Photolytic generation of benzhydryl cations and radicals from quaternary phosphonium salts: How highly reactive carbocations survive their first nanoseconds

Article abstract of DOI:10.1021/ja3017522

UV irradiation (266 or 280 nm) of benzhydryl triarylphosphonium salts Ar₂CH-PAr₃⁺X^- yields benzhydryl cations Ar₂CH⁺ and/or benzhydryl radicals Ar ₂CH^?. The efficiency and mechanism of the photo-cleavage were studied by nanosecond laser flash photolysis and by ultrafast spectroscopy with a state-of-the-art femtosecond transient spectrometer. The influences of the photo-electrofuge (Ar₂CH ⁺), the photo-nucleofuge (PPh₃ or P(p-Cl-C ₆H₄)₃), the counterion (X^- = BF ₄^-, SbF₆^-, Cl^-, or Br^-), and the solvent (CH₂Cl₂ or CH ₃CN) were investigated. Photogeneration of carbocations from Ar ₂CH-PAr₃⁺BF₄^- or -SbF₆^- is considerably more efficient than from typical neutral precursors (e.g., benzhydryl chlorides or bromides). The photochemistry of phosphonium salts is controlled by the degree of ion pairing, which depends on the solvent and the concentration of the phosphonium salts. High yields of carbocations are obtained by photolyses of phosphonium salts with complex counterions (X^- = BF₄^- or SbF₆ ^-), while photolyses of phosphonium halides Ar₂CH-PPh ₃⁺X^- (X^- = Cl^- or Br ^-) in CH₂Cl₂ yield benzhydryl radicals Ar ₂CH^? due to photo-electron transfer in the excited phosphonium halide ion pair. At low concentrations in CH₃CN, the precursor salts are mostly unpaired, and the photo-cleavage mechanism is independent of the nature of the counter-anions. Dichloromethane is better suited for generating the more reactive benzhydryl cations than the more polar and more nucleophilic solvents CH₃CN or CF₃CH ₂OH. Efficient photo-generation of the most reactive benzhydryl cations (3,5-F₂-C₆H₃)₂CH⁺ and (4-(CF₃)-C₆H₄)₂CH⁺ was only achieved using the photo-leaving group P(p-Cl-C₆H ₄)₃ and the counter-anion SbF₆^- in CH₂Cl₂. The lifetimes of the photogenerated benzhydryl cations depend greatly on the decay mechanisms, which can be reactions with the solvent, with the photo-leaving group PAr₃, or with the counter-anion X^- of the precursor salt. However, the nature of the photo-leaving group and the counterion of the precursor phosphonium salt do not affect the rates of the reactions of the obtained benzhydryl cations toward added nucleophiles. The method presented in this work allows us to generate a wide range of donor- and acceptor-substituted benzhydryl cations Ar ₂CH⁺ for the purpose of studying their electrophilic reactivities.

Full text of DOI:10.1021/ja3017522

Article
pubs.acs.org/JACS
Photolytic Generation of Benzhydryl Cations and Radicals from
Quaternary Phosphonium Salts: How Highly Reactive Carbocations
Survive Their First Nanoseconds
Johannes Ammer,† Christian F. Sailer,‡ Eberhard Riedle,‡ and Herbert Mayr*,†
†Department Chemie, Ludwig-Maximilians-Universitat Munchen, Butenandtstrasse 5-13 (Haus F), 81377 Munchen, Germany
̈
̈
̈
‡Lehrstuhl fur BioMolekulare Optik, Ludwig-Maximilians-Universitat Munchen, Oettingenstrasse 67, 80538 Munchen, Germany
̈
̈
̈
̈
S
* Supporting Information
ABSTRACT: UV irradiation (266 or 280 nm) of benzhydryl
+
triarylphosphonium salts Ar2CH-PAr3 X− yields benzhydryl cations
Ar2CH+ and/or benzhydryl radicals Ar2CH•. The efficiency and
mechanism of the photo-cleavage were studied by nanosecond laser
flash photolysis and by ultrafast spectroscopy with a state-of-the-art
femtosecond transient spectrometer. The influences of the photo-
electrofuge (Ar2CH+), the photo-nucleofuge (PPh3 or P(p-Cl-
C6H4)3), the counterion (X− = BF4 , SbF6 , Cl−, or Br−), and the
solvent (CH2Cl2 or CH3CN) were investigated. Photogeneration of
−
−
+
−
−
carbocations from Ar2CH-PAr3 BF4 or -SbF6 is considerably more efficient than from typical neutral precursors (e.g.,
benzhydryl chlorides or bromides). The photochemistry of phosphonium salts is controlled by the degree of ion pairing, which
depends on the solvent and the concentration of the phosphonium salts. High yields of carbocations are obtained by photolyses
of phosphonium salts with complex counterions (X− = BF4 or SbF6 ), while photolyses of phosphonium halides Ar2CH-
−
−
PPh3 X− (X− = Cl− or Br−) in CH2Cl2 yield benzhydryl radicals Ar2CH• due to photo-electron transfer in the excited
+
phosphonium halide ion pair. At low concentrations in CH3CN, the precursor salts are mostly unpaired, and the photo-cleavage
mechanism is independent of the nature of the counter-anions. Dichloromethane is better suited for generating the more reactive
benzhydryl cations than the more polar and more nucleophilic solvents CH3CN or CF3CH2OH. Efficient photo-generation of
the most reactive benzhydryl cations (3,5-F2−C6H3)2CH+ and (4-(CF3)-C6H4)2CH+ was only achieved using the photo-leaving
−
group P(p-Cl-C6H4)3 and the counter-anion SbF6 in CH2Cl2. The lifetimes of the photogenerated benzhydryl cations depend
greatly on the decay mechanisms, which can be reactions with the solvent, with the photo-leaving group PAr3, or with the
counter-anion X− of the precursor salt. However, the nature of the photo-leaving group and the counterion of the precursor
phosphonium salt do not affect the rates of the reactions of the obtained benzhydryl cations toward added nucleophiles. The
method presented in this work allows us to generate a wide range of donor- and acceptor-substituted benzhydryl cations Ar2CH+
for the purpose of studying their electrophilic reactivities.
stability,36 even in alcoholic and aqueous solution, with a high
preference for heterolytic cleavage and low tendency to produce
1. INTRODUCTION
The photolytic generation of carbocations by heterolytic
radicals.14−20,32−34,38 While we have recently reported several
examples for the photolytic generation of carbocations from
quaternary phosphonium salts,14−19 there was no systematic
investigation about the relationship between the structure of the
precursor salt and the yields of the generated carbocations. The
lack of information became obvious when we failed to generate
benzhydrylium ions with empirical electrophilicity parameters E
> 7 by laser flash photolysis of phosphonium salts Ar2CH-
cleavage of neutral (R-X) and charged (R-X+) precursors has
been employed not only for studying the rates of fast reactions of
carbocations with nucleophiles1−19 but also for the photo-
initiation of carbocationic polymerizations.20 Furthermore,
photogenerated carbenium ions are the initial cleavage products
of the photolysis of certain photoacid and photobase
generators.21−26 Common precursors for such applications are
halides R-Hal,2−5,21a,27 acetates R-OAc,6−10,28 aryl ethers,6−9 and
onium salts such as halonium,25,29 sulfonium,21b,25,30 ammo-
nium,2,11,14,22,31 and phosphonium salts.14−20,23,32−34 Hetero-
lytic bond cleavages are often accompanied by formation of
radicals via homolytic processes, particularly when the resulting
carbocations are not highly stabilized and less polar solvents are
employed.35
+
−
PPh3 BF4 . As the photolytic generation of highly electrophilic
carbocations is of general importance, we have now examined
how the efficiency of carbocation formation can be influenced by
the reactivity of the photo-electrofuge (carbocation-to-be), the
photo-nucleofuge (photo-leaving group), and the counterion of
Among the many photo-leaving groups, phosphines turned
out to be particularly advantageous, because they combine high
Received: February 22, 2012
Published: May 15, 2012
© 2012 American Chemical Society
11481
dx.doi.org/10.1021/ja3017522 | J. Am. Chem. Soc. 2012, 134, 11481−11494

Journal of the American Chemical Society
Article
from 290 to 700 nm and polarized at the magic angle was used as probe
light.
the phosphonium salt. To gain insight into the ultrafast dynamics
of these processes, the nanosecond laser flash photolysis
experiments are supplemented by experiments on a state-of-
the-art femtosecond transient spectrometer.39
The use of benzhydryl derivatives is advantageous for these
investigations because of the clearly assignable distinct spectra of
the resulting cations and radicals.40 Benzhydryl cations
furthermore do not have β-protons and therefore cannot
eliminate H+, which reduces the number of subsequent
processes.15 Moreover, a systematic variation of the reactivity
of the generated carbocations is achieved by using substituted
benzhydryl cations Ar2CH+ (E+) whose electrophilic reactivities
are quantified accurately by the empirical electrophilicity
parameters E.41
The time-dependent transient spectra were recorded with temporal
resolutions between 100 and 400 fs, depending on the thickness of the
sample layer, but always well below all observed decay rates. The decay
kinetics were followed at λmax of the benzhydryl cations or radicals. The
absorbances were converted to quantum yields using the absorption
coefficients of the benzhydryl cations40 and the accurately determined
excitation probability.39 A least-squares fit of the time-dependent
absorbances A(t) to the sum of exponential curves and a constant
provided the decay rate constants and amplitudes from which the
quantum yields and rate constants for the different processes were
derived (see Supporting Information for details).
3. RESULTS AND DISCUSSION
3.1. General. Scheme 1 shows a mechanism for the photo-
generation of benzhydryl cations E+ and benzhydryl radicals E•
In the following, we will first investigate how the yields and
dynamics of the photoproducts change with variations of the
benzhydryl (section 3.2) and phosphine moieties (section 3.3) of
the precursor molecules, of the solvent (section 3.4), and of the
counter-anions of the phosphonium salts (section 3.5). We will
then show how this information can be employed to generate
highly reactive carbocations such as E(18−20)+ (section 3.6).
After discussing how the reaction conditions affect the lifetimes
of the carbocations on the >10 ns time scale (section 3.7), we will
finally demonstrate that the method presented in this work is well
suitable for the study of bimolecular reactions of the generated
benzhydryl cations (section 3.8).
Scheme 1. Generation of Benzhydryl Cations E+ and
Benzhydryl Radicals E• by Photolysis of Phosphonium Ions E-
+
PPh3
2. EXPERIMENTAL SECTION
2.1. Materials. Solvents. For the nanosecond laser flash photolysis
experiments, p.a. grade dichloromethane (Merck) was subsequently
treated with concentrated sulfuric acid, water, 10% NaHCO3 solution,
and again water. After predrying with anhydrous CaCl2, it was freshly
distilled over CaH2. Commercially available CH2Cl2 (Merck,
spectrophotometric grade) was used for the ultrafast measurements
because reactions with impurities are too slow to be observed on the
picosecond time scale. Acetonitrile (HPLC grade, VWR or
spectrophotometric grade, Sigma-Aldrich) and 2,2,2-trifluoroethanol
(99%, Apollo) were used as received.
+
Phosphonium Salts. The phosphonium salts E-PAr3 X− were
prepared by heating Ar2CH-OH with Ph3PH+X− or by treating
Ar2CH-Br with PAr3 and subsequent anion metathesis.42
2.2. Laser Flash Photolysis. Nanosecond Laser Flash Photolysis.
+
from benzhydryl phosphonium ions E-PPh3 in line with
+
Solutions of E-PAr3 X− (A266nm ≈ 0.1−1.0, ca. 5 × 10−5 to 10−4 M) were
previously proposed mechanisms for similar systems.35 The
excited precursor molecules can either undergo heterolytic bond
cleavage to the carbocation/triphenylphosphine pair [E+ PPh3]
(red pathway) or homolytic bond cleavage to the radical pair
[E•PPh3•+] (blue pathway). Both pairs can then either undergo
geminate recombination to the starting material or diffusional
separation, which results in the free benzhydryl cations E+ or
radicals E•. Only the UV/vis-absorbing species which escape the
geminate solvent cage (E+, E•, and PR3•+; bottom line of Scheme
1) have sufficient lifetimes (>10 ns) to be observed spectroscopi-
cally with a nanosecond laser flash setup.
irradiated with a 7-ns laser pulse from the fourth harmonic of a Nd/YAG
laser (λexc = 266 nm, 30−60 mJ/pulse, diameter ∼1 cm). A xenon lamp
was used as probe light for UV/vis detection (d = 1 cm). Transient
spectra were obtained as difference spectra from subsequent
determinations with and without laser pulse using an ICCD camera
with 10 ns gate width; i.e., the ns time scale spectra published in this
work show the average of the transient signals during the first 10 ns after
the laser pulse. To reduce noise, 4−16 such spectra were averaged,
whereby the solutions of the substrates were replaced completely
between subsequent laser pulses by using a flow cuvette.
Decay kinetics of the benzhydryl cations E+ were measured by
following the absorbances A at λmax; typically ≥64 individual decay
curves were averaged. The decay rate constants kobs (s−1) were obtained
from the averaged curve by least-squares fitting to the single-exponential
3.2. Effect of the Photo-electrofuge (i.e., Structure of
the Benzhydrylium Ion). Nanosecond Spectroscopy in
CH2Cl2. The transient spectra which we obtained by irradiation
obst
function A(t) = A0 e−k + C. The non-exponential decays were
evaluated with the software Gepasi.43
of 10−5−10−4 M solutions of E(1−20)-PPh3 BF4 in CH2Cl2
with a 7-ns laser pulse (266 nm, 30−60 mJ/pulse) are compiled
in section S3 of the Supporting Information; four characteristic
examples are shown in Figure 1. The transient spectra feature
+
−
Femtosecond UV/Vis Transient Absorption Measurements. The
employed broadband transient absorption spectrometer is described in
+
detail in ref 39. Solutions of E-PAr3 X− (A280 nm ≈ 0.1−0.3, ca. 4 × 10−4
to 6 × 10−3 M) were pumped through a flow cell of 120 μm or 1 mm
thickness and irradiated with 35-fs pulses (λexc = 280 nm, 200 nJ/pulse)
from the frequency-doubled output of a noncollinear optical parametric
amplifier (NOPA). The pulses were focused down to a diameter of
about 100 μm inside the sample. A CaF2 white light continuum spanning
three types of absorption bands: (i) broad bands with λmax
=
426−535 nm, which can be assigned to the cations E+ by
comparison with the previously reported spectra of benzhy-
drylium ions,40 and the cation-like reactivities (see below); (ii)
11482
dx.doi.org/10.1021/ja3017522 | J. Am. Chem. Soc. 2012, 134, 11481−11494

Journal of the American Chemical Society
Article
Table 1. Electrophilicity Parameters E of the Benzhydryl
Cations E(1-20)+ and Absorption Maxima λmax (nm) of
Benzhydryl Radicals E• and Benzhydryl Cations E+ in CH2Cl2
a
Electrophilicity parameters E of the benzhydryl cations E+; from ref
b
41a unless noted otherwise. Ratio of absorbances at λmax (E+) and at
Figure 1. Transient spectra obtained after irradiation (7-ns pulse, λexc
=
λmax (E•) obtained by laser flash photolysis (7-ns pulse, λexc = 266 nm)
266 nm, gate width = 10 ns) of CH2Cl2 solutions of benzhydryl
of E(1−20)-PPh3 BF4− in CH2Cl2. Due to the overlap with the PPh3
+
•+
+
triphenylphosphonium tetrafluoroborates: (a) E3-PPh3 BF4− (A266 nm
=
band, absorbances at λmax (E•) overestimate the amount of radicals
0.16), (b) E12-PPh3 BF4− (A266 nm = 0.49), (c) E17-PPh3 BF4− (A266 nm
+
+
c
d
present. No radicals detected. New or revised E parameters from
unpublished work. Approximate values.
+
= 0.90), (d) E20-PPh3 BF4− (A266 nm = 0.64).
e
sharp bands with λmax = 328−344 nm, which closely resemble the
published spectra of benzhydryl radicals E• in CH3CN;40 and
(iii) a shoulder at ca. 350−360 nm, which we assign to the
triphenylphosphine radical cation PPh3•+, in agreement with its
reported spectrum in CH2Cl2 solution (with λmax ≈ 330 nm).44
false color representation of the ps transient absorption data
+
obtained by irradiating E12-PPh3 BF4− in CH2Cl2 with a ∼35-fs
pulse (280 nm, 200 nJ/pulse): The wavelength is plotted on the
horizontal axis and the time after the laser pulse on the vertical
axis. Blue color indicates low absorbance and red color high
absorbance.
+
The photo-cleavage of the phosphonium ions E(1−9)-PPh3
in CH2Cl2 yields the stabilized benzhydrylium ions E(1−9)+
exclusively. When we irradiated solutions of E(10−20)-
The plot features three types of bands: (i) a broad absorption
band below 400 nm, which disappears in the first 30 ps, is
assigned to the excited state absorption (ESA) of the
phosphonium salt precursor; (ii) the band of the benzhydryl
cation E12+ (λmax ≈ 443 nm), which reaches a maximum within
the first 25 ps; and (iii) a small band of the benzhydryl radical
E12• (λmax ≈ 332 nm), which becomes visible after the decay of
the excited state of the phosphonium ion. The graph to the right
of the color plot shows the dynamics of the absorbance at λmax of
the carbocation E12+ (red) and at 332 nm (blue), λmax of the
radical E12• overlapping with the excited-state absorption in the
first tens of picoseconds.
+
−
PPh3 BF4 in CH2Cl2, the ratios of the absorbances of the
benzhydryl cations E+ and benzhydryl radicals E• decreased with
increasing electrophilicity E of the carbocations (Table 1). The
least stable carbocations in the series, E(18−20)+ are hardly
+
−
detectable after photolysis of E(18−20)-PPh3 BF4 , and the
radicals E(18−20)• are obtained almost exclusively (Figure 1d
and Figure S3 in the Supporting Information).
Picosecond Dynamics in CH2Cl2. The processes which lead to
the formation of E+ and E• are too fast to be followed with the
nanosecond laser flash photolysis instrument. A closer look at
these processes is provided by transient absorption measure-
ments with sub-100-fs time resolution which we performed for
selected benzhydryl triarylphosphonium salts. Figure 2 shows a
The intense short-lived (<0.1 ps) signal directly after the laser
pulse is a coherent artifact that is also observed in the pure
11483
dx.doi.org/10.1021/ja3017522 | J. Am. Chem. Soc. 2012, 134, 11481−11494

Journal of the American Chemical Society
Article
After the ESA has disappeared and the absorbance of E12+ has
reached its maximum, the population of E12+ decreases as a
result of the geminate recombination of E12+ with the photo-
leaving group PPh3; in part, the photo-fragments diffuse away
from each other and these E12+ persist on this time scale. Once
the band of the radical E12• is clearly developed, it does not show
noticeable dynamics. After 1.8 ns, we observe the spectrum
shown in the graph above the color plot (Figure 2), which is
essentially the same as the spectrum obtained by the 7-ns laser
pulse (Figure 1b).
The corresponding plot for the photolysis of the tetrafluoro-
+
−
substituted benzhydryl phosphonium ion E20-PPh3 BF4 in
CH2Cl2 is shown in Figure S4.1 in the Supporting Information.
The heterolysis of this precursor is much less effective and the
radical E20• is formed predominantly. In addition, the small
initial concentration of carbocations E20+ decays rapidly so that
only a very low concentration can be observed on the
nanosecond time scale (Figure 1d; also see Table 2).46
Figure 2. Transient absorptions observed after irradiating a 5.2 × 10−3
+
M solution of E12-PPh3 BF4− in CH2Cl2 by a 35-fs pulse (λexc = 280 nm,
A280 nm = 0.2, d = 120 μm). The graph above the color plot shows the
spectrum after 1.8 ns (black). The graph on the right shows the
dynamics of the absorbances at selected wavelengths: Absorbance of
benzhydryl cation E12+ (445 nm, red); and absorbance of the excited
state (ESA) of the phosphonium ion and the benzhydryl radical E12•
(332 nm, blue). The time scale is linear between −1 and +1 ps and
logarithmic above 1 ps.
Picosecond Dynamics in CH3CN. Figure S4.2 in the
Supporting Information illustrates that in CH3CN essentially
the same kind of photo-processes occur after irradiation of E12-
+
−
PPh3 BF4 as in CH2Cl2 (Figure 2). Figure 3 shows the time-
dependent quantum yields of the substituted benzhydryl cations
E+ during the first 1.6 ns after the excitation pulse; the numeric
values are listed in Table 2. It is evident that the quantum yields
of the initial heterolytic photo-cleavage, Φhet, decrease with
increasing electrophilicity of the generated benzhydryl cations.
At the same time, homolytic bond cleavage becomes more
favorable although the overall efficiency of bond cleavage
decreases (not shown). Due to the overlap of the bands of E•
solvent and will be ignored in the following.39 A discussion of the
absorption changes during the first 2 ps (shaded area) which
include relaxation, planarization, and solvation effects is beyond
the scope of this paper and is treated in detail elsewhere.45
The ESA disappears during the first 30 ps, accompanied by a
simultaneous increase of the absorbance of E12+ which suggests
that E12+ is formed by direct heterolysis of the excited precursor
salt. It does not, however, exclude the generation of the
benzhydryl cations E12+ by homolytic bond cleavage and
subsequent considerably faster single electron transfer (SET) in
the geminate radical pair (Scheme 1, dashed arrow).
•+
with the ESA and the PAr3 band we could not evaluate the
radical yields on the early picosecond time scale.
As illustrated by Figure 3, the concentrations of the benzhydryl
cations E+ decrease considerably during the first 300 ps after their
formation which is rationalized by the geminate recombination
with the photo-leaving group PPh3. Immediately after C−P bond
cleavage, the two photofragments are in close vicinity (ion pairs).
Table 2. Yields and Rate Constants Associated with the Dynamics of E+ after Irradiation of E-PAr3 BF4− with Ar = Ph or p-Cl-C6H4
+
a
(Bold) in CH3CN or CH2Cl2 with a 35-fs Laser Pulse (λexc = 280 nm)
b
c
d
e
f
g
E+
E(E+)
PAr3
solvent
Φhet (%)
Yrecomb (%)
Φfree (%)
krecomb (s−1
)
kesc (s−1
)
E8+
3.63
4.43
5.47
PPh3
PPh3
PPh3
CH3CN
CH3CN
CH3CN
CH2Cl2
CH3CN
CH2Cl2
CH3CN
CH3CN
CH3CN
CH3CN
CH3CN
CH2Cl2
CH3CN
34
30
24
11
25
21
25
25
19
21
27
23
18
9
1.9 × 109
2.5 × 109
2.7 × 109
1.1 × 109
7.9 × 108
7.2 × 109
7.5 × 109
8.3 × 109
4.9 × 109
3.0 × 109
E9+
E12+
P(p-Cl-C6H4)3
20
14
h
h
h
h
(∼16)
6−10
5−8
4−7
(∼10)
(6 × 108)
3.6 × 109
3.6 × 109
3.8 × 109
(5 × 109)
8.5 × 109
7.5 × 109
6.9 × 109
i
i
E14+
E17+
E18+
6.23
6.87
7.52
PPh3
29
32
35
36
42
4−7
3−5
3−4
5−7
i
i
PPh3
i
i
PPh3
i
i
P(p-Cl-C6H4)3
PPh3
8−12
8.9 × 109
6.6 × 109
1.6 × 1010
9.3 × 109
i
i
E20+
(8.02)
3−4
∼2
ij
j
ij
j
j
(∼1) ,
(29)
(≤1) ,
(2 × 109)
(5 × 109)
i
i
P(p-Cl-C6H4)3
8−12
36
5−8
8.8 × 109
1.6 × 1010
a
b
See section S5 in the Supporting Information for details. Electrophilicity parameters E of the benzhydryl cations E+; see Table 1 for references.
c
Quantum yield of heterolytic bond cleavage (including the possibility of initial homolytic bond cleavage and subsequent fast electron transfer).
d
e
Yield of geminate recombination of E+ with the phosphine PAr3. Overall quantum yield of free E+ (at ∼2 ns) after diffusional separation of the
f
g
photo-leaving group. First-order rate constant for the geminate recombination of E+ with PAr3. First-order rate constant for the diffusional
h
separation of E+ and PAr3. The different behavior of this photo-leaving group in the early photo-dissociation process in CH2Cl2 reduces the
i
accuracy of our fit and we give only approximate values for this system. To calculate the quantum yields, absorbance coefficients of (5.0−7.5) × 104
M−1 cm−1 were assumed for the benzhydryl cations E(14,17,18,20)+ in analogy to reported values for similar benzhydrylium ions.40 The values have
j
to be considered approximate because of the low absorbance of E20+.
11484
dx.doi.org/10.1021/ja3017522 | J. Am. Chem. Soc. 2012, 134, 11481−11494

Journal of the American Chemical Society
Article
Figure 4. Transient spectra obtained after irradiation (7-ns pulse, λexc
=
266 nm, gate width: 10 ns) of CH2Cl2 solutions of benzhydryl
+
−
triarylphosphonium tetrafluoroborates E18-PAr3 BF4 with different
photoleaving groups PAr3 = PPh3 (black, A266 nm = 1.0) and PAr3 = P(p-
Cl-C6H4)3 (orange, A266 nm = 1.0).
Figure 3. Time-dependent quantum yields Φ of E+ observed after
+
irradiation of E-PPh3 BF4− solutions in CH3CN with a 35-fs laser pulse
(λexc = 280 nm).
They can either undergo geminate recombination or the
fragments diffuse apart. After complete diffusional separation
of E+ from the photo-leaving group, bond formation is no longer
possible (more precisely: is too slow to be observable on this time
scale) and the absorbances of E+ reach a plateau (Figure 3). The
yields for the geminate recombination of the benzhydryl cations
E+ with the phosphine PPh3, Yrecomb, increase with the
electrophilicity E of the carbocations E+, and diminish the final
quantum yields of the diffusionally separated (free) benzhydryl
cations, Φfree (Table 2).
The rate constants listed in Table 2 for the geminate
recombination of E+ with PPh3, krecomb (s−1), and for the
diffusional separation of E+ from PPh3, kesc (s−1), can be derived
from Yrecomb = krecomb/(krecomb+kesc) and the observed rate
constants for the decrease of the benzhydrylium ions (see section
S5 in the Supporting Information for details). With increasing
electrophilicity E of the benzhydryl cations E+, the recombina-
tion rate constant krecomb increases steadily while kesc remains
almost constant at (7−9) × 109 s−1.
3.3. Effect of the Photo-nucleofuge (i.e., Triarylphos-
phine). It was already observed by Modro and co-workers that
the use of a more nucleophilic phosphine as photo-leaving group
decreased the amount of cation-derived photo-products in
photolyses of phosphonium salts.32 When we employed tris(4-
chlorophenyl)phosphine P(p-Cl-C6H4)3 as a photo-leaving
group instead of PPh3, the formation of benzhydryl cations E+
was considerably more efficient and even allowed us to generate
highly electrophilic benzhydrylium ions.
Two reasons might account for the increased carbocation
yields with P(p-Cl-C6H4)3 as photo-leaving group: First, the
oxidation potentials of the two phosphines (E0 = 1.06 V for
ox
PPh3 and 1.28 V for P(p-Cl-C6H4)3 in CH3CN)48 indicate a
higher thermodynamic preference of E+/PAr3 pairs over E•/
•+
PAr3 pairs in the case of P(p-Cl-C6H4)3 than in the case of
PPh3. Thus, the preference for the heterolytic over the homolytic
+
+
pathway should be larger for E-P(p-Cl-C6H4)3 than for E-PPh3 .
Furthermore, P(p-Cl-C6H4)3 is less nucleophilic than PPh3 (ΔN
= 1.75)37 and, therefore, allows more carbocations to undergo
diffusional separation instead of geminate recombination.
The data from the ultrafast spectroscopic measurements
illustrate that both effects contribute to the better overall
quantum yields Φfree when P(p-Cl-C6H4)3 is used as photo-
leaving group instead of PPh3 (Table 2, bold entries): For this
leaving group, the observed initial quantum yields of the
heterolytic bond cleavage, Φhet, are higher and the yields of the
geminate recombination, Yrecomb, are lower. While the differences
+
−
are small for the photolysis of E12-PAr3 BF4 in CH3CN, the
effects are more important in CH2Cl2 and crucial for the
generation of the most reactive benzhydryl cations (Table 2).
3.4. Effect of Solvent on the Picosecond Dynamics. The
overall quantum yields of the free carbocations, Φfree, are
considerably lower in CH2Cl2 than in CH3CN (Table 2), which
is a consequence of the decreased quantum yields for the
heterolytic bond cleavage, Φhet. The lower solvent nucleophil-
icity of CH2Cl2 compared to CH3CN only becomes relevant at
longer time scales (see below).
Figure 4 shows the transient spectra obtained after irradiation
of the benzhydryl triarylphosphonium tetrafluoroborates E18-
The rate constants for the cage escape, kesc, are of comparable
magnitude in CH2Cl2 and CH3CN for both photo-nucleofuges
PAr3 (Table 2). In contrast, the diffusional separation of E+ from
Cl− is very slow after the photolysis of E-Cl in CH2Cl2 due to the
Coulombic attraction between the charged photo-fragments,4
+
PAr3 BF4− with PAr3 = PPh3 (black curves) and PAr3 = P(p-Cl-
C6H4)3 (orange curves). Considerably higher concentrations of
the carbocations E18+ and lower amounts of the radicals E18•
were obtained when P(p-Cl-C6H4)3 was employed as photo-
+
which explains why photolyses of E-PAr3 give much higher
+
−
leaving group. Similarly, irradiation of E20-P(p-Cl-C6H4)3 BF4
yields of carbocations in CH2Cl2 than photolyses of E-Cl.
3.5. Effect of the Counterion in the Precursor
Phosphonium Salt. Transient Spectra in CH3CN and
CH2Cl2. At low phosphonium salt concentrations (∼1 × 10−4
gave higher yields of E20+ and lower yields of E20• than
irradiation of E20-PPh3 BF4 , but the absorbance of E20+ was
still too low (A < 0.04) to study its reaction rates on the
nanosecond time scale. The shoulders of the radical bands
(PAr3•+) are weaker and red-shifted to 350−380 nm when P(p-
Cl-C6H4)3 is used as photo-leaving group (Figure 4), in
agreement with the fact that the absorbance maxima of the
+
−
+
M), the association equilibria of E12-PPh3 X− in acetonitrile are
entirely on the side of the free (unpaired) ions.42 Since the
lifetime of the excited state is only a few ps, which is too short for
the diffusive approach of external X−, the photochemistry of E12-
•+
+
tris(4-chlorophenyl)phosphine radical cation P(p-Cl-C6H4)3
PPh3 is not affected by the counter-anion X− under these
47
•+
are slightly red-shifted compared to PPh3
.
conditions.
11485
dx.doi.org/10.1021/ja3017522 | J. Am. Chem. Soc. 2012, 134, 11481−11494

Journal of the American Chemical Society
Article
Accordingly, ∼1.2 × 10−4 M solutions of E12-PPh3 X− with
different counter-anions X− in CH3CN gave very similar
transient spectra upon irradiation with a 7-ns laser pulse (Figure
5a): Irrespective of the counter-anion X−, the predominant
obtained from precursors with X− = Cl− or Br− (Figure 5a) result
from the diffusion-controlled trapping of E12+ by the halide ions
(see below),7 which can already be noticed on this time scale
(first 10 ns).
+
In CH2Cl2, the phosphonium salts E12-PPh3 X− have a
+
considerably higher tendency to form ion pairs. At concen-
trations of 1 × 10−4 M, which we typically used in the
nanosecond laser flash photolysis experiments, ∼57% of the Cl−,
−
∼43% of the Br−, ∼40% of the BF4 , and roughly (10−30)% of
the SbF6− salts exist as ion pairs in CD2Cl2.42 When we irradiated
+
solutions of E12-PPh3 BF4− in CH2Cl2, we obtained mostly the
benzhydrylium ion E12+ (λmax ≈ 443 nm) together with small
amounts of the radical E12• (λmax ≈ 332 nm) (Figure 5b, black
+
curve). Irradiation of CH2Cl2 solutions of E12-PPh3 SbF6− gave
almost the same concentrations of E12+ and E12• as the
tetrafluoroborate precursor (Figure 5b, blue curve). In contrast,
the concentration ratio of E12+ and E12• was reversed when we
used the phosphonium bromide E12-PPh3 Br− as precursor
+
(Figure 5b, red curve). Irradiation of the phosphonium chloride
E12-PPh3+ Cl− gave an intermediate amount of E12• while the
concentration of E12+ was almost the same as with the
phosphonium bromide precursor (Figure 5b, green curve).
Transient spectra obtained by analogous experiments with E12-
PPh3+ BPh4− are difficult to interpret because of the overlap with
−
the absorbances of photoproducts derived from BPh4 and are
discussed in section S6 in the Supporting Information.
Mechanism. The reduced yield of carbocations E12+
obtained from the phosphonium halides in CH2Cl2 (Figure
5b) can in part be explained by the immediate combination of
E12+ with Br− or Cl− which are in close vicinity if they have been
generated from the paired phosphonium halides. However, the
increased yields of the radicals obtained from the phosphonium
halides cannot be explained by the mechanism in Scheme 1 and
subsequent reactions with the counterions, because Cl− and Br−
do not reduce the benzhydrylium ions in the dark. Scheme 1,
therefore, has to be extended as depicted in Scheme 2.
+
Figure 5. Transient spectra obtained by irradiation of E12-PPh3 X−
(A266 nm = 0.5, (1.0−1.2) × 10−4 M) with different counterions X− =
BF4− (black), SbF6− (blue), Br− (red) or Cl− (green) in CH3CN (a) and
CH2Cl2 (b) with a 7-ns laser pulse (λexc = 266 nm, gate width: 10 ns).
photoproduct was the benzhydrylium ion E12+ (λmax ≈ 436 nm)
together with small amounts of the radical E12• (λmax ≈ 329 nm).
Since the absorption coefficients of E12+ and E12• are similar,40
the absorbance ratios in Figure 5 directly translate to
concentration ratios. The slightly lower concentrations of E12+
Scheme 2. Generation of Benzhydryl Cations E+ and Benzhydryl Radicals E• by Photolysis of Phosphonium Salts E-PR3 X− (R =
+
Ph or p-Cl-C6H4): (a) Reactions of Unpaired Phosphonium Ions (Predominant Mechanism in CH3CN) and (b) Reactions of
a
Paired Phosphonium Ions (Predominant Mechanism in CH2Cl2)
a
b
For the sake of simplicity, the geminate recombination reactions for the radical pairs are not shown. Radical combination or electron transfer.4
11486
dx.doi.org/10.1021/ja3017522 | J. Am. Chem. Soc. 2012, 134, 11481−11494

Journal of the American Chemical Society
Article
+
Figure 6. Transient absorptions obtained after irradiating solutions of E12-PPh3 X− with different counterions X− in CH3CN (a−c) and CH2Cl2 (d−f)
by a 35-fs laser pulse (λexc = 280 nm). The graphs above the color plots show the spectra after 1.8 ns (black). The graphs on the right show the dynamics
of the absorbances at certain wavelengths: absorbance of benzhydryl cation E12+ (436 or 445 nm, red) and absorbance of the excited state (ESA) and the
benzhydryl radical E12• (329 or 333 nm, blue). (a) X− = SbF6− in CH3CN; (b) X− = Br− in CH3CN; (c) X− = Br− in CH3CN with 4.8 × 10−3 M added
NEt4+ Br−; (d) X− = SbF6− in CH2Cl2; (e) X− = Cl− in CH2Cl2; (f) X− = Br− in CH2Cl2. The color scales in the false color plots are comparable in Figures
(a-f), but the absorbances in the small graphs (spectra and dynamics) were scaled to the available space and cannot be compared directly. The time scale
is linear between −1 and +1 ps and logarithmic above 1 ps. For the reasons discussed in context of Figure 2, the absorption changes during the first 2 ps
(shaded area) are discussed elsewhere.45 Experimental conditions: (a,d−f) (5−7) × 10−3 M precursor (A280 nm = 0.2), d = 120 μm; (b,c) 4 × 10−4
precursor (A280 nm = 0.1), d = 1 mm.
M
The phosphonium precursors can exist as free phosphonium
ions or paired with the counter-anions. Like the unpaired
phosphonium ions (Scheme 2a), the ion pairs can undergo
heterolytic bond cleavage to the benzhydryl cations E+ (Scheme
2b, red pathway) or homolytic bond cleavage to the benzhydryl
radicals E• (Scheme 2b, blue pathway). If the counter-anion is
oxidizable, there is the additional possibility of a photo-electron
transfer (PET) in the excited phosphonium ion pair (Scheme 2b,
11487
dx.doi.org/10.1021/ja3017522 | J. Am. Chem. Soc. 2012, 134, 11481−11494

Journal of the American Chemical Society
Article
green pathway). Such a PET was already proposed by Griffin et
al.49 and further substantiated by Modro and co-workers who
suggested a mechanism similar to Scheme 2 for the photolysis of
arylmethyl phosphonium salts with oxidizable counterions.32,50
As expected for an electron transfer mechanism, the yields of the
accumulation of a “dark” intermediate with a relatively low
absorption coefficient which cannot be detected within the large
probe range from 290 to 700 nm. This intermediate might be the
phosphoranyl/chlorine radical pair [E12-PPh3 Cl•] (Scheme 2b,
•
green pathway), which can either dissociate to E12• and PPh3 or
+
radicals E12• obtained from E12-PPh3 X− increase with
undergo a back electron transfer to regenerate the phosphonium
decreasing oxidation potentials of the counterions X− (Br− <
chloride E12-PPh3 Cl−. After ∼800 ps, the “dark” intermediate is
+
Cl− ≪ BF4− and SbF6 ). Related to the degree of ion pairing of
completely consumed and the formation of E12• ceases.
−
the precursor salts in these solvents,42 Scheme 2a is the
predominant pathway in CH3CN and Scheme 2b predominates
in CH2Cl2.
Due to the lower oxidation potential of bromide, electron
transfer reactions from Br−, which generate Br•, are more
favorable than the corresponding reactions of Cl−. Thus, the PET
pathway yielding the radical E12• from the excited state is
extremely effective when X− = Br−. As a result, the ESA
disappears almost instantaneously and the band of E12• appears
within a few ps (Figure 6f). Accordingly, only a very small
Our results agree with those of Johnston, Scaiano, and co-
workers, who studied the photolyses of arylmethyl triphenyl-
phosphonium chlorides in CH3CN and other solvents under
conditions where ion-pairing is not negligible.33 They had
already noticed that the concentrations of transient arylmethyl
cations increased and the concentrations of radicals decreased
when inorganic salts of non-nucleophilic anions (e.g., LiClO4,
NaBF4) were added to the phosphonium chloride solutions,
because these anions replace the Cl− in the initial phosphonium
salt ion pairs. As expected, this effect is larger in less polar
solvents.33
amount of carbocation E12+ is generated from E12-PPh3 Br− in
+
CH2Cl2; again the decay of E12+ is very effective due to
combination with Br−. In contrast to the observations with the
chloride precursor, the radical band keeps rising with an observed
rate constant of kobs = 6.3 × 108 s−1 throughout the whole time
scale (Figure 6f, blue curve), i.e., the radical formation is slower
but more effective in the case of X− = Br−. This effect is also
explained by the lower oxidation potential of Br−: After
formation of the not observable phosphoranyl radical, homolytic
Picosecond Dynamics in CH3CN and CH2Cl2. The data from
the ultrafast measurements corroborate this interpretation.
Figure 6 shows the false color representations of the ps transient
•
cleavage of E12-PPh3 yields Ph2CH• (green pathway in Scheme
absorptions obtained after irradiation of E12-PPh3 X− with
2b). On the other hand, the concurrent back electron transfer
depends greatly on the reduction potential of X• and is much less
important with Br• than with Cl•. As the back electron transfer
decay pathway for the “dark” state is suppressed, this state
becomes longer-lived and benzhydryl radicals E12• keep forming
over the whole time scale (Figure 6f) in a more effective and
longer-ongoing51 process. Furthermore, many benzhydryl
radicals E12• survive due to the less important electron transfer
and radical combination reactions between E12• and Br•.
Ion-Pairing and UV/Vis Spectra of the Photo-generated
Benzhydryl Cations. It should be noted that the carbocations E+
which are obtained by the heterolytic cleavage from the paired
+
different counter-anions (X− = SbF6 +, Cl−, and Br−) in CH3CN
−
−
or CH2Cl2. The plots for E12-PPh3 SbF6 in CH3CN (Figure
6a) and CH2Cl2 (Figure 6d) are very similar to that of the
tetrafluoroborate precursor (Figure 2) and can be interpreted
analogously (see above). Likewise, the color plot obtained with 4
+
× 10−4 M E12-PPh3 Br− in CH3CN (Figure 6b) closely
+
−
resembles that of E12-PPh3 SbF6 (Figure 6a), because ion
pairing is negligible at these concentrations42 and the PET
mechanism depicted in Scheme 2b (green pathway) cannot
occur. In all these cases, E12+ is the predominant photo-product,
and only very small amounts of E12• are obtained.
precursor salts [E-PR3 X−] (Scheme 2b, red pathway) may
+
At larger precursor concentrations or in the presence of added
bromide, however, the association equilibrium of the precursor
phosphonium salt is shifted toward the ion pairs, and the PET
pathway becomes available also in CH3CN. Figure 6c shows the
remain paired with the negatively charged counterions during
escape of PR3 from the solvent cage [E+PR3X−]. Thus, if the
association equilibrium of the benzhydrylium salt E+X− is
favorable enough and X− is a weakly nucleophilic counter-ion
+
false color plot obtained after irradiation of E12-PPh3 Br− in the
presence of 4.8 × 10−3 M added NEt4+ Br−. We now observed a
significant amount of benzhydryl radicals E12• while the yield of
the benzhydryl cations E12+ decreased.
(e.g., SbF6 ), photolysis of [E-PR3 X−] yields long-lived ion pairs
−
+
[E+X−].
Figure 5b shows that the absorption maxima of the
carbocations in CH2Cl2 vary slightly with the counter-anions.
The absorption maxima λmax of the benzhydryl cations E12+
which were generated from the phosphonium halide precursors
are at slightly higher wavelengths (λmax ≈ 450 nm, Figure 5b, red
and green curves) than those of the benzhydrylium ions
generated from the phosphonium tetrafluoroborate or hexa-
fluoroantimonate precursors (λmax ≈ 443 and 445 nm, Figure 5b,
black and blue curves). It has previously been reported that the
absorption maxima of the paired benzhydrylium tetrachlorobo-
The false color representations of the transient absorption data
recorded after irradiation of E12-PPh3 X− with different counter-
+
anions (X− = SbF6 , Cl−, and Br−) in CH2Cl2 are shown in Figure
−
+
6d−f. As already discussed, the plot for E12-PPh3 SbF6− (Figure
6d) is similar to that observed in CH3CN. Irradiation of E12-
PPh3 Cl− gave the color plot shown in Figure 6e, which is an
+
intermediate case between the SbF6 and the Br− salts. At any
−
time, only a small amount of carbocations E12+ is present. In
addition, most of the initially generated E12+ decay during the
first 1.8 ns due to the combination reaction of E12+ with Cl−. The
decay of the ESA is not associated with an increase of the
carbocation absorbance, indicating that the excited state
disappears predominantly by the PET mechanism and not by
heterolytic bond cleavage. The dynamics at 332 nm is most
interesting (Figure 6e, blue curve), because the ESA decreases
within ∼30 ps, while the benzhydryl radicals E12• appear with a
marked delay (kobs = 3.8 × 109 s−1). The dent between the
decrease of the ESA and the formation of E12• implies the
−
rates [E(3−8)+BCl4 ] are at ∼2 nm shorter wavelengths than
those of the free ions.52 Thus, the lower λmax of the
−
benzhydrylium ions E12+ which were generated from BF4 or
SbF6− salts are in agreement with the presence of benzhydrylium
ion pairs. The same λmax as shown in Figure 5b (determined with
10 ns gate width) are also observed by the ultrafast measure-
ments after ∼1 ns and then remain constant during the whole
lifetime (μs time scale) of E12+ (see Figure S7 in the Supporting
Information). As the diffusional approach of external anions is
11488
dx.doi.org/10.1021/ja3017522 | J. Am. Chem. Soc. 2012, 134, 11481−11494

Journal of the American Chemical Society
Article
comparably slow, the paired benzhydrylium ions observed after
∼1 ns must originate from paired phosphonium salts.
The higher λmax for the benzhydryl cations E12+ obtained from
the halide precursors (Figure 5b, red and green curves) can be
explained by the fact that carbocations which originate from the
paired fraction of the phosphonium halide precursors are
immediately trapped by the halide anions. Thus, the carbocations
which we can observe spectrophotometrically on the >10 ns time
scale53 are only the free (unpaired) benzhydrylium ions E12+
which are obtained from the unpaired fraction of the
phosphonium halides. External halide ions subsequently
consume the unpaired benzhydrylium ions in a diffusion-
controlled reaction (see below) which does not affect λmax of
the remaining unpaired benzhydrylium ions but only reduces the
signal intensity.
In CH3CN, the precursor salts as well as the benzhydryl
cations are mostly unpaired in the concentration range employed
in our experiments, and we observe the unpaired carbocations
E12+ predominantly. Thus, the absorption bands of E12+ feature
identical absorption maxima (λmax ≈ 436 nm) irrespective of the
counterions in this solvent (Figure 5a).
Figure 7. Transient spectra obtained after irradiation of CH2Cl2
solutions of benzhydryl tris(p-chlorophenyl)phosphonium salts with
different counter-anions with a 7-ns laser pulse (λexc = 266 nm, gate
Counterion Effects in the Photochemistry of Other Onium
Salts. The counterion effects discussed in the preceding
paragraphs should also be relevant for the photochemistry of
other onium salts. For example, we have previously shown that
one can generate E12+ in CH2Cl2 by laser flash photolysis of the
+
−
width: 10 ns): (a) E18-P(p-Cl-C6H4)3 X− with X− = BF4 (black,
A266nm = 1.0), X− = SbF6− (blue, A266nm = 1.0) and X− = Br− (red, A266nm
+
= 1.0); (b) E20-P(p-Cl-C6H4)3 X− with X− = BF4− (black, A266nm = 1.0)
and X− = SbF6− (blue, A266 nm = 0.9).
−
quaternary ammonium tetrafluoroborate E12-DABCO+BF4
(DABCO = 1,4-diazabicyclo[2.2.2]octane) but not from the
corresponding quaternary ammonium bromide.14 Benzhydryl
trimethylammonium iodide has also been used as photo-base-
When we irradiated CH2Cl2 solutions of the phosphonium
hexafluoroantimonate E18-P(p-Cl-C6H4)3 SbF6 , however, the
+
−
intensity of the cation band was doubled compared with that
+ −
−
generator because irradiation of E12-NMe3 I yields trimethyl-
obtained from the corresponding BF4 salt, while that of the
amine but not the benzhydryl cation E12+ which would trap the
amine.22 To account for the formation of NMe3 and the absence
of E12+, Jensen and Hanson discarded the PET mechanism and
favored a photo-SN1 reaction in CH3CN and CH3OH which
involves photoheterolysis of the precursor with subsequent
trapping of the benzhydryl cations E12+ by the I− anions or the
nucleophilic solvent.22b Our results with the phosphonium
analogues suggest that the PET mechanism may well be a
relevant pathway for the generation of tertiary amines in less
polar solvents.
radical band remained virtually unchanged (Figure 7a, blue line).
Similarly, the absorbance of E20+ more than tripled when we
+
−
used the hexafluoroantimonate E20-P(p-Cl-C6H4)3 SbF6
(Figure 7b, blue curve) instead of the corresponding
tetrafluoroborate (Figure 7b, black curve), while the yield of
the benzhydryl radical E20• was unaffected. The combination of
the P(p-Cl-C6H4)3 photo-leaving group and the SbF6− counter-
ion hence finally allowed us to generate E20+ in sufficient
concentrations to study its kinetics with nucleophiles in CH2Cl2.
Similarly, we could also obtain the highly electrophilic 4,4′-
bis(trifluoromethyl)benzhydrylium ion E19+ from E19-P(p-Cl-
3.6. Laser Flash Photolytic Generation of Highly
Electrophilic Benzhydrylium Ions. The information on the
influence of photo-nucleofuges (PAr3) and counter-ions X− on
carbocation yields derived in the previous sections have
subsequently been used to generate highly reactive carbocations
in order to study their reactivities in bimolecular reactions on the
>10 ns time scale. For these investigations, we were restricted to
the solvent CH2Cl2, because CH3CN reacts fast with highly
electrophilic benzhydrylium ions such as E(14−20)+ (see
below). In section 3.3 we have already demonstrated that the
use of P(p-Cl-C6H4)3 as photo-nucleofuge gives higher yields of
carbocations than when PPh3 is employed. Figure 7a shows the
transient spectra obtained by irradiating solutions of the
+
C6H4)3 SbF6− (Figure S8 in the Supporting Information).
−
Apparently, the BF4 anions trap a significant portion of the
−
carbocations E(18−20)+ within the [E+BF4 ] ion pairs that are
generated by the laser pulse. This is not the case for the parent
benzhydryl cation E12+ which was obtained in similar
concentrations from the hexafluoroantimonate and the tetra-
fluoroborate precursor (Figure 5b, black and blue curves); on the
−
microsecond time scale we also see a faster decay with the BF4
−
counterion than with SbF6 (see below). The trapping of the
carbocation by BF4− within the ion pair becomes less efficient as
the electrophilicity of the carbocations is reduced. The higher
−
−
reactivity of BF4 compared to SbF6 is in agreement with the
phosphonium salts E18-P(p-Cl-C6H4)3 X− with different
+
calculated enthalpies of fluoride abstractions in the gas phase,
counter-anions X− in CH2Cl2 with a 7-ns laser pulse (λexc
=
which are 151 kJ mol−1 more exothermic for BF4 than for
−
SbF6 .54 Furthermore, the photoinitiation efficiencies of onium
−
266 nm). As discussed in section 3.3, the phosphonium
+
−
tetrafluoroborate E18-P(p-Cl-C6H4)3 BF4 gave a moderate
yield of E18+ along with significant amounts of E18• (Figure 7a,
black curve). In view of the results presented in section 3.5 it is no
surprise that we could not observe any carbocation E18+ but only
the radical E18• when we irradiated the phosphonium bromide
salts in cationic polymerizations generally depend on the nature
−
of the anions and decrease in the order SbF6− > AsF6− > PF6
>
BF4 .26 This is usually explained by the different nucleophilicities
of the anions which are considered to be relevant for the stability
of the active center in the propagation step of cationic
polymerizations.26
−
+
E18-P(p-Cl-C6H4)3 Br− (Figure 7a, red curve).
11489
dx.doi.org/10.1021/ja3017522 | J. Am. Chem. Soc. 2012, 134, 11481−11494

Journal of the American Chemical Society
Article
3.7. Lifetimes of Benzhydrylium Ions in CH2Cl2, CH3CN,
and CF3CH2OH. Not only the yields of the carbocations E+ on
the ≤10 ns time scale but also their lifetimes on the μs time scale
depend greatly on the experimental conditions. Figure 8 shows
reaction which summarizes all (pseudo-)first-order reactions
which may occur (Figure 9). Details and more examples of such
Figure 9. Decay of the concentration of E12+ observed after irradiation
+
of E12-PPh3 SbF6− (A266 nm = 0.53, 1.03 × 10−4 M) in CH2Cl2 with a 7-
○
ns laser pulse: Superposition of experimental data ( ), an exponential fit
(black), and a fit calculated by Gepasi (red) according to the kinetic
model with variable [PPh3] shown in this Figure (kPPh = (1.31 0.003)
3
× 1010 M−1 s−1 and k0 = (6.49 0.02) × 103 s−1).
fits can be found in section S9 of the Supporting Information. As
expected, second-order rate constants kPPh ≈ 1 × 1010 M−1 s−1
3
are found for the combinations of E8+, E9+, and E12+ with PPh3,
indicating diffusion-controlled reactions. The obtained rate
constants k0 (s−1) for the first-order background decay reactions
agree with the trends discussed below.
Figure 8. Time-dependent absorbances of E12+ obtained after 7-ns
+
irradiation of E12-PPh3 X− (A266 nm = 0.5, (1.0−1.2) × 10−4 M) with
At typical concentrations of the photofragments E+ and PAr3
in our experiments (∼10−6−10−5 M), we find such non-
exponential decay kinetics for all systems in which the
benzhydryl cations E+ have lifetimes >10 μs. The recombination
reaction with the photo-leaving group thus sets a lower limit for
measuring pseudo-first-order kinetics of the benzhydryl cations
E+ with external nucleophiles: Only pseudo-first-order rate
constants larger than (1−5) × 105 s−1 can be determined reliably
by fitting the data to an exponential decay curve; otherwise the
decay kinetics will be dominated by the second-order reaction
with the photo-leaving group.
different counter-anions X− = BF4− (black), SbF6− (blue), Br− (red) or
Cl− (green) with a 7-ns laser pulse: (a) in CH3CN and (b) in CH2Cl2
(inset: enlarged decay curves for E12+ from precursors with halide
counterions).
the time-dependent absorbances of the parent benzhydrylium
ion E12+, which we observed when we generated this
carbocation from precursors E12-PPh3 X− with different
+
counter-anions (X− = BF4 , SbF6 , Br−, or Cl−) in CH3CN or
CH2Cl2. The lifetime of E12+ depends on the decay mechanism
of the carbocation, which can be (i) recombination with the
photo-leaving group PPh3, (ii) reaction with the counter-anions
X− of the phosphonium salt precursor, or (iii) reaction with the
solvent.
−
−
Reaction with the Counter-anion of the Precursor
+
Phosphonium Salt. When precursors E-PPh3 X− with halide
counter-anions were used for the generation of benzhydryl
cations E+, we observed exponential decays of the carbocations
which were significantly faster than the decays of carbocations
Recombination with the Photo-nucleofuge. A general
limitation of the laser flash photolysis technique is entailed by
the recombination reactions of the carbocations with the free
(diffusionally separated) photo-nucleofuges. Photo-nucleofuges
−
−
generated from phosphonium salts with X− = BF4 or SbF6
(Figure 8). Halide ions undergo diffusion-controlled reactions
with reactive carbocations (E > −2) in aprotic solvents7 and the
reactions follow pseudo-first-order kinetics since [E+] ≪ [X−]
(e.g., Hal−, NR3, PR3, RCO2 ) typically undergo diffusion-
−
+
(only a small fraction of Ar2CH-PPh3 X− is cleaved to the
controlled recombination reactions with highly electrophilic
carbocations (E > 0) in solvents of low nucleophilicity.55
Exceptions to this rule are anionic photo-leaving groups in
fluorinated alcohols which stabilize anions very well (e.g., acetate
or p-cyanophenolate in CF3CH2OH).8 In our case, the
triarylphosphines (N ≥ 12.58, sN = 0.65)37 undergo diffusion-
controlled or almost diffusion-controlled reactions with the
benzhydrylium ions E(1−20)+.
carbocations). For example, irradiation of 1.2 × 10−4 M solutions
of E12-PPh3 X− with X− = Br− or Cl− in CH3CN (Figure 8a, red
+
and green curves) gave pseudo-first-order rate constants kobs ≈ 6
× 106 s−1 for the decay of E12+. Irradiation of 1.0 × 10−4
M
solutions of the same precursors in CH2Cl2 (Figure 8b, red and
green curves) yielded similar rate constants (kobs ≈ 7 × 106 s−1).
These values correspond to second-order rate constants of k2 ≈ 5
× 1010 M−1 s−1 (CH3CN) and k2 ≈ 7 × 1010 M−1 s−1 (CH2Cl2)
for the diffusion-controlled reactions of E12+ with Br− and Cl−.
An almost mono-exponential decay of E12+ was also observed
The blue curve in Figure 8b shows that the recombination
reaction with PPh3 can be observed when E12+ is generated by
irradiation of E12-PPh3 SbF6− in CH2Cl2. Since E12+ and PPh3
+
in CH2Cl2 when we irradiated E12-PPh3 X− with X− = BF4
+
−
are generated in equimolar amounts by the laser pulse, the
observed decay of the absorbance is not mono-exponential.
Using the software Gepasi,43 we could fit the observed decay to a
kinetic model which takes into account the second-order
recombination reaction with PPh3 and a general first-order
(Figure 8b, black curve, kobs ≈ 2.6 × 105 s−1). This decay is much
slower than the decays for X− = Cl− or Br− but significantly faster
than the non-exponential decay observed for X− = SbF6− (Figure
−
8b, blue curve) indicating that E12+ reacts with BF4 . Similar
11490
dx.doi.org/10.1021/ja3017522 | J. Am. Chem. Soc. 2012, 134, 11481−11494

Journal of the American Chemical Society
Article
Table 3. First-Order Rate Constants k1 (s−1) for the Decay of
Benzhydryl Cations E+ in CH3CN and 2,2,2-Trifluoroethanol
(TFE) at 20 °C
mono-exponential decays were found for the benzhydrylium ions
E(10−17)+ which were generated from the phosphonium
+
−
tetrafluoroborates E(10−17)-PAr3 BF4 ; the decay rate con-
stants increase with the electrophilicities E of the carbocations
(see section S10 in the Supporting Information for details). As
the yields of the more reactive benzhydryl cations E(18−20)+
−
obtained from the BF4 salt precursors were lower than those
−
from the SbF6 salt precursors (see above), one can conclude
that the reactions of E(18−20)+ with BF4− already occur on time
scales <10 ns. For the highly reactive carbocations E(18−20)+
we also observed mono-exponential decays when they were
generated from the corresponding hexafluoroantimonate salts
+
−
E(18−20)-P(p-Cl-C6H4)3 SbF6 . As the background decay
rates k0 of the carbocations E(18−20)+ observed on the >10
ns time scale after irradiation of E(18−20)-P(p-Cl-C6H4)3 X−
+
with X− = BF4− and SbF6− also become similar (Figure S10.2 in
the Supporting Information), we assume that now the reactions
of E+ with solvent impurities such as residual water in CH2Cl2 are
dominating.
a
Electrophilicity parameters E of the benzhydryl cations E+; see Table
1 for references. Photolysis of E-Cl in CH3CN.40 Photolysis of E-
b
c
d
e
+
−
PPh3 BF4 , this work. Not determined. Photolysis of E12-Cl in
CH3CN gave a value of 2.5 × 106 s−1.40 Photolysis of benzhydryl p-
f
We have already discussed in section 3.5 that the
cyanophenyl ether in TFE gave a value of 3.2 × 106 s−1.8
−
benzhydrylium tetrafluoroborates E+BF4 predominantly exist
as ion pairs in CH2Cl2 solutions (in the presence of ∼1 × 10−4
M
phosphonium tetrafluoroborate). Accordingly, a further increase
of the concentration of BF4− has little effect on the kinetics. Thus,
the decay rate constant of E18+ increased only slightly (factor
1.5) when we irradiated CH2Cl2 solutions of E18-P(p-Cl-
carbocations were generated from E(18,20)-P(p-Cl-
+
−
C6H4)3 SbF6 (Figure S10.2 in the Supporting Information).
However, these values are probably due to impurities and do not
reflect the reactivity of CH2Cl2 (see above). They just represent
an upper limit for the nucleophilic reactivity of CH2Cl2. Anyway,
the lifetimes of the benzhydrylium ions in highly purified CH2Cl2
(see Experimental Section) are much longer than in anhydrous
CH3CN and TFE and allow us to study the electrophilic
reactivities of E(14−20)+ toward a variety of nucleophiles.
3.8. Counter-ion Effects on Bimolecular Reactions. As
discussed in section 3.5, the usual assumption that only free ions
are observed in nanosecond laser flash photolysis experiments1a,b
does not hold when the carbocations are generated by photolysis
of certain onium salts with low-nucleophilicity counter-ions (e.g.,
C6H4)3 BF4− in the presence of 1.4 × 10−2 M KBF4/18-crown-6.
+
−
The high concentration of BF4 ions reduced the initial
absorbance of E18+ by less than 30%, i.e., the effect is much
smaller than when exchanging 5.7 × 10−5 M SbF6− for the same
concentration of BF4− (Figure 7a).
Reactions with the Solvent. In CH3CN or 2,2,2-trifluor-
oethanol (TFE), which are typical solvents for the laser-flash-
photolytic generation of carbocations,1 the characterization of
highly electrophilic carbocations is hampered by the nucleophil-
icity of these solvents. In CH3CN, for example, the parent
benzhydryl cation E12+ decays mono-exponentially with a first-
order rate constant of k1 = 2.52 × 106 s−1 when it is generated
−
−
BF4 , SbF6 ). Consequently, the question arises whether
bimolecular reactions of the photolytically generated carboca-
tions are affected by the nature of the counter-anion in the
precursor salt. Previous results already showed that the rate
constants for the reactions of moderately stabilized benzhy-
drylium ions such as methoxy- or methyl substituted
benzhydrylium ions E(3−9)+ with neutral nucleophiles like π-
nucleophiles3d,56,57 or hydride donors58 in CH2Cl2 are
independent of the nature of the counter-anion and the degree
of ion pairing. We now investigated the influence of the counter-
anion on the reactions of the considerably more electrophilic
benzhydrylium ions E12+ and E18+ with π-nucleophiles (Table
4).
+
from E12-PPh3 BF4− or SbF6− (Figure 8a), that is, it decays at
least 1 order of magnitude faster than in CH2Cl2. A slightly larger
value (k1 = 3.21 × 106 s−1) was observed for the decay of E12+ in
trifluoroethanol (TFE). These rate constants are independent of
the choice of the photo-leaving groups (Table 3). As solvation
effects have a relatively small influence on the reactivities of
carbocations,56 the ∼440-fold increase of the decay rate of E12+
in CH3CN and TFE compared with CH2Cl2 (k0 ≈ 6.5 × 103 s−1,
Figure 9) must result from reactions of E12+ with these
solvents.8,9,40
The most reactive benzhydryl cations of this series, E18+,
E19+, and E20+, decay too fast in CH3CN or TFE (τ < 10 ns) to
be observed with the nanosecond laser flash photolysis setup. It is
possible, however, to generate the acceptor-substituted benzhy-
drylium ions E(14−17)+ in these solvents and to follow the
exponential decays of their UV/vis absorbances. Since the first-
order rate constants for their reactions with CH3CN and TFE are
≥1 × 107 s−1 (Table 3), it is difficult to characterize the
electrophilic reactivities of these benzhydrylium ions toward
other nucleophiles in these solvents, because only nucleophiles
that react with rate constants close to the diffusion limit can
efficiently compete with these solvents.
When we generated the benzhydrylium ions E+ from different
precursors E-PAr3 X− with different counter-anions X− in the
+
presence of a large excess of π-nucleophiles, we observed
exponential decays of the UV/vis absorbances of the benzhydryl
cations E+ from which we obtained the pseudo-first-order rate
constants kobs (s−1). Plots of kobs versus the nucleophile
concentrations were linear in all cases, as exemplified in Figure
10 for the reaction of E12+ with allyltrimethylsilane.
The intercepts of these plots vary with the counter-anion X− of
the phosphonium salts and correspond to the rate constants k0
for the background decay reactions discussed in section 3.7. For
X− = Br− and Cl−, we find quite large intercepts of k0 ≈ 7 × 106
s−1 due to the diffusion-controlled reactions of E+ with the halide
Dichloromethane is considerably less nucleophilic than
CH3CN or TFE. First-order decay rate constants of ∼2 × 106
s−1 (E18+) and ∼3 × 106 s−1 (E20+) were measured when these
−
anions. For X− = BF4− and SbF6 , the intercepts are substantially
11491
dx.doi.org/10.1021/ja3017522 | J. Am. Chem. Soc. 2012, 134, 11481−11494

Journal of the American Chemical Society
Article
Table 4. Second-Order Rate Constants k2 (M−1 s−1) for the
Reactions of π-Nucleophiles with Benzhydryl Cations E+
carbocation (E = 7.52) in our series that could be obtained from
precursors with different counterions. When we generated E18+
+
−
+
Obtained from Different Precursors E-PAr3 X− in CH2Cl2 at
from either E18-P(p-Cl-C6H4)3 BF4 or E18-P(p-Cl-
+
−
C6H4)3 SbF6 , we again measured the same rate constants
within experimental error (Table 4). Moreover, we also obtained
20 °C
precursor E-PAr3 X−
+
+
−
the same rate constant when we used E18-PPh3 BF4 as
precursor, which shows that the photo-leaving group does not
have any effect on the carbocations’ reactivities either. As
discussed in section 3.6, we could not generate E18+ from the
phosphonium bromide precursor in CH2Cl2 (radical formation),
and therefore we cannot compare the reactivities of free and
paired carbocations in this case. The counterion independence of
the experimental rate constants for carbocation alkene
combination reactions implies that ion pairing stabilizes the
transition states to about the same extent as the reactant
carbocations.
E+
E12+
PAr3
X−
k2/M−1 s−1
Reaction with allyltrimethylsilane
−
PPh3
PPh3
PPh3
PPh3
BF4
SbF6
Br−
Cl−
1.60 × 107
1.43 × 107
1.42 × 107
1.53 × 107
−
E18+
Reaction with 2,3-dimethyl-1-butene
−
−
PPh3
BF4
BF4
8.22 × 107
8.24 × 107
8.27 × 107
P(p-Cl-C6H4)3
P(p-Cl-C6H4)3
−
SbF6
4. CONCLUSION
The efficiencies of the photo-generation of benzhydrylium ions
+
and benzhydryl radicals from phosphonium salts E-PAr3 X−
depend not only on the photo-electrofuge (E+) and the photo-
leaving group PAr3, but also on the counter-ion X−, the solvent,
and the concentration of the precursor molecules. Depending on
the reaction conditions, benzhydryl radicals E• or benzhydryl
cations E+ may be obtained almost exclusively. The results
presented in this work should also be relevant for the
photochemistry of other onium salts. Spectroscopic investiga-
tions on the fs to ps time scale like those performed in this and
related work4,5,16,45 provide a complete microscopic under-
standing of the photo-generation and the dynamics of reactive
intermediates in the geminate solvent cage. With the knowledge
of phosphonium salt photochemistry acquired from the present
study, we can now select the proper precursor salts for the
efficient generation of highly reactive carbocations which are not
easily accessible by conventional methods. The method
described in this work will subsequently be used to characterize
the electrophilic reactivities of the acceptor-substituted benzhy-
drylium ions E(14−20)+ in CH2Cl2 at 20 °C, which provides a
further extension of our long-ranging electrophilicity scale.41
Figure 10. Plots of the pseudo-first-order rate constants kobs (s−1) for the
reactions of E12+ with allyltrimethylsilane in CH2Cl2 when E12+ was
generated by irradiation of 1.0 × 10−4 M solutions of the precursors E12-
−
PPh3X− with different counter-anions X− = BF4− (black squares), SbF6
(blue squares), Br− (red squares), or Cl− (green squares) against the
concentration of allyltrimethylsilane. The small graphs show the
ASSOCIATED CONTENT
absorbance decays of E12+ in presence of 5.4 × 10−2
M
■
−
allyltrimethylsilane (black curve, X− = BF4 ; red curve, X− = Br−).
S
* Supporting Information
Additional transient spectra, details of the experimental
procedures and data evaluation, details of the kinetic measure-
ments. This material is available free of charge via the Internet at
http://pubs.acs.org.
lower and their origin has been discussed above. The slopes of
the four plots are independent of the counter-anion and provide
the second-order rate constants k2 (M−1 s−1) for the reaction of
E12+ with allyltrimethylsilane listed in Table 4. We thus
measured the same rate constants within experimental error
AUTHOR INFORMATION
■
for the reactions of E12+ with allyltrimethylsilane when E12+ was
Corresponding Author
+
generated from different precursors E12-PPh3 X− with X−
=
Herbert.Mayr@cup.uni-muenchen.de
BF4 , SbF6 , Br−, or Cl− (Table 4). As discussed in section 3.5,
the benzhydryl cations obtained from precursors with halide ions
are the free (unpaired) cations because [E12+ Hal−] pairs
−
−
Notes
The authors declare no competing financial interest.
collapse to covalent E12-Hal in less than 10 ns. Since E12+BF4
−
ACKNOWLEDGMENTS
and E12+SbF6− are significantly paired, on the other hand, we can
conclude that paired and unpaired benzhydrylium ions E12+
react with the same rate constants in bimolecular reactions and
can be characterized by an anion-independent electrophilicity
parameter (E = 5.47).
■
We thank Prof. Shinjiro Kobayashi for setting up the nanosecond
laser flash working station, Dr. Armin Ofial, Dr. Igor Pugliesi,
Sebastian Thallmair, and Konstantin Troshin for helpful
discussions, Christoph Grill for early experimental work, and
the Deutsche Forschungsgemeinschaft (SFB 749) for financial
support.
The same situation has been observed for the reactions of 2,3-
dimethyl-1-butene with E18+, which is the most electrophilic
11492
dx.doi.org/10.1021/ja3017522 | J. Am. Chem. Soc. 2012, 134, 11481−11494

Journal of the American Chemical Society
Article
(19) (a) Kanzian, T.; Lakhdar, S.; Mayr, H. Angew. Chem. 2010, 121,
9717−9720; Angew. Chem., Int. Ed. 2010, 49, 9526−9529. (b) Streidl,
N.; Branzan, R.; Mayr, H. Eur. J. Org. Chem. 2010, 4205−4210.
(c) Lakhdar, S.; Ammer, J.; Mayr, H. Angew. Chem. 2011, 123, 10127−
10130; Angew. Chem., Int. Ed. 2011, 50, 9953−9956.
(20) For examples of carbocations generated by photolysis of
phosphonium precursors acting as initiating species in carbocationic
polymerizations, see: (a) Takata, T.; Takuma, K.; Endo, T. Makromol.
Chem., Rapid Commun. 1993, 14, 203−206. (b) Takuma, K.; Takata, T.;
Endo, T. J. Photopolym. Sci. Technol. 1993, 6, 67−74.
(21) Examples of photoacid generation mechanisms with heterolytic
cleavage of carbon−heteroatom bonds: (a) Pohlers, G.; Scaiano, J. C.;
Step, E.; Sinta, R. J. Am. Chem. Soc. 1999, 121, 6167−6175. (b) Sanrame,
C. N.; Brandao, M. S. B.; Coenjarts, C.; Scaiano, J. C.; Pohlers, G.;
Suzuki, Y.; Cameron, J. F. Photochem. Photobiol. Sci. 2004, 3, 1052−
1057.
REFERENCES
■
(1) Reviews: (a) McClelland, R. A. In Reactive Intermediate Chemistry;
Moss, R. A., Platz, M. S., Jones, M. J., Eds.; Wiley: Hoboken, NJ, 2004, p
3−40. (b) McClelland, R. A. Tetrahedron 1996, 52, 6823−6858.
(c) Fleming, S. A.; Pincock, J. A. Mol. Supramol. Photochem. 1999, 3,
211−281. (d) Das, P. K. Chem. Rev. 1993, 93, 119−144.
(2) McClelland, R. A.; Chan, C.; Cozens, F. L.; Modro, A.; Steenken, S.
Angew. Chem. 1991, 103, 1389−1391; Angew. Chem., Int. Ed. 1991, 30,
1337−1339.
(3) (a) Bartl, J.; Steenken, S.; Mayr, H. J. Am. Chem. Soc. 1991, 113,
7710−7716. (b) Johnston, L. J.; Kwong, P.; Shelemay, A.; Lee-Ruff, E. J.
Am. Chem. Soc. 1993, 115, 1664−1669. (c) Kobayashi, S.; Hori, Y.;
Hasako, T.; Koga, K.-i.; Yamataka, H. J. Org. Chem. 1996, 61, 5274−
5279. (d) Mayr, H.; Schimmel, H.; Kobayashi, S.; Kotani, M.;
Prabakaran, T. R.; Sipos, L.; Faust, R. Macromolecules 2002, 35,
4611−4615. (e) Phan, B. T.; Nolte, C.; Kobayashi, S.; Ofial, A. R.; Mayr,
H. J. Am. Chem. Soc. 2009, 131, 11392−11401.
(22) Examples for the photogeneration of tertiary amines from
benzhydryl derivatives: (a) Hanson, J. E.; Jensen, K. H.; Gargiuolo, N.;
Motta, D.; Pingor, D. A.; Novembre, A. E.; Mixon, D. A.; Kometani, J.
M.; Knurek, C. In Microelectronics Technology. Polymers for Advanced
Imaging and Packaging; Reichmanis, E., Ober, C. K., MacDonald, S. A.,
Iwayanagi, T., Nishikubo, T., Eds.; ACS Symposium Series 614;
American Chemical Society: Washington, DC, 1995. (b) Jensen, K. H.;
Hanson, J. E. Chem. Mater. 2002, 14, 918−923.
(4) Sailer, C. F.; Fingerhut, B. P.; Thallmair, S.; Nolte, C.; Ammer, J.;
Mayr, H.; de Vivie-Riedle, R.; Pugliesi, I.; Riedle, E. J. Am. Chem. Soc.
2012, submitted.
(5) Sailer, C. F.; Fingerhut, B. P.; Ammer, J.; Nolte, C.; Pugliesi, I.;
Mayr, H.; de Vivie-Riedle, R.; Riedle, E. In Ultrafast Phenomena XVII;
Chergui, M., Jonas, D., Riedle, E., Schoenlein, R. W., Taylor, A., Eds.;
Oxford University Press: New York, 2011; pp 427−429.
(6) (a) McClelland, R. A.; Banait, N.; Steenken, S. J. Am. Chem. Soc.
1986, 108, 7023−7027. (b) McClelland, R. A.; Kanagasabapathy, V. M.;
Banait, N. S.; Steenken, S. J. Am. Chem. Soc. 1989, 111, 3966−3972.
(c) McClelland, R. A.; Kanagasabapathy, V. M.; Banait, N. S.; Steenken,
S. J. Am. Chem. Soc. 1991, 113, 1009−1014. (d) McClelland, R. A.;
Kanagasabapathy, V. M.; Banait, N.; Steenken, S. J. Am. Chem. Soc. 1992,
114, 1816−1823. (e) Van Pham, T.; McClelland, R. A. Can. J. Chem.
2001, 79, 1887−1897.
(23) Examples for the photogeneration of PPh3 from quaternary
̈
phosphonium salts: (a) Onen, A.; Arsu, N.; Yagci, Y. Angew. Makromol.
Chem. 1999, 264, 56−59. (b) Kasapoglu, F.; Aydin, M.; Arsu, N.; Yagci,
Y. J. Photochem. Photobiol., A 2003, 159, 151−159.
(24) For reviews about photoinitiators, see refs 25 and 26 and the
following: (a) Yagci, Y.; Reetz, I. Prog. Polym. Sci. 1998, 23, 1485−1538.
(b) Yagci, Y.; Durmaz, Y. Y.; Aydogan, B. Chem. Rec. 2007, 7, 78−90.
́ ́
(c) Allonas, X.; Croutxe-Barghorn, C.; Fouassier, J.-P.; Lalevee, J.;
Malval, J.-P.; Morlet-Savary, F. In Lasers in Chemistry; Lackner, M., Ed.;
Wiley-VCH: Weinheim, 2008; Vol. 2, pp 1001−1027. (d) Suyama, K.;
Shirai, M. Prog. Polym. Sci. 2009, 34, 194−209. (e) Yagci, Y.; Jockusch,
S.; Turro, N., J. Macromolecules 2010, 43, 6245−6260.
(7) Minegishi, S.; Loos, R.; Kobayashi, S.; Mayr, H. J. Am. Chem. Soc.
2005, 127, 2641−2649.
(8) McClelland, R. A.; Kanagasabapathy, V. M.; Steenken, S. J. Am.
Chem. Soc. 1988, 110, 6913−6914.
(25) (a) Toba, Y. J. Photopolym. Sci. Technol. 2003, 16, 115−118.
(b) Crivello, J. V. J. Photopolym. Sci. Technol. 2008, 21, 493−497.
(26) (a) Fouassier, J.-P. In Photoinitiation, Photopolymerization, and
Photocuring: Fundamentals and Applications; Hanser: Munich, 1995; pp
102−144. (b) Lazauskaite, R.; Grazulevicius, J. V. In Handbook of
Photochemistry and Photobiology; Nalwa, H. S., Ed.; American Scientific
Publishers: Stevenson Ranch, CA, 2003; Vol. 2, pp 335−392.
(9) Minegishi, S.; Kobayashi, S.; Mayr, H. J. Am. Chem. Soc. 2004, 126,
5174−5181.
(10) (a) Hallett-Tapley, G.; Cozens, F. L.; Schepp, N. P. J. Phys. Org.
Chem. 2009, 22, 343−348. (b) Horn, M.; Mayr, H. Chem.Eur. J. 2010,
16, 7478−7487. (c) Horn, M.; Mayr, H. Eur. J. Org. Chem. 2011, 6470−
6475.
(11) Baidya, M.; Kobayashi, S.; Brotzel, F.; Schmidhammer, U.; Riedle,
E.; Mayr, H. Angew. Chem. 2007, 119, 6288−6292; Angew. Chem., Int.
Ed. 2007, 46, 6176−6179.
́
(27) (a) Miranda, M. A.; Perez-Prieto, J.; Font-Sanchis, E.; Scaiano, J.
C. Acc. Chem. Res. 2001, 34, 717−726. (b) Kropp, P. J. In CRC
Handbook of Organic Photochemistry and Photobiology, 2nd ed.;
Horspool, W., Lenci, F., Eds.; CRC Press: Boca Raton, 2004; pp 1-1−
1-32. (c) Kitamura, T. In CRC Handbook of Organic Photochemistry and
Photobiology, 2nd ed.; Horspool, W., Lenci, F., Eds.; CRC Press: Boca
Raton, 2004; pp 11-1−11-10. (d) Peters, K. S. Chem. Rev. 2007, 107,
859−873.
(12) (a) Cozens, F.; Li, J.; McClelland, R. A.; Steenken, S. Angew.
Chem. 1992, 104, 753−755; Angew. Chem., Int. Ed. 1992, 31, 743−745.
(b) Mladenova, G.; Chen, L.; Rodriquez, C. F.; Siu, K. W. M.; Johnston,
L. J.; Hopkinson, A. C.; Lee-Ruff, E. J. Org. Chem. 2001, 66, 1109−1114.
(c) Loos, R.; Kobayashi, S.; Mayr, H. J. Am. Chem. Soc. 2003, 125,
14126−14132. (d) Schaller, H. F.; Schmidhammer, U.; Riedle, E.; Mayr,
H. Chem.Eur. J. 2008, 14, 3866−3868.
(28) Pincock, J. A. Acc. Chem. Res. 1997, 30, 43−49.
(29) Dektar, J. L.; Hacker, N. P. J. Org. Chem. 1991, 56, 1838−1844.
(30) Dektar, J. L.; Hacker, N. P. J. Am. Chem. Soc. 1990, 112, 6004−
6015.
(13) See refs 14−18 for kinetic studies using quaternary
triarylphosphonium salts and refs 18 and 19 for kinetic studies with
other quaternary phosphonium salts as precursors.
(31) Appleton, D. C.; Bull, D. C.; Givens, R. S.; Lillis, V.; McKenna, J.;
McKenna, J. M.; Thackeray, S.; Walley, A. R. J. Chem. Soc., Perkin Trans.
2 1980, 77−82.
(14) Ammer, J.; Baidya, M.; Kobayashi, S.; Mayr, H. J. Phys. Org. Chem.
2010, 23, 1029−1035.
(15) Ammer, J.; Mayr, H. Macromolecules 2010, 43, 1719−1723.
(16) Sailer, C. F.; Singh, R. B.; Ammer, J.; Riedle, E.; Pugliesi, I. Chem.
Phys. Lett. 2011, 512, 60−65.
(32) Imrie, C.; Modro, T. A.; Rohwer, E. R.; Wagener, C. C. P. J. Org.
Chem. 1993, 58, 5643−5649.
(33) Alonso, E. O.; Johnston, L. J.; Scaiano, J. C.; Toscano, V. G. Can. J.
Chem. 1992, 70, 1784−1794.
(17) (a) Shi, L.; Horn, M.; Kobayashi, S.; Mayr, H. Chem.Eur. J.
2009, 15, 8533−8541. (b) Troshin, K.; Schindele, C.; Mayr, H. J. Org.
Chem. 2011, 76, 9391−9408.
(34) (a) Alonso, E. O.; Johnston, L. J.; Scaiano, J. C.; Toscano, V. G. J.
Am. Chem. Soc. 1990, 112, 1270−1271. (b) Imrie, C.; Modro, T. A.;
Wagener, C. C. P. J. Chem. Soc., Perkin Trans. 2 1994, 1379−1382.
(35) For general reviews, see refs 1 and 26. Similar photocleavage
mechanisms were previously discussed for halides (refs 4 and 27),
carboxylates (ref 28), halonium salts (ref 29), sulfonium salts (ref 30),
ammonium salts (ref 31), and phosphonium salts (refs 32 and 33).
(18) (a) Baidya, M.; Kobayashi, S.; Mayr, H. J. Am. Chem. Soc. 2010,
132, 4796−4805. (b) Duan, X.-H.; Maji, B.; Mayr, H. Org. Biomol. Chem.
2011, 9, 8046−8050. (c) Nigst, T. A.; Ammer, J.; Mayr, H. Angew. Chem.
2011, 124, 1381−1385; Angew. Chem., Int. Ed. 2011, 51, 1353−1356.
(d) Nolte, C.; Ammer, J.; Mayr, H. J. Org. Chem. 2012, 77, 3325−3335.
11493
dx.doi.org/10.1021/ja3017522 | J. Am. Chem. Soc. 2012, 134, 11481−11494

Journal of the American Chemical Society
Article
(36) For equilibrium constants of stabilized benzhydryl cations with
triarylphosphines, see ref 37. In cases where highly stabilized
carbocations do not give stable triarylphosphonium salts, one may use
a more Lewis-basic phosphine such as P(n-Bu)3 as the photo-leaving
group (refs 18 and 19).
107, 99−110. (d) Burfeindt, J.; Patz, M.; Muller, M.; Mayr, H. J. Am.
̈
Chem. Soc. 1998, 120, 3629−3634.
(58) (a) Mayr, H.; Basso, N.; Hagen, G. J. Am. Chem. Soc. 1992, 114,
3060−3066. (b) Mayr, H.; Lang, G.; Ofial, A. R. J. Am. Chem. Soc. 2002,
124, 4076−4083.
(37) Kempf, B.; Mayr, H. Chem.Eur. J. 2005, 11, 917−927.
(38) For a review of the literature up to 1994, see: Dankowski, M. In
The chemistry of organophosphorus compounds; Hartley, F. R., Ed.; Wiley:
Chichester, 1994; Vol. 3, pp 325−343.
(39) Megerle, U.; Pugliesi, I.; Schriever, C.; Sailer, C. F.; Riedle, E. Appl.
Phys. B: Laser Opt. 2009, 96, 215−231.
(40) Bartl, J.; Steenken, S.; Mayr, H.; McClelland, R. A. J. Am. Chem.
Soc. 1990, 112, 6918−6928.
(41) (a) Mayr, H.; Bug, T.; Gotta, M. F.; Hering, N.; Irrgang, B.; Janker,
B.; Kempf, B.; Loos, R.; Ofial, A. R.; Remennikov, G.; Schimmel, H. J.
Am. Chem. Soc. 2001, 123, 9500−9512. (b) Mayr, H.; Kempf, B.; Ofial,
A. R. Acc. Chem. Res. 2003, 36, 66−77. (c) For a comprehensive
database of nucleophilicity and electrophilicity parameters, see: http://
www.cup.lmu.de/oc/mayr/DBintro.html.
(42) Syntheses, structural investigations, and ion-pairing phenomena
of benzhydryltriarylphosphonium salts will be reported in a subsequent
publication.
(43) (a) Mendes, P. Comput. Appl. Biosci. 1993, 9, 563−571.
(b) Mendes, P. Trends Biochem. Sci. 1997, 22, 361−363. (c) Mendes,
P.; Kell, D. B. Bioinformatics 1998, 14, 669−883. (d) Further
information about Gepasi: www.gepasi.org.
(44) Alfassi, Z. B.; Neta, P.; Beaver, B. J. Phys. Chem. A 1997, 101,
2153−2158.
(45) Fingerhut, B. P.; Sailer, C. F.; Ammer, J.; Riedle, E.; de Vivie-
Riedle, R. J. Phys. Chem. A 2012, DOI: 10.1021/jp300986t.
(46) Exclusive formation of E8+ is observed after irradiation of E8-
+
PPh3 BF4− in CH3CN; the plot for this system is shown in Figure S6 in
the Supporting Information of ref 16.
(47) (a) Nakamura, M.; Miki, M.; Majima, T. J. Chem. Soc., Perkin
Trans. 2 2000, 1447−1452. (b) Tojo, S.; Yasui, S.; Fujitsuka, M.;
Majima, T. J. Org. Chem. 2006, 71, 8227−8232.
(48) Fukuzumi, S.; Shimoosako, K.; Suenobu, T.; Watanabe, Y. J. Am.
Chem. Soc. 2003, 125, 9074−9082.
(49) Griffin, C. E.; Kaufman, M. L. Tetrahedron Lett. 1965, 12, 773−
775.
(50) Both groups have derived their conclusions from product studies
after extended irradiation of benzyl triphenylphosphonium salts. It
should be noted that product studies do not necessarily give the same
results as transient measurements. Multiple irradiation of the same
starting material molecules in preparative photolyses will distort the
product ratios if the geminate recombination of one or both types of
photofragment pair is a relevant process. Further complications arise if
the photoproducts (e.g., arylmethyl halides) can undergo subsequent
photolysis reactions.
(51) The observed rate constant kobs for formation of E12• is the sum
of the rate constants for phosphoranyl radical dissociation and back
electron transfer. The derivation of this relationship is analogous to the
case of the carbocation dynamics in the geminate solvent cage, cf. section
S5 in the Supporting Information.
(52) Schneider, R.; Mayr, H.; Plesch, P. H. Ber. Bunsen-Ges. 1987, 91,
1369−1374.
(53) The absorbance maximum of the short-lived (<300 ps) [E+ Cl−]
ion pairs is also blue-shifted compared to the free E+. The dependence of
λmax on the distance between E+ and Cl− is investigated in detail in ref 4.
(54) Krossing, I.; Raabe, I. Chem.Eur. J. 2004, 10, 5017−5030.
(55) For nucleophilicity parameters of common photo-nucleofuges
such as halides, carboxylates, tertiary amines, pyridines, tertiary
phosphines, and others, see ref 41c.
(56) Mayr, H.; Schneider, R.; Schade, C.; Bartl, J.; Bederke, R. J. Am.
Chem. Soc. 1990, 112, 4446−4454.
(57) (a) Mayr, H.; Schneider, R.; Irrgang, B.; Schade, C. J. Am. Chem.
Soc. 1990, 112, 4454−4459. (b) Hagen, G.; Mayr, H. J. Am. Chem. Soc.
1991, 113, 4954−4961. (c) Mayr, H.; Patz, M. Macromol. Symp. 1996,
11494
dx.doi.org/10.1021/ja3017522 | J. Am. Chem. Soc. 2012, 134, 11481−11494