporting
the idea that both reactions could be concerted, and that neither
process involves a change in mechanism nor rate limiting step
over the series investigated.
It is interesting to speculate as to why these catalysts are more
effective (relative to methoxide) toward substrates with better
leaving groups. Based on the widely held notion that catalysis
in general must result from better stabilization of the TS than
ground state,36 the metal containing systems must preferentially
bind to the ‘dissociated’ or exploded five-coordinate TS as shown
s
s
maximally operative at pH values near neutrality (8.34) in
methanol. For the most reactive substrate 6a a solution contain-
ing 1 mmol dm−3 of catalyst (La3+(−OCH3))2, 5:Zn2+(−OCH3)
and 5:Cu2+(−OCH3) accelerates the methanolysis relative to
s
s
=
in 7 or 8 relative to the corresponding four-coordinate P O
the background methoxide reaction at pH optimum values
substrate. Shown in Fig. 3 is a reaction coordinate diagram37
for the Mx+–−OCH3 catalyzed reaction where the lower left and
upper right corners represent the starting and final states, and the
upper left and lower right corners represent the Mx+-complexed
of 9.1 and 8.75 by 8.5 × 107-fold, 1.7 × 106-fold and 1.4 ×
107-fold respectively, leading to t1/2 values of 0.026, 1.33 and
0.33 s, respectively. For the least reactive substrate, 6f, under
the same conditions, a 1 mmol dm−3 solution of (La3+(−OCH3))2
or 5:Zn2+(−OCH3) promote the methanolysis reaction by 7.8 ×
105-fold and 7.8 × 104-fold relative to the background reaction
at sspH 9.1, giving t1/2 values of 1.37 and 13.7 h respectively.
x+
+
=
five-coordinate phosphorane intermediate and the M :O P <
+ −OAr separated state respectively. The left/right axes represent
the nucleophile–P distance as defined by bnuc/beq while the
upper/lower axes represent the P-leaving group separation as
defined by −blg/beq, the value of which is 0.7–0.84 in the
present work. Also shown on Fig. 3 is a hypothetical methoxide
concerted reaction (structures with the encircled Mx+ removed)
for which the Leffler index is −blg/beq = 0.5 in the present case.
Acknowledgements
The authors gratefully acknowledge the financial assistance
of the Natural Sciences and Engineering Research Council of
4 0 8 6
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 4 0 8 2 – 4 0 8 8

View Article Online
Canada, the Canada Foundation for Innovation, Queen’s Uni-
versity and the United States Department of the Army, Army
Research Office, Grant No. W911NF-04-1-0057 and the Defense
Threat Reduction Agency, Joint Science and Technology Office
(06012384BP).39 Roxanne Lewis thanks the Natural Sciences
and Engineering Council of Canada for an Undergraduate
Summer Research Award at Queen’s University.
10 For the designation of pH in non-aqueous solvents we use the forms
described by Bosch and co-workers11 based on the recommendations
of IUPAC, in the Compendium of Analytical Nomenclature, Definitive
Rules 1997 3rd ed., Blackwell, Oxford, UK 1998. If one calibrates the
measuring electrode with aqueous buffers and then measures 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 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.
11 (a) E. Bosch, F. Rived, M. Roses and J. Sales, J. Chem. Soc., Perkin
Trans. 2, 1999, 1953; (b) F. Rived, M. Rose´s and E. Bosch, Anal.
Chim. Acta, 1998, 374, 309; (c) E. Bosch, P. Bou, H. Allemann and
M. Rose´s, Anal. Chem., 1996, 68, 3651; (d) I. Canals, J. A. Portal, E.
Bosch and M. Rose´s, Anal. Chem., 2000, 72, 1802.
12 It has been pointed out by a referee that, for methanolysis of the
G-agents, the fluoride product may bind to the metal ion and
be inhibitory and so the reactions could be stoichiometric. This
is a distinct possibility that further research will have to confirm
and, if present, circumvent. On the other hand, since our previous
work has shown that the La3+ catalysts promote phosphorothioate
methanolysis reactions for at least 100 turnovers,7c thiolate inhibition
of that system does not appear to present a problem.
References
1 (a) Y.-C. Yang, J. A. Baker and J. R. Ward, Chem. Rev., 1992, 92,
1729; (b) Y.-C. Yang, Acc. Chem. Res., 1999, 32, 109; (c) Y.-C. Yang,
Chem. Ind., 1995, 334.
2 Conference on Disarmament. The Convention of the Develop-
ment, Production, Stockpiling and Use of Chemical Weapons
and on Their Destruction. CD/1170. Geneva, August, 1992. (see
http://www.dfait-maeci.gc.ca/nndi-agency/cwc_1-en.asp).
3 H. Morales-Rojas and R. A. Moss, Chem. Rev., 2002, 102, 2497, and
references therein.
4 There is increasing concern that these relatively simple-to-make
materials could become preferred weapons of terrorist organizations.
Indeed, March 20, 2005 was the 10th anniversary of the sarin attack
in the Tokyo subway system by the Aum Shinrikyo cult, now known
as Aleph. That attack, one of the worst in Japan’s postwar history,
killed 12 people and left more than 5500 others ill and would have
been far more serious had Aum used pure materials.
13 P. Malatesta and G. Ceccarini, Boll. Chim. Farm., 1982, 121, 183.
14 (a) V. A. Palm, Tables of Rate and Equilibrium Constants of Hetero-
cyclic Reactions, vol. 1 (1), Proizbodstvenno–Izdatelckii Kombinat
Biniti, Moscow, 1975; (b) V. A. Palm, Tables of Rate and Equilibrium
Constants of Heterocyclic Reactions, suppl. vol. 1, issue 3, Tartu,
1985, Tartuskii Gosudarsvennii Universitet.
5 (a) L. Barr, C. J. Easton, K. Lee, S. F. Lincoln and J. S. Simpson,
Tetrahedron Lett., 2002, 7797; (b) T. Koike and E. Kimura, J. Am.
Chem. Soc., 1991, 113, 8935; (c) M. M. Ibrahim, K. Ichikawa and
M. Shiro, Inorg. Chim. Acta, 2003, 353, 187; (d) T. Itoh, H. Hisada,
Y. Usui and Y. Fujii, Inorg. Chim. Acta, 1998, 283, 51; (e) F. M.
Menger and T. Tsuno, J. Am. Chem. Soc., 1989, 111, 4903; (f) P.
Scrimin, P. Tecilla and U. Tonellato, J. Org. Chem., 1991, 56, 161 and
references therein; (g) F. Tafesse, Inorg. Chim. Acta, 1998, 269, 287;
(h) P. Scrimmin, G. Ghinlanda, P. Tecilla and R. A. Moss, Langmuir,
1996, 12, 6235; (i) C. A. Bunton, P. Scrimmin and P. Tecilla, J. Chem.
Soc., Perkin Trans. 2, 1996, 419; (j) Y. Fujii, T. Itoh and K. Onodera,
Chem. Lett., 1995, 305; (k) S. J. Oh, C. W. Yoon and J. W. Park,
J. Chem. Soc., Perkin Trans. 2, 1996, 329; (l) T. Berg, A. Simeonov
and K. Janda, J. Comb. Chem., 1999, 1, 96; (m) J. R. Morrow and
W. C. Trog l e r , Inorg. Chem., 1989, 28, 2330; (n) R. W. Hay and N.
Govan, J. Chem. Soc., Chem. Commun., 1990, 714; (o) D. Kong, A. E.
Martell and J. Reibenspies, Inorg. Chim. Acta, 2002, 333, 7; (p) R. W.
Hay and N. Govan, Polyhedron, 1998, 17, 463, 2079; (q) B. L. Tsao,
R. J. Pieters and J. Rebek, Jr., J. Am. Chem. Soc., 1995, 117, 2210;
15 The overall rate constant for the 31P NMR experiment is somewhat
slower than observed in the UV-vis kinetics and probably stems from
a non-optimal sspH setting in d1-methanol and a low initial concen-
tration of catalyst, neither of which favours complete conversion to
the (La3+(−OCH3)2) dimer.16b
.
16 (a) G. T. T. Gibson, A. A. Neverov and R. S. Brown, Can. J. Chem.,
2003, 81, 495; (b) We have determined in ref.7a and 17 that the maxi-
mal activity for the La3+ catalyzed methanolysis of both carboxylate
esters and phosphate triesters lies in the sspH region between 8.7–9.1,
where 95% of the activity is attributed to La3+2(−OCH3)2. Thus, the
conditional observed kL2 a rate constants reported in Table1 and 2 are
ca. 1/2 of the k2 of the La3+2(−OCH3)2 active form.
17 (a) A. A. Neverov, N. E. Sunderland and R. S. Brown, Org. Biomol.
Chem., 2005, 3, 65; (b) A. A. Neverov, T. McDonald, G. Gibson and
R. S. Brown, Can. J. Chem., 2001, 79, 1704; (c) A. A. Neverov, G.
Gibson and R. S. Brown, Inorg. Chem., 2003, 42, 228.
18 S.-B. Hong and F. M. Rauchel, Biochemistry, 1996, 35, 10904.
19 S. A. Ba-Saif and A. Williams, J. Org. Chem., 1988, 53, 2204.
20 N. Bourne, E. Chrystiuk, A. M. Davis and A. Williams, J. Am. Chem.
Soc., 1988, 110, 1890.
¯
(r) M. Yamami, H. Furutachi, T. Yokoyama and H. Okawa, Inorg.
Chem., 1998, 37, 6832; (s) C. M. Hartshorn, A. Singh and E. L.
Chang, J. Mater. Chem., 2002, 12, 602; (t) M. Rombach, C. Maurer,
K. Weis, E. Keller and H. Vahrenkamp, Chem. Eur. J., 1999, 5, 1013.
6 (a) R. W. Hay, N. Govan and K. E. Parchment, Inorg. Chem.
Commun., 1998, 1, 228; (b) P. R. Norman, A. Tate and P. Rich,
Inorg. Chim. Acta, 1988, 145, 211; (c) R. W. Hay, N. Govan and P. R.
Norman, Transition Met. Chem. (London), 1998, 23, 133; (d) R. J.
Whithey, Can. J. Chem., 1969, 476, 4383; (e) R. S. Brown and M.
Zamkanei, Inorg. Chim. Acta, 1985, 108, 201.
21 S. A. Khan and A. J. Kirby, J. Chem. Soc. B, 1970, 1172.
22 R. Rowell and D. G. Gorenstein, J. Am. Chem. Soc., 1981, 103,
5894.
23 S. A. Ba-Saif, M. A. Waring and A. Williams, J. Am. Chem. Soc.,
1990, 112, 8115.
24 E. Buncel, K. G. Albright and I. Onyido, Org. Biomol. Chem., 2004,
2, 601.
25 K. T. Douglas and A. Williams, J. Chem. Soc., Perkin Trans. 2, 1976,
7 (a) J. S. W. Tsang, A. A. Neverov and R. S. Brown, J. Am. Chem.
Soc., 2003, 125, 7602; (b) T. Liu, A. A. Neverov, J. S. W. Tsang and
R. S. Brown, Org. Biomol. Chem., 2005, 3, 1525; (c) J. S. W. Tsang,
A. A. Neverov and R. S. Brown, Org. Biomol. Chem., 2004, 2, 3457;
(d) W. Desloges, A. A. Neverov and R. S. Brown, Inorg. Chem., 2004,
43, 6752; (e) A. A. Neverov and R. S. Brown, Org. Biomol. Chem.,
2004, 2, 2245.
8 (a) We have presented a complete evaluation of the lanthanide
catalyzed ethanolysis of paroxon and while all Ln3+ give some
reaction, La3+ is certainly the best. R. S. Brown, A. A. Neverov, J. S. W.
Tsang, G. T. T. Gibson and P. J. Montoya-Pela´ez, Can. J. Chem.,
2004, 82, 1791; (b) A preliminary screening of the methanolysis of
paraoxon promoted by lanthanides and the transition metal ions
Zn2+, Co2+, Ni2+ and Cu2+ indicates that all the lanthanides are
active at a Ln3+/(−OCH3) ratio of unity, but La3+ is the best. For the
transition metals, Zn2+ 7b and Cu2+ are the best catalysts, but these tend
to dimerize to inactive forms. Addition of phenanthroline ligands
modifies the activities, but inactive dimer formation is still a problem
which can be completely alleviated by complexation to a tridentate
ligand such 1,5,9 triazacyclododecane or 1,4,7-triazacyclononane
ligand, with the former having the highest activity. N. E. Sunderland,
A. A. Neverov and R. S. Brown, unpublished results.
515.
26 A. Williams and R. A. Naylor, J. Chem. Soc. B, 1071, 1967.
27 (a) W. P. Jencks and M. Gilchrist, J. Am. Chem. Soc., 1968, 90, 2622;
(b) D. J. Hupe and W. P. Jencks, J. Am. Chem. Soc., 1977, 99, 451;
(c) J. M. Sayer and W. P. Jencks, J. Am. Chem. Soc., 1977, 99, 464;
(d) M. J. Gresser and W. P. Jencks, J. Am. Chem. Soc., 1977, 99, 6963.
28 (a) S. Thea and A. Williams, Chem. Soc. Rev., 1986, 15, 125; (b) A.
Williams, Acc. Chem. Res., 1984, 17, 425; (c) A. Williams, Concerted
Organic and Bio-organic Mechanisms, CRC Press, Boca Raton,
2000.
29 Given that the autoprotolysis constant of methanol is 10−16.77
2
s
s
(mol dm−3
)
(ref. 11) one can compute a pKa for methanol itself
of 18.13 on the mol dm−3 scale.
30 Buncel24 analyzes the Douglas and Williams25 data for the transfer
of (CH3)2P(O) group between oxyanion nucleophiles as giving a beq
of 0.88, suggesting that the OAr has a net starting charge of −0.12.
If one makes a correlation between the net OAr oxygen charge and
the RrI values of the EtO, Ph and CH3 groups on the P, i.e. charge
O = n(RrI), then n = 1.43 0.13, and the net positive charge for
the OAr group attached to a EtO(CH3)P(O) moiety is predicted to
be 0.5 0.05.
31 C. Maxwell, A. A. Neverov and R. S. Brown, submitted.
32 J. Suh, S. J. Son and M. P. Suh, Inorg. Chem., 1998, 37, 4872
have suggested that the hydrolysis of carboxylate esters promoted
9 Z.-L. Liu, A. A. Neverov and R. S. Brown, Org. Biomol. Chem., 2005,
3, 3379–3387.
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 4 0 8 2 – 4 0 8 8
4 0 8 7

View Article Online
by 5:Zn2+ involves external attack of hydroxide on a coordinated
substrate.
constant, a factor of 10 difference still makes the reaction exceed the
diffusion limit.
35 (a) R. Schurhammer, V. Erhart, L. Troxler and G. Wipf, J. Chem.
Soc., Perkin Trans. 2, 1999, 2423 and references therein; (b) D. M.
Rackham, Spectrosc. Lett., 1980, 13, 513; (c) R. Du Preez and J. S. S.
Preston, S. Afr. J. Chem., 1986, 39, 89, 137.
36 M. I. Page, in Enzyme Mechanisms, ed. M. I. Page and A. Williams,
Royal Society of Chemistry, London, 1987, pp. 1–13.
37 (a) W. P. Jencks, Chem. Rev., 1972, 72, 705; (b) D. A. Jencks and W. P.
Jencks, J. Am. Chem. Soc., 1977, 99, 7948.
38 (a) M.-Y. Yang, O. Iranzo, J. P. and J. R. Morrow, J. Am. Chem. Soc.,
2005, 127, 1064; (b) E. Buncel, R. Nagelkerke and G. R. Thatcher,
Can. J. Chem., 2003, 81, 53; (c) D. Herschlag and W. P. Jencks, J. Am.
Chem. Soc., 1987, 109, 4664.
33 (a) K. N. Dalby and W. P. Jencks, J. Am. Chem. Soc., 1997,
119, 7271 and references therein; (b) R. A. McClelland, V. M.
Kanagasabapathy, N. S. Banait and S. Steenken, J. Am. Chem. Soc.,
1991, 113, 1009; (c) R. S. McClelland, F. L. Cozens, S. Steenken,
T. L. Amyes and J. P. Richard, J. Chem. Soc., Perkin Trans. 2, 1993,
1717.
34 It is customarily assumed32 that the equilibrium binding constant
=
=
metal ion complexes with neutral C O or P O substrates is ca.
1 mol−1dm3. We are also assuming this number based on the fact
that we have not observed any saturation behaviour of the reaction
kinetics at concentrations of catalyst up to 10 mmol dm−3. In the case
of 1 mmol dm−3 of [5:Cu2+] at sspH 8.75, the [−OCH3] = 10−8 mol dm−3
,
the computed second order rate constant for methoxide attack would
be 2.1 × 1011 mol−1 dm3 s−1, a value that exceeds the diffusion limit by
roughly 50-fold. While one might question the value of the binding
39 The content of the information does not necessarily reflect the
position or the policy of the federal government, and no official
endorsement should be inferred.
4 0 8 8
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 4 0 8 2 – 4 0 8 8