Chemistry of singlet oxygen with arylphosphines

Article abstract of DOI:10.1016/j.tet.2006.07.112

The chemistry of singlet oxygen with a variety of arylphosphines has been studied. Rates of singlet oxygen removal by para-substituted arylphosphines show good correlation with the Hammett σ parameter (ρ=-1.53 in CDCl₃), and with the Tolman electronic parameter. The only products for the reactions of these phosphines with singlet oxygen are the corresponding phosphine oxides. Conversely, for ortho-substituted phosphines with electron-donating substituents, there are two products, namely a phosphinate formed by intramolecular insertion and phosphine oxide. Kinetic analyses demonstrate that both products are formed from the same intermediate, and this allows determination of the rate ratios for the competing pathways. Increasing the steric bulk of the phosphine leads to an increase in the amount of insertion product. VT NMR experiments show that peroxidic intermediates can only be detected for very hindered and very electron-rich arylphosphines.

Full text of DOI:10.1016/j.tet.2006.07.112

Tetrahedron 62 (2006) 10729–10733
Chemistry of singlet oxygen with arylphosphines
Dong Zhang,a Bin Ye,a David G. Ho,a Ruomei Gaob and Matthias Selkea,
*
aDepartment of Chemistry and Biochemistry, California State University, Los Angeles, CA 90032, USA
bDepartment of Chemistry, Jackson State University, 1325 J. R. Lynch St., P.O. Box 17910, Jackson, MS 39217, USA
Received 1 June 2006; revised 21 July 2006; accepted 27 July 2006
Available online 26 September 2006
Abstract—The chemistry of singlet oxygen with a variety of arylphosphines has been studied. Rates of singlet oxygen removal by para-
substituted arylphosphines show good correlation with the Hammett s parameter (r¼ꢀ1.53 in CDCl3), and with the Tolman electronic
parameter. The only products for the reactions of these phosphines with singlet oxygen are the corresponding phosphine oxides. Conversely,
for ortho-substituted phosphines with electron-donating substituents, there are two products, namely a phosphinate formed by intramolecular
insertion and phosphine oxide. Kinetic analyses demonstrate that both products are formed from the same intermediate, and this allows
determination of the rate ratios for the competing pathways. Increasing the steric bulk of the phosphine leads to an increase in the amount
of insertion product. VT NMR experiments show that peroxidic intermediates can only be detected for very hindered and very electron-
rich arylphosphines.
Ó 2006 Published by Elsevier Ltd.
1. Introduction
represent an interesting target for studying reactive interme-
diates and reaction channels, since their electronic and steric
parameters are well defined and can be varied in a systematic
manner. Electronic parameters that can be used to assess the
donor abilities of arylphosphines include Hammett constants
as well as the Tolman electronic parameter.14 The latter pa-
rameter is given by the CO stretching frequency of the car-
bonyl group trans to the phosphine in a Ni complex of the
type Ni(CO)3(PR3). The stronger the electron-donating abil-
ity of the phosphine, the stronger the backbonding to the trans
CO ligand, and hence the lower the CO stretching frequency.
Chemical properties of arylphosphines are also influenced by
the steric bulkof the aryl groups, and the cone angle of aphos-
phine ligand has long been used as a measure of its bulk.14
The cone angle is obtained by taking a space-filling model
of the MP(R3) group. The metal is the apex of the cone,
and the entire ligand forms the actual cone.4,5 We have re-
ported the special case of tris(ortho-methoxyphenyl)phos-
phine: the exceptionally large steric bulk of this phosphine
(cone angle of 205ꢁ)15 slows down the reactions of peroxidic
intermediate so that it can be directly observed at very low
temperatures by VT NMR. 31P and 17O NMR measurements
at ꢀ80ꢁ strongly suggest that the peroxidic intermediate is
a three-membered ring containing a phosphorus and two
oxygen atoms, that is, a phosphadioxirane.10 We report
herein, the products for the reaction of a variety of arylphos-
phines with singlet oxygen, kinetic data, as well as attempts
to observe and characterize products and peroxidic interme-
diates of these reactions. Electronic parameters for the differ-
ent arylphosphines were simply varied by using different
substituents in the para position; the cone angle is the same
for all of these para-substituted phosphines, namely 145ꢁ.
The chemistry of singlet dioxygen with heteroatoms, i.e.,
organic sulfides1–5 and, more recently, phosphines5–9 has
been the subject of numerous papers, as these reactions
posses rather complicated mechanisms. Elucidation of the
nature of reactive intermediates, and, ideally, their direct
observation, have been a major goal of these studies. Such
intermediates are of fundamental interest, as they should
be even better oxidants than singlet dioxygen itself. In par-
ticular, many kinetic and trapping studies have been carried
out to determine the nature of the peroxidic intermediates
formed during the photooxidation of organic sulfides. The
oxidation of phosphines has attracted somewhat less atten-
tion. An early paper by Bolduc and Goe11 suggested an open
zwitterionic species formed as a primary intermediate, but
trapping studies by Sawaki and co-workers7 indicated that
the primary intermediate in the photooxidation of triphenyl
phosphines and tributyl phosphite is electrophilic. This is
in contrast with the nucleophilic behavior of the persulfoxide
intermediate in the photooxidation of organic sulfides.1 Ab
initio calculations by Foote and co-workers8 suggested a
cyclic phosphadioxirane intermediate, which would indeed
be expected to behave as an electrophile. Arylphosphines
have also been investigated as fuel-stabilizers for jet fuels,
and very recently Beaver and co-workers have provided
strong evidence that phosphadioxiranes may play a key
role during this antioxidant activity.12,13 Arylphosphines
Keywords: Singlet oxygen; Cone angle; Arylphosphine; Phosphadioxirane.
* Corresponding author. Tel.: +1 323 343 2347; fax: +1 323 343 6490;
e-mail: mselke@calstatela.edu
0040–4020/$ - see front matter Ó 2006 Published by Elsevier Ltd.
doi:10.1016/j.tet.2006.07.112

10730
D. Zhang et al. / Tetrahedron 62 (2006) 10729–10733
OCH3
Conversely, steric parameters were changed by employing
different substituents in the ortho position. All of these aryl-
phosphines are unreactive with triplet oxygen, thus allowing
kinetic investigations of their reactions with singlet oxygen.
0.4
0
y = -1.5276x - 0.0939
R2 = 0.9826
H
-0.4
-0.2
0
0.2
0.
4
0.6
F
-0.4
-0.8
-1.2
Cl
2. Results and discussion
ρ = -1.53
CF3
2.1. Photooxidation of para-substituted arylphosphines
σ
Reaction of singlet oxygen at room temperature with para-
substituted arylphosphines 1–5 leads to the corresponding
phosphine oxides in quantitative yield (Eq. 1).
Figure 1. Hammett plot for the reaction of para-substituted arylphosphines
with singlet oxygen.
phosphines 1–4 (the Tolman electronic parameter for phos-
phine 5 has not been determined). Since the Tolman elec-
tronic parameter is a measure of the s-donating ability of
the phosphine, it is apparent that in the absence of steric
variation, the s-donation of the arylphosphine to the empty
degenerate p* MO of the singlet oxygen molecule deter-
mines the reaction rate of singlet oxygen with the para-
substituted arylphosphines.
1O2
O
X
P
2
X
P
2
3
3
1
2
3
4
5
X = MeO
X = H
ð1Þ
X = F
X = Cl
X = CF3
In the absence of physical quenching, the value of kT is the
rate of formation of the peroxidic intermediate (i.e., the
phosphadioxirane). For all para-substituted arylphosphines,
the corresponding phosphine oxide was the only major
product observed by 31P NMR (other products such as the
corresponding phosphinates7 were 1% or less). In order to
determine whether or not all of the peroxidic intermediates
lead to the formation of the reaction products or whether
loss of dioxygen from the intermediate (indirect physical
quenching) is a major process, competition experiments
have been carried out. These competition experiments are
done under continuous irradiation of a solution containing
a singlet oxygen acceptor with a known chemical reaction
rate constant and the arylphosphine that is being studied.
Thus the two substrates compete for one reactive intermedi-
ate (i.e., singlet oxygen) that is produced under steady-state
conditions. 9,10-Dimethylanthracene (DMA) was used as
a singlet oxygen acceptor in these experiments. DMA is
known to interact with singlet oxygen by chemical reaction
only. As reference values, we have used kT values for the
reaction of DMA with singlet oxygen from the recent
literature,9 namely 2.5ꢂ0.1ꢃ107 Mꢀ1 sꢀ1 in CDCl3 and
2.9ꢂ0.2ꢃ107 Mꢀ1 sꢀ1 in deuterated benzene. Loss of
Total rate constants of singlet oxygen removal (kT) by phos-
phines 1–5 have been measured by luminescence quenching
experiments in a deuterated benzene and chloroform. The kT
values and the corresponding s values as well as the Tolman
electronic parameters for various phosphines are summa-
rized in Table 1.
The logs of the kT values (kSubst/kH) yield a good correlation
with the Hammett s parameters; the Hammett plot gives
a negative slope of ꢀ1.53 in CDCl3 (Fig. 1), and ꢀ1.93 in
C6D6 (r2¼0.987). Since there are three aryl rings per phos-
phine, the actual r values for the electronic effect of one
aryl ring are ꢀ0.51 and ꢀ0.64, respectively. The reason
for the larger value in benzene is unclear at the moment,
although it may simply be caused by the smaller dielectric
constant of benzene, similar to what has recently been pre-
dicted by ab initio calculations for solvent effects on r values
in benzhydryl cations.16 The negative r value is in agreement
with the electrophilic character of the initial singlet oxygen
attack on the arylphosphine. The magnitude of r for one
arylphosphine is smaller than that for the corresponding
attack of singlet oxygen on aryl sulfides in thioanisoles.17
1
DMA and compounds 1–5 was monitored by H and 31P
A plot of the kT values versus the Tolman electronic para-
meter also gives a good linear correlation (r2¼0.971) for
NMR. The relative reaction rate ratio for the kr values of
Table 1. Summary of kinetic data, electronic, and steric parameters for phosphines
Phosphine
kT in CDCl3
kr in CDCl3
kT in C6D6
kr in C6D6
ko/ki in
CDCl3
sp
Cone angle
(deg)d
Tolman electronic
parameter (cmꢀ1 d
(Mꢀ1 sꢀ1)ꢃ106
(Mꢀ1 sꢀ1)ꢃ106
(Mꢀ1 sꢀ1)ꢃ106
(Mꢀ1 sꢀ1)ꢃ106
)
(p-CH3OC6H4)3Pa
(C6H5)3Pb
14.9ꢂ1.6
8.1ꢂ0.3
4.9ꢂ0.2
3.1ꢂ0.2
0.9ꢂ0.05
5.4ꢂ0.5
0.1ꢂ0.01
2.8ꢂ0.4
32.4ꢂ1.2
16.2ꢂ1.6
11.0ꢂ1.0
7.1ꢂ0.7
2.2ꢂ0.3
33.1ꢂ 4.3
10.5ꢂ 0.6
5.5ꢂ0.2
59.4ꢂ0.9
21.5ꢂ1.1
11.0ꢂ0.8
8.0ꢂ0.8
2.6ꢂ0.6
ꢀ0.27
0
0.06
0.23
0.54
145
145
145
145
145
194
205
205
2066.1
2068.9
2071.3
2072.8
(p-FC6H4)3Pb
(p-ClC6H4)3Pb
(p-CF3C6H4)3Pc
(o-CH3C6H4)3Pb
(o-CF3C6H4)3Pc
3.3ꢂ0.3
0.9ꢂ0.06
8.0ꢂ0.2
80ꢂ7
25ꢂ3
2066.6
2058.3
Not measurable
Not measurable
a
(o-CH3OC6H4)3
5.0ꢂ0.2
a
Ref. 9.
This work.
Ref. 10.
Ref. 14.
b
c
d

D. Zhang et al. / Tetrahedron 62 (2006) 10729–10733
10731
the phosphine and the singlet oxygen acceptor (i.e., DMA)
was obtained from the equation of Higgins and co-workers.18
corresponding meta and para isomers, which do not form
the insertion product.9 We now report that the behavior of
the tris(ortho-tolyl)phosphine (9) is quite similar: reaction
of singlet oxygen with 9 leads to the formation of tris(ortho-
tolyl)phosphine oxide (10) and ortho tolyl di(ortho-tolyl)-
phosphinate (11).
n
o
f
0
log ½Arylphosphineꢄ=½Arylphosphineꢄ
n
o
f
0
log ½DMAꢄ=½DMAꢄ
If the phosphine oxides and phosphinates are formed from
the same intermediate (i.e., a phosphadioxirane), we can
use a simple steady-state treatment to obtain the rate ratio
for the intramolecular (ki) and intermolecular (ko) reaction
rates: according to Scheme 1, the rate of formation of the
phosphinates (compounds 7 and 10) is given by Eq. 3:
krðArylphosphineÞ
¼
ð2Þ
krðDMAÞ
For all of the para-substituted arylphosphines (compounds
1–5), the rate of formation of the corresponding phosphine
oxides (kr) is twice the rate of singlet oxygen removal kT,
both in benzene and chloroform. Thus indirect physical
quenching (loss of dioxygen from the periodic intermediate)
is not detectable, and all of the phosphadioxirane intermedi-
ates are converted to product. We also conducted competi-
tion experiments in benzene (Table 1), and found that the
reaction rates (i.e., kr values) follows the same trend, but
are somewhat larger than in chloroform. The reason for
this increase is unclear at this time.
d½Phosphinateꢄ=dt ¼ ki½Phosphadioxiraneꢄ
ð3Þ
Likewise, the rate of formation of the phosphine oxides
(compounds 8 and 11) can be expressed by Eq. 4:
d½Phosphine oxideꢄ=dt ¼ 2ko½Phosphadioxiraneꢄ½Phosphineꢄ
ð4Þ
2.2. ortho-Substituted arylphosphines
Assuming a steady-state concentration of the phosphadioxir-
ane, and working at low conversion of the starting phosphine,
we can combine Eqs. 3 and 4 to obtain an integrated expres-
sion for the rate ratio of intramolecular versus intermolecular
oxidation:
Generally, kT values are lower for these phosphines com-
pared to their para-substituted counterparts. It is particularly
instructive to compare the kT values for tris(ortho-methoxy-
phenyl)phosphine (6), tris(ortho-tolyl)phosphine (9), and
triphenylphosphine (2). Despite the fact that the electron
density at the phosphorus atom is decreasing from 6 to 9
to 2 (cf. the Tolman electronic parameters), the correspond-
ing kT values are increasing. This illustrates the strong sen-
sitivity of reactions of singlet oxygen with arylphosphines
to steric effects, similar to what has been observed for
organic sulfides.
½Phosphine oxideꢄ=½Phosphinateꢄ ¼ ð2ko½PhosphineꢄÞ=ki
ð5Þ
According to Eq. 5, a plot of the product ratio versus starting
material concentration should be linear with a slope of twice
the rate ratio ko/ki. A plot of the photooxidation product ratio
(11/10) for tris(ortho-tolyl)phosphine versus starting concen-
tration of tris(ortho-tolyl)phosphine (in CDCl3) indeed gives
a good linear correlation with a slope of 161ꢂ14 (Fig. 2).
We have recently reported an unusual intramolecular oxida-
tion pathway during the photooxidation of tris(ortho-
methoxyphenyl)phosphine (6). In addition to tris(ortho-
methoxyphenyl)phosphine oxide (7), significant amounts
of ortho methoxyphenyl di(ortho-methoxyphenyl)phosphi-
nate (8) are obtained. In fact, at low concentration, this intra-
molecular oxidation product becomes the major reaction
product. We have demonstrated that there exists a phosphine
cone angle dependence for this reactive channel of the
putative peroxidic intermediate by comparison with the
The rate ratio of intermolecular versus intramolecular
oxidation for tris(ortho-tolyl)phosphine is thus 80ꢂ7.
This ratio is considerably larger than that of the reaction of
phosphine 6 with singlet oxygen where ko/ki is 25ꢂ3 in
CDCl3. Intramolecular oxidation is thus less favorable for
R
O
P
R
O
R
O
R
R
R
O
ki
R
1O2
kT
P
P
7: R = OMe
10: R = Me
R
R
PR3
ko
O
R
R
P
6: R = OMe
9: R = Me
2
R
8: R = OMe
11: R = Me
Scheme 1. Reaction of electron-rich ortho-substituted arylphosphines with singlet oxygen.

10732
D. Zhang et al. / Tetrahedron 62 (2006) 10729–10733
(26 ppm at ꢀ80 ꢁC).9 Compound 8 is formed due to the re-
9
8
7
6
5
4
3
2
1
0
y = 160.77x - 0.3577
2 = 0.9603
action of phosphadioxirane with unreacted starting material.
Formation of 8 relative to phosphadioxirane is diminished in
CH2Cl2/toluene mixtures compared to neat CH2Cl2 solution.
In the present work, we noted that tris(ortho-tolyl)phos-
phine (9) shows similar behavior compared to 6 at room tem-
perature. We, therefore, reasoned that phosphine 9 might
also form a phosphadioxirane that can be directly detected
at low temperature. However, despite numerous attempts,
we were unable to observe any peroxidic intermediate for
phosphine 9 at ꢀ80 ꢁC, and only the corresponding phos-
phine oxide 11 was obtained, regardless of solvent. It ap-
pears that the relatively small decrease in the cone angle
of 9 compared to 6 and possibly the decrease in the electron
density at the phosphorus atom are sufficient to make the
phosphadioxirane intermediate too short-lived for direct ob-
servation. Interestingly, formation of the phosphinates 7 and
10 is greatly diminished if the photooxidation is carried out
at low temperature and completely suppressed below
ꢀ60 ꢁC. This appears to be in agreement with the very recent
observation by Beaver and co-workers13 who noted that the
phosphinate–phosphine oxide ratio is actually increased at
elevated temperatures.
R
0
0.0
1
0
.0
2
0
.0
3
0.0
4
0.05
[(o-CH3C6H4)3P] M
Figure 2. Product ratio of phosphine oxide (intermolecular product)/phos-
phinate (intramolecular product) versus starting phosphine concentration
for the reaction of tris(ortho-tolyl)phosphine with singlet oxygen.
tris(ortho-tolyl)phosphine as compared to tris(ortho-
methoxyphenyl)phosphine (6). This could be due to the
smaller cone angle of this phosphine (194ꢁ compared to
205ꢁ for phosphine 6, see Table 1) or due to a combination
of the smaller cone angle and an electronic effect, i.e., de-
creased electron density at the phosphorus atom. The latter
effect can be quite drastic, as illustrated by the behavior of
tris(ortho-trifluoromethyl)phenyl phosphine (12). The cone
angle of phosphine 12 is similar to that of 6, but tris(ortho-
trifluoromethyl)phenyl phosphine obviously is highly elec-
tron deficient. Phosphine 12 does quench singlet oxygen
with a rather slow rate constant kT¼1ꢃ105 Mꢀ1 sꢀ1, as
determined by a singlet oxygen luminescence quenching
experiment. However, we were unable to detect any products
either at room temperature or temperatures as low as
ꢀ80 ꢁC. Phosphine 12, therefore, represents the first known
example of physical quenching of singlet oxygen by a phos-
phine. It is quite possible that the quenching mechanism
involves a transient dioxygen–phosphine complex, i.e., indi-
rect physical quenching.19 The lack of reactivity of 12 is not
merely an electronic effect, since the corresponding para
isomer tris(para-trifluoromethyl)phenyl phosphine (5) re-
acts with singlet oxygen to form phosphine oxide, albeit at
a low rate (kT¼9ꢃ105 Mꢀ1 sꢀ1, see Table 1), and without
any concomitant physical quenching. It, therefore, appears
that electron-rich substituents in the ortho position are
needed for the intramolecular oxygen atom insertion reac-
tion to become a major pathway.
3. Experimental
3.1. Materials
All solvents and materials were the purest commercially
available products and used as received, except for tris-
(ortho-trifluoromethylphenyl)phosphine, which was synthe-
sized according to a literature method.20
3.2. Instrumentation
1H and 31P NMR spectra were obtained on a Bruker AC300
and DRX400 instrument. Low temperature spectra were ob-
tained by standard VT NMR techniques. Absorption spectra
were recorded on a Carey 300 UV–vis spectrophotometer.
3.3. Time-resolved singlet oxygen measurements
2.3. Low temperature photooxidation of arylphosphines
A nanosecond Nd:YAG laser (model MiniLaseII/10 Hz;
New Wave Research Inc., USA) doubled (532 nm) or tripled
(355 nm) in frequency is used as an excitation source.
Singlet oxygen luminescence is monitored at right angle.
The detector is a cryogenic germanium photodiode detector
(model 403HS; Applied Detector Corp., USA) cooled by liq-
uid nitrogen and specialized for detection of near-infrared
radiation. Three different filters are used to remove unde-
sired radiation. A Schott color glass filter (model RG850;
cut-on 850 nm; Newport, USA) is taped to the sapphire en-
trance of the detector to block any additional ultraviolet and
visible light from entering. The port opening to the detector
contains the remaining filters. The long wave pass filter (sil-
icon filter model 10LWFw1000; Newport, USA) transmits in
the range of 1100–2220 nm and blocks in the range of 800–
954 nm. A band pass filter (model BP-1270-080-B*; CWL
1270 nm; Spectrogen, USA) blocks in the UV, visible, and
IR regions and only transmits in the range of 1200–
1310 nm with maximum transmission of 60% at 1270 nm.
Signals were digitized on a LeCroy 9350CM 500 MHz
We were unsuccessful in detecting any peroxidic intermedi-
ates at temperatures as low as ꢀ80 ꢁC for the para-
substituted phosphines 1–5. It appears that the cone angle
of these phosphines (145 ꢁC) is simply too small to prevent
the reaction with starting material even at such low temper-
atures from being too fast to allow spectroscopic detection of
any peroxidic intermediate.
In contrast to the para-substituted arylphosphines 1–5, the
reaction of tris(ortho-methoxyphenyl)phosphine (6) and sin-
glet dioxygen at ꢀ80 ꢁC in methylene chloride or methylene
chloride/toluene mixtures leads to the rapid formation of a
new phosphine species, namely tris(ortho-methoxyphenyl)-
phosphadioxirane.10 This phosphadioxirane can be identi-
fied by its 31P NMR signal at ꢀ45 ppm. The only other
peak observed in the 31P NMR spectrum during the low
temperature photooxidation of 6 is the signal for the corre-
sponding tris(ortho-methoxyphenyl)phosphine oxide (8)

D. Zhang et al. / Tetrahedron 62 (2006) 10729–10733
10733
oscilloscope and analyzed using Origin software. Generally,
four to six runs were conducted per phosphine, and errors are
reported as one standard deviation.
acetone–20% methylene chloride/liquid N2 were used
for controlling the irradiation temperatures at ꢀ80 and
ꢀ90 ꢁC, respectively. After the photooxidation, the NMR
tubes were frozen in liquid N2 and subsequently placed
into the precooled NMR spectrometer.
3.4. General procedures for arylphosphines photo-
oxidation by singlet oxygen
Photooxidation reaction mixtures of 0.01–0.10 M in tris-
(ortho-methoxyphenyl)phosphine and singlet oxygen sensi-
tizers (tetraphenylporphyrin or C70, max absorption was
generally kept between 0.5 and 0.6) were prepared in test
tubes or NMR tubes. For NMR analyses, deuterated solvents
were employed during the photooxidation. The samples
were presaturated with oxygen for 1–2 min and then irradi-
ated under a constant stream of oxygen with a 250 W tung-
sten–halogen lamp. A 492 nm cutoff filter was used so that
the arylphosphines themselves were not irradiated.
Acknowledgements
This research was supported by a Henry Dreyfus Teacher–
Scholar Award. Support by the NIH-NIGMS MBRS
program (Award number GM08101) is also gratefully ac-
knowledged.
References and notes
1. Clennan, E. L. Acc. Chem. Res. 2001, 34, 875 and references
therein.
2. Jensen, F.; Greer, A.; Clennan, E. L. J. Am. Chem. Soc. 1998,
120, 4439.
3.5. Competition experiments between phosphines and
9,10-dimethylanthracene
The phosphine compounds were irradiated in the presence of
a singlet oxygen acceptor namely 9,10-dimethylanthracene
(DMA), which removes singlet oxygen by chemical reaction
only so that kT¼kr¼2.5ꢃ107 Mꢀ1 sꢀ1 in CDCl3.9 We used
tetraphenylporphyrin (TPP) or C70 as sensitizers. Phosphine
concentration ranged from 0.0005 to 0.003 M, and DMA
concentration ranged from 0.0005 to 0.004 M. Irradiation
times ranged from 5 to 30 min (for compound 5). Conversion
of each compound was generally kept between 20 and 80%.
3. Liang, J.-J.; Gu, C.-L.; Kacher, M. L.; Foote, C. S. J. Am. Chem.
Soc. 1983, 105, 4717.
4. Gu, C.-L.; Foote, C. S.; Kacher, M. L. J. Am. Chem. Soc. 1981,
103, 5949.
5. Clennan, E. L.; Pace, A. Tetrahedron 2005, 61, 6665.
6. Nahm, K.; Foote, C. S. J. Am. Chem. Soc. 1989, 111, 909.
7. Tsuji, S.; Kondo, M.; Ishiguro, K.; Sawaki, Y. J. Org. Chem.
1993, 58, 5055.
8. Nahm, K.; Li, Y.; Evanseck, J. D.; Houk, K.; Foote, C. S. J. Am.
Chem. Soc. 1993, 115, 4879.
1
DMA concentrations were monitored by H NMR while
phosphine concentrations were monitored by 31P NMR.
All of the NMR peaks used to determine concentrations
are summarized below.
9. Gao, R.; Ho, D. G.; Dong, T.; Khuu, D.; Franco, N.; Sezer, O.;
Selke, M. Org. Lett. 2001, 3, 3719.
10. Ho, D. G.; Gao, R.; Chung, H.-Y.; Celaje, J.; Selke, M. Science
2003, 302, 259.
Phosphine
31P NMR in
CDCl3, ppm
1H NMR in CDCl3
(methyl) ppm
11. Bolduc, P. R.; Goe, G. L. J. Org. Chem. 1974, 39, 3178.
12. Beaver, B. D.; Gao, L.; Fedak, M.; Coleman, M. M.;
Sobkowiak, M. Energy Fuels 2002, 16, 1134.
13. Beaver, B. D.; Burgess-Clifford, C.; Fedak, M. G.; Gao, L.;
Iyer, P. S.; Sobkowiak, M. Energy Fuels 2006, 20, 1639.
14. Tolman, C. A. Chem. Rev. 1977, 77, 313.
15. Hirsivaara, L.; Guerricabeitia, L.; Haukka, M.; Suomalainen,
P.; Laitinen, R. H.; Pakkanen, T. A.; Pursiainen, J. Inorg.
Chim. Acta 2000, 307, 47. Ab initio calculations by these
authors indicate that smaller cone angles, which would make
the phosphorous atom more exposed to intermolecular attack,
correspond to higher energy conformers.
(p-CH3OC6H4)3P (1)
(p-CH3OC6H4)3PO
(C6H5)3P (2)
(C6H5)3PO
(p-FC6H4)3P (3)
(p-FC6H4)3PO
ꢀ17.8
21.4
3.81
3.84
ꢀ13.0
21.2
ꢀ16.6
18.9
(p-ClC6H4)3P (4)
(p-ClC6H4)3PO
(p-CF3C6H4)3P (5)
(p-CF3C6H4)3PO
DMA
ꢀ16.0
19.0
ꢀ13.7
17.7
3.11
2.16
DMAO2
(o-CH3C6H4)3P (9)
(o-CH3C6H4)3PO (10)
(o-CH3C6H4)3PO2 (11)
(o-CH3OC6H4)3P (6)
(o-CH3OC6H4)3PO (7)
(o-CH3OC6H4)3PO2 (8)
ꢀ37.2
29.1
22.2
16. Kim, K. C.; Lee, K. A.; Sohn, C. K.; Sung, D. D.; Oh, H. K. J.;
Lee, I. J. J. Phys. Chem. A 2005, 109, 2978.
17. Bonesi, S. M.; Fagnoni, M.; Albini, A. J. Org. Chem. 2004, 69,
928.
18. Higgins, R.; Foote, C. S.; Cheng, H. Adv. Chem. Ser. 1968, 77,
102.
ꢀ39.2
3.76
3.58
3.60, 3.73
26.0
27.0
Generally, four to six runs were conducted per phosphine,
and errors are reported as one standard deviation. A control
experiment demonstrated that the DMA endoperoxide does
not oxidize the arylphosphines during the timeframe of the
competition experiments (i.e., 1 h or less).
19. We use the term indirect physical quenching to describe deac-
tivation of singlet oxygen via the formation of an unstable
intermediate, which then loses a dioxygen molecule. Direct
physical quenching refers to physical deactivation without
the formation of any unstable adduct (i.e., energy transfer).
This terminology has been used for the reaction of singlet
oxygen with organic sulfides. Foote, C. S.; Peters, J. W.
IUPAC Congr., 23rd, Spec. Lec. 1971, 4, 129.
20. Eapen, K. C.; Tamborski, C. J. Fluorine Chem. 1980, 15,
239.
3.6. Low temperature NMR studies
Samples were directly irradiated in NMR tubes placed
in a transparent Dewar flask. Acetone/liquid N2 and 80%