Stereochemical inversion of phosphonothioate methanolysis by La(III) and Zn(II): Mechanistic implications for the degradation of organophosphate neurotoxins

Article abstract of DOI:10.1021/ic2016897

The utility of phosphonothioate methanolysis to degrade organophosphate neurotoxins has prompted the stereochemical investigation of this useful transformation. The methanolysis of enantiomerically pure O,S-diethyl phenylphosphonothioate (5) was studied b

Full text of DOI:10.1021/ic2016897

Article
pubs.acs.org/IC
Stereochemical Inversion of Phosphonothioate Methanolysis by
La(III) and Zn(II): Mechanistic Implications for the Degradation of
Organophosphate Neurotoxins
Louis Y. Kuo* and Sara K. Glazier
Department of Chemistry, Lewis & Clark College, Portland, Oregon 97219, United States
S
* Supporting Information
ABSTRACT: The utility of phosphonothioate methanolysis to degrade
organophosphate neurotoxins has prompted the stereochemical investigation
of this useful transformation. The methanolysis of enantiomerically pure O,S-
diethyl phenylphosphonothioate (5) was studied both in the presence and in the
absence of metal ions known to catalyze the phosphonothioate → phosphonate
transformation. This report outlines the syntheses of enantiomerically pure 5
and its methanolysis product O-ethyl O-methyl phenylphosphonate (7).
Compound 7 results from exclusive P−S scission of 5, which is the desired
mode of phosphonothioate methanolysis (Ea = 14.5
0.5 kcal/mol). The
stereochemical analysis of the phosphonothioate methanolysis was done for
the first time with β-cyclodextrin, and it shows complete inversion on the
phosphorus center upon methoxide displacement of ethanethiolate. The
presence of La(III) or Zn(II) complexes do not alter this SN2-like substitution
which sheds new light on the mechanism of methanolysis of phosphonothioates.
Scheme 1
INTRODUCTION
■
Phosphonothioates are neurotoxins, and there are societal
benefits for degrading them through methanolysis.1,2 As acetyl-
choline esterase inhibitors,3,4 phosphonothioates are used as
pesticides for agricultural use,5 but their greatest notoriety is
found as chemical warfare agents.3 As such, a major effort has
been devoted toward the chemical degradation of these
compounds through oxidative6−13 and hydrolytic transforma-
tions.14 A problem with alkaline hydrolysis of phosphono-
thioates is the production of both phosphonothioate and
phosphonate anions3 that are resistant to further nucleophilic
attack. Frustratingly, in some cases the phosphonothioate
product is just as toxic as the parent neurotoxin.15 To that end,
alcoholysis is a promising route especially when coupled with
metal complexes that serve as catalysts. Brown and co-workers16,17
have used many metal complexes to catalyze phosphonothioate
methanolysis reactions with up to 109-fold rate enhancements.
The methoxide-bridged lanthanum(III) system18 has been used to
degrade live nerve agents such as the infamous VX. The proposed
mechanism (Scheme 1) starts from the La-μ-(OMe)2-La
dimer16,17 (I), and it involves delivery of the La(III)-bound
methoxide onto a coordinated phosphonothioate (II) leading to
displacement of the alkythiolate. The methoxide-bridged
lanthanum dimer is regenerated with the addition of the methanol
solvent that also releases the phosphonate product. Subsequent
alkoxide displacement was also found to occur from this released
phosphonate.
there are two possible routes that yield exclusive P−S scission
of a phosphonothioate (Scheme 2).19 One is a concerted
SN2(P) in which the nucleophile has a facial selectivity to attack
opposite the alkylthiolate leaving group (SR). In the other
route the nucleophile has the opposite facial selectivity as it
attacks opposite the alkoxide (OR′) to initially yield a trigonal
bipyramid intermediate (A). This is followed by a low-energy-
barrier pseudorotation (Ψ)20,21 to intermediate B that places
the departing ethanethiolate leaving group in the axial (apical)
position.
The P−S bond scission is the desired route, for the
phosphonothioate → phosphonate transformation leads to
a less toxic product.16−18 In terms of the alcoholysis mechanism
Received: August 4, 2011
Published: December 19, 2011
© 2011 American Chemical Society
328
dx.doi.org/10.1021/ic2016897 | Inorg. Chem. 2012, 51, 328−335

Inorganic Chemistry
Article
Scheme 2
The phosphorus stereocenter of chiral phosphonothioates
allows one to interrogate the stereochemistry of alcoholysis
with enantiomerically pure substrates. Therefore, a phospho-
nate product with inverted stereochemistry results from a
SN2(P) pathway in which the nucleophile has a facial selectivity
opposite the alkythiol group. A phosphonate product with
retention of the stereocenter proceeds through the trigonal
bipyramid intermediate that undergoes the low-energy A → B
isomerization (Ψ). DeBruin and co-workers19 showed that
nucleophilic attack of ethoxide on O,S-dimethyl phenyl-
phosphonothioate has a facial selectivity that places the
methylthiolate ligand axial and opposite the alkoxide (EtO)
nucleophile. This leads to inversion of configuration (SN2(P))
that is consistent with earlier findings which used alkoxide
nucleophiles.22−32 Interestingly, Grignard nucleophiles resulted
in retention of configuration.27,33−39
RESULTS AND DISCUSSION
■
The investigation on the stereochemistry of DEPP methanol-
ysis (P−S scission) proceeds through the following three
points: (1) the synthesis of enantiomerically-pure starting
phosphonothioate and product phosphonate, (2) enantiomeric
resolution with 31P NMR using a chiral encapsulating host, and
(3) the stereochemical outcome of the methoxide attack on the
chiral phosphonothioate with and without metal complexes.
The two title compounds involved in this study are the
starting O,S-diethyl phenylphosphonothioate (5) and the
product, O-ethyl O-methyl phenylphosphonate (7). The
enantiomeric synthesis of these key compounds has been
described by DeBruin and co-workers.19
In terms of making the (S)-O,S-diethyl phenylphosphono-
thioate ((S)-5), the enantiomeric resolution occurs with O-
ethyl phenylphosphonothioic acid (3). This was made from
sulfur addition to O-ethyl phenylhydrogenphosphinate (1) in
the presence of dicyclohexylamine followed by HCl acid-
ification (Scheme 3). The racemic phosphonothioic acid 3 was
resolved with the addition of brucine that formed the
phosphonothioate anion 4. The crystals isolated from the
acetone solution of this brucine salt were confirmed by single
crystal X-ray crystallography to be the S enantiomer of O-ethyl
phenylphosphonothioate ((S)-4) as shown in Figure 1.
Acidification of this brucinium salt gave the optically pure
thioic acid (S)-3, which was converted to and stored as the
dicyclohexylammonium salt ((S)-2). Finally, addition of ethyl
iodide to this (S)-2 salt gave the enantiomerically pure (S)-O,S-
diethyl phenylphosphonothioate ((S)-5) starting compound
(Scheme 3).
The methanolysis of (S)-5 would give either the S or R
enantiomer of O-ethyl O-methylphosphonate ((S)-7 or (R)-7).
We modified a prior literature protocol19 for making (S)-7
shown in Scheme 4b that starts from the dicyclohexylammo-
nium salt of (S)-O-ethyl phenylphosphonothioate ((S)-2)
made in Scheme 3. Addition of methyl iodide (Scheme 4b)
to this salt yielded the corresponding S-methylated phospho-
nothioate, (S)-6. The final conversion to the phosphonate was
done in methanol with AgNO3, which is known to promote
the replacement of alkylthiolates with the alkoxide (CH3O)
with inversion.29 The (S)-6 → (S)-7 conversion represents an
inversion, due to a priority-assignment change (circled numbers
in Schemes 4a and 4b) when the leaving group (MeS; priority 1)
was replaced by MeO (priority 2) in (S)-7.
Patterson and co-workers20,40 have carried out several
computational studies on the hydrolysis and perhydrolysis on
model phosphonothioates. Initial attack by HO− or HOO− had
a facial selectivity where the nucleophile attacked the face
opposite the alkoxy ligand. This resulted in a trigonal bipyramid
intermediate (A) in which the alkoxide group occupies the
apical position. Accordingly, P−O scission would result which
contradicts experimental results that favor P−S over P−O bond
scission by 87:13 for alkaline hydrolysis41 and 100:0 for
perhydrolysis.10 Their computational work revealed a low-
energy isomerization (Ψ) that interchanged the apical alkoxide
for the alkylthiolate (A → B in Scheme 2) which rationalized
the preferential P−S bond scission. This facial selectivity for
attack opposite the alkoxy ligand followed by the pseudoro-
tation was also seen in the calculation of ethoxide attack on
O,S-dimethyl phenylphosphonothioate.21 In this work, Patter-
son and Menke found that the pseudorotation of the initial
trigonal bipyramind (A) will lead to both P−O and P−S
scission, but P−O cleavage was reversible and slightly
endothermic while the P−S cleavage pathway was irreversible
and highly exothermic. Their calculation work was in
disagreement with the results of DeBruin in terms of the facial
selectivity of the ethoxide nucleophilic attack. As the authors
concluded, “Further studies may be required to resolve this
discrepancy.”21
In line with prior alcoholysis work, we report methanolysis
results on the chiral substrate O,S-diethyl phenylphosphono-
thioate (DEPP) that show a SN2(P) route (eq 1) during the
initial reaction period wherein exclusive P−S scission yields O-
ethyl O-methyl phenylphosphonate (EMPP). Evidence for this
result comes from 100% inversion of the starting DEPP
configuration. In addition, we also show that the La(III) and
Zn(II)-promoted methanolysis (yielding P−S scission) proceed
through this SN2(P) pathway.
Therefore, if (S)-5 underwent nucleophilic substitution with
inversion, (S)-7 would be formed as shown in Scheme 4a. This
329
dx.doi.org/10.1021/ic2016897 | Inorg. Chem. 2012, 51, 328−335

Inorganic Chemistry
Article
Scheme 3
equivalence was attributed to enantiomeric discrimination by
the host CDs.
We applied the same enantiomeric discrimination strategy to
discern the stereochemistry of DEPP methanolysis. Initially, a
racemic mixture of DEPP yielded two distinct and equal 31P
NMR signals in a D2O solution when β-CD was added. We
identified the upfield 31P signal to be the R enantiomer when
authentic (R)-DEPP ((R)-5) was added (Figure 3). This
chemical shift nonequivalence is consistent with a guest−host
interaction between DEPP and β-CD that results from an
optimal fit of the phenyl group of DEPP into the β-CD cavity.
In this connection, the addition of α- or γ-CD to racemic DEPP
yielded no enantiomeric discrimination as seen by 31P NMR;
the smaller and larger CDs did not have the optimal fit for an
effective guest−host interaction. Moreover, when isopropanol
was added to the β-CD + racemic-DEPP mixture, both 31P
signals merged back to one (Figure 3d). Confirmation that only
one enantiomer (S) of 5 (DEPP) and 7 was made in Schemes 3
and 4b comes from the observation that both compounds
yielded only one 31P signal in the presence aqueous β-CD.
The methanolysis of (R)-DEPP ((R)-5) was monitored with
31P NMR until the reaction reached completion (∼30 min at
room temperature in 0.6 M NaOMe) with only P−SEt scission.
This was evident with the production of only one product 31P
NMR signal at ∼22 ppm attributed to O-ethyl O-methyl phos-
phonate (7). We rule out ethoxide displacement during this
time period as there was no 46.7 ppm signal due to O-methyl-S-
ethyl phenylphosphonothioate (8), which was independently
made. Interestingly, the measured activation energy barrier for
this process (Supporting Information, S1) was 14.5 0.5 kcal/mol,
Figure 1. Crystal structure of (S)-O-ethyl phenylphosphonothioate
((S)-4) isolated in Scheme 3 showing just the phosphonothioate
without the brucine cation for clarity purposes. The phenyl group is
placed pointing away to show the (S) configuration with the following
priority ranking in decreasing order: S (yellow) → OEt → O (red) →
Ph. R = 4.28%.
would be confirmed by addition of the authentic optically pure
(S)-O-ethyl O-methyl phenylphosphonate enantiomer ((S)-7)
made in Scheme 4a. Likewise if methanolysis of (S)-5
underwent retention, then the (R)-7 product would be formed,
which upon enantiomeric resolution would be distinct from the
synthesized (S)-7.
Enantiomeric resolution by NMR spectroscopy was done
with a cyclodextrin (CD) encapsulating agent. Inclusion com-
plexation of small molecules into the CD cavity (Figure 2) in
aqueous media is driven by a combination of chiral recognition
and hydrophobic interactions,42 and they have been proposed
as potential chiral shift reagents for NMR spectroscopy.43
Recently, Talebpour and Molaabasi44 showed that cyclo-
dextrins successfully resolve enantiomeric mixtures of chiral
organophosphates that possess a hydrophobic group. Specifi-
cally when the phenyl-containing pesticide fenamiphos
(racemic) was added to various β-CDs, there were two distinct
and equal 31P NMR signals, and this chemical shift non-
330
dx.doi.org/10.1021/ic2016897 | Inorg. Chem. 2012, 51, 328−335

Inorganic Chemistry
Article
Scheme 4
Figure 2. Structure of β-CD showing the cyclic polysaccharide that forms a conical shape with hydrophobic cavity. β-CD forms a guest−host
interaction with the phenyl groups of DEPP (R enantiomer of DEPP shown docked).
indicated a stereospecific process. If this product was (S)-7
phosphonate (retention, in Scheme 4a), then the addition of
authentic (S)-7 yields only one enantiomer that would show up
as one 31P signal in the presence of β-CD. Instead (Figure 4a−d),
addition of the independently made (S)-7 to this methanolysis
product yielded a second 31P signal in the presence of β-CD; the
original O-ethyl O-methyl phenylphosphonate (7) product was
the R enantiomer. This indicated a (R)-5 → (R)-7 methanolysis.
Further confirmation of this stereochemical transformation
was undertaken for the methanolysis of (S)-5. This enantiomer
was made from the (S)-4 in Scheme 3, and it contained a small
amount of the R enantiomer. Therefore, the starting (S)-5 was
not optically pure, for it contained a minor amount of the (R)-5
enantiomer when assayed with β-CD (Supporting Information, S2).
Nevertheless, when the (S) -5 phosphonothioate underwent
methanolysis as described above, the (S)-7 product was formed;
addition of authentic (S)-7 (containing some (R)-7 impurity) to
the methanolysis-product/β-CD mixture yielded mainly one
31P signal as shown in Figure 4e−h. Both the aforementioned
Figure 3. 31P NMR spectra of β-cyclodextrin in D2O (a) with 3 μL of
(S)-5 → (S)-7 and the (R)-5 →(R)-7 conversion indicate an
inversion (Scheme 4b) for the methanolysis process.
racemic DEPP (5), (b) with 3 μL of (R)-DEPP ((R)-5), (c) with 3 μL
of racemic DEPP and (R)-DEPP added, and (d) with the addition of
propanol (the varying chemical shifts are due to changes in
concentrations and environments).
This methanolysis result is consistent with prior stereochem-
ical findings of alkoxide attack on phosphonothioates.22−32
Therefore, this analysis was extended to two metal-catalyzed
methanolyses of phosphonothioates. Specifically, La(OTf)3 and
Zn/2,9-dimethylphenanthroline were investigated to see how
the stereochemical outcome of methanolysis is influenced by
these metal ions/complexes. Brown and co-workers found a
109-fold rate enhancement for the methanolysis of aryl phono-
sphonothioates in the presence of La(III),16 and a 106-fold rate
which was almost identical to the calculated enthalpy (14.4 kcal/mol)
of activation for ethoxide attack on O,S-dimethyl methyl-
phosphonothioate.21
Upon completion of the (R)-5 methanolysis, the methanol
was removed in vacuo and replaced with D2O and β-CD. The
31P NMR showed only one product signal at 22.2 ppm which
331
dx.doi.org/10.1021/ic2016897 | Inorg. Chem. 2012, 51, 328−335

Inorganic Chemistry
Article
46.7 ppm (vide supra) that would result from ethoxide release.
The measured activation energy barrier of 15.6 1.1 kcal/mol
(Supporting Information, S3) for the La(III)-catalyzed
methanolysis of 5 is close to the Ea for the methanolysis
without any metal ions (14.5 0.5 kcal/mol in the Supporting
Information, S1). However, these two processes were done
under radically different solvent conditions wherein the former
was buffered with N-ethylmorpholine (pH 9.80HOMe) and the
latter was done in 0.6 M NaOMe. This precludes making
reasonable assessments on the activation effects the La(III) ion
makes on the methanolysis. Removal of La(OTf)3 from the
N-ethylmorpholine and NaOMe solution yielded a methanolysis
process too slow to measure.
Figure 4. 31P NMR spectra of the product of the methanolysis of (S)-
DEPP and (R)-DEPP. The methanolysis of (R)-DEPP: (a) the
methanolysis product, (b) methanolysis product + β-CD, (c) 1.3 μL of
(S)-7 added, (d) another 1.3 μL of (S)-7 added. 31P NMR solvent was
D2O. The methanolysis of (S)-DEPP: (e) the methanolysis product,
(f) the methanolysis product + β-CD, (g) 1.3 μL of (S)-7 added, (h)
another 1.3 μL of (S)-7 added. The small upfield signals in (f) and (g)
result from the optical impurity of the starting (S)-DEPP. The
authentic (S)-7 contained some R enantiomer impurity.
Moreover, we saw 100% stereospecificity when (R)-5 was
used in the presence of La(OTf)3. Specifically, when β-CD was
added to the product (after replacement of MeOH with D2O)
in Figure 5 (time = 1 h), we saw the identical behavior
as in Figure 4a−d (Supporting Information, S5). This indicated
complete inversion (SN2(P)) for the (R)-5 → (R)-7
methanolysis. At room temperature, the 7 product showed
no (S)-7 enantiomer at time 60 min. Only the (R)-5 was
used as the starting material, for it was 100% optically pure
(vide supra).
enhancement when the Zn/2,9-dimethylphenanthroline system
was applied to catalytically degrade (i.e., methanolysis)
phosphate triesters.45 The Zn(II) coordination system resulted
from the complexation of Zn(OTf)2 and one equivalent of 2,9-
dimethylphenanthroline (neocuproine) to form a monomer−
dimer mixture (eq 2).45
Over time, the initial (R) − 7 product undergoes racemi-
zation to yield the (S)-7 enantiomer which is seen in Figure 5.
This is due to a second methanolysis on the initial (R)-7
product as shown in eq 3. This secondary transformation is
further substantiated with the appearance of O,O-dimethyl
phenylphosphonate (i.e., ethoxide leaving group) with a 31P
signal at 23.5 ppm (* in Figure 5).
Interestingly, the Zn(II)−neocuproine system (eq 2) yielded
the same stereochemical results when (R)-5 underwent
methanolysis. Complete conversion to the phosphonate (7)
The methanolysis of 5 by stoichiometric La(OTf)3 was
complete after one hour (room temperature), and this isolate
contained no O-methyl-S-ethyl phenylphosphonothioate (8) at
Figure 5. Methanolysis (31P NMR) of of (R)-5 with La(OTF)3 catalyst at room temperature over extended time to show production of both O,O-
dimethyl phenylphosphonate (* at 23.5 ppm) and (S)-7. Methanolysis product was dried and combined with aqueous β-CD. Verification of these
two products was done with authentic addition (Supporting Information, S4). The production of both compounds indicates a second methanolysis
on the initial (R)-7 product as shown in eq 3.
332
dx.doi.org/10.1021/ic2016897 | Inorg. Chem. 2012, 51, 328−335

Inorganic Chemistry
Article
1
alcohol. H NMR (CDCl3): δ 1.4 (t, 3H, O−CH2−CH3), 4.2 (q, 2H,
O−CH2−CH3), 7.2 (m, 2H, meta) 7.5 (d, 2H, ortho), 7.6 (s, 1H,
P−H), 7.8(d, 1H, para). 13C NMR (CDCl3): δ 16.7 (d, JPC =7.1 Hz,
at room temperature took six hours with no sign of 8 (i.e.,
ethoxide leaving group). β-Cyclodextrin addition to this
product followed by authentic (S)-7 (Supporting Information,
S6) gave the same results as La(OTf)3-catalyzed methanolysis
and sodium methoxide degradation (Figure 4a−d). Like the
case with La(OTf)3-catalyzed methanolysis, this showed that
the Zn(II)−neocuprine coordination complex promoted a (R)-
5 → (R)-7 conversion with inversion on the phosphorus center.
O−CH2−CH3), 62.4 (d, JPC = 6.4 Hz, O−CH2−CH3), 129.2 (d, JPC
=
13.7 Hz, meta), 131.3 (d, JPC = 12 Hz, ortho), 133.5 (d, JPC = 3 Hz,
para), 132.3 (d, JPC = 150 Hz, ipso). 31P NMR (CDCl3): δ 24.6.
Synthesis of O-Methyl and O-Ethyl Phenylphosphono-
thioate Dicyclohexylammonium Salt (2). Elemental sulfur (6.15 g,
190 mmol) was added to a solution of 1 (29.94 g, 190 mmol) and
dicyclohexylamine (34.45 g, 190 mmol) in diethyl ether (300 mL)
slowly over 30 min and then stirred for 4 h. The resulting solid was
filtered off, dried and recrystallized with ethyl acetate. The resulting salt
crystals of 2 were obtained (44.3 g, 120 mmol) with a 63% yield. The
same protocol was used to make the dicyclohexylammonium salt of
O-methyl phenylphosphonothioate (2b). 31P NMR (CDCl3): δ 67.43.
Interconversion of O-Alkyl Phenylphosphonothioate Dicy-
clohexylammonium Salt with Corresponding Thioacid (3). The
2 salt (44.3 g, 0.12 mol) was added to a stirring solution of sodium
hydroxide (1.5 M, 230 mL), stirred for 45 min and then washed with
toluene (3 × 100 mL). The aqueous layer was then acidified with
sulfuric acid (6 N, 80 mL), which was then saturated with sodium
chloride. The organic layer that formed was extracted with ether
(4 × 200 mL), and the ether layer was dried over sodium sulfate and
filtered. The ether solution was finally concentrated under reduced
pressure to the resulting 3 thioacid oil (22.42 g, 120 mmol) with a
CONCLUSION
■
These results indicate that methanolysis either with or without
La(III) or Zn(II) catalysis proceeds with inversion of stereo-
chemistry that is consistent with an “SN2-like” pathway (Scheme 2).
In the methanolysis of diethyl S-aryl phosphonothioates catalyzed
by La(III),17 Brønsted plot measurements suggest that the alkoxide
nucleophile attacks opposite the leaving group in most likely a
concerted process. Interestingly the metal ions did not stabilize the
hypothesized intermediate/transition state to promote a stepwise
pathway. Recent calculation studies46 show that an inversion of
stereochemistry could occur through either a concerted or stepwise
mechanism. In the case for the latter pathway, a poor leaving group
(pKa > 8) favors a nonconcerted route that proceeds through a
phosphorane (trigonal bipyramid) intermediate with no pseudoro-
tation. As this is the first case of employing cyclodextrins to probe
the stereochemical outcome of methanolysis, future work is focused
on using these oligosaccharides to promote hydrolysis of
phosphonothioates.
1
100% yield. H NMR (CDCl3): δ 1.4 (t, 3H, O−CH2−CH3), 4.2
(q, 2H, O−CH2−CH3), 7.5 (m, 3H, meta and para), 7.9 (d, 2H, ortho),
4.5 (br, 1H, SH). 31P NMR (CDCl3): δ 80.2.
Resolution of O-Ethyl Hydrogen Phenylphosphonothioate
(3) with Brucine. A solution of the racemic 3 thioacid (20 g,
104 mmol) in acetone (50 mL) was added dropwise over 30 min to a
boiling and stirring solution of acetone (550 mL) and brucine (41 g,
96 mmol). The volume was reduced below 600 mL with boiling and
then cooled to room temperature and allowed to sit for 5 days. The
resulting solid was then filtered from solution and recrystallized
from methanol (300 mL) giving the white (S)-4 brucine salt (26.04 g,
41.7 mmol; mp 198−210 °C). The remaining acetone solution was
concentrated to an oil, which was converted to a solid upon the
addition of methanol (120 mL). The mixture was then heated to
boiling and filtered, giving the (R)-4 brucine salt (22.75 g, 36.4 mmol;
mp 110−115 °C). The combined yield of the brucine salts ((S)-4 and
(R)-4) was 75%. 31P NMR (CDCl3): δ 70.9.
The (S)-4 brucine salt (25.0 g, 40.0 mmol) was dissolved in a
solution of sodium hydroxide in 35% v/v methanol (0.900 M, 50 mL)
and upon the addition of water (65 mL) it solidified. The resulting
solid was then washed with methylene chloride (4 × 40 mL). The
aqueous layer was then acidified with hydrochloric acid (6 M, 8 mL),
and the resulting organic layer was extracted with methylene chloride
(4 × 15 mL) and concentrated under reduced pressure to give the (S)-
3 thioacid oil (7.02 g, 34.6 mmol). The thioacid products are unstable
and require transformation into the corresponding dicyclohexylammo-
nium salt for long-term storage. Dicyclohexylamine (7.23 mL,
34.6 mmol) was added to a solution of this (S)-3 thioacid (7.02 g,
34.6 mmol) in ether (70 mL), stirred for 2.5 h and filtered. The resulting
(S)-2 salt was recrystallized from ethyl acetate. Starting from the 25.0 g
of brucine (S)-4 salt, the recovered (S)-2 cyclohexylammonium salt
EXPERIMENTAL SECTION
■
Materials and Methods. 31P, 1H and 13C NMR spectra were
obtained on a Bruker Avance-300 at 121 MHz, 300 MHz, and
75 MHz, respectively. High resolution mass spectral analyses were
done by U. Illinois-Champaign Urbana Mass Spectrometry Lab
Services. All cyclodextrins and reagents for the synthesis of
phosphonothioates were purchased from TCI (Portland, OR) and
used without further purification. In a typical methanolysis reaction
without the La(III) catalyst, 2.6 μL of DEPP (5) was added to a 1 mL
0.6 M NaOMe/MeOH solution at room temperature. Samples were
periodically removed and monitored by 31P NMR in CDCl3 until the
reaction reached completion (∼30 min). Upon completion, the
methanolic/methoxide solution was neutralized with HCl, dried in
vacuo and then redissolved in 800 μL of D2O. Cyclodextrin (70 mg)
was added to this aqueous solution for enantiomeric resolution with
31P NMR. For temperature dependent studies the methanolysis of
DEPP was monitored (31P) in a 0.6 M NaOMe/CD3OD solution.
Methanolysis with La(III) was done according to the procedure of
Tsang and co-workers16 except three times more La(OTf)3 (40 mg
versus 12.9 mg) was used. The methanolysis of DEPP by Zn(II)
complexes followed the procedure of Desloges and co-workers45
wherein the reaction took place in 1.0 mL of dry MeOH (with 20%
CD3OD as a NMR lock) containing 1 mM each of Zn(OTf)2,
neocuproine, NaOMe and 5. The methanolysis was buffered in 0.5 mM
tetrabutylammonium hydroxide.
Synthesis of O-Methyl or O-Ethyl Phenylphosphinate
(PhP(O)(H)OEt) or PhP(O)(H)OMe) (1 or 1b). A solution of
ethanol (22.1 mL, 540 mmol) (or appropriate alcohol), pyridine
(26.2 mL, 325 mmol), and toluene (36 mL) was added dropwise over
30 min to a solution of dichlorophenylphosphine (34 mL, 250 mmol)
in toluene (175 mL). The mixture was stirred for 1.5 h and allowed to
sit without stirring for 1 day. The solution and resulting white solid
were washed with saturated sodium bicarbonate (80 mL), and the
aqueous layer was back extracted with methylene chloride (70 mL).
The toluene and methylene chloride layers were combined, dried over
magnesium sulfate, filtered and then concentrated down to the O-ethyl
(1) oil product (29.94 g, 190 mol) with a 76% yield. The O-methyl
product (1b) was made with a similar procedure with methanol as the
(5.49 g, 0143 mol) represented a 34% yield (mp 153−154 °C; [α]D
+9.3°, methanol)
=
The (R)-4 brucinium salt was converted to the thioacid ((R)-3) and
then to the dicyclohexylammonium salt, (R)-2, with the same
procedure (4.01 g, 11 mmol; 30% yield from the brucine salt; mp
152−153 °C; [α]D = −7.7°, methanol).
Synthesis of O,S-Diethyl Phenylphosphonothioate, DEPP (5)
and O-Methyl-S-Ethyl Phenylphosphonothioate (8). The dicy-
clohexylammonium salt of compound 2 (5.0 g, 13 mmol) was slowly
added to a stirring solution of distilled toluene (100 mL) and ethyl
iodide (2.4 mL, 30 mmol). The mixture was stirred for 3 days (but less
time may be used). The resulting suspension was filtered and washed
with anhydrous hexanes, which were concentrated under reduced
pressure. This oil was washed repeatedly with a minimal amount of
333
dx.doi.org/10.1021/ic2016897 | Inorg. Chem. 2012, 51, 328−335

Inorganic Chemistry
Article
anhydrous hexanes to remove residual salt. The hexane washes were
concentrated down to the resulting 5 oil (1.5 g, 6.3 mmol). Kugelrohr
distillation was used if extraction with hexanes did not remove
impurities. The enantiomerically pure (S)- and (R)-5 were obtained
from the resolved dicyclohexylammonium salts with a similar
procedure with 53% and 61% yields respectively. 1H NMR
(CDCl3): δ 1.3 (t, 3H, O−CH2−CH3), 1.4 (t, 3H, S−CH2−CH3),
2.8 (q, 2H, S−CH2−CH3), 4.2 (q, 2H, O−CH2−CH3), 7.5 (m, 3H,
meta and para), 7.9 (d, 2H, ortho). 13C NMR (CDCl3): δ 16.6 (d,
JPC = 5.9 Hz, S−CH2−CH3), 16.8 (d, JPC = 6.9 Hz, O−CH2−CH3),
25.3 (d, JPC = 3.1 Hz, S−CH2−CH3), 62.5 (d, JPC = 6.9 Hz, O−CH2−
CH3), 128.9 (d, JPC = 14.75 Hz, meta), 131.6 (d, JPC = 10.6 Hz, ortho),
132.9 (d, JPC = 3.1 Hz, para), 132.3 (d, JPC = 150 Hz, ipso). 31P NMR
(CDCl3): δ 44.65. HRMS: calculated for C10H15O2SP 230.05304
(EI+); found 230.05364.
Table 1. Crystallographic Data for Compound 4
chemical formula
formula weight
temperature
C31H37N2O6PS
596.66
199(2) K
wavelength
0.71073 Å, Mo Kα
0.080 × 0.300 × 0.400 mm
colorless plate
monoclinic
crystal size
crystal habit
crystal system
space group
P 1 21 1
unit cell dimensions
a = 8.608(2) Å
b = 14.461(3) Å
c = 12.669(4) Å
1484.3(7) Å3
2
α = 90°
β = 109.748°
γ = 90°
volume
The O-methyl-S-ethyl phenylphosphonothioate (8) was made in
the same manner starting with the dicyclohexylammonium O-methyl
Z
density (calcd)
adsorption coefficient
F(000)
1.335 Mg/cm3
0.210 mm−1
632
1
phenylphosphonothioate salt, 2b. H NMR (CDCl3): δ 3.89 (s 3H,
O−CH3), 2.77 (q, 2H, S−CH2−CH3), 1.29 (t, 3H, S−CH2−CH3), 7.9
(d, ortho, 2H), 7.6 (m, meta and para, 3H). 13C NMR (CDCl3): δ
16.6 (d, JPC = 5.6 Hz, S−CH2−CH3), 25.3 (d, JPC = 3.0 Hz, S−CH2−
CH3), 52.6 (d, JPC = 7.2 Hz, O−CH3), 128.9 (d, JPC = 14.9 Hz, meta),
131.6 (d, JPC = 10.7 Hz, ortho), 132.9 (d, JPC = 3.3 Hz, para), 132.3 (d,
JPC = 150 Hz, ipso). 31P NMR (CDCl3, ppm): δ 46.7.
theta range
2.52 to 25.02°
23505
reflections collected
independent reflections
max. and min transmission
function minimized
data/restraints/parameters
5180 [R(int) = 0.0706]
0.9834 and 0.9208
∑w(Fo − Fc2)2
2
Synthesis of S-Methyl O-Ethyl Phenylphosphonothioate
(6). The dicylcohexylammonium salt of racemic O-ethyl phenyl-
phosphonothioate (2) (500 mg, 1.3 mmol) was added to stirring
distilled methyl iodide (3.5 mL, 56 mmol). The solution was stirred
for 24 h and filtered, and the precipitate was washed with anhydrous
hexanes. The filtrate and hexane washes were concentrated under
reduced pressure and then Kugelrohr distilled at 70−85 °C (15−
20 μm) giving the resulting colorless 6 oil (90 mg, 0.40 mmol) with a
5180/1/377
a
final R indices
4240 data; I > 2σ(I) R1 = 0.0428, wR2 =
0.1023
all data R1 = 0.0581, wR2 = 0.1098
a
2
2
w = 1/[σ2(Fo ) + (0.0508P)2 + 0.2209P] where P = (Fo + 2Fc2)/3.
ASSOCIATED CONTENT
1
■
32% yield. H NMR (CDCl3): δ 1.4 (t, 3H, O−CH2−CH3), 2.2 (s,
S
* Supporting Information
3H, S−CH3), 4.2 (m, 2H, O−CH2−CH3), 7.5 (m, 3H, meta and
para), 7.9 (d, 2H, ortho). 13C NMR (CDCl3): δ 12.4 (d, JPC = 3.43 Hz,
S−CH3), 16.8 (d, JPC = 6.7 Hz, O−CH2−CH3), 62.6 (d, JPC = 7.2 Hz,
O−CH2−CH3), 128.9 (d, JPC = 13.35 Hz, meta), 131.6 (d, JPC = 11.5
Hz, ortho), 132.9 (d, JPC = 3.3 Hz, para), 133.2 (d, JPC = 150 Hz, ipso).
31P NMR (CDCl3): δ 45.5. HRMS: calculated for C9H13O2PS (EI+)
Detailed 31P NMR spectra, activation measurements, and
crystal structure report for LewClark_brucine. This material is
available free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*E-mail: kuo@lclark.edu.
■
216.03739; found 216.03835.
Synthesis of Enantiomerically Pure (S)-O-Ethyl O-Methyl
Phenylphosphonothioate (7). (S)-6 (1.17 g, 5.0 mmol), made
from (S)-2 as indicated above, was added to a suspension of silver
nitrate (1.68 g, 10 mmol) in 10 mL of methanol at 0 °C. The mixture
was then stirred for 24 h. This suspension was filtered, reduced down
to 5 mL, and then diluted with methylene chloride (50 mL). The
solution was washed with saturated sodium bicarbonate (4 × 25 mL),
dried over sodium sulfate, filtered and concentrated under reduced
pressure to the 7 oil (625 mg, 3.0 mmol) with a 62.5% yield. No
ACKNOWLEDGMENTS
■
This work was funded by NSF-RUI award CHE-0956740, and
S.K.G. thanks the Lewis & Clark College John S. Rogers
Science Foundation for financial support. We thank R. S.
Brown and Y. Shariati for useful discussion and help, and Bruce
Noll at Bruker (Madison WI) for solving the structure of 4.
The activation data from Allison Akagi is gratefully acknowl-
edged.
1
further purification was performed. H NMR (CDCl3): δ 1.36 (t, 3H,
O−CH2−CH3), 3.7 (s, 3H, O−CH3), 4.15 (m, 2H, O−CH2−CH3),
7.53 (m, 3H, meta and para), 7.8 (d, 2H, ortho). 13C NMR (CDCl3):
δ 16.8 (d, JPC = 6.5 Hz, O−CH2−CH3), 52.9 (d, JPC = 5.2 Hz, O−
CH3), 62.7 (d, JPC = 5.5 Hz, O−CH2−CH3), 128.3 (d, JPC = 14.9 Hz,
meta), 131.6 (d, JPC = 10.9 Hz, ortho), 132.3 (d, JPC =3.3 Hz, para),
133.2 (d, JPC = 150 Hz, ipso). 31P NMR (CDCl3): δ 20.23. HRMS:
calculated for C9H13O3P (EI+) 200.05898; found 200.06024.
X-ray Structure Analysis of Brucinium Salt of (S)-O-Ethyl
Phenylphosphonothioate. A colorless platelike crystal of dimen-
sions 0.080 mm × 0.300 mm × 0.400 mm was used for X-ray crystallo-
graphic analysis on a Bruker SMART X2S benchtop diffractometer
equipped with a Mo Kα microfocus source (λ = 0.7107 Å). A total of
2160 frames were collected that were integrated (Bruker SAINT
v7.68) using a monoclinic unit cell in the range of 2.52° ≤ θ ≤ 25.02°.
The structure was solved and refined using the Bruker SHELXTL
Software Package with space group P1 21 1 and Z = 2. The final
anistropic full-matrix least-squares refinement converged at R1 =
4.28%. Table 1 summarizes the crystal data and final residuals, and
more extensive tables including specific details of data collection and
structure refinement are in the Supporting Information.
REFERENCES
■
(1) Larson, R. A.; Weber, E. J. Reaction Mechanisms in Environmental
Organic Chemistry; Lewis Publishers: Boca Raton, FL, 1994.
(2) Somani, M. Chemical Warfare Agents; Academic Press: San Diego,
CA, 1992.
(3) Yang, Y. C.; Baker, J. A.; Ward, J. R. Chem. Rev. 1992, 92, 1729−
1743.
(4) Battershill, J. M.; Edwards, P. M.; Johnson, M. K. Food Chem.
Toxicol. 2004, 42 (8), 1279−1285.
(5) Thompson, C. M.; Berkman, C. E.; Ryu, S.; Jackson, J. A.; Quinn,
D. A.; Larsen, A. Rev. Pestic. Toxicol. 1993, 2, 133−148.
(6) Kennedy, R. J.; Stock, A. M. J. Org. Chem. 1960, 25, 1901−1906.
(7) Zhu, W.; Ford, W. T. J. Org. Chem. 1991, 56, 7022−7026.
(8) Webb, K. S. Tetrahedron lett. 1994, 35, 3457−3460.
(9) Bunton, C. A.; Foroudian, H. J.; Kumar, A. J. Chem. Soc., Perkin
Trans. 1995, 2, 33−39.
334
dx.doi.org/10.1021/ic2016897 | Inorg. Chem. 2012, 51, 328−335

Inorganic Chemistry
Article
(10) Yang, Y.-C.; Szafraniec, L. L.; Beaudry, W. T. J. Org. Chem.
1993, 58, 25.
(11) YangY.-C.BergF. J.SzafraniecL. L.BeaudryW. T.BuntonC.
(44) Molaabasi, F.; Talebpour, Z. J. Agric. Food Chem. 2011, 59, 803−
808.
(45) Desloges, W.; Neverov, A. A.; Brown, R. S. Inorg. Chem. 2004,
43, 6752−6761.
(46) Tarrat, N. THEOCHEM 2010, 941, 56−60.
A.KumarA.J. Chem. Soc., Perkin Trans. 21997
(12) Yang, Y.-C.; Szafraniec, L. L.; Beaudry, W. T.; Rohrbaugh, D. K.;
Procell, L. R.; Samuel, J. B. J. Org. Chem. 1996, 61, 8407−8413.
(13) Blasko, A.; Bunton, C. A.; Kumar, A. J. Phys. Org. Chem. 1997,
10, 427−434.
(14) Morales-Rojas, H.; Moss, R. A. Chem. Rev. 2002, 102 (7),
2497−2521.
(15) Yang, Y. C. Acc. Chem. Res. 1999, 32, 109−115.
(16) Tsang, J. S.; Neverov, A. A.; Brown, R. S. J. Am. Chem. Soc. 2003,
125 (25), 7602−7607.
(17) Liu, T.; Neverov, A. A.; Tsang, J. S. W.; Brown, R. S. Org.
Biomol. Chem. 2005, 3, 1525−1533.
(18) Melnychuk, S. A.; Neverov, A. A.; Brown, R. S. Angew. Chem.,
Int. Ed. 2006, 45, 1767−1770.
(19) Debruin, K. E.; Tang, C.-l. W.; Johnson, D. M.; Wilde, R. L.
J. Am. Chem. Soc. 1989, 111 (15), 5871−5879.
(20) Seckute, J.; Menke, J. L.; Emnett, R. J.; Patterson, E. V.; Cramer,
C. J. J. Org. Chem. 2005, 70 (22), 8649−8660.
(21) Menke, J. L.; Patterson, E. V. THEOCHEM 2007, 811, 281−
291.
(22) Hall, C. R.; Inch, T. D.; Williams, N. E. Phosphorus Sulfur 1983,
18 (1−2−3), 213−216.
(23) Hall, C. R.; Inch, T. D.; Peacock, G.; Pottage, C.; Williams, N. E.
J. Chem. Soc., Perkin Trans. 1 1984, No. 4, 669−674.
(24) Aaron, H. S.; Uyeda, R. T.; Frack, H. F.; Miller, J. I. J. Am. Chem.
Soc. 1962, 84, 617−621.
(25) Michalski, J.; Okruszek, A.; Stec, W. J. Chem. Soc. D 1970,
No. 22, 1495−1497.
(26) Farnham, W. B.; Mislow, K.; Mandel, N.; Donohue, J. J. Chem.
Soc., Chem. Commun. 1972, No. 3, 120−121.
(27) Hall, C. R.; Williams, N. E. 1,3,2-Thiazaphospholidin-2-ones
derived from ephedrine. Preparation and stereochemistry of ring-opening
reactions; Chem. Def. Establ.: 1981; pp 2746−2750.
(28) Hall, C. R.; Inch, T. D. Tetrahedron Lett. 1976, No. 40, 3645−
3648.
(29) Cooper, D. B.; Hall, C. R.; Harrison, J. M.; Inch, T. D. J. Chem.
Soc., Perkin Trans. 1 1977, No. 17, 1969−1980.
(30) Inch, T. D.; Lewis, G. J. Carbohydr. Res. 1975, 45 (1), 65−72.
(31) Inch, T. D.; Lewis, G. J.; Wilkinson, R. G.; Watts, P. J. Chem.
Soc., Chem. Commun. 1975, No. 13, 500−501.
(32) Hall, C. R.; Inch, T. D.; Pottage, C.; Williams, N. E. Tetrahedron
1985, 41 (21), 4909−4917.
(33) Benschop, H. P.; van den Berg, G. R.; Boter, H. L. Recl. Trav.
Chim. Pays-Bas 1968, 87 (5), 387−395.
(34) Benschop, H. P.; Platenburg, D. H. J. M. J. Chem. Soc. D 1970,
No. 17, 1098−1099.
(35) Van den Berg, G. R.; Platenburg, D. H. J. M.; Benschop, H. P.
Recl. Trav. Chim. Pays-Bas 1972, 91 (7), 929−934.
(36) Farnham, W. B.; Murray, R. K. Jr.; Mislow, K. J. Am. Chem. Soc.
1970, 92 (19), 5809−5810.
(37) Mislow, K.; Donohue, J.; Mandel, N.; Farnham, W. B.; Murray,
R. K.; Benschop, H. P. J. Am. Chem. Soc. 1971, 93 (15), 3792−3793.
(38) DeBruin, K. F.; Johnson, D. M. J. Chem. Soc., Chem. Commun.
1975, No. 18, 753−754.
(39) Moriyama, M.; Bentrude, W. G. J. Am. Chem. Soc. 1983, 105
(14), 4727−4733.
(40) Daniel, K. A.; Kopff, L. A.; Patterson, E. V. J. Phys. Org. Chem.
2008, 21 (4), 321−328.
(41) In Review and Evaluation of Alternative Chemical Disposal
Technologies; National Research Council, Washington, DC; National
Academy Press: Washington, DC, 1996.
(42) Dugas, H., Cyclodextrins. In Bioorganic Chemistry: A Chemical
Approach to Enzyme Action, 2nd ed.; Cantor, P. C. R., Ed.; Springer-
Verlag: New York, 1989; pp 350−366.
(43) MacNicol, D. D.; Rycroft, D. S. Tetrahedron Lett. 1977, No. 25,
2173−2176.
335
dx.doi.org/10.1021/ic2016897 | Inorg. Chem. 2012, 51, 328−335