Alkali metal ion catalysis in nucleophilic displacement by ethoxide ion on p-nitrophenyl phenylphosphonate: Evidence for multiple metal ion catalysis1

Article abstract of DOI:10.1139/v02-202

In continuation of our studies of alkali metal ion catalysis and inhibition at carbon, phosphorus, and sulfur centers, the role of alkali metal ions in nucleophilic displacement reactions of p-nitrophenyl phenylphosphonate (PNPP) has been examined. All al

Full text of DOI:10.1139/v02-202

53
Alkali metal ion catalysis in nucleophilic
displacement by ethoxide ion on p-nitrophenyl
phenylphosphonate: Evidence for multiple metal
ion catalysis1
Erwin Buncel, Ruby Nagelkerke, and Gregory R.J. Thatcher
Abstract: In continuation of our studies of alkali metal ion catalysis and inhibition at carbon, phosphorus, and sulfur
centers, the role of alkali metal ions in nucleophilic displacement reactions of p-nitrophenyl phenylphosphonate (PNPP)
has been examined. All alkali metal ions studied acted as catalysts. Alkali metal ions added as inert salts increased the
rate while decreased rate resulted on M+ complexation with 18-crown-6 ether. Kinetic analysis indicated the interaction
of possibly three potassium ions, four sodium ions, and five lithium ions in the transition state of the reactions of
ethoxide with PNPP. Pre-association of the anionic substrate with two metals ions in the ground state gave the best fit
to the experimental data of the sodium system. Thus, the study gives evidence of the role of several metal ions in
nucleophilic displacement reactions of ethoxide with anionic PNPP, both in the ground state and in the transition state.
Molecular modeling of the anionic transition state implies that the size of the monovalent cation and the steric require-
ment of the pentacoordinate transition state are the primary limitations on the number of cations that can be brought to
bear to stabilize the transition state and catalyze nucleophilic substitution at phosphorus. The bearing of the present
work on metal ion catalysis in enzyme systems is discussed, in particular enzymes that catalyze phosphoryl transfer,
which often employ multiple metal ions. Our results, both kinetic and modeling, reveal the importance of electrostatic
stabilization of the transition state for phosphoryl transfer that may be effected by multiple cations, either monovalent
metal ions or amino acid residues. The more such cations can be brought into contact with the anionic transition state,
the greater the catalysis observed.
Key words: alkali metal ion catalysis, nucleophilic displacement at phosphorus, multiple metal ion catalysis, phosphoryl
transfer. 63
Résumé: Dans le cadre de nos études sur la catalyse par les ions de métaux alcalins et sur l’inhibition au niveau des
centres carbonés, phosphorés et sulfurés, on a examiné le rôle des métaux alcalins sur les réactions de substitutions nu-
cléophiles du phénylphosphonate de p-nitrophényle (« PNPP »). Tous les ions de métaux alcalins étudiés agissent
comme catalyseurs. Les ions de métaux alcalins ajoutés sous la forme de sels inertes augmentent la vitesse alors que
l’on observe des diminutions de vitesse lorsque les ions M+ sont complexés par l’éther 18-couronne-6. L’analyse ciné-
tique indique qu’il pourrait exister une interaction de trois ions potassium, quatre ions sodium et cinq ions lithium dans
l’état de transition des réactions de l’éthanolate avec le « PNPP ». Une préassociation du substrat anionique avec deux
ions métalliques dans l’état fondamental permet de réaliser le meilleur ajustement des données expérimentales du sys-
tème du sodium. Cette étude permet donc de mettre en évidence le rôle de plusieurs ions métalliques dans les réactions
de substitutions nucléophiles de l’éthanolate avec le « PNPP » anionique, tant dans l’état fondamental que dans l’état
de transition. Des calculs de modélisation moléculaire de l’état de transition anionique impliquent que la taille du ca-
tion monovalent et les nécessités stériques de l’état de transition pentacoordiné sont les limitations primaires du nombre
de cations qui peuvent être amenés à stabiliser l’état de transition et à catalyser la substitution nucléophile au niveau
du phosphore. On discute des implications du présent travail sur la catalyse par des ions métalliques sur les systèmes
enzymatiques, particulièrement ceux qui catalysent le transfert de groupes phosphoryles qui font souvent appel à de
multiples ions métalliques. Nos résultats, tant cinétiques que théoriques, révèlent l’importance de la stabilisation élec-
trostatique de l’état de transition pour le transfert de phosphoryle qui peut être affecté par de plusieurs cations, ions
métalliques monovalents ou résidus d’acides aminés; plus on peut amener de tels cations en contact avec l’état de tran-
sition anionique meilleure est la catalyse observée.
Received 25 September 2002. Published on the NRC Research Press Web site at http://canjchem.nrc.ca on 29 January 2003.
E. Buncel,2 R. Nagelkerke,3 and R.J. Thatcher. Department of Chemistry, Queen’s University, Kingston, ON K7L 3N6, Canada.
1Part 10 in a series on Metal ion catalysis and inhibition in nucleophilic displacement reactions at carbon, phosphorous, and sulfur
centers. Part 9: ref. (1).
2Corresponding author (e-mail: buncele@chem.queensu.ca).
3Present address: Department of Chemistry, Centralia College, Centralia, WA 98531, U.S.A.
Can. J. Chem. 81: 53–63 (2003)
doi: 10.1139/V02-202
© 2003 NRC Canada

54
Can. J. Chem. Vol. 81, 2003
Mots clés : catalyse par des ions métalliques alcalins, substitution nucléophile au niveau du phosphore, catalyse par des
ions métalliques multiples, transfert de phosphoryle.
[Traduit par la Rédaction]
Introduction
Buncel et al.
have applied in the previous cases (i.e., simple ion pair and
dissociated ion reaction pathways) are no longer applicable
here, but that multiple ion association has to be invoked.
Both ground state and transition state interactions are dis-
cussed.
The role of metal ions in the mechanism of phosphoryl
transfer reactions (2, 3) has been under extensive investiga-
tion owing to the importance of these processes in biological
systems (4, 5). The ubiquitous presence of alkali metal ions
in biological systems calls for intensive investigation of ca-
talysis by these monovalent ions, but such studies have only
rarely been undertaken. The majority of studies have been
directed towards the catalytic effects of divalent ions such as
Mg2+, Ca2+, Mn2+, Fe2+, Cu2+, and Zn2+ in the hydrolysis of
phosphate esters (6–8).
Results
The reaction under investigation is shown below with the
assumption that the substrate is fully ionized in basic etha-
nol.
We have initiated systematic studies of the influence of al-
kali metal ions in nucleophilic displacement reactions at car-
bon, phosphorus, and sulfur centers (9–20). Our previous
reports have shown that in the reaction of alkali metal ethox-
ides and phenoxides with p-nitrophenyl diphenylphosphinate
1 (PNPDP), all the alkali metals functioned as catalysts,
with the MOEt reactivity order LiOEt > NaOEt > KOEt >
CsOEt > EtO– (13, 14). However, with p-nitrophenyl ben-
zenesulfonate 2 (PNPBS), both catalysis and inhibition was
observed, the reactivity order now being KOEt > CsOEt >
NaOEt > EtO– > LiOEt (15, 16).
The rate of the reaction in anhydrous ethanol at 25°C was
followed spectrophotometrically by monitoring the forma-
tion of p-nitrophenoxide ion under pseudo-first-order condi-
tions with the alkoxide in at least 20-fold excess. Excellent
first-order kinetics were observed up to 90% reaction and
rate constants were obtained from linear log (A∞ – At) vs.
time plots. Slower reactions were conducted for only one
half-life or less, and rate constants were obtained by an ini-
tial rate method. Results are presented in Tables 1–3 and dis-
played in Fig. 1.
The effect of added complexing agent has been examined;
Fig. 2 and Table 4 show the effect of added 18-crown-6 on
the rate of reaction of PNPP with KOEt at a constant base
concentration. The rate decreased until 1 equiv of complex-
ing agent had been added, after which point increasing the
crown concentration had no effect on the reaction rate. The
constant rate observed after 1 equiv of complexing agent had
been added corresponds to the reaction of free ethoxide with
the phosphonate ester on the assumption that cations are
fully complexed by the added 18-crown-6 and are unreac-
tive. The average rate constant for the free ethoxide so ob-
tained is 5 2 × 10–7 s–1 M–1.
The role of the metal ion in the reaction mechanism was
further investigated by adding metal ions in the form of
unreactive salts to reactions of alkali metal ethoxides with 3.
The effect of added LiCl and LiNO3 on the LiOEt reaction is
shown in Table 5. To achieve a substantial [salt]:[base] ratio,
a low concentration of base was used for these experiments.
The reactions were so slow that plots of absorbance vs. time
did not show the curvature expected for first-order reactions.
Instead, the absorbance varied linearly with time and zero-
order rates were observed. The reported rate constants were
determined in the same manner as for initial rates,
In the present work we have extended our investigations
to p-nitrophenyl phenylphosphonate 3 (PNPP), as the first
monoanionic phosphorus ester in this series, noting the im-
portance of charged phosphate esters in biological systems.
In the reaction of 3 with alkali metal ethoxides in ethanol
the order of catalytic effectiveness was determined to be
LiOEt > NaOEt > KOEt > CsOEt > > EtO–, i.e., qualita-
tively similar to the situation with PNPDP (1). However, in-
depth analysis shows that the kinds of treatment that we
© 2003 NRC Canada

Buncel et al.
55
Table 1. Kinetic data for the reaction of
PNPP with LiOEt in EtOH at 25.0°C.
Table 3. Kinetic data for the reaction of
PNPP with KOEt in EtOH at 25.0°C.
[MOEt] (M)
kobs (s–1)
[MOEt] (M)
kobs (s–1)
0.100
0.106
0.140
0.140
0.200
0.240
0.300
0.320
0.400
0.400
0.400
0.500
0.560
0.600
0.700
0.800
5.18 × 10–6
3.93 × 10–6
6.67 × 10–6
1.25 × 10–5
1.65 × 10–5
2.32 × 10–5
3.95 × 10–5
4.12 × 10–5
7.49 × 10–5
7.31 × 10–5
6.46 × 10–5
1.39 × 10–4
1.83 × 10–4
2.34 × 10–4
3.76 × 10–4
6.34 × 10–4
0.189
0.245
0.283
0.300
0.377
0.415
0.500
0.604
0.698
0.792
3.52 × 10–6
3.86 × 10–6
7.47 × 10–6
6.56 × 10–6
1.32 × 10–5
1.28 × 10–5
1.94 × 10–5
2.97 × 10–5
4.57 × 10–5
6.15 × 10–5
Fig. 1. Kinetic data for the reaction of PNPP with LiOEt, KOEt,
and NaOEt, and with KOEt in the presence of excess 18-crown-6
in EtOH at 25°C.
Table 2. Kinetic data for the reaction of
PNPP with NaOEt in EtOH at 25.0°C.
[MOEt] (M)
kobs (s–1)
0.170
0.239
0.242
0.298
0.401
0.501
0.601
0.700
0.800
6.64 × 10–6
1.32 × 10–5
1.28 × 10–5
2.05 × 10–5
3.82 × 10–5
6.73 × 10–5
1.09 × 10–4
1.70 × 10–4
2.61 × 10–4
converting the change in absorbance per run time, ∆A/∆t, to
a zero-order rate constant and dividing the rate constant by
the substrate concentration to get kobs
.
Discussion
Catalytic role of alkali metal ions
The order of reactivity exhibited (Fig. 1) in the present
work with PNPP is LiOEt > NaOEt > KOEt > > EtO–, which
is qualitatively similar to that observed in our previous work
with PNPDP (1), but the rates of reaction are much smaller
of alkali – metal ethoxides in ethanol are known (21), one
could calculate the contribution of each species to the two
separate reaction pathways (for the case of PhO–M+ the re-
spective contributions could be evaluated by a mathematical
treatment) (14). For the phosphinate 1 it was found that for
all the alkali metal ethoxides and phenoxides, associated
RO–M+ ion pairs were more reactive than free RO– ions.
This could be rationalized in terms of a more effective stabi-
lization of the transition state by M+ than of the ground
state. Stabilization energies of the transition state and of the
ground state could be evaluated by a method developed orig-
inally by Kurz (22) and applied by Mandolini and coworkers
(23), Tee (24), and others (25–28). For the sulfonate (2) –
ethoxide reaction the higher alkali metal ions acted as
in the present work: kEtO− (PNPP)/k
(PNPDP) is ca. 10–6.
EtO−
This is in accord with nucleophilic attack occurring in the
present case on an anionic substrate. The increase in rate on
addition of the unreactive salts LiCl and LiNO3 to the LiOEt
reaction, while moderate (Table 5), is in accord with a cata-
lytic effect of Li+ (see also ref. 1).
Catalysis by alkali metal ethoxide ion pairs
In our previous studies with p-nitrophenyl diphenylphos-
phinate (1) as well as with p-nitrophenyl benzenesulfonate
(2) results were analyzed in terms of reactivity by dissoci-
ated EtO– (or PhO–) ions and by the associated EtO–M+ (or
PhO–M+) ion pairs (13–17). Since the association constants
© 2003 NRC Canada

56
Can. J. Chem. Vol. 81, 2003
Table 4. Kinetic data for the reaction of PNPP with KOEt in the presence of 18-crown-6 in EtOH at 25.0°C.
[KOEt] (M)
[18C6] (M)
kobs (s–1)
k2 (s–1M–1)
[18C6]/[KOEt]
0.1008
0.1008
0.1008
0.1008
0.1008
0.1008
0.1008
0.1008
0.1008
0.1008
0.1008
0
0
6.61 × 10–7
8.35 × 10–7
5.96 × 10–7
2.81 × 10–7
3.22 × 10–7
4.01 × 10–7
1.49 × 10–7
7.49 × 10–8
5.82 × 10–8
3.27 × 10–8
3.98 × 10–8
6.55 × 10–6
8.28 × 10–6
5.91 × 10–6
2.79 × 10–6
3.20 × 10–6
3.97 × 10–6
1.48 × 10–6
7.43 × 10–7
5.77 × 10–7
3.24 × 10–7
3.95 × 10–7
0
0
0.030
0.061
0.061
0.061
0.091
0.101
0.127
0.134
0.210
0.301
0.602
0.602
0.602
0.903
1.01
1.26
1.33
2.08
Note: Average kEtO− value for kEtO− ([18C6]/[KOEt] > 1.0) = 5 2 × 10–7 s–1M–1.
Table 5. (a) Kinetic data for the reaction of PNPP with LiOEt in
the presence of LiCl in EtOH at 25.0°C; and (b) kinetic data for
the reaction of PNPP with LiOEt in the presence of LiNO3 in
EtOH at 25.0°C.
Fig. 2. Kinetic data for the reaction of PNPP with KOEt in the
presence of 18-crown-6 in EtOH at 25°C.
(a) Ratio LiCl/LiOEt
kobs (s–1)
0a
0
9.47
23.7
37.7
52.1
66.3
6.2 × 10–7
6.7 × 10–7
7.0 × 10–7
7.8 × 10–7
8.0 × 10–7
kobs (s–1)
(b) Ratio LiNO3/LiOEt
0
0a
9.26
23.2
37.1
51.0
64.9
6.4 × 10–7
7.5 × 10–7
7.8 × 10–7
8.4 × 10–7
9.0 × 10–7
Note: [LiOEt] = 0.00398 M.
aNo reaction was observed without added salt.
Scheme 1. Metal ethoxide ion pair catalysis.
catalysts but Li+ acted as inhibitor, which is indicative of
greater stabilization of the ground state than of the transition
state (17–19).
Applying the free ethoxide – metal ethoxide ion pair mo-
del to the present system, with reaction pathways for both
species (Scheme 1), leads to the rate eqs. [1] and [2] where
S– refers to the substrate
−
PNPP with metal ethoxides. The upward curvature in Fig. 3
suggests that higher involvement in M+ should be consid-
ered.
[1]
rate = kEtO− [EtO ][S−] + kMOEt[MOEt][S−]
−
−
[2]
kobs/[EtO ] = kEtO− + kMOEtKa[EtO ]
According to eq. [2], plots of kobs/[EtO–] vs. [EtO–] calcu-
lated (see Tables S1–S3)4 using the known (21) values of Ka
should be linear. However, it may be seen from the curved
nature of the plot in Fig. 3 for LiOEt and the very similar
plots for NaOEt and KOEt (not shown) that the ion pair
mechanism does not adequately describe the reaction of
Catalysis by pre-association mechanism
One can consider a pre-association pathway in which one
metal ion associates with the substrate and the consequent
reaction of ethoxide with S– M+, along with the parallel reac-
tions of S– with EtO– and S– with MOEt, as shown in
Scheme 2.
4 Supplementary data may be purchased from the Depository of Unpublished Data, Document Delivery, CISTI, National Research Council
Canada, Ottawa, ON K1A 0S2, Canada (http://www.nrc.ca/cisti/irm/unpub_e.shtml for information on ordering electronically).
© 2003 NRC Canada

Buncel et al.
57
Fig. 3. Ion pairing treatment of the data for the reaction of
Scheme 3. Ion pair and pre-association catalysis.
PNPP with LiOEt in EtOH at 25°C.
given in Scheme 3, leads to the rate law in eq. [5] from
which eq. [6] can be derived where kN refers to the “new”
rate constant for attack of an ion pair on a metal-complexed
substrate.
−
[5]
rate = kEtO− [EtO ][S−] + kM+ [EtO–][M+S−]
+ kMOEt[MOEt][S−] + kN[MOEt][M+S−]
−
−
[6]
kobs = kEtO− [EtO ] + (kM+ Kc + kMOEtKa)[EtO ]2
−
+ kNKcKa[EtO ]3
Scheme 2. Pre-association catalysis.
The relationship given in eq. [6] is more complex, having
[EtO–], [EtO–]2, and [EtO–]3 terms, and the previously used
plot of kobs vs. an independent variable (for the PNPDP sys-
tem) is not possible for this relationship. In an attempt to try
to fit the experimental data using the relationship presented
in eq. [6], a spreadsheet (Lotus 123 (Release 2.01)) was
used. Arbitrary values were input for the unknown constants
(kN, kM+ , and Kc) and these values were optimized iteratively
until a best fit to the experimental data was observed. Once a
relationship was determined, the fit of the calculated values
of kobs to the experimental value of kobs was measured using
standard linear regression methods present in the Lotus pro-
gram.
However, no combination of values for the unknown con-
stants could be found for eq. [6] that would satisfactorily fit
all the kobs experimental data, especially the data for the
higher base concentrations (0.06–0.09 M). Equation [6] con-
tains first-, second-, and third-order terms in [EtO–]. How-
ever, it will be shown in the following section that a plot of
kobs vs. [EtO–] has curvature that is adequately fit by a fifth-
order [EtO–] term and this curvature cannot be fit by second-
and third-order terms. This result is taken to indicate that the
proposed mechanism in Scheme 3, involving just two metal
ions, is not adequate to describe the reaction of metal ethox-
ides with PNPP.
The rate equation corresponding to Scheme 2 is given in
eq. [3] from which eq. [4] follows, which is of the same
form as eq. [2].
−
−
[3]
rate = kEtO− [EtO ][S−] + kM+ [EtO ][M+S−]
+ kMOEt[MOEt][S−]
−
−
[4]
kobs/[EtO ] = kEtO− + (kM+ Kc + kMOEtKa)[EtO ]
As before, it follows from eq. [4] that plots of kobs/[EtO–]
vs. [EtO–] should be linear, but as has already been noted
these plots (e.g., Fig. 3) are curved upward thus ruling out
the pre-associative mechanism of Scheme 2.
Catalysis by both pre-associative and ion pair
mechanism
To explore a pathway involving two metal ions, we con-
sider the possibility of attack on S– M+ by MOEt in parallel
with the pathways shown in Scheme 2. This mechanism,
Multiple ion catalysis
From attempts to fit the kinetic data to third-order terms
in [EtO–] it became apparent that higher order terms in
[EtO–] were needed to fit kobs over the entire [EtO–] range
studied (0–0.1 M). Given that [EtO–] is equal to [M+],
© 2003 NRC Canada

58
Can. J. Chem. Vol. 81, 2003
Fig. 4. Curve fitting of the data for the reaction of PNPP with
LiOEt in EtOH at 25°C with [Li+]n, n = 4, 5, 6, and 7 (see Ta-
ble 6).
Fig. 5. Curve fitting of the data for the reaction of PNPP with
NaOEt in EtOH at 25°C with [Na+]n, n = 4, 5, and 6 (see Ta-
ble 6).
attempts were made to fit the observed kinetic data to [M+]n.
The relationship being examined is shown in eq. [7], where
kp refers to the rate constant for reaction of free ethoxide
with the substrate as mediated by n metal ions. The results
are shown in Figs. 4, 5, and 6. Table 6 gives the coefficients
and the goodness of the fits (as measured by r2) for the best
fit lines shown in these figures.
Fig. 6. Curve fitting of the data for the reaction of PNPP with
KOEt in EtOH at 25°C with [K+]n, n = 3, 4, 5, and 6 (see Ta-
ble 6).
The rate increase from reaction of the mono-anion at rela-
tively high concentrations (0.05–0.1 M) of alkali metal ion is
consistent with rate constants that are approximately fourth-
order in K+, fifth-order in Na+, and sixth-order in Li+, or
their kinetic equivalents, as shown below.
kp
−
[7]
[8]
[9]
EtO + S− + n[M+] → products
rate = kobs[S−]
kobs = kp[M+]n
−
[10] kobs = 1.42 × 104 [Li+]6 = 1.42 × 104[Li+]5[EtO ]
−
[11] kobs = 60.8 [Na+]5 = 60.8 [Na+]4[EtO ]
−
[12] kobs = 0.963[K+]4 = 0.963[K+]3[EtO ]
The results are interpreted as approximately three K+, four
Na+, and five Li+ ions being involved in the transition states
of the respective reactions, assuming there is no pre-
association with the ground state. Consideration of metal ion
association with the substrate will not affect the calculation
of the number of metal ions present in the transition state, as
shown below. The number of metal ions appears to be re-
lated to the size of the ions, which suggests that size limits
the number of ions that may participate with the transition
state.
sumes a “one-step”, or elementary, reaction, which would
preclude pre-association of these metal ions with the anionic
substrate followed by subsequent attack of ethoxide, a two-
step reaction. Pre-association of the substrate with metal
ions would increase the number of metal ions present in the
transition state.
The role of the metal ions is not discernible from the
above treatment, which only gives an estimate of the number
of metal ions that associate with the transition state. It
should be noted that the use of eq. [9] (based on eq. [7]) as-
© 2003 NRC Canada

Buncel et al.
59
Table 6. Table of the coefficients used to fit [M+]n terms to the experimental values of kobs for the reaction of MOEt on PNPP in etha-
nol at 25°C.
Coefficient (r2)
Metal ethoxide system
LiOEt
[M+]3
[M+]4
[M+]5
783
(0.978)
60.8
(0.996)
11.7
(0.986)
[M+]6
1.42 × 104
(0.995)
762
(0.993)
138
(0.951)
[M+]7
2.53 × 105
41.7
(0.928)
4.73
(0.970)
0.963
(0.990)
(0.991)
NaOEt
KOEt
7.63 × 10–2
(0.934)
Note: Equation to be fit: kobs = (coefficient)[M+]n. See Figs. 4, 5, and 6. The correlation coefficients (r2) are shown and the terms with the best fit are in
bold type. The data for kobs are given in supplementary tables S1, S2, and S3.
Scheme 4. Na+ catalyzed reaction of AcP with Pi.
To further analyze the kinetic data, a detailed reaction
mechanism needs to be proposed, which requires assump-
tions about how many ions interact with the ground state.
The data for sodium, for example, could be accounted for by
association of one metal ion with the ethoxide and one with
the anionic substrate, or, alternatively, by the kinetically
equivalent free ethoxide attack on the substrate pre-
associated by two metal ions in the ground state, both with
possibly four additional ions in the transition state. There are
several kinetically equivalent possibilities, which will be dis-
cussed below.
charge of –4 delocalized over seven monovalent and two
divalent oxygens. There is thus ample scope for the four ad-
ditional bound Na+ ions, not shown, to coordinate with sev-
eral of these oxygens in the transition state.
In an attempt to investigate the possible role of the metal
ions in the ground state of the present system, the associa-
tion of one and two metal ions with mono-anionic PNPP in
the ground state was considered. In support for the pre-
association of two metal ions in the present system, theoreti-
–
cal ab initio studies of the anion PO2H2 by Streitwieser et
al. (29) indicate that the best simple representation of the di-
–
sodium complex of PO2H2 in the gas phase is the structure
+
Na-O-PH2 -O-Na, tetrahedral at phosphorus and containing
Pre-association and multiple ion catalysis — Possible
rate laws
co-linear P—O—Na bonds.
An example of the complexity of the analysis of multiple
ion catalytic participation is given by Herschlag and Jencks
who studied the Na+ catalyzed reaction of the anions acetyl
phosphate (AcP) and orthophosphate (Pi) to yield acetate
and pyrophosphate (PPi) anions (7). In their analysis they
noted that the data was well fit by a fourth-order term in Na+
concentration. When they considered a mechanism that in-
cluded pre-association of the dianionic substrate they were
able to re-analyze their data in terms of the associated rate
laws. They interpreted their data in terms of two Na+ ions
that bind in the ground state and five additional ions that
bind in the transition state. They noted, however, that the
data could be interpreted by the kinetically equivalent asso-
ciation of two Na+ ions with AcP dianion, two Na+ ions with
Pi dianion, or one Na+ ion with each dianion, all in the
ground state. On the basis of the kinetic dependence alone,
Herschlag and Jencks have speculated on a possible transi-
tion state structure for reaction of the dianions of acetyl
phosphate and orthophosphate, with three bridging Na+ ions
and with four additional bound Na+ ions not shown
(Scheme 4) (7). The transition state for this reaction has a
If one metal ion is presumed to associate with the anionic
substrate with free ethoxide as the nucleophile, then the fol-
lowing mechanism and associated rate laws may be pro-
posed (Scheme 5). Ka refers to the association of the
substrate with either one or two metal ions, kc refers to the
rate constant for the attack of free ethoxide on an uncom-
′
plexed substrate, and kc refers to the rate constant for the at-
tack of free ethoxide on a metal-complexed substrate.
[13] rate = kobs[S]total
[14] rate = kobs[S + S·M] = kc[S·M] + kc [M]n[S·M]
′
[15] Ka = [S·M]/[S−][M+]
n
′
[16] kobs = (kc[S·M] + kc [M] [S·M]) / ([S + S·M])
[17] kobs = (kcKa[M+])/(1+ Ka[M+])
+ n
+
+
′
+ kc [M ] (Ka[M ])/(1 + Ka[M ])
© 2003 NRC Canada

60
Can. J. Chem. Vol. 81, 2003
Scheme 5. Pre-association with one Na+.
Scheme 6. Pre-association with two Na+.
If two metal ions are presumed to associate with the an-
ionic substrate and free ethoxide is the nucleophile, then the
following mechanism and associated rate laws may be pro-
posed (Scheme 6).
Table 7. Parameters used in curve fitting (Figs. 7–9).
Set 1 Set 2 Set 3
1.65 × 10–5 5.0 × 10–6 3.9 × 10–6 3.8 × 10–6
Set 4
kc(M–1s–1)
–6 –1
′
kc (M s )
136
10
70
60
64
120
67
140
[18] rate = kobs[S]total
Ka(M–2)
Fit (r2)
0.9987
0.9964
0.9957
0.9957
[19] rate = kobs[S + S·M2] = kc2 [S·M2]
Set 5
4.5 × 10–4
kc2 (M–6s–1) 850
Set 6
Set 7
Set 8
+ kc2 [M]n[S·M2]
′
8.0 × 10–5 4.8 × 10–5 3.5 × 10–5
190
60
0.9993
kc2(M–1s–1)
[20] ka2 = [S·M2]/[S−][M+]2; [S·M2] = Ka2[S−][M+]2
′
142
100
0.9993
110
160
0.9994
Ka2(M–2)
Fit (r2)
10
0.9982
[21] kobs = (kc2[S·M2] + kc2 [M]n[S·M2])/(S + S·M2])
′
[22] kobs = (kc2Ka2[M+]2)/(1 + Ka2[M+]2)
+ kc2 [M+]n(Ka2[M+]2)/(1 + Ka2[M+]2)
′
[Na+]5 term was used to fit the kinetic data. It may be seen
that inclusion of a term for the pre-association of the sub-
strate with two metal ions in the rate law gives a better fit
over the low range of sodium ion concentration (0.03–
0.06 M).
It was possible to fit the experimental data to the relation-
ships presented in eqs. [17] and [22] using a spreadsheet
program. The data for the Na+ system was arbitrarily used to
fit to the models presented above. There are several un-
known parameters to be placed in the relationships, namely
In the curve fitting, the values assigned to the parameters
′
kc2, kc2 , and Ka2 were arbitrary. Sets 1–4 and 5–8 are just a
′
kc2, kc2 , Ka2, and n, the number of metal ions associated
few of the many possible values that could be used to fit the
data. However, it is important to note that the magnitude of
kc(2), the rate constant for the unassociated ethoxide for both
sets of parameters (≈10–6 – 10–4) is much smaller in magni-
with the transition state. By putting in arbitrary values for
these constants and optimizing them iteratively, it was possi-
ble to find several sets of values that fit the experimental val-
ues of kobs well, as measured by a linear regression
correlation coefficient (r2). In the previous section a good fit
to the experimental data for the Na+ system was given by
n = 5, so this was the value used for this parameter.
The best fit to the experimental data was found via
eq. [22] with n = 5 and values from set 8 in Table 7 for the
mechanism that had two metal ions pre-associated with the
substrate in the ground state (Fig. 7, r2 = 0.9994). It is inter-
esting to compare Fig. 7, in which a pre-association term is
included in the data treatment, with Fig. 5, in which only the
tude than that for the corresponding values of kc2 (≈101 –
′
102), the rate constant for the associated ethoxide. This is in-
terpreted as further evidence for multiple ion catalysis.
Proceeding to Li+ and K+ catalysis, Figs. 8 and 9 show
sample fits for the plots of kobs vs. [M+], which were calcu-
lated based on the rate laws associated with the pre-
association of two metal ions with the substrate using the
stoichiometry from eq. [22] and n = 4 (K+) and 6 (Li+). The
arbitrary values for the parameters used are given in the
© 2003 NRC Canada

Buncel et al.
61
Fig. 7. Multi-ion catalysis treatment, including pre-association of
the substrate with two sodium ions, of the data for the reaction
of PNPP with NaOEt in EtOH at 25°C. (Data for kobs in supple-
mentary table S2 and parameters from set 8). Note the goodness
of the fit for the entire range of data.
Fig. 8. Multi-ion catalysis treatment, including pre-association of
the substrate with two Li+ ions, of the data for the reaction of
PNPP with LiOEt in EtOH at 25°C (r2 = 0.997). (Data for kobs
in supplementary table S1 and values for the parameters used are
–1 –1
kc2 = 1 × 10–5 M s , kc2 = 2000 M–7s–1, and Kc2 = 1000 M–2).
′
observed for this reaction. A full kinetic analysis was under-
taken for sodium based on this multiple ion mechanism, and
good fits to the experimental data for all three systems stud-
ied were possible with arbitrary values used for the un-
known parameters. The unknown parameters were the rate
constants for the uncomplexed and complexed ground states
and the association constant for the ground state with either
one or two metal ions, as well as the number of metal ions
added in the transition state.
The mechanism proposed on the basis of the kinetic anal-
ysis is just one of several that are possible, but the good
agreement with the experimental data is evidence for multi-
ple metal ion catalysis of nucleophilic substitution at phos-
phorus and suggests possible roles for metal ions in both the
ground states and transition states of the reaction.
It is interesting to speculate on the location of the four
Na+ ions stabilizing the probable trigonal bipyramidal (TBP)
transition state structure for PNPP ethanolysis. This transi-
tion state bears a charge of –2 and possesses relatively bulky
ligands at one equatorial and both apical positions. In con-
trast, the TBP transition state structure postulated by
Herschlag and Jencks (7) for pyrophosphate formation from
acetyl phosphate bears a charge of –4, is stabilized by seven
Na+ ions, and possesses no equatorial ligands and no bulky
hydrocarbon ligands (Scheme 4). The apical–equatorial met-
al ion coordination proposed for the TBP intermediate in al-
kaline phosphatase catalysis and in other metaloenzyme-
catalyzed phosphoryl transfer reactions would support the
transition state structure shown in Fig. 10a. However,
figure captions, as well as the correlation coefficients. Good
fits (r2 > 0.99) were obtained for both systems.
Pre-association and multiple ion catalysis – Possible
mechanisms
It has been shown that alkali metal catalysis of the reac-
tion of ethoxide with PNPP cannot be attributed simply to
catalysis by one metal ion, by either ion pair catalysis or a
pre-associative mechanism. Even catalysis by two metal
ions, as described by a combination of the two modes —
namely ion pair catalysis on a single metal-complexed sub-
strate — was not supported by the data treatment. The re-
sults of curve fitting the experimental rate data are consistent
with the participation of possibly 3–5 metal ions in the tran-
sition state, in addition to possible pre-association of metal
ions with the ground state. The largest ion (K+) has probably
three and the smallest ion (Li+) has probably five ions in the
transition state, while Na+ is intermediate, having probably
four. The high sensitivity of the observed reaction rates to
the concentration of the metal ion suggests that the metal
ions lower charge repulsion between the anionic ethoxide
nucleophile and the PNPP substrate. Direct binding of metal
ions to the reactants by pre-association or ion pairing could
lower charge repulsion between the reactants (7).
The data treatment suggests that pre-association of an-
ionic PNPP with probably two metal ions in the ground
state, in addition to multiple cations binding in the transition
state, may account for the multiple alkali metal ion catalysis
© 2003 NRC Canada

62
Can. J. Chem. Vol. 81, 2003
Fig. 9. Multi-ion catalysis treatment, including pre-association of
the substrate with two K+ ions, of the data for the reaction of
PNPP with KOEt in EtOH at 25°C (r2 = 0.996). (Data for kobs in
supplementary table S3 and values for the parameters used are
kc2 = 1 × 10–4 M–1s).
Fig. 10. Possible transition state structures for Na+ catalyzed
ethanolysis of PNPP (net charge +2): (a) idealized geometry,
with structure (b) obtained by semi-empirical calculations with
alkyl and aryl groups removed for ease of visualization.
excess of base was used (relative to the phosphonate ester
concentration) to ensure pseudo-first-order conditions for all
reactions. Rate constants were calculated from at least 30
absorbance readings. Rate constants for reactions involving
added metal ions were calculated as the slope of a plot of
ln(A∞ – At) vs. time. Rate constants for reactions involving
complexing agents were calculated using initial rate methods
(zero-order kinetics) and the data spanned less than one half-
life. Plots of absorbance vs. time were linear; therefore, the
reported rate constants were determined by converting the
observed change in absorbance to a change in concentration
of the product and dividing through by the initial concentra-
tion of the substrate (kzero = k1 [substrate]).
molecular orbital calculations on a putative TBP intermedi-
ate for PNPP ethanolysis stabilized by four Na+ ions show
that the steric requirement of the aryl, aryloxy, and alkoxy
ligands prevents such an ideal geometry (Fig. 10a). In this
crowded pentacoordinate TBP transition state, only a smaller
number of the larger metal ions can be brought into contact
with the anionic oxygens surrounding phosphorus. Our mo-
lecular modeling thus provides support for the stoichiometry
of the transition states for catalysis by Li+, Na+, and K+ ob-
tained from the kinetic analysis.
Molecular modeling
An equilibrium geometry was obtained from geometry op-
timization at the PM3 level of a putative TBP intermediate
([p-NO2PhOPPh(O)2OEt]2–). After addition of four sodium
atoms and constraint of the four atoms consisting of phos-
phorus, apical oxygens, and equatorial carbon, an equilib-
rium geometry was obtained by iterative relaxation of
molecular fragments, followed by relaxation of all atoms but
the four constrained atoms (Fig. 10b).
Conclusions
Experimental section
Metal ion catalysis is important in many enzyme systems
including those that catalyze phosphoryl transfer (4, 31). In
the catalysis of reactions at phosphorus, enzymes often em-
ploy multiple metal ions; for example, phospholipase C and
nuclease P1 employ two or three catalytic Zn2+ ions at the
active site (4). Whilst divalent metal ions and in particular
Zn2+ are the most important biological Lewis acid catalysts,
catalytic roles for monovalent cations have been suggested
in enzymic phosphoryl transfer. The location of two alkali
metal ions (Na+ or K+) in addition to Mg2+ at the ATP bind-
ing site of Hsc70 has been explained in terms of a catalytic
role in the ATPase activity of Hsc70 involving stabilization
of the pentacoordinate transition state (32). Interestingly, the
relative size of Na+ versus K+ is reflected in quite different
coordination in the protein crystal structures and presumably
different catalytic effects. Similarly, the active site of fruc-
tose-1,6-bisphosphatase contains three metal ion binding
sites, one of which is a monovalent cation binding site at
Materials
The general procedure of Williams and coworkers (30)
was followed for the synthesis of the phosphonate PNPP (1).
The crude product was recrystallized from ethanol to a con-
stant melting point (116–118°C, lit. (31) value mp 117–
119°C). 18-Crown-6 (Aldrich) was recrystallized from ace-
tonitrile, then dried over P2O5 in vacuo. LiCl (Aldrich, Gold
Label) and LiNO3 (Merck Suprapur) were dried over P2O5
in vacuo.
Kinetic method
The reaction of p-nitrophenyl phenylphosphonate (PNPP)
with alkali metal ethoxides was studied in anhydrous ethanol
at 25°C. The rates of the reactions were determined by mon-
itoring of the formation of p-nitrophenoxide spectrophoto-
metrically using either a Varian Cary3 or a Hewlett-Packard
8452A Diode Array spectrophotometer. An at least 20-fold
© 2003 NRC Canada

Buncel et al.
63
which K+ is postulated to exert Lewis acid catalysis of nu-
cleophilic substitution at phosphorus (whereas the smaller
size of Li+ results in inhibition rather than activation) (33).
In some proteins, cationic amino acid residues replace
monovalent metal ions, suggesting that monovalent cations
mimic the properties of cationic amino acids in stabilizing
anionic transition states.
Enzyme models have been reported that demonstrate bi-
functional catalysis by two metal ions (34). However, the
best metal-containing systems for acceleration of phosphoryl
transfer employ lanthanides (35), exchange-inert Co3+ sys-
tems, and ligand-coordinated Cu2+ (36).
10. R. Nagelkerke, M.J. Pregel, E.J. Dunn, G.R.J. Thatcher, and E.
Buncel. Org. React. Tartu, 102, 11 (1995).
11. E. Buncel, E.J. Dunn, R.A.B. Bannard, and J.G. Purdon. J.
Chem. Soc., Chem. Commun. 162 (1984).
12. E. Buncel and M.J. Pregel. J. Chem. Soc., Chem. Commun.
1566 (1989).
13. E.J. Dunn and E. Buncel. Can. J. Chem. 67, 1440 (1989).
14. E.J. Dunn, R.Y. Moir, and E. Buncel. Can. J. Chem. 69, 1837
(1990).
15. M.J. Pregel, E.J. Dunn, and E. Buncel. Can. J. Chem. 68, 1846
(1990).
16. M.J. Pregel, E.J. Dunn, and E. Buncel. J. Am. Chem. Soc. 113,
3545 (1991).
The simple system reported herein reveals multiple metal
ion catalysis of nucleophilic substitution at phosphorus in a
phosphonate monoester monoanion by the monovalent cat-
ions Na+, K+, and Li+. More unexpectedly, the detailed ki-
netic treatment of the data requires possibly three, four, or
five cations in the transition state, even though the putative
pentacoordinate transition state is only dianionic. The pri-
mary limitation to the number of cations involved in the
transition state appears to be the size of the cation and steric
limitations on the number of cations that can access the oxy-
gen ligands. Our results, both kinetic and modeling, reveal
the importance of electrostatic stabilization of the transition
state for phosphoryl transfer that may be effected by multi-
ple cations, either monovalent metal ions or amino acid resi-
dues. The more such cations can be brought into contact
with the anionic transition state, the greater the catalysis ob-
served.
17. M.J. Pregel and E. Buncel. J. Am. Chem. Soc. 115, 10 (1993).
18. M.J. Pregel and E. Buncel. J. Org. Chem. 56, 5583 (1991).
19. M.J. Pregel, E.J. Dunn, R. Nagelkerke, G.R.J. Thatcher, and E.
Buncel. Chem. Soc. Rev. 24, 445 (1995).
20. E. J. Dunn and E. Buncel. In Metal ions in biology and medi-
cine. Edited by Ph. Collery, L.A. Poirier, N.A. Littlefield and
J.C. Etienne. John Libbey Eurotext, Paris. 1994.
21. (a) J.R. Jones. Prog. React. Kinet. 7, 1 (1973); (b) A. Papoutsis,
G. Papanastasiou, J. Jannakoudakis, and C. Georgulis. J.
Chim. Phys. 82, 913 (1985).
22. J.L. Kurz. J. Am. Chem. Soc. 85, 987 (1963).
23. R. Cacciapaglia and L. Mandolini. Chem. Soc. Revs. 22, 221
(1993).
24. O.S. Tee. Adv. Phys. Org. Chem. (1994).
25. R. Wolfenden. Acc. Chem. Res. 5, 10 (1972).
26. E. Buncel and H. Wilson. Acc. Chem. Res. 12, 42 (1979).
27. R.L. Schowen. In Transition states of biochemical processes.
Edited by R.D. Gandour and R.L. Showen. Plenum, New York.
1978.
Acknowledgment
28. A. Fersht. Enzyme structure and mechanism. 2nd. ed. W.H.
Freeman and Company, New York. 1985.
29. (a) A. Streitwieser, A. Rajca, R.S. McDowell, and R. Glaser. J.
Am. Chem. Soc. 109, 4184 (1987); (b) A. Rajca, J.E. Rice, A.
Streitwieser, and H.F. Schaeffer. J. Am. Chem. Soc. 109, 4189
(1987).
This research was supported by the Natural Sciences and
Engineering Research Council of Canada (NSERC). The
award to R.N. of a Postgraduate Scholarship by NSERC is
gratefully acknowledged.
30. J.S. Loran, R.A. Naylor, and A. Williams. J. Chem. Soc.
References
Perkin Trans. 2, 418 (1977).
31. D.E. Wilcox. Chem. Rev. 96, 2375 (1996).
32. S.M. Wilbanks and D.B. McKay. J. Biol. Chem. 270, 2251
(1995).
1. R. Nagelkerke, G.R.J. Thatcher, and E. Buncel. Org. Biomol.
Chem. 1, 163 (2003).
2. G.R.J. Thatcher and R. Kluger. Adv. Phys. Org. Chem. 25, 99
(1989).
3. F.H. Westheimer. Chem. Rev. 81, 313 (1981).
4. W.N. Lipscomb and N. Slater. Chem. Rev. 96, 2375 (1996).
5. W.P. Jencks. Catalysis in chemistry and enzymology. Wiley,
New York. 1969.
6. R. Breslow. Adv. Enzymol. 58, 1 (1986).
7. D. Herschlag and W.P. Jencks. J. Am. Chem. Soc. 109, 4665
(1987).
33. R. Zhang, V. Villeret, W.N. Lipscomb, and H.J. Fromm. Bio-
chemistry, 35, 3038 (1996).
34. (a) S. Liu and A.D. Hamilton. Bioorg. & Med. Chem. Lett. 7,
1779 (1997); (b) R. Breslow and B. Zhang. J. Am. Chem. Soc.
116, 7893 (1994)
35. (a) W.W. Butcher and F.H. Westheimer. J. Am. Chem. Soc. 77,
2420 (1955); (b) A.A. Neverov and R.S. Brown. Inorg. Chem.
40, 3588 (2001); A.A. Neverov and R.S. Brown. J. Chem. Soc.
Perkin Trans. 2, 1039 (2002).
36. (a) M. Wall, R.C. Hynes, and J. Chin. Angew. Chem. 105,
1696 (1993); (b) D. Wahnon, A.-M. Lebuis, and J. Chin.
Angew. Chem., Int. Ed. 34, 2412 (1995).
8. R.A. Kenley, R.H. Fleming, R.M. Lane, D.S. Tse, and J.S.
Winterle. Inorg. Chem. 23, 1870 (1984).
9. M.J. Pregel, E.J. Dunn, and E. Buncel. Atual. Fis. Quim. Org.
Conf. Latinoam. Fis. Quim. Org. 1st, 110 (1991).
© 2003 NRC Canada