On the tautomerism of secondary phosphane oxides

Article abstract of DOI:10.1002/ejoc.201000037

The tautomeric behaviour of five secondary phosphane oxides (SPOs) with different electronic properties has been investigated by NMR and IR spectroscopy, density functional theory calculations and X-ray structural analysis. Proof is given that only with s

Full text of DOI:10.1002/ejoc.201000037

FULL PAPER
DOI: 10.1002/ejoc.201000037
On the Tautomerism of Secondary Phosphane Oxides
Andrea Christiansen,[a,b] Chuanzhao Li,[c] Marc Garland,[c] Detlef Selent,[a] Ralf Ludwig,[d]
Anke Spannenberg,[a] Wolfgang Baumann,[a] Robert Franke,[b] and Armin Börner*[a,d]
Dedicated to Uwe Rosenthal on the occasion of his 60th birthday
Keywords: Phosphorus / Tautomerism / NMR spectroscopy / IR spectroscopy
The tautomeric behaviour of five secondary phosphane ox-
ides (SPOs) with different electronic properties has been in-
vestigated by NMR and IR spectroscopy, density functional
theory calculations and X-ray structural analysis. Proof is
given that only with strong electron-withdrawing groups on
the phosphorus atom can the relevant trivalent phosphinous
acid be observed in the equilibrium, whereas in the case of
electron-rich phosphane oxides the pentavalent species
dominates. For a detailed analysis of the IR spectra, the de-
convolution programme BTEM has been successfully ap-
plied.
Introduction
(HA)SPOs A are weak acids.[3] Therefore, they can tauto-
merize in solution to phosphinous acids B.[4] The latter are
trivalent, and thus they are able to bind to a metal centre
to give metal complexes C. As a result, we recently sug-
gested the term preligands.[5] Indeed, several examples have
been reported in the literature in which (HA)SPOs have
been successfully used as ancillary ligands in homogeneous
metal-catalyzed reactions.[6,7]
In spite of these applications in catalysis little is known
about the chemical and physical properties of the desmo-
trope depicted in Scheme 1 even though this information is
crucial for an assessment of their coordination behaviour
and ultimately their catalytic properties. In general, (HA)-
SPOs are reputed to be stable towards oxygen, which has
been explained by the dominance of the pentavalent tauto-
mer A in solution.[8] It is assumed that the equilibrium is
only shifted towards the trivalent tautomer B in the pres-
ence of a metal centre.
As a result of their versatility and modular synthesis, tri-
valent phosphorus compounds such as phosphanes, phos-
phinites, phosphonites and phosphites represent the most
commonly used classes of ancillary ligands in homogeneous
catalysis.[1] Owing to the presence of the free electron pair
they are able to coordinate to a metal centre. However,
pentavalent phosphorus compounds can also be considered
as potential ligands, provided there is an equilibrium with
trivalent phosphorus species as constituents.
In this respect, secondary phosphane oxides, abbreviated
as SPOs (R = alkyl, aryl), or heteroatom-substituted sec-
ondary phosphane oxides, abbreviated as HASPOs (at least
one R = NRЈЈ2 or ORЈЈ), are of particular interest (A,
Scheme 1).[2]
Already in the 1960s, secondary phosphonates, phos-
phinates and phosphane oxides were being examined in
deuterium exchange[9] and oxidation[9c,10] reactions.[3a] The
results of these investigations suggested that the pentavalent
species is dominant in most instances. Preliminary IR spec-
troscopic measurements confirmed the formal P=O nature
of alkyl- and aryl-substituted SPOs.[11] Pietrusiewicz and
Duddeck and their co-workers suggested that the tauto-
meric equilibrium forms rapidly because the migration of
an (acidic) proton is involved.[12] Recent DFT calculations
by Ustynyuk and Babin revealed that an intramolecular
proton transfer is not favoured due to high activation barri-
ers.[13] In the absence of a proton carrier, as in non-polar
media, the rearrangement proceeds as a synchronous trans-
Scheme 1. Tautomeric equilibrium of pentavalent and trivalent
phosphorus compounds and the formation of metal complexes.
[a] Leibniz-Institut für Katalyse an der Universität Rostock e.V.,
A.-Einstein-Str. 29a, 18059 Rostock, Germany
Fax: +49-381-1281-51202
E-mail: armin.boerner@catalysis.de
[b] Evonik Oxeno GmbH,
Paul-Baumann-Straße 1, 45772 Marl, Germany
[c] Institute of Chemical and Engineering Sciences,
1 Pesek Road, Jurong Island, 627833 Singapore
[d] Institut für Chemie der Universität Rostock,
A.-Einstein-Str. 3a, 18059 Rostock, Germany
Eur. J. Org. Chem. 2010, 2733–2741
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2733

A. Börner et al.
FULL PAPER
fer of two protons in dimeric associates. However, system-
atic studies on this tautomerism have not been available un-
til now.
For our study we chose five related SPOs 1–5 possessing
P substituents with electron-donating or -withdrawing
properties. Recently, the pertinent equilibria for (OH)2P-
(O)H[14] and (tBu)PhP(O)H[15] were studied theoretically.
The seminal work was performed by Hoge et al. who inves-
tigated electronically poor SPOs and found a solvent-de-
pendent shift of the tautomeric equilibria.[16] In addition,
for compound 5 the zero-point-corrected energy difference
∆EZPE between both tautomers was calculated.
Figure 2. Plot of the P–H coupling constants of the P=O tautomers
1–5 in the 31P NMR spectrum: experimentally determined (in
CD2Cl2) (top) and calculated at the B3LYP/6-31+G* level of
theory (bottom).
To widen the scope of these investigations, herein we
wish to present our studies on the tautomeric equilibria of
SPOs 1–5 based on NMR spectroscopic characterization,
investigations on the influence of solvent and temperature,
DFT calculations, X-ray structural analysis and IR spectro-
scopic characterization.
this trend. In general, the electronic differences between the
phenyl-substituted compounds are small. This is also re-
vealed in Figure 3.
Results and Discussion
NMR Spectroscopic Investigations
Correlation of the 31P NMR Shifts and 31P–1H Coupling
Constants
Investigations of the SPOs 1–5 in CD2Cl2 by 31P NMR
spectroscopy revealed that with the exception of the perflu-
orinated compound 5 at ambient temperature all other
SPOs exist exclusively in the pentavalent tautomeric form A
(Scheme 1). The experimentally determined chemical shifts
(δ31P) and 31P–1H couplings (JP–H) are depicted in Fig-
ures 1 and 2. For comparison, the calculated values are also
given.[17]
Figure 3. Correlation of the P–H coupling constants versus chemi-
cal shifts in the 31P NMR spectrum of the P=O tautomers of 1–5:
experimentally determined (in CD2Cl2) (top) and calculated at the
B3LYP/6-31+G* level of theory (bottom).
Influence of the Solvent
As the tautomeric equilibrium of electron-poor SPO 5 in
dichloromethane displayed besides the phosphane oxide A
also the phosphinous acid B (ratio 59:41), this SPO was
first chosen to investigate the effect of solvent on the shift
of the tautomeric equilibrium. Although the pentavalent
tautomer A predominates in the equilibrium in [D8]toluene
and CD2Cl2 with a ratio of 59:41, in more polar solvents
like [D8]THF and CD3OD the equilibrium is shifted com-
pletely to the phosphinous acid B (Table 1).
Clearly the trivalent tautomer 5B is “stabilized” by polar
solvents. Surprisingly, this remarkable solvent dependency
was not observed for SPOs 1–4. The relevant trivalent tau-
tomers B were not detected in toluene, THF, dichlorometh-
ane or trifluoroethanol. In each case only a single signal
characterizing the pentavalent tautomer was found. In de-
Figure 1. Plot of the 31P NMR chemical shifts (δ) of the P=O tau-
tomers of 1–5: experimentally determined in CD2Cl2 (top) and cal-
culated at the B3LYP/6-31+G* (middle) and B3LYP/6-311+G*
levels of theory (bottom).
As can clearly be seen, electron-withdrawing substituents, uteriated methanol as well as in trifluoroethanol, exchange
which decrease the Lewis basicity of the phosphorus atom, of the P-bonded proton by deuterium took place. It is as-
cause a shift to higher field and an enhancement in the P– sumed that this reaction is reversible.[9f] Esterification with
H couplings. The quantum chemical calculations confirm the alcoholic solvent did not take place with SPOs 1–4. In
2734
www.eurjoc.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 2733–2741

On the Tautomerism of Secondary Phosphane Oxides
Table 1. Tautomeric equilibrium of bis(tetrafluoropyridyl)phos- Influence of Temperature
phane oxide (5) at room temperature according to a 31P NMR
To evaluate the temperature dependency of the equilib-
analysis.
ria, dynamic NMR measurements were performed. SPOs
1–3 and 5 were investigated in a temperature range of –80
to +55 °C. Although at room temperature for compounds
1–3 only the corresponding phosphane oxide could be de-
tected (see above), we assume an infinitesimally small
amount of the phosphinous acid to be present and to con-
tribute to the P–H coupling constant. Thus, the measured
coupling constants represent an average value derived from
Solvent
Ratio A/B
[D8]Toluene
CD2Cl2
[D8]THF
CD3OD
[D8]Toluene/CD3OD (3:1)
[D8]Toluene/CF3CD2OD (3:1)
CF3CD2OD
59:41
59:41
0:100
0:100[a]
0:100
81:19
87:13
[a] Measurement after 60 min; P–OD tautomer B corresponds to Equation (1).
only 8.8% of the overall signal intensity due to degradation reac-
tions with CD3OD.
Javerage = x2J(P–OH) + (1–x)1J(P–H)
(1)
This equation is only valuable under the precondition
that a fast exchange between the P–H and P–OH species
takes place. According to Equation (1), a change in the cou-
pling constant can be attributed to a shift in the tautomeric
equilibrium. With THF as solvent for SPOs 1–3 only a
small decrease in the P–H coupling was observed (in the
range of 6–9 Hz) by increasing the temperature, which indi-
cates a slight shift in the equilibria in favour of the trivalent
tautomer. Unfortunately, the latter could not be detected in
the spectrum, apparently because of the poor sensitivity of
the analytical method applied.
contrast, in deuteriated methanol or in mixtures of meth-
anol/toluene, SPO 5 was converted rapidly. The correspond-
ing methyl ester and several decomposition products were
observed, characterized by signals in the 31P NMR spec-
trum between δ = +10 and –10 ppm. This observation indi-
cates that the solvent does not only influence the state of
the tautomeric equilibrium, but may also contribute to the
degradation of SPOs.
Clearly the polarity of the solvent is not the only reason
for the shift in the equilibrium observed. Thus, although
the polarity of THF is classified as being between toluene
and dichloromethane,[18] the ratio of tautomers A/B does
not reflect this order. Moreover, in trifluoroethanol, which
displays a much higher polarity than methanol, surpris-
ingly, the pentavalent tautomer 5A dominates the equilib-
rium. This effect can be rationalized by taking the hydro-
gen-bonding acceptor properties of the solvent into con-
sideration. Whereas methanol and THF exhibit strong hy-
drogen-bonding acceptor properties, trifluoroethanol, tolu-
ene and CH2Cl2 are rather poor hydrogen-bonding ac-
ceptors.[18] Hence, more than the polarity of solvents, their
hydrogen-bonding acceptor properties affect the state of the
tautomeric equilibria of secondary phosphane oxides.
When diphenylphosphane oxide (3) was dissolved in a
mixture of CDCl3 and [D6]acetone we observed, besides the
pentavalent P=O tautomer A (δ = 22.8 ppm), the formation
of a second species characterized by a resonance at δ =
32.4 ppm. Analysis of the 13C NMR spectrum, the reaction
with non-deuteriated acetone and a comparison with litera-
ture data[19] revealed that the trivalent tautomer 3B was not
formed,[20] but P-(2-hydroxyisopropyl)diphenylphosphane
oxide (6; Scheme 2). This compound derives from the ad-
dition of 3B to acetone, similar to that recently described
by Hoge et al. for the reaction of (C6F5)2P(O)H.[16]
The question arises as to whether there is a temperature
dependency of the P–H coupling constants as a matter of
principle. Thus,
a
sample of bis(p-tolyl)phosphane,
(CH3C6H4)2PH, was cooled from room temperature to
–80 °C, and an increase in JP–H of 2 Hz was observed. This
means that the previously measured data with ∆JP–H = 6–
9 Hz for SPOs 1–3 have to be corrected on average by 2–
4 Hz. The resulting differences in the coupling constants are
only marginal so that an influence of the temperature on
the equilibria can be neglected.[21]
When the equilibrium of electron-poor SPO 5 was inves-
tigated in toluene (in which both tautomers coexist) in the
temperature range –80 to +55 °C, a shift in the tautomeric
equilibrium in favour of the trivalent tautomer took place,
as demonstrated by a shift in the signal intensities of both
tautomers by about 6%. Quantum chemical calculations of
the equilibrium constant carried out for –100 and +55 °C
at the B3LYP/6-31+G* level of theory also indicated a
change of 7%, which agrees well with the experimentally
determined value. However, taking the large temperature
interval into consideration this contribution can be ne-
glected.
DFT Calculations and X-ray Structural Analysis
As NMR spectroscopy is not powerful enough to quan-
tify the state of the tautomeric equilibria of ligands 1–4,
DFT calculations were performed by using the Gaussian 03
software package[22] with the B3LYP functional[23] to deter-
mine whether electronic variations at the para position of
phenyl substituents in secondary phosphane oxides influ-
ence the tautomeric equilibria or not. The results are listed
in Table 2. In a first step, the individual structures of the
Scheme 2. Reaction of SPO 3 with [D6]acetone.
Eur. J. Org. Chem. 2010, 2733–2741
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2735

A. Börner et al.
FULL PAPER
tautomers were optimized at the B3LYP/6-31G* level of
Calculations of the structures of the trivalent phosphi-
theory. For the optimized structures, frequency calculations nous acids B of 1–5 gave two minima in each case, which is
were performed to ensure that they are local minima and in accordance with the existence of rotational P–OH iso-
to yield zero-point vibrational energies and Gibbs free ener- mers. Already in 1971, Dobbie and Straughan observed two
gies that include their thermal corrections at 298 K. Sub- OH stretching vibrations for gaseous (CF3)2POH in the IR
sequently, the structures obtained were further optimized at spectrum and suggested the coexistence of rotational iso-
the B3LYP/6-311+G* level of theory. The free enthalpy (G) mers.[25] These observations were recently confirmed by
was finally calculated by using the electronic energies (Etot
)
DFT calculations by Hoge et al.[26]
at the B3LYP/6-311+G* level and by employing zero-point
Thereupon we calculated the potential energy curves for
vibrational energies as well as statistical thermodynamic compounds 1–5 along the C–P–O–H dihedral angle and
contributions obtained at the B3LYP/6-31G* level of obtained two minima in each case. Exemplary results for
theory.
diphenylphosphinous acid (3B) and bis(tetrafluoropyridyl)
phosphinous acid (5B) are depicted in Figures 5 and 6.[27]
The energy difference calculated for Ph2POH (3B) is
14.18 kJmol–1, whereas the difference in the minima calcu-
lated for (C5NF4)2POH (5B) is much smaller
(6.50 kJmol–1).[28]
Table 2. Calculated differences in the free enthalpy of the tautomers
of 1–5 in the relevant equilibria in the gas phase.
Tautomer A
1
2
3
4
5
∆G(A–B) [kJ/mol]
Share of A [%]
–31.17 –16.53 –12.64 –10.88 +16.74
Ͼ99.99 Ͼ99.87 Ͼ99.27 Ͼ98.78 Ͻ0.01
The state of the equilibrium between the pentavalent and
trivalent isomers can be calculated from the free enthalpies
according to ∆G = –RTlnK in which K = [A]/[B] and [B] =
100 – [A].
In agreement with the experimental data, the pentavalent
tautomer A in SPOs 1–4 is much more stable than the tri-
valent form, and therefore it dominates the equilibria of 1–
4. In contrast, for the electron-poor compound 5, the phos-
phinous acid B is stabilized by 16.74 kJmol–1 with respect
to the corresponding oxide form. Hence, the equilibrium is
shifted towards the trivalent P–OH form. The calculated
state of the equilibria of SPOs 1–5 conform to those deter-
mined by NMR spectroscopy in a THF solution.
For SPO 2 the structure of the pentavalent tautomer was
additionally evidenced by X-ray structural analysis. The so-
lid-state structure as well as distances and angles are given
in Figure 4. The calculated data fit well with those obtained
by structural analysis. According to both methods the P–O
distance is characteristic of a formal P=O double bond.
Figure 5. Torsion potential of Ph2POH (3B) along the C–P–O–H
dihedral angle (calculated at the B3LYP/6-31+G* level).
Figure 4. Molecular structure of 2A with selected bond lengths and
Figure 6. Torsion potential of (C5NF4)2POH (5B) along the C–P–
O–H dihedral angle (calculated at the B3LYP/6-31+G* level).
angles and comparison with the calculated data.[24]
2736
www.eurjoc.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 2733–2741

On the Tautomerism of Secondary Phosphane Oxides
IR Spectroscopic Studies
bands, as exemplarily shown for di-tert-butylphosphane ox-
ide (1) in Figure 9. Therefore, the state of the equilibrium
The tautomeric equilibria of the SPOs can also be deter-
mined by IR spectroscopy by taking advantage of the char-
acteristic P–H, O–H and P–O vibrational bands. We started
by calculating the IR spectra using DFT. The calculated
absorption bands of the relevant pentavalent tautomers of
type A are depicted in Figure 7.
cannot be determined just by integrating the ν and ν
˜
˜
PH
OH
bands.
Figure 9. IR spectra of the calculated rotational isomers (top and
middle) as well as of the pentavalent tautomer (bottom) of desmo-
trope 1 (at the B3LYP/6-31+G* level).
Figure 7. Calculated IR spectra of the pentavalent tautomers A of
SPOs 1–5 (at the B3LYP/6-31+G* level).
However, a reliable conclusion can be derived by com-
Clearly, the P–H vibration (νPH in the range of ca. 2300–
˜
parison of the signal area of the ν
bands. To identify
2500 cm–1) is particularly valuable for kinetic studies be-
cause there are no overlapping signals in this region in con-
trast to the P–O area.
˜
PO
these pivotal resonances deconvolution of the experimen-
tally measured IR spectra is required owing to overlapping
signals. Exemplarily, the BTEM algorithm was applied to
deconvolute the spectra of bis(tetrafluoropyridyl)phos-
phane oxide (5) and diphenylphosphane oxide (3).[31]
Figure 8 shows a comparison between the experimentally
obtained and calculated P–H vibrational bands of SPOs 1–
5.[29,30] There is a difference in the values obtained by the
two methods because the calculations were carried out in a
harmonic approximation, whereas in experiments anhar-
monic contributions also play a role. Irrespective of this fea-
ture, a general trend similar to the P–H coupling constants
in the NMR analysis can be observed: strong electron-ac-
Bis(tetrafluoropyridyl)phosphane Oxide (5)
Full-range deconvolutions of (C5NF4)2P–OH (5) in tolu-
ene and [D6]benzene are shown in Figure 10. The BTEM
spectral estimate in toluene (Figure 10a) clearly shows the
O–H vibration, but not the P–O band.[32] In contrast, the
spectral estimate in [D6]benzene (Figure 10b) shows not
only the O–H band, but also the very intense P–O vibration
at 1456 cm–1.
cepting substituents at the phosphorus atom enhance ν
.
˜
PH
As discussed above, the NMR spectroscopically deter-
mined tautomeric equilibrium for perfluorinated SPO 5 in
toluene is 59:41 [P(O)H/P–OH] at room temperature. DFT
calculations predicted the presence of two rotational iso-
mers, which could not be observed in the NMR spectra
even at 200 K. Accordingly, BTEM analysis was reapplied
to the spectroscopic data in a narrower range, specifically
3250–3800 cm–1. Two spectral estimates were obtained that
have band maxima at 3470 (Figure 11a) and 3434 cm–1
(Figure 11b). The difference ν
– ν
= 36 cm–1 is in
OH_2
˜
˜
OH_1
good agreement with the value of 38 cm–1 obtained by DFT
calculations. To the best of our knowledge, the present de-
convolution has enabled the first observation of two simul-
taneous P–OH rotational isomers in the liquid phase by IR
spectroscopy.
Figure 8. Experimentally obtained (bottom; in KBr, compound 5
measured by means of ATR) and calculated (top; calculated at the
B3LYP/6-31+G* level) P–H vibrational bands of compounds 1–5.
The DFT calculations described above reveal that the
BTEM analysis was reapplied to the spectral range 2086–
phosphinous acids B exist as two rotational isomers that 3800 cm–1 to identify the P–H stretching vibration of the
exhibit different interactions between the OH groups and pentavalent tautomer 5A. As shown in Figure 12, one pure-
the other P-substituents. This feature leads to distinct OH component spectrum was recovered in which a P–H vi-
Eur. J. Org. Chem. 2010, 2733–2741
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2737

A. Börner et al.
FULL PAPER
bration is correlated with the O–H vibration at 3470 cm–1.
Owing to the weak intensity of the P–H vibrations, a second
spectral estimate could not be obtained.
Figure 12. A BTEM spectral estimate showing correlated O–H and
P–H regions for the desmotrope of SPO 5.
According to the NMR spectroscopic investigations the
pentavalent tautomer A of 3 dominates the equilibrium. A
full-range deconvolution of 3 by using BTEM is shown in
Figure 13a.
Figure 10. Vibrational spectrum of (C5NF4)2P–OH (5B) after de-
convolution with BTEM in (a) toluene and (b) [D6]benzene by
using the full range of spectroscopic data.
Figure 13. IR spectra of (C6H5)2P(O)H (3) in toluene after decon-
volution with BTEM (target νP–H): (a) whole range and (b) section
showing the O–H and P–H vibrations.
Figure 11. BTEM spectral estimates of the ν
band of the two
˜
OH
rotational isomers of (C5NF4)2P–OH (5B) in toluene.[33]
2738
www.eurjoc.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 2733–2741

On the Tautomerism of Secondary Phosphane Oxides
troscopy. The yields listed below are isolated yields. NMR spectra
were recorded with Bruker AV 300 and AV 400 spectrometers.
Chemical shifts are reported in ppm relative to deuteriated solvents.
TMS was used as the standard in 1H and 13C NMR spectra and
H3PO4 in 31P NMR spectra. Mass spectra were recorded with an
AMD 402 spectrometer and IR spectra with a Nicolet Magna-IR
550 spectrometer. Elemental analyses were performd with a LECO
CHNS-932. Melting points were determined with Stuart SMP 3.
In contrast to the results obtained with SPO 5, the spec-
tral estimate for 3 shows a pronounced P–H vibration at
2312 cm–1. A subsequent analysis with BTEM was per-
formed in the narrower region 2086–3800 cm–1. The spec-
tral estimate (Figure 13b) displays an intense P–H vibration
as well as vibrations in the O–H region. Therefore, this re-
gion was analysed with BTEM to look for the rotational
isomers of the trivalent tautomer. However, this gave rise to
the same spectrum as that obtained for the pentavalent
P(O)–H tautomer in Figure 13a. Apparently, bands of the
phosphane oxide and the phosphinous acid are superim-
posed in the same spectrum, as already shown for com-
pound 5.
BTEM Analysis: Around 300 spectra were measured for each ex-
perimental run. Each set was analysed individually by using the
BTEM algorithm to provide pure-component spectra of the com-
ponents present. Details of the BTEM algorithm are provided else-
where.[34] Because the toluene and benzene solvents absorb strongly
in some regions, spectroscopic data in these regions were deleted
prior to BTEM analyses. BTEM analysis was then conducted on
three regions of data: (1) 500–3800 cm–1 for full-range deconvol-
utions, (2) 2086–3800 cm–1 for deconvolutions focused on the P–
H and O–H vibrations and (3) 3250–3800 cm–1 for deconvolutions
focused on O–H vibrations.
Conclusions
The tautomeric equilibria of five secondary phosphane
oxides have been investigated by NMR and IR spectroscopy
Di-tert-butylphosphane Oxide (1): Strem, 98%. 1H NMR
as well as by quantum chemical calculations. Solvent depen- (300 MHz, CD2Cl2): δ = 6.17 (d, JP-H = 449.8 Hz, 1 H, PH), 1.29
[d, J = 16.0 Hz, 18 H, C(CH3)3] ppm. 13C NMR (75 MHz,
dency based on the hydrogen-bond acceptor properties
rather than on the polarity of solvents was demonstrated
for SPO 5. The solvent not only influences the state of the
CD2Cl2): δ = 33.8 [d, J = 58 Hz, C(CH3)3], 25.4 [d, J = 2 Hz,
C(CH3)3] ppm. 31P NMR (121 MHz, CD2Cl2): δ = 72.4 ppm. IR
(KBr): ν = 2284 (P–H) cm–1.
˜
tautomeric equilibrium, but may also contribute to the de-
gradation of SPOs, for example, esterification. A strong in-
fluence of temperature on the state of the equilibria can be
ruled out.
The influence of electron-withdrawing substituents,
which decrease the Lewis basicity of the phosphorus atom,
is reflected by the phosphorus chemical shifts in the NMR
spectra. Electron-withdrawing substituents cause an en-
hancement of the P–H couplings and a shift to higher fields.
Similarly, in IR spectroscopy, electron-accepting substitu-
Di(p-tolyl)phosphane Oxide (2): Usually, SPO 2 is synthesized by
Grignard reaction of di(p-tolyl)magnesium bromide with diethyl
phosphite and HCl.[36] For small amounts the following procedure
is more useful. 2-Propanol (25 mL) was added under constant stir-
ring to bis(4-methylphenyl)phosphane (2.145 g, 932 mmol) at am-
bient temperature. The solution was kept at 50 °C and stirred in an
open flask for 90 min. The solvent was distilled in vacuo and the
residue dissolved in dry, boiling diethyl ether (30 mL). On cooling
to room temperature the product precipitated as a white solid. The
product was recrystallized from diethyl ether to remove unreacted
phosphane. Colourless needles (1.42 g, 62%) suitable for X-ray
crystallography were obtained, which proved to be air-stable for
several weeks. M.p. 95–96 °C (Et2O). 1H NMR (300 MHz,
CD2Cl2): δ = 7.98 (d, JP–H = 477.3 Hz, 1 H, PH), 7.54 (m, 4 H,
C6H4), 7.30 (m, 4 H, C6H4), 2.39 (s, 6 H, CH3) ppm. 13C NMR
(75 MHz, CD2Cl2): δ = 143.4 (d, J = 2.8 Hz, C6H4, CCH3), 130.8
(d, J = 11.7 Hz, CH, C6H4), 129.8 (d, J = 13.0 Hz, CH, C6H4),
129.3 (d, J = 103.2 Hz, C6H4, CP), 21.6 (d, J = 1.2 Hz, CH3) ppm.
31P NMR (121 MHz, CD2Cl2): δ = 20.5 ppm. 1H NMR (300 MHz,
[D8]THF): δ = 7.99 (d, JP–H = 477.8 Hz, 1 H, PH), 7.61–7.48 (m,
ents enhance νPH. The state of the equilibria between the
˜
pentavalent and trivalent isomers of the SPOs 1–5 was cal-
culated from the free enthalpies. They conform to those de-
termined by NMR spectroscopy in a THF solution.
For SPO 2 the structure of the pentavalent tautomer was
additionally evidenced by X-ray structural analysis.
DFT calculations revealed that phosphinous acids of
type B exist as two rotational isomers. Indeed, the presence
of rotational isomers for (C5NF4)2P–OH (5B) and Ph2P–
OH (3B) has been evidenced by IR spectroscopy for the
first time in the liquid phase. The identification of both tau-
tomers of compounds 3 and 5 by IR spectroscopy is the
precondition for the quantitative evaluation of their tauto-
meric equilibria. Next, we will focus on the establishment
of time-dependent concentration profiles with the aim of
tracking the equilibria during the complexation of SPOs to
metal centres.
4 H, C6H4), 7.35–7.26 (m, 4 H, C6H4), 2.39 (s, 6 H, CH3) ppm. 31
P
NMR (121 MHz, [D ]THF): δ = 20.6 ppm. IR (KBr): ν = 2337 (P–
˜
8
H) cm–1. C14H15OP (230.25): calcd. C 73.03, H 6.57, P 13.45; found
C 73.1, H 6.55, P 13.58.
Diphenylphosphane Oxide (3): Aldrich, Ͼ97%. 1H NMR
(300 MHz, CD2Cl2): δ = 8.05 (d, JP–H = 480.1 Hz, 1 H, PH), 7.69
(m, 4 H, C6H5), 7.54 (m, 6 H, C6H5) ppm. 13C NMR (76 MHz,
CD2Cl2): δ = 132.7 (d, J = 2.8 Hz, C6H5, C-4), 132.4 (d, J =
100.6 Hz, CP), 130.8 (d, J = 11.4 Hz, C6H5, C-2), 129.1 (d, J =
12.5 Hz, C6H5, C-3) ppm. 31P NMR (122 MHz, CD2Cl2): δ =
1
20.4 ppm. 31P NMR (122 MHz, CD3OD): δ = 24.5 (d, JP–H
=
495.4 Hz), 23.9 (t, JP–D = 75.7 Hz) ppm. 31P NMR (162 MHz,
1
Experimental Section
1
1
CF3CD2OD): δ = 29.6 (d, JP–H = 498.5 Hz), 29.0 (t, JP–D
=
General Procedures: Unless stated otherwise, all reactions were car-
ried out under argon 5.0 by using standard Schlenk techniques.
Commercially available compounds were used as purchased. Sol-
vents were dried by conventional procedures and distilled under
argon. Where possible, reactions were monitored by NMR spec-
76.3 Hz) ppm. IR (KBr): ν = 2368 (P–H) cm–1.
˜
Bis(4-fluorophenyl)phosphane Oxide (4): tert-Butyl alcohol (20 mL)
was added to bis(4-fluorophenyl)chlorophosphane (5 mL, Digital
Speciality Products, 98.5%) under constant stirring.[37] The Schlenk
Eur. J. Org. Chem. 2010, 2733–2741
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2739

A. Börner et al.
FULL PAPER
tube was sealed. After 6 min of stirring at room temperature, the
volatiles were removed in vacuo, and the product was obtained
quantitatively as a viscous liquid. It should be stored below 10 °C
to prevent disproportionation into R2PH and R2P(O)OH (R = 4-
supporting the IR experiments and the application of BTEM. We
thank PD Dr. B. Hoge (University of Cologne) for a generous gift
of SPO 5 and valuable discussions.
F-C6H4). 1H NMR (400 MHz, CD2Cl2): δ = 8.06 (d, JP–H
=
488.60 Hz, 1 H, PH), 7.67 (m, 4 H, C6H4), 7.09 (m, 4 H, C6H4) [1] a) R. H. Crabtree, The Organometallic Chemistry of Transition
ppm. 13C NMR (101 MHz, CD2Cl2): δ = 165.1 (dd, J = 3.3, J =
252.9 Hz, CF), 133.1 (dd, J = 8.9, J = 13.1 Hz, C6H4), 128.0 (dd,
J = 3.4, J = 103.3 Hz, CP), 116.1 (dd, J = 13.9, J = 21.6 Hz, C6H4)
ppm. 19F NMR (282 MHz, CD2Cl2): δ = –106.0 (m) ppm. 31P
Metals, Wiley, New York, 2001; b) A. Börner (Ed.), Phosphorus
Ligands in Asymmetric Catalysis, Wiley-VCH, Weinheim, 2008,
vol. I–III.
[2] a) D. M. Roundhill, P. R. Sperline, W. B. Beaulieu, Coord.
Chem. Rev. 1978, 26, 263–279; b) B. Walther, Coord. Chem.
Rev. 1984, 60, 67–105.
[3] a) M. Grayson, C. E. Farley, C. A. Streuli, Tetrahedron 1967,
23, 1065–1078; b) L. D. Quin, A Guide to Organophosphorus
Chemistry, Wiley-Interscience, New York, 2000.
NMR (162 MHz, CD Cl ): δ = 17.5 ppm. IR (KBr): ν = 2327 (P–
˜
2
2
H) cm–1. MS (EI, 70 eV): m/z = 237 [M]+. C12H9F2OP (238.17):
calcd. C 60.51, H 3.81, P 13.00; found C 60.53, H 3.86, P 13.03.
Bis(tetrafluoropyridyl)phosphane Oxide (5): This compound was
[4] J. Chatt, B. T. Heaton, J. Chem. Soc. A 1968, 2745–2757.
[5] N. V. Dubrovina, A. Börner, Angew. Chem. 2004, 116, 6007–
6010; Angew. Chem. Int. Ed. 2004, 43, 5883–5886.
[6] For reviews, see: a) L. Ackermann, Synthesis 2006, 10, 1557–
1571; b) L. Ackermann, Synlett 2007, 4, 507–526; c) L. Acker-
mann, R. Born, J. H. Spatz, A. Althammer, C. J. Gschrei, Pure
Appl. Chem. 2006, 78, 209–214; d) L. Ackermann in Phospho-
rus Ligands in Asymmetric Catalysis (Ed.: A. Börner), Wiley-
VCH, Weinheim, 2008, vol. 2, pp. 831–847.
[7] For recent work, compare: a) L. Ackermann, Synlett 2007,
507–526; b) A. Pfaltz, Y. Ribourdouille, X. Feng, B. Ramalin-
gam, B. Pugin, F. Spindler WO2007135179 A1 20071129, 2007
(Chem. Abstr. 2007, 148, 11345); c) J. Bigeault, L. Giordano, I.
de Riggi, Y. Gimbert, G. Buono, Org. Lett. 2007, 9, 3567–3570;
d) J. Bigeault, I. de Riggi, Y. Gimbert, L. Giordano, G. Buono,
Synlett 2008, 1071–1075; e) B. Pugin, H. Landert, B.
Gschwend, A. Pfaltz, F. Spindler WO 2009065783 A1
20090528, 2009 (Chem. Abstr. 2009, 150, 563987); f) L. Acker-
mann, S. Barfuesser, Synlett 2009, 808–812.
[8] D. E. C. Corbridge, Phosphorus: An Outline of its Chemistry,
Biochemistry and Uses, 5th ed., Elsevier, Amsterdam, 1995, p.
336.
[9] a) Z. Luz, B. L. Silver, J. Am. Chem. Soc. 1961, 83, 4518–4521;
b) J. Reuben, D. Samuel, B. L. Silver, J. Am. Chem. Soc. 1963,
85, 3093–3096; c) D. Samuel, Pure Appl. Chem. 1964, 8, 449–
457; d) W. J. Bailey, R. B. Fox, J. Org. Chem. 1963, 28, 531–
534; e) W. J. Bailey, R. B. Fox, J. Org. Chem. 1964, 29, 1013–
1017; f) T. L. Emmick, R. L. Letsinger, J. Am. Chem. Soc. 1968,
90, 3459–3465.
synthesized by Hoge et al.[16] and generously provided for our re-
1
search. H NMR (400 MHz, [D8]THF): δ = 9.00 (br., 1 H, POH)
ppm. 13C NMR (101 MHz, [D8]THF): δ = 144.5–141.9 (m, C5NF4)
ppm. 19F NMR (282.4 MHz, [D8]THF): δ = –92.9 (m), –137.1 (m)
ppm. 31P NMR (162 MHz, [D8]THF): δ = 71.8 (quint, JP–F
=
26.9 Hz, POH) ppm. 1H NMR (300 MHz, [D8]toluene): δ = 7.87
(d, JP–H = 585.5 Hz, PH), 4.78 (br., POH) ppm. 19F NMR
(282.4 MHz, [D8]toluene): δ = –86.6 [m, P(O)H], –89.9 (m, POH),
–135.1 [m, P(O)H], –136.4 (m, POH) ppm. 31P NMR (121 MHz,
[D8]toluene): δ = 69.1 (br., POH), –22.8 [br., P(O)H] ppm. 31P
NMR (162 MHz, CD3OD): δ = 70.2 (quint, JP–F = 26.7 Hz, POD)
ppm. 31P NMR (162 MHz, CF3CD2OD): δ = 72.8 (quint, JP–F
=
27.6 Hz, POD), –15.3 [t, J = 92.4 Hz, P(O)D], –15.0 [d, J =
606.0 Hz, P(O)H] ppm. IR (ATR): ν = 2402 (P–H) cm–1.
˜
2-(Diphenylphosphanyl)-2-propanol (6): A solution of diphenylphos-
phane oxide (3; 60 mg, 0.3 mmol) in acetone (1.5 mL) was concen-
trated to dryness in vacuo after 36 h of stirring at room tempera-
ture. Then the residue was dissolved in [D8]THF (0.4 mL) at 50 °C.
On cooling to room temperature, the product precipitated as
colourless needles (20 mg, 26%). M.p. 125–127 °C. 1H NMR
(300 MHz, CD2Cl2/[D6]acetone): δ = 7.88 (m, 4 H, C6H5), 7.28 (m,
6 H, C6H5), 3.01 (br., 1 H, OH), 1.19 (d, J = 13.4 Hz, 6 H, CH3)
ppm. 13C NMR (75 MHz, CD2Cl2/[D6]acetone): δ = 132.5 (d, JP–
= 7.9 Hz, o-C6H5), 132.0 (d, JP–C = 89.9 Hz, i-C6H5), 131.5 (d,
C
JP–C = 2.7 Hz, p-C6H5), 128.2 (d, JP–C = 10.7 Hz, m-C6H5), 24.8
(d, JP–C = 6.7 Hz, CH3) ppm; C–OH signal not found. 31P NMR
[10] P. Nylen, Z. Anorg. Allg. Chem. 1938, 235, 161–182.
[11] N. B. Colthup, L. H. Daly, S. E. Wiberley, Introduction to Infra-
red and Raman Spectroscopy, Academic Press, New York, 1964,
pp. 299–300, 402.
[12] D. Magiera, A. Szmigielska, K. M. Pietrusiewicz, H. Duddeck,
Chirality 2004, 16, 57–64.
(121 MHz, CD2Cl2/[D6]acetone):
δ = 31.3 ppm. C15H17O2P
(260.27): calcd. C 69.22, H 6.58; found C 69.3, H 6.23.
X-ray Crystal Structure Analysis of 2: Data were collected with a
STOE IPDS II diffractometer by using graphite-monochromated
Mo-Kα radiation. The structure was solved by direct methods
(SHELXS-97)[35] and refined by full-matrix least-squares tech-
niques on F2 (SHELXL-97).[35] XP (Bruker AXS) was used for
graphical representations. C14H15OP, Mr = 230.23, orthorhombic,
space group P212121, a = 5.6391(11), b = 7.583(2), c = 28.459(6) Å,
V = 1217.0(4) Å3, Z = 4, ρcalcd. = 1.257 gcm–3, µ = 0.201 mm–1, T
= 200 K, 17197 reflections measured, 2570 independent reflections
(Rint = 0.0394) of which 2313 were observed [IϾ2σ(I)], R1 = 0.0279
[IϾ2σ(I)], wR2 = 0.0778 (all data), 151 refined parameters.
[13] Yu. A. Ustynyuk, Yu. V. Babin, Ross. Khim. Zh. 2007, 51, 130–
138.
[14] D. B. Chesnut, Heteroat. Chem. 2000, 11, 73–80.
[15] F. Wang, P. L. Polavarapu, J. Org. Chem. 2000, 65, 7561–7565.
[16] B. Hoge, S. Neufeind, S. Hettel, W. Wiebe, C. Thoesen, J. Or-
ganomet. Chem. 2005, 690, 2382–2387.
[17] The curve in the middle of Figure 1 was calculated by using
B3LYP/6-31+G* [σ(H3PO4) = 362.13 ppm, scaled onto δ(31P)
of H3PO4 (δ = 0 ppm)] and the lower curve by using GIAO-
B3LYP/6-311+G*//B3LYP/6-311+G* [σ(PPh3) = 294.61 ppm,
scaled onto δ(31P) of PPh3 (δ = –6.0 ppm)].
[18] For a review, see: I. A. Shuklov, N. V. Dubrovina, A. Börner,
Synlett 2007, 2925–2943.
Acknowledgments
[19] a) A. N. Chekhlov, Zh. Strukt. Khim. 2001, 42, 152–155 (russ.);
b) A. Y. Garner, US 3346647, 1967 (Chem. Abstr. 1967, 67,
116943).
Financial support was provided by Evonik Oxeno GmbH. One of
us (A. C.) is grateful for a research scholarship from the Degussa
Stiftung. We thank Dr. O. Zayas, Mrs. H. Borgwald, Mrs. M. Gei-
sendorf and Ms. K. Romeike for their skilled technical assistance.
We also thank the group of Prof. Dr. M. Garland at the Institute
for Chemical and Engineering Sciences (ICES) in Singapore for
[20] Originally, this signal was attributed to the corresponding tri-
valent P–OH tautomer 3B; see ref.[12]
[21] A similar conclusion was drawn by Hoge et al. from experi-
ments considering the equilibrium between (C6F5)2P(O)H and
2740
www.eurjoc.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 2733–2741

On the Tautomerism of Secondary Phosphane Oxides
1
(C6F5)2POH carried out in the range between –70 °C and
+50 °C in THF; see ref.[16]
both P–OH rotational isomers, but is caused by a H–31P cou-
pling (2J). Owing to a fast exchange of the acid proton, this
phenomenon is not observable at room temperature, but only
at low temperatures when the mobility is decelerated. The same
phenomenon is observed for many alcohols, e.g., methanol: H.
Günther, NMR-Spektroskopie – Eine Einführung, Thieme,
Stuttgart, 1973.
The measurement of the less stable SPOs 4 and 5 caused some
problems. Thus, the absorption of νPH is very small. Also, other
authors reported on this issue during the IR characterization
of electron-poor phosphane oxides, for which no unequivocal
bands for P–OH, P–H, or P=O vibrations could be observed:
D. D. Magnelli, G. Tesi, I. U. Lowe, W. E. McQuistion, Inorg.
Chem. 1966, 5, 457–461.
[22] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, J. A. Montgomery, Jr., T.
Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar,
J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N.
Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K.
Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y.
Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P.
Hratchian, J. B. Cross, C. Adamo, J. Jaramillo, R. Gomperts,
R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pom-
elli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P.
Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich,
A. D. Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D.
Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui,
A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu,
A. Liashenko, P. Piskorz, I. Komaromi, R. L. Martin, D. J.
Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara,
M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W.
Wong, C. Gonzalez, J. A. Pople, Gaussian 03, Revision D.01,
Gaussian Inc., Wallingford, CT, 2004.
[29]
[30]
˜
The experimentally determined value of diphenylphosphane
oxide (3) (νPH = 2368 cm–1) is in accord with the value reported
˜
in the literature of ν = 2370 cm–1: J. Chatt, B. T. Heaton, J.
˜
PH
Chem. Soc. A 1968, 2745–2757.
[31]
[32]
[33]
See the Exp. Sect. for further discussions on the pretreatment
of data and BTEM analysis.
This phenomenon is caused by intense solvent bands. For more
details, see the Exp. Sect.
[23] a) C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785–
789; b) A. D. Becke, Phys. Rev. A 1988, 38, 3098–3100; c) A. D.
Becke, J. Chem. Phys. 1992, 96, 2155–2160; d) A. D. Becke, J.
Chem. Phys. 1992, 97, 9173–9177; e) A. D. Becke, J. Chem.
Phys. 1993, 98, 5648–5652.
The shoulder of ν
and the bands at 3594 and 3685 cm–1
˜
OH_1
neither characterize compound 5B nor traces of water. We at-
tribute them to impurities and decomposition products.
[24] Calculated at the B3LYP/6-311+G* level of theory.
[25] R. C. Dobbie, B. P. Straughan, Spectrochim. Acta, Part A 1971,
27, 255–260.
[26] B. Hoge, P. Garcia, H. Willner, H. Oberhammer, Chem. Eur. J.
2006, 12, 3567–3574.
[34] a) W. Chew, E. Widjaja, M. Garland, Organometallics 2002,
21, 1982–1990; b) E. Widjaja, C. Li, M. Garland, Organometal-
lics 2002, 21, 1991–1997; c) M. Garland, Mechanisms in Homo-
geneous Catalysis: A Spectroscopic Approach, VCH-Wiley,
Weinheim, 2005.
[27] The abrupt changes at 340 and 60°, respectively, can be ration-
alized by a slight change in the arrangement of the aromatic
substituents as a result of a change in the dihedral angle.
[28] At room temperature a solution of (C5NF4)2POH (5B) in THF
exhibited only a broad resonance for the OH group at δ =
9.03 ppm in the 1H NMR spectrum, but with a decrease in
temperature to 233 K a slight shift to a higher field was ob-
served, and the signal was split into a doublet with J = 18.5 Hz.
However, this doublet cannot be attributed to the existence of
[35] G. M. Sheldrick, Acta Crystallogr., Sect. A 2008, 64, 112–122.
[36] C. A. Busacca, J. C. Lorenz, N. Grinberg, N. Haddad, M.
Hrapchak, B. Latli, H. Lee, P. Sabila, A. Saha, M. Sarvestani,
S. Shen, R. Varsolona, X. Wei, C. H. Senanayake, Org. Lett.
2005, 7, 4277–4280.
[37] For the hydrolysis of P–Cl bonds with tBuOH, see: M. T.
Reetz, T. Sell, R. Goddard, Chimia 2003, 57, 290–292.
Received: January 12, 2010
Published Online: April 7, 2010
Eur. J. Org. Chem. 2010, 2733–2741
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2741