Oxidation of phosphorus centers by ferrate(VI): Spectral observation of an intermediate

Article abstract of DOI:10.1021/ic3001812

The kinetics and mechanism for the oxidation of phosphite, hypophosphite, phenylphosphite, and trimethylphosphite by ferrate(VI) are reported. Hypophosphite is rapidly oxidized to phosphite which is slowly oxidized to phosphate, trimethylphosphite is oxidized to trimethylphosphate, and phenylphosphite is oxidized phenylphosphate. ¹⁸O induced shifts of the ³¹P NMR signals support oxygen transfer from ferrate(VI) to the phosphorus center during the oxidation process. Deuteration of the hydridic hydrogens in hypophosphite and phosphite resulted in significant kinetic isotope effects on the reaction rates. It is proposed that ferrate(VI) acts as a two-electron oxidant in conjunction with oxide transfer coupled with phosphorus hydrogen bond breaking for phosphite and hypophosphite and simple oxygen transfer for trimethylphosphite and phenylphosphite.

Full text of DOI:10.1021/ic3001812

Article
pubs.acs.org/IC
Oxidation of Phosphorus Centers by Ferrate(VI): Spectral
Observation of an Intermediate
Shaun M. Hightower, Bret B. Lorenz, Jonathan G. Bernard, and Michael D. Johnson*
Department of Chemistry and Biochemistry, New Mexico State University, Las Cruces, New Mexico 88003, United States
S
* Supporting Information
ABSTRACT: The kinetics and mechanism for the oxidation
of phosphite, hypophosphite, phenylphosphite, and trimethyl-
phosphite by ferrate(VI) are reported. Hypophosphite is
rapidly oxidized to phosphite which is slowly oxidized to
phosphate, trimethylphosphite is oxidized to trimethylphos-
phate, and phenylphosphite is oxidized phenylphosphate. 18O
induced shifts of the 31P NMR signals support oxygen transfer
from ferrate(VI) to the phosphorus center during the
oxidation process. Deuteration of the hydridic hydrogens in hypophosphite and phosphite resulted in significant kinetic
isotope effects on the reaction rates. It is proposed that ferrate(VI) acts as a two-electron oxidant in conjunction with oxide
transfer coupled with phosphorus hydrogen bond breaking for phosphite and hypophosphite and simple oxygen transfer for
trimethylphosphite and phenylphosphite.
They proposed that an oxygen bridged intermediate is formed
and that the large deuterium isotope effect points to the
cleaving of a P−H bond in the rate-determining step. On the
basis of this evidence, Haight concluded that the removal of H+
from the P−H bond by H2O or H2PO3 precedes an internal
redox step involving the oxygen bridged intermediate.
INTRODUCTION
■
Over the past several years it has been established that iron
undergoes catalytic reactions that involve oxidation states above
+3.1 Unfortunately, within these oxidation states the aqueous
ions are generally unstable, and relatively little information
about the structures or reactions of these oxidation states is
known. Because of this, we have carried out a series of studies
of the reactions of iron(VI), also known as the ferrate ion,
−
A kinetic study of Cr(V) and hypophosphorous acid was
carried out by Ghosh and Gould between pH 2 and 4.12 The
oxidation of the formally P(I) center was shown to be a 2e−
process to form phosphite. They proposed that an oxygen-
bridged intermediate, in which the bridge is formed from the
hypophosphite oxygen, is the primary reaction pathway.
The oxidation of both P(I) and P(III) centers by
permanganate was studied by Zahonyi-Budo.13 They found
2‑
FeO4 , with a variety of substrates. The present study presents
the kinetics and proposed mechanisms for the oxidations of a
series of phosphorus centers by iron(VI).
The oxidations of phosphorus compounds have been studied
with a variety of oxidants such as Bi(V),2 chloramine T,3 OH
radicals,4 Ce(V),5 Ag(I),6 Ag(II),7 V(V),8 and AuCl4 .
9
−
−
that H3PO3 is oxidized by MnO4 approximately 250 times
Formation of M−O−P intermediates as well as outer-sphere
redox processes has been proposed.
more slowly than H3PO2. For hypophosphorous acid a kinetic
isotope effect of kH/kD of 4.6 was observed. As for
chromate(VI), this suggests that cleavage of the P−H bond
occurs in the rate determining step. Zahonyi-Budo proposed a
mechanism involving hydride abstraction via a bridged
intermediate with simultaneous loss of a proton.
Mehrotra also reported the permanganate oxidation of
phosphorous acid and found the reaction was first order with
respect to both reactants.14 The hydrogen ion dependence was
shown to be independent of ionic strength. He interpreted this
to indicate that the reaction proceeds via an intramolecular
pathway as opposed to a reaction between ions.
Haight et al. concluded that H2PO2− is not oxidized nor does
it form an anhydride with the chromate ion.10 It is interesting,
however, that hypophosphorous acid does form a spectrally
observable anhydride intermediate with chromate that sub-
sequently undergoes oxidation to form phosphorous acid. This
evidence led Haight to conclude that an OH group on the
phosphorus center must be present for anhydride formation.
Haight et al. also studied the chromate(VI) oxidation of
phosphorous acid at high P(III) concentrations, 0.1 M <
[H3PO3] < 1.0 M.11 A kinetic isotope effect of kH/kD = 4.0 was
observed, and they reported a rate law for the Cr(VI) and
phosphorous acid reaction as follows:
Oxidations involving iron(VI) continue to be an important
area of investigation.15 In our earlier work on the iron(VI)
oxidation of several different nitrogen containing substrates, we
rate = [Cr(VI)][H3PO3](k1[H+] + k2[H3PO3])
/(1 + Kf[H3PO3])
Received: January 24, 2012
Published: June 4, 2012
© 2012 American Chemical Society
6626
dx.doi.org/10.1021/ic3001812 | Inorg. Chem. 2012, 51, 6626−6632

Inorganic Chemistry
Article
with chemical shifts of known samples under identical conditions. A
400 MHz Varian NMR was used to collect data. These experiments
were carried out at different pHs (9 and 10), and no difference in
products was observed.
were forced to invoke the formation of iron−substrate
intermediates based on kinetic arguments which often, but
not always, involved oxygen transfer.16 Related intermediates
have also been proposed by Lee in the oxidation of alcohols
and by Johnson and Reed in the oxidation of sulfur containing
ions.17 Thus far, however, no direct observations have been made of
any iron(VI) intermediate during its oxidation of a substrate.
As part of our ongoing investigations into the chemistry of
Gas chromatography was used to identify the product for the
oxidation of trimethylphosphite to trimethylphosphate. A spent
aqueous reaction mixture was extracted with methlenechloride, and
the extraction mixture was injected on a Superox GC column to
separate starting materials from reaction products. Assignment of
retention times was carried out using known standards. Again, these
were carried out at different pHs, and no differences were observed.
In order to determine whether oxygens from ferrate(VI) were
transferred to the phosphorus center during oxidation, 18O labeling
studies were carried out. A sample of potassium ferrate(VI) was
dissolved into 90% 18O labeled water, the oxygens were allowed to
equilibrate (∼15 min), and solvent water was removed via vacuum
drying.19a This solid sample was rapidly dissolved in an aqueous
solution of the phosphorus reducing agent, which resulted in bleaching
of the violet color within 30 s, much faster than the solvent exchange
rate of the ferrate oxygens.19a The pH of the spent mixture was
approximately 9. The spent reaction mixture was treated as described
above for the NMR product analysis procedure, and the transfer of
oxygen was determined using the 18O induced shift in the 31P signal.19b
To check for induced oxygen exchange, normal sodium phosphate,
sodium phosphite, and sodium hypophosphite were treated by the
product isolation procedure described above in the presence of
enriched solvent, and no O-18 enrichment was observed in any of
these compounds. This confirms that any incorporation of O-18 into
the final products arises from transfer from the iron center.
2‑
FeO4 , we have studied the reactions of ferrate(VI) with
hypophosphite, phosphite, and organic phosphorus com-
pounds. While the original intent of this work was simply to
study the heretofore unexplored reactions, the unexpected
formation of observable intermediates makes this an important
piece of iron(VI) chemistry. The present study of the ferrate
oxidation of phosphorus centers has led to the proposal of a
mechanism that involves oxygen transfer coupled with the
breaking of a P−H bond in the transition state. The product
analysis, kinetics, and mechanism of these reactions are
described along with the first spectral observations of bridged
iron(VI)−substrate intermediates.
EXPERIMENTAL SECTION
■
Materials. Sodium hypophosphite and phosphorous acid (normal
and deuterated) and phenylphosphorous acid were purchased from
Aldrich Chemicals and used without purification. Trimethylphosphite
was purchased from Fischer Scientific. All other chemicals were
reagent grade and used without further purification.
The potassium ferrate was prepared by the method of Thompson,
Schreyer, and Ockerman.18 Purity of the ferrate was determined
spectrophotometrically as previously described.16 Typical purities of
>90% were obtained, and the reaction rates were independent of
sample.
Stock solutions of reductants were prepared using water from a
Barnstead Nanopure system. The ionic strength was maintained at 1.0
M with sodium perchlorate, and sodium phosphate buffers and sodium
hydroxide were used to control pH. The efficacious selection of
phosphate buffers also prevented precipitation of the iron(III)
hydroxides produced upon reduction of ferrate(VI).
Kinetic Measurements. The reaction rates were measured
spectrophotometrically using a D110 Dionex stopped-flow spectro-
photometer interfaced with an OLIS data collection and processing
system. The reactions were monitored at 505 nm under pseudo-first-
order conditions, >10-fold excess reductant, throughout. The substrate
concentrations were in the millimolar range, and the pH range studied
was between 8 and 11. Oxidations were also carried out at 1.0 M
NaOH for phosphite, which used an HP8452A diode array
spectrometer. The exponential absorbance changes were fitted using
the OLIS fitting routines. Each rate constant is an average of 3−5
trials. Temperature was controlled to within 0.1 °C using a constant
temperature bath.
Rapid scanning experiments were carried out using an OLIS RSM-
1000 spectrophotometer equipped with an OLIS stopped-flow.
Conventional spectrophotometric experiments used an HP8452
system equipped with a thermostatted cell holder. Absorbance versus
time changes were fitted using OLIS fitting routines to obtain rate
constants.
RESULTS AND DISCUSSION
■
Stoichiometry and Product Analysis. The ionic
phosphorus reaction products were identified using NMR
whereas GC analysis was used to observe trimethylphosphate.
For hypophosphite, phosphite, and phenylphosphite, the
oxidations were carried out with a slight excess of these
reductants over ferrate, the 31P NMR spectrum contained peaks
of the parent compound, and its oxidation product as
confirmed with known samples of these compounds.
Trimethylphosphite was oxidized with ferrate, and the solution
was extracted with methylene chloride. Only one product peak
was observed with a retention time identical to that of an
authentic trimethylphosphate sample. The final iron product
was identified by standard qualitative analytical techniques
described elsewhere, and in each case, iron(III) was observed.20
With hypophosphite and trimethylphosphite the reaction
stoichiometries were determined by direct titration with ferrate,
and for these a (1.6 0.2)/1, (1.5 0.3)/1 ratio of reductant/
ferrate was observed, respectively. The pH varied between 8
and 9 for the stoichiometry determinations. Redox for these
two reactions was rapid, so a clear end point (the purple
ferrate(VI) ion remained) was observed. Unfortunately, a
precise direct titration with phosphite or phenylphosphite was
not practical as the reactions were too slow and the ferrate
decomposed significantly during the titration. However, rough
estimates of the stoichiometry support a 1.5:1 ratio. On the
basis of these results, a 3:2 reductant to ferrate ratio is proposed
for all the reactions studied, which is congruent with the
reaction products observed. The following reactions can be
written:
Product Analysis and Tracer Studies. The final iron product
was determined using colormetric tests described previously.16 The
final phosphorus oxidation products were identified by 31P NMR using
an external standard (conc H3PO4). Spent reaction mixtures (with no
phosphate buffer) were drawn through a 0.22 μm nylon filter to
remove precipitated ferric hydroxides. The filtrate was passed through
a 1 cm × 5 cm CM25-Sephadex column to remove any remaining
dissolved iron(III). This was followed by elution with one column
volume of distilled water. The eluted samples were rotary evaporated
to dryness, redissolved in a minimum amount of water, and placed in
an NMR tube. The chemical shifts were determined and compared
−
3H2PO2 + 2FeO 2− + 7H+
4
→ 2Fe3+ + 3HPO3 + 5H2O
2−
(1)
6627
dx.doi.org/10.1021/ic3001812 | Inorg. Chem. 2012, 51, 6626−6632

Inorganic Chemistry
Article
2−
3HPO3 + 2FeO 2− + 7H+
In order to check for the production of radicals, these
reactions were carried out in the presence of acrylonitrile which
is well-known to polymerize when radicals are produced. For
each of these systems, no polymerization was observed which
suggests that radicals are not formed. We have used this test for
systems described in ref 16. The lack of observation of radicals
suggests that a 2-electron redox reaction occurs for these
reactions and is supported by our oxygen transfer experiments
and our spectral observation of an intermediate in these
oxidations.
A general mechanism consistent with this rate law can be
written as follows. In this mechanism we have assumed that the
process occurs via a 2-electron oxidation process, and on the
basis of our work with sulfite, this occurs via oxygen transfer.
4
→ 2Fe3+ + 3PO 3− + 5H2O
(2)
(3)
(4)
4
3(CH3O)3P + 2FeO 2− + 10H+
→ 2Fe3+ + 3(CH3O)3PO + 5H2O
4
2−
3PhPO2 + 2FeO 2− + 10H+
4
→ 2Fe3+ + 3PhPO3 + 5H2O
2−
Reaction Kinetics. Absorbance versus time plots for the
disappearance of ferrate (505 nm) fit nicely to a single
exponential decrease to provide the observed pseudo-first-order
rate constants. Plots of these rate constants versus reductant
concentration, P (P denotes phosphite, hypophosphite, phenyl-
phosphite, or trimethylphosphite where appropriate), were
linear, thereby establishing first order terms with respect to P
and ferrate, see Figure 1 inset. The slopes of these lines (the
Ka
HFeO − ⇌ H+ + FeO 2−
(6)
(7)
4
4
kH
HFeO − + P → Fe(IV) + PO
4
ko
FeO 2− + P → Fe(IV) + PO
(8)
(9)
4
Fe(IV) + PO → Fe(II) + PO
fast
Fe(II) + Fe(VI) → Fe(V) + Fe(III)
fast
(10)
Fe(V) + P → Fe(III) + PO
fast
(11)
On the basis of this mechanism, k1 from eq 5 equals kHKa−1
where Ka is the acid dissociation constant for the ferrate ion.
−
Using Sharma’s value for the HFeO4 acid dissociation
constant, 5.01 × 10−8 at 25.0 °C,22 a value for kH was
calculated for each substrate, see Table 1. Details of this are
given in Supporting Information.
Another scheme can be written to account for the production
of Fe(III) in the fast steps, eqs 9−11, that involves the rapid
disproportionation of Fe(IV) followed by a rapid Fe(V)
oxidation.
Figure 1. Hydrogen ion dependence of reaction rates. Inset:
Representative substrate dependence on reductant concentrations.
For both plots: trimethylphosphate, left y-axis; other reductants, right
y-axis. Conditions: I = 1.0 M (NaClO4), sodium phosphate buffer =
0.05 M, T = 25.0 °C.
2Fe(IV) → Fe(III) + Fe(V)
(12)
Fe(V) + P → Fe(III) + PO
(13)
This alternative gives the same stoichiometry for the overall
reaction, but the reaction rates for the Fe(IV) and Fe(V)
oxidations of these substrates are unknown. Previous work by
Bielski et al. on other substrates has shown that the +4 and +5
oxidation states react at least 2−3 orders of magnitude faster
than does Fe(VI), so there is no kinetic basis to eliminate this
pathway.23
To differentiate between these two possibilities requires the
estimation and comparison of rate constants for reactions 10
and 12. The following gives a brief analysis of these two
scenarios. Although the reaction rate constant for eq 10, the
Fe(II)−Fe(VI) reaction, has not been measured using
hexaquairon(II) or a corresponding “aqua−hydroxo” complex,
the aquapentacyanoferrate(II)/ferrate(VI) rate constant has
second order rate constant) were determined between pH 8
and 11 and are shown in Figure 1 as plots versus the hydrogen
ion concentration. The observed linear plot with a positive y-
intercept is best described by the following rate law, eq 5, where
ko is the intercept and k1 is the slope of the line.
−d([Fe(VI)]/dt = (ko + k1[H+])[FeO 2−][P]
(5)
4
The pKas for hypophosphorous acid, phosphorous acid, and
phenylphosphorus acid are low (1.2, 1.43, 2.1, respectively).21
Since the experiments were conducted above pH 8, protonation
is assumed to involve the ferrate ion, pKa 7.322 at 25 °C. Since
trimethylphosphite possesses no acidic hydrogens, protonation
must occur at ferrate(VI).
a
Table 1. Proton Independent and Proton Dependent Rate Constants for Ferrate(VI)−Substrate Reactions
−
2−
H2PO2
0.93 0.11
13
HPO3
(CH3O)3P
PhPO2−
ko (M−1 s−1
)
0.021 0.005
1.9 0.1
4900 300
(2.2 0.2) × 105
0.036 0.003
1.2 0.6
b
kH (M−1 s−1
)
1
a
b
Conditions: T = 25 °C, 0.05 M sodium phosphate buffer, 1.0 M (NaClO4) ionic strength. Calculated from Kaka using Ka values in text and
Supporting Information.
6628
dx.doi.org/10.1021/ic3001812 | Inorg. Chem. 2012, 51, 6626−6632

Inorganic Chemistry
Article
a
Table 2. Thermodynamic Parameters for the Ferrate(VI) Oxidation
−
2−
−
H2PO2
HPO3
(CH3O)3P
1
PhPO2
kH
45
−73 10
kO
32
−140 30
kH
61
−35 11
kO
100 10
−77
kH
18
kO
22
−99 15
kH
50
−130 30
kO
ΔH⧧ (kJ/mol)
ΔS⧧ (J/mol K)
1
2
5
3
1
91
2
5
9
−170 20
−98
a
Conditions: I = 1.0 M NaClO4, 0.05 M sodium phosphate.
been reported equal to 108 M−1 s−1.24 This would suggest that
reaction 10 would be very fast and not rate determining.
To probe eq 12, if we assume 103 fold higher rate constant
−
for Fe(IV) compared to Fe(VI) for the oxidation of H2PO2 ,
then in the reductant concentration range in this study, kcalc in 1
M OH− could vary between 0.1 and 100 s−1. Bielski has
reported the disproportionation rate constant for Fe(IV) to
equal 2 s−1 in 1 M OH−,25 which is competitive with the
hypophosphite/Fe(IV) reaction under these conditions. Of
course, whether these calculated rates present a true picture
over the pH range used in this study is uncertain, and further
kinetic experiments are required to directly measure these rates.
We have initiated studies of iron(V) reactions with phosphite
and hypophosphite using pulsed radiolysis to generate this
oxidation state. Preliminary results indicate that both of these
reactions are slower than the disproportionation of iron(V)
with estimates of the bimolecular rate constants to be less than
102 M−1 s−1 at pH 9.26
Values for both kO and kH were determined at three
temperatures (10, 25, and 35 °C), and standard Eyring plots
were linear. Using these data, the activation parameters were
calculated, Table 2. The temperature dependence of the
monohydrogenferrate(VI) pKa was used in these calculations
using Sharma’s data.22 As can be seen, a significant activation
enthalpy is observed, and more importantly, a very negative
activation entropy was calculated for the reaction of both
protonated and deprotonated ferrate with each substrate. These
large negative entropy values imply that significant organization
occurs during each reaction, which supports the formation of an
intermediate between ferrate and the phosphorus centers
during the redox process. Although an oxygen bridge between
the iron and phosphorus center is most likely, oxygen labeling
studies were carried out to confirm this.
Rapid Scanning and O-18 Labeling Experiments. In
conjunction with other studies in this laboratory, it was noted
that reaction mixtures of ferrate(VI) and oxyions in 1.0 M
NaOH produced spectra where the absorption maxima were
shifted from ferrate(VI) in 1.0 M NaOH. When solutions of
ferrate(VI) (λmax = 505 nm, see Supporting Information for
spectrum) and phosphite were mixed in 1.0 M NaOH, a new
peak at 560 nm formed rapidly, see Figure 2, and then slowly
disappeared over a few minutes. The rate of appearance of this
new peak was first order with respect to ferrate and excess
phosphite, see Figure 2 inset, and plots of kobs versus
[phosphite] had a significant y-intercept. This was interpreted
as an equilibrium where the slope/intercept (0.86 M−1 s−1/0.29
s−1) gives Keq = 3.0 M−1. The disappearance of this peak fits a
first order absorbance decrease (at 560 nm) where kdiss was
0.001 s−1 and was independent of the phosphite concentration.
If we assume Keq*k determined at 1.0 M NaOH for the
phosphite−ferrate(VI) reaction should equal ko determined
between pHs 8 and 10, then this product equals 0.003 M−1 s−1.
The observed value for ko equals 0.021 M−1 s−1. Considering
the vastly different conditions by which these two numbers
were obtained, a factor of 7 difference seems acceptable
Figure 2. Spectrum of ferrate(VI)−phosphite complex. Conditions:
1.0 M NaOH, 25 mM phosphite, 0.5 mM ferrate(VI), T = 25 °C.
agreement and supports our proposed mechanism. A
reasonable explanation for why we observe this in 1.0 M
NaOH and not around pH 9 is that iron(VI) is exists
2‑
completely in its deprotonated form, FeO4 , and in this form it
is much slower (about 100×) to oxidize the bound substrate.
It is well-known that chromate(VI) reacts with similar
oxyanion substrates to form chromium esters and the
equilibrium constant for the ferrate/phosphite system is
comparable with similar chromate systems.27 Since oxygen
transfer has been shown to occur during the oxidation of sulfite
to sulfate by ferrate and a similar spectral shift is observed when
solutions of ferrate(VI) and sulfite are mixed in 1.0 M NaOH, it
is reasonable to assume that these spectra are of a ferrate(VI)−
substrate ester, i.e., [O3Fe−O−PO3H2]4‑. The two reactions
observed for phosphite then are ester formation followed by an
intramolecular 2-electron transfer that involves an oxygen
transfer.
Bielski has reported the spectrum of the iron(IV)−
pyrophosphate complex in alkaline media.28 His work shows
a peak at 420 nm, which differs little from the “aqueous”
alkaline iron(IV) spectrum. Iron(V) has two absorptions
(approximately 490 and 390 nm) which are the same for the
aqueous and carbonate species. Both of these systems have
spectral characteristics that are very different from our
observation of a peak at 560 nm for the iron(VI)−phosphite
reaction intermediate which is a 55 nm shift from aqueous
iron(VI). The significant differences from our absorption
maximum for the iron(VI)−phosphite system and those of
alkaline “aqueous” iron(IV) and iron(V) along with the spectral
data for the iron(IV)−pyrophosphate species support, but
admittedly do not eliminate, our assumption that we are
observing an iron(VI)−phosphite intermediate. Unfortunately,
we cannot obtain a pure solid for this intermediate to better
probe its nature. Its subsequent decay in solution also hinders
6629
dx.doi.org/10.1021/ic3001812 | Inorg. Chem. 2012, 51, 6626−6632

Inorganic Chemistry
Article
formation of a frozen sample where the only species present is
the intermediate. In addition, work by Sharma on the question
of one versus two electron transfer reactions for ferrate(VI)
shows that for sulfite, selenite, and arsenite a two electron
pathway is favored.29 Since we have seen similar shifts in the
UV−vis spectrum of ferrate(VI) with these reductants in 1 M
NaOH, we favor the formation of bridged iron(VI)−
phosphorus intermediate.
Assuming that an iron(VI)−phosphite dimer is formed and
subsequently undergoes intramolecular electron transfer, it
should should show transfer of an oxygen from the ferrate(VI)
ion to the final phosphate product. Oxygen labeling experi-
ments were carried out using 18O labeled potassium ferrate(VI).
It is clear from the 18O induced shift of the 31P NMR signal
shown in Figure 3 that oxygen is transferred from ferrate to the
Table 3. It is clear that a P−H bond breaks to reach the
transition state.
a
Table 3. Kinetic Isotope Effects
kH/kD in D2O with kH/kD in H2O with kH/kD in D2O with
P−H
P−D
P−D
−
H2PO2
1.0
1.1
9.6
8.3
10
2−
HPO3
P(OCH3)3
8.5
b
1.0
a
Conditions: T = 25 °C, 0.05 M sodium phosphate buffer (pH or pD
b
= 9), 1.0 M (NaClO4) ionic strength. While the substrate contains no
P−H bond, the reaction was run to give a baseline for D2O effects on
reaction rates.
Proposed Ferrate(VI) Reaction Intermediates. On the
basis of the above observations, a picture of the intermediate in
the four reactions may be formulated. For each system an
oxygen bridged species arises via ester formation with the
ferrate(VI) center. While direct coordination to the iron(VI)
center by a phosphorus oxygen is possible, it is not favored in
these systems since the oxygen from the iron(VI) was observed
to be transferred to the phosphorus center. This suggests
formation of an intermediate with an Fe−O−P bridge using an
iron based oxygen. Such bridging should be quite facile for
trimethylphosphite or phenylphosphite to form a 4-coordinate
phosphorus center, and both of these substrates are observed to
“rapidly” reduce ferrate(VI). The two steps are shown below.
HFeO − + P(OMe)3 ⇄ HO3Fe − O − P(OMe)3−
(14)
4
As a control for later studies, these reactions were also carried
out in 99 atom % D2O and showed no solvent isotope effect, i.e.,
k
H O/kD O = 1.0 0.1.
2 2
Figure 3. 31P NMR of the phosphite formed from the oxidation of
hypophosphite with 18O labeled ferrate(VI). The resonance at 4.141
ppm is due to 18O shift of the 31P signal at 4.168 for phosphite without
18O labeling.
In contrast to the trimethylphosphite reduction, the ferrate
phosphorus center during oxidation of phosphite. Since
phosphate and phosphite are slow to exchange their oxygens
with solvent water and the redox reaction occurs more rapidly
than the exchange rate of the ferrate oxygens with solvent, the
formation of an iron−oxygen−phosphorus bridge is invoked to
account for this transfer. The same oxygen transfer was
observed for all of the phosphorus complexes in this study
using 31P NMR.
Kinetic Isotope Effects. Since many investigations into
phosphite and hypophosphite oxidation reactions have involved
hydride abstraction from the phosphorus center, the effects of
deuteration of the substrate on the reaction rates were
examined. As a control, the reactions were run in D2O using
normal phosphite, hypophosphite, and trimethylphosphite,
which either do not have hydridic hydrogens or exchange
their hydrogens very slowly with D2O. In each case there was
no effect of using D2O over H2O as the solvent; i.e., no solvent
based isotope effect was observed. When the hydridic
hydrogens on the phosphorus centers were deuterated,
however, substantial kinetic isotope effects were observed; see
oxidations of phosphite, hypophosphite, and phenylphosphite
are much slower. This could be due to two factors: valence shell
expansion of the phosphorus center (to form a 5-coordinate
species) and/or abstraction of a hydridic hydrogen from the
phosphorus. For these centers, bridging via a ferrate(VI)
oxygen accounts for the transfer of oxygen to the final
phosphorus product, so in this sense, they are the same as to
the first two systems described. However, when these reactions
−
are carried out with deuterated substrates, i.e., D2PO2 and
DPO32−, the reactions are substantially slower and showed
kinetic isotope effects, kH/kD, of 10 and 8.5, respectively. This
indicates that the breaking of a P−H bond occurs in the
transition state. This can be accounted for as illustrated with
hypophosphite and presumably also occurs with phosphite.
The pathway in eq 16 is similar to that proposed by Gould et
al. for electron transfer via an intramolecular hydride
abstraction in a cobalt(III)−hypophosphite complex.12
The reaction mechanism proposed is also similar to that
suggested by Haight et al. for the chromate(VI) oxidation of
hypophosphite.10 He proposed that coordination of the
6630
dx.doi.org/10.1021/ic3001812 | Inorg. Chem. 2012, 51, 6626−6632

Inorganic Chemistry
Article
REFERENCES
■
(1) Sawyer, D. T.; Hage, J. P.; Sobkowiak, A. J. Am. Chem. Soc. 1995,
117, 106. Sawyer, D. T.; Kang, C.; Llobet, A.; Redman, C. J. Am. Chem.
Soc. 1993, 115, 5817. Sawyer, D. T.; Tung, H.; Kang, C. J. Am. Chem.
Soc. 1992, 114, 3445. Barton, D. H. R.; Motherwell, W. S. B. Chem.
Commun. 1983, 41. Barton, D. H. R.; Haey, F.; Ozballik, N.; Young, E.;
Balavoine, G.; Gref, A.; Boivin, J. J. Nouv. Chim. 1989, 13, 177. Que,
L.; Nam, W. Chem. Rev. 1996, 96, 2607. Pierre, J. L.; Fontercave, M.
BioMetals 1999, 12, 195. Solomon, E.; et al. Chem. Rev. 2000, 100, 235.
Rohde, J.; In, J.; Lim, M.; Brennessel, W.; Bukowski, M.; Stubna, A.;
Munck, E.; Nam, W.; Que, L. Science 2003, 299, 1037. Rohde, J.; In, J.;
Lim, M.; Brennessel, W.; Bukowski, M.; Stubna, A.; Munck, E.; Nam,
W.; Que, L. Science 2003, 299, 1037. Pestovsky, O.; Bakac, A. J. Am.
Chem. Soc. 2004, 126, 13757. Pestovsky, O.; Bakac, A. Inorg. Chem.
2006, 45, 814. Pestovsky, O.; Bakac, A. ACS Symp. Ser. 2008, 985
(Ferrates), 167.
(2) Inani, K. M.; Sharma, P. D.; Gupta, U. K. J. Chem. Soc., Dalton
Trans. 1985, 2571. Cooper, J. J. Phys. Chem. 1970, 74, 955. Griffith, R.
O.; McKeown, A.; Taylor, R. P. Trans. Faraday Soc. 1934, 30, 530.
Jenkins, W. A.; Yost, D. M. J. Inorg. Nucl. Chem. 1959, 11, 297.
(3) Bhatt, L.; Sharma, P. D.; Gupta, Y. K. J. Chem. Soc., Dalton Trans.
1984, 2653.
(4) Schafer, K.; Asmus, K.-D. J. Phys. Chem. 1981, 84, 852.
(5) Indrayan, A. K.; Mishra, S. K.; Gupta, Y. K. Indian J. Chem. 1983,
21A, 1066.
(6) Indrayan, A. K.; Mishra, S. K.; Gupta, Y. K. Inorg. Chem. 1981, 20,
1924.
(7) Indrayan, A. K.; Mishra, S. K.; Gupta, Y. K. Inorg. Chem. 1981, 20,
450.
(8) Mehrotra, R. N. Can. J. Chem. 1985, 63, 663.
(9) Sengupta, K. K.; Basu., B.; Sengupta, S.; Nandi., S. Polyhedron
1983, 2, 983.
(10) Haight, G. P.; Rose, M.; Preer, J. J. Am. Chem. Soc. 1968, 90,
4809.
(11) Frennesson, S.; Beattie, J. K.; Haight, G. P. Acta Chem. Scand.
hypophosphite followed by deprotonation of the P center
resulted in an internal redox step. This was believed to be
general base catalyzed. In order to check for a general base
mechanism for ferrate−hypophosphite, the reaction was carried
out using a sodium borate buffer instead of sodium phosphpate
at pH 9. No difference in rate was observed between these two
buffers. Therefore, in the above mechanism the ferrate oxygen
is involved in the hydrogen abstraction since there is no
observed general base catalysis, as might be expected if simple
deprotonation of the phosphorus center were involved.
To further probe this mechanism, labilization of a P−H bond
with respect to exchange to solvent water was attempted by
coordination with a non-redox-active center such as molyb-
date.30 Solutions of phosphite and molybdate were mixed in
1
D2O and adjusted to pD ∼ 9 and H NMR spectra collected
over time. Even after exposure times of several hours there were
no decreases of the P−H signals in the NMR spectra. This
indicates, though not unambiguously, that coordination of a
phosphorus hydride moiety to high oxidation state metals does
not labilize the P−H bond. This could mean that the
chromate(VI) oxidation mechanisms reported by Haight et al.
require modification to reflect an intramolecular abstraction/
breaking of the P−H bond.
1969, 23, 3277.
(12) Gould, E.; Ghosh, S. K.; Bose, R. N.; Laai, K. Inorg. Chem. 1986,
25, 4737.
CONCLUSIONS
■
For the first time in iron(VI) oxidation chemistry, a quasistable
intermediate that involves an oxygen bridge ester has been
observed. The subsequent electron transfer process involves
transfer of an oxygen from the iron(VI) center to the final
phosphorus product. In addition, for those phosphorus
substrates containing hydridic hydrogens, the breaking of the
P−H bond occurs in the transition state and is presumably via
abstraction with an iron(VI) oxygen.
(13) Zahonyi-Budo, E.; Simandi, L. I. Inorg. Chim. Acta 1991, 181,
149. Zahonyi-Budo, E.; Simandi, L. I. Inorg. Chim. Acta 1993, 205, 207.
(14) Mehrotra, R. N. J. Chem. Soc., Dalton Trans. 1984, 1531.
(15) Shan, J.; Yang, Y.; Shan, J. Eur. J. Chem. 2011, 2, 342. Zboril, R.;
Andrle, M.; Oplustil, F.; Machala, L.; Tucek, J.; Filip, J.; Marusak, Z.;
Sharma, V. K. J. Hazard. Mater. 2012, 211. Villanueva-Rodriguez, M.;
Bandala, E. R.; Quiroz, M. A.; Guzman-Mar, J. L.; Perata-Hernandez, J.
M.; Hernandez-Ramirez, A. Sustainable Environ. Res. 2011, 21, 337.
Jiang, J.; Zhou, Z.; Ole, P. Sep. Purif. Technol. 2012, 88, 95.
Zimmermann, S. G.; Schmukat, A.; Schulz, M.; Benner, J.; von
Gunten, U.; Ternes, T. A. Environ. Sci. Technol. 2012, 46, 876. Zhang,
P.; Zhang, G.; Dong, J.; Fan, M.; Zeng, G. Sep. Purif. Technol. 2012, 84,
46. Shan, J.; Zhang, J. E-J. Chem. 2011, 8, 903. Sharma, V. K.; Luther,
G. W.; Millero, F. J. Chemosphere 2011, 82, 1083. Shan, J. H.; Zang, J.
Y.; Shen, H. X.; Xiao, Q. W. Asian J. Chem. 2011, 23, 701. Sharma, V.
K. Chemosphere 2008, 73, 1379. Read, J. F.; Bewick, S. A.; Oikle, S. E.;
Schaubel, C.; Taylor, S. N.; Theriault, A.; Watson, K. J. Inorg. React.
Mech. 2005, 5, 265. Read, J. F.; Griffiths, B. H. B.; Lynch, L. L.;
MacCormick, K. J.; Tam, N. C. Inorg. React. Mech. 2004, 5, 135. Read,
J. F.; MacCormick, K. J.; McBain, A. M.. Transition Met. Chem. 2004,
29, 149. Read, J. F.; Boucher, K. D.; Mehlman, S. A.; Watson, K. J.
Inorg. Chim. Acta 1998, 267, 159. Ohta, T.; Kamachi, T.; Shiota, Y.;
Yoshizawa, K. J. Org. Chem. 2001, 66, 4122. Read, J. F.; John, J.;
MacPherson, J.; Schaubel, C.; Theriault, A. Inorg. Chim. Acta 2001,
315, 96. Read, J. F.; Bewick, S. A.; Graves, C. R.; MacPherson, J. M.;
Salah, J. C.; Theriault, A.; Wyand, A. E. H. Inorg. Chim. Acta 2000, 303,
244. Read, J. F.; Bewick, S. A. Inorg. React. Mech. 2005, 5, 305. Read, J.
F.; Bewick, S. A.; Donaher, S. C.; Eelman, M. D.; Oakey, J.; Schaubel,
C.; Tam, N. C.; Theriault, A.; Watson, K. J. Inorg. React. Mech. 2005, 5,
281.
ASSOCIATED CONTENT
* Supporting Information
Additional figure, table, and details. This material is available
■
S
free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*E-mail: johnson@nmsu.edu. Phone: (575) 646-3627. Fax:
(575) 646-2394.
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors wish to acknowledge the help of Dr. Dennis
Johnson in conducting the NMR experiments and Dr. David
Richens for his comments. We also wish to thank the referees
for very useful comments. This work was funded in part by an
ACS-PRF grant (#35481-AC3).
6631
dx.doi.org/10.1021/ic3001812 | Inorg. Chem. 2012, 51, 6626−6632

Inorganic Chemistry
Article
(16) Johnson, M. D.; Hornstein, B. J. Inorg. Chim. Acta 1994, 225,
145. Johnson, M. D.; Hornstein, B. J. J. Chem. Soc., Chem. Commun.
1996, 965. Johnson, M. D.; Hornstein, B. J.; Wischnewsky, J. ACS
Symp. Ser. 2008, 985 (Ferrates), 177−188.
(17) Lee, D. G.; Gai, H. Can. J. Chem. 1993, 71, 1394. Johnson, M.
D.; Read, J. F. Inorg. Chem. 1996, 35, 6795.
(18) Thompson, G. W.; Ockerman, L. T.; Schreyer, J. M. J. Am.
Chem. Soc. 1951, 73, 1379.
(19) (a) Goff, H.; Murmann, R. K. J. Am. Chem. Soc. 1971, 93, 6586.
(b) Hendry, P.; Sargeson, A. M. J. Am. Chem. Soc. 1989, 111, 2521.
(20) Johnson, M. D.; Bernard, J. Inorg. Chem. 1992, 31, 5140.
(21) Albert, A.; Serjeant, E. P. The Determination of Ionization
Constants, 3rd ed.; Chapman and Hall: New York, 1984; pp 148, 163.
(22) Sharma, V. K.; Burnett, C. R.; Millero, F. Chem. Phys. 2001, 3,
2059.
(23) Bielski, B. J. H.; Thomas, M. J. Inorg. Chem. 1989, 28, 3947.
(24) Johnson, M. D.; Sharma, K. D. Inorg. Chim. Acta 1999, 293, 229.
(25) Bielski, B. Methods Enzymol. 1990, 186, 108.
(26) Johnson, M. D.; Cabelli, D. Unpublished results.
(27) Keq values for chromate with dihydrogen phosphate, sulfate, and
chloride are 2.9, 4.1, and 17.1, respectively. Haight, G. P. Inorg. Chem.
1964, 3, 1777.
(28) Melton, J. D.; Bielski, B. H. J. Radiat. Phys. Chem. 1990, 36, 725.
(29) Sharma, V. Environ. Sci. Technol. 2010, 44, 5148. Johnson, M.
D.; Lorenz, B. B. Unpublished results for the ferrate(VI)/arsenite,
selenite, antimonite dimers.
(30) Roper, G. C.; Haas, T. E.; Gillman, H. D. Inorg. Chem. 1970, 9,
1049. Hargis, L. G. Anal. Chem. 1969, 41, 597.
6632
dx.doi.org/10.1021/ic3001812 | Inorg. Chem. 2012, 51, 6626−6632