Fingerprints of singlet and triplet phenyl cations

Article abstract of DOI:10.1002/ejoc.200700339

The photolyses of seven phenyl cation precursors in acetonitrile in the presence of anisole resulted in four distinct product patterns. These patterns are due to the chemoselective and regioselective chemistry of various phenyl cation isomers. This spin-selective chemistry provides a tool with which to fingerprint the singlet/triplet nature of any phenyl cation. Wiley-VCH Verlag GmbH & Co. KGaA, 2007.

Full text of DOI:10.1002/ejoc.200700339

FULL PAPER
DOI: 10.1002/ejoc.200700339
Fingerprints of Singlet and Triplet Phenyl Cations
Micha Slegt,[a] Hermen S. Overkleeft,[a] and Gerrit Lodder*[a]
Keywords: Photochemistry / Phenyl cations / Spin multiplicity / Regioselectivity / Chemoselectivity
The photolyses of seven phenyl cation precursors in acetoni- which to fingerprint the singlet/triplet nature of any phenyl
trile in the presence of anisole resulted in four distinct prod-
uct patterns. These patterns are due to the chemoselective
and regioselective chemistry of various phenyl cation iso-
mers. This spin-selective chemistry provides a tool with
cation.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2007)
Introduction
The ability to steer reactions along singlet or triplet path-
ways by electronic excitation has led to the development of
valuable chemo-, regio-, and stereoselective reactions.[1–4]
For example, triplet phenyl cations, generated by photolysis,
show different chemoselectivity toward nucleophiles[5] than
the photogenerated singlet parent phenyl cation.[6] Whereas
singlet phenyl cations are fairly unselective, triplet phenyl
cations display a selectivity for unsaturated compounds (π
nucleophiles) over compounds with lone pairs (n nucleo-
philes).[7,8] Practical implementations of these results are
found in the photolysis of fluoroquinolones, a major class
of antibiotics that react with cellular tissue through triplet
aryl cations,[9] and in the photochemistry of the organo-
phosphorous pesticide Fenthion, which degenerates via a
singlet phenyl cation intermediate,[10] both under the influ-
ence of (sun)light.
The singlet and triplet natures of phenyl cations, which
are among the most reactive intermediates known,[11] have
since long been subjects of intensive study.[12,13] It was not
until the parent phenyl cation was isolated in an argon ma-
trix that its singlet closed-shell ground state (I1, Figure 1)
was ascertained.[14] Isomers of this cation are possible. One
electron of the aromatic sextet may be transferred, with spin
retention, into the empty sp2 orbital, yielding a species with
a π5σ1 electronic configuration: the open-shell singlet
phenyl cation (I2). If the transfer occurs with spin inversion
the triplet phenyl cation (I3) is formed, which is an open-
shell species by nature.
Figure 1. Singlet and triplet phenyl cations.
Whether or not phenyl cations are actually generated
through solvolysis of their diazonium salt precursors – in
SN1 reactions – remains unclear.[17] This is not so if light is
used to trigger the loss of the N2 leaving group[5,6,18] In
the photolysis of (substituted) phenyldiazonium salts, both
singlet and triplet phenyl cations are generated.[5,6,7a,7b] An-
other photochemical route to phenyl cations is the photoly-
sis of 4-chloro- (or 4-fluoro-)anilines[7c,7d,8] or 4-chloro-
phenols.[7e] These precursors yield triplet phenyl cations.
Whether photogenerated phenyl cations are singlet or
triplet species depends on two factors: namely a) the singlet
or triplet nature of the reactive electronically excited
state,[19] and b) the relative stabilities of the singlet and trip-
let cations.[20] As depicted in Scheme 1, electronic excitation
converts the ground state of the photolabile compound into
a vibrationally excited singlet excited state. After vibrational
relaxation (VR) to the vibrationally relaxed first singlet ex-
cited state, expulsion of the leaving group yields a singlet
phenyl cation that can be trapped or can convert into a
triplet cation. If the rate of intersystem crossing (ISC) is
larger than the rate of bond fission, the singlet excited state
is converted into an isoenergetic triplet excited state. After
vibrational relaxation to the first triplet excited state, the
leaving group is cleaved off, forming a triplet phenyl cation.
This ion is either trapped or undergoes conversion to a sing-
let cation, if that is energetically feasible.
The textbook methods for thermal generation of phenyl
cations are β-decay of tritiated precursors[15] and solvolysis
of phenyldiazonium salts.[6] In all cases studied but one, the
tritium decay method yields singlet phenyl cations.[16]
[a] Leiden Institute of Chemistry, Gorlaeus Laboratories, Leiden
University,
P. O. Box 9502, 2300 RA Leiden, The Netherlands
E-mail: lodder@chem.leidenuniv.nl
5364
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 5364–5375

Fingerprints of Singlet and Triplet Phenyl Cations
Scheme 1. General scheme for photochemical singlet or triplet phenyl cation formation.
A time-honoured method to characterise cationic reac-
In all cases the major, or even exclusive, photoproducts
tive intermediates is the determination of the ortho/meta/ are the 2-methoxy-, 3-methoxy-and 4-methoxybiphenyls 1–
para ratios of aromatic substitution product mixtures. This 7o,m,p (Table 1). The ortho isomers of these Friedel–Crafts
method turns out to be useful for the type of spin-chemistry products are produced more abundantly than the para iso-
discussed above. In this paper we offer a method by which mers, while low to very low yields of the meta-methoxy iso-
to fingerprint the natures of photogenerated phenyl cations mers are obtained (Table 2). However, the o/m/p ratios vary
of different background and spin multiplicity based on their with the precursor. In addition to the C-alkylation products
different chemo- and regioselectivities toward the nucleo- 1–7o,m,p in three cases (Entries 1, 6 and 7 of Table 1) O-
phile methoxybenzene (= anisole) in acetonitrile. In the ac- alkylation products – the diphenyl ethers 1e, 6e and 7e –
companying paper the method is used to probe the singlet/ are formed, while in the photolysis of 1 and 6, F-alkylation
triplet natures of photogenerated naphthyl cations.[21]
(Schiemann) products 1s and 6s are also produced. The
Ritter reaction, alkylation on the nitrogen atom of acetoni-
trile, yields acetanilide 1r (= 7r) (Entries 1 and 7). Finally,
small amounts of 1–7h, where the leaving group is replaced
by hydrogen, are found in most reaction mixtures.
Results
Discussion
Compounds 1–7, all proposed or expected precursors of
phenyl cations, were photolysed in acetonitrile/anisole
(1:1, v/v) (Scheme 2). Excesses of anisole were used to allow
alkylation to take place in high yield. Acetonitrile was used
Phenyldiazonium Tetrafluoroborate (1)
The methoxybiphenyls 1o,m,p produced in the photolysis
as solvent because it is a poor nucleophile and a poor elec- of 1 in acetonitrile/anisole are proposed to result from at-
tron donor in comparison with, for instance, methanol. The tack of the photogenerated phenyl cation on the π-electrons
latter property precludes the formation of products derived of anisole, as depicted for the ortho product in Scheme 3.
from photoinduced electron transfer from the solvent to the The formation of diphenyl ether 1e occurs through attack
substrate.[22] The reactions were carried out at 2 °C; at that on the n-electrons of the oxygen atom of anisole.[23] The
–
temperature, during the time of irradiation, there is no sig- abstraction of F– from the counterion BF4 (yielding the
nificant thermolytic product formation (Ͻ5%). The results Schiemann product 1s) is also a cation-mediated reac-
of the photolyses are summarised in Table 1, while Table 2 tion.[5,24] Attack of the phenyl cation on the nitrogen atom
reports the ortho/meta/para ratios of the methoxybiphenyls of the acetonitrile solvent is responsible for the formation
1–7o,m,p.
of acetanilide 1r (Ritter product).[25]
Eur. J. Org. Chem. 2007, 5364–5375
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5365

M. Slegt, H. S. Overkleeft, G. Lodder
FULL PAPER
Scheme 2. Photolysis of compounds 1–7.
The ortho/meta/para ratio of 68:13:19 observed for phenyl radical was studied, and b) the wavelength depen-
1o,m,p produced in the photolysis of 1 is very similar to the dence of the photolysis reaction of 1 was investigated.
ratio observed in the radiolytical phenylation of anisole, but
a) Photolysis of iodobenzene, which gives homolytic
not similar to the ratios in the thermolyses of various phen- cleavage of the C–I bond,[33] in acetonitrile/anisole under
yldiazonium salts (Table 3). The similarity in o/m/p ratios in the same reactions conditions as used for 1–7 yields 1o,m,p
the photochemical and radiochemical phenylation of ani- in a ratio of 75:13:12. The product ratio of 1o,m,p and 1h
sole implies that the Friedel–Crafts products in both reac- is 2.1:1, quite different from the 5.3:1 ratio observed in the
tions are produced via the same product-forming reactive photolysis of 1.
intermediate: the (closed-shell) singlet phenyl cation I1.[26]
b) Irradiation of 1 at λexc = 300 nm and λexc = 350 nm
Clearly, β-decay of tritium, producing helium as leaving yields the biphenyls 1o,m,p in exactly the same o/m/p ratio
group, best resembles the interaction of a photon with a as upon irradiation at λexc = 254 nm. The relative amount
molecule and subsequent extrusion of nitrogen. The dif- of benzene increases slightly with the wavelength of exci-
ferent o/m/p ratios in the thermal solvolysis of phenyldia- tation: the 1o,m,p/1h ratio is 5.1:1 at λexc = 300 and 350 nm.
zonium salts in relation to the photochemical and radio-
Thus, we conclude that reactions involving the phenyl
chemical results support the view on the thermal solvolysis radical are not a major product-forming route in the pho-
of phenyldiazonium cations in water expressed in a recently tolysis of 1.
published paper, which argues against a discrete SN1
mechanism.[17]
4-Chloroaniline (2)
In theory, the phenylation products 1o,m,p might also
result from attack on anisole by a photogenerated phenyl
The photolysis of 2 in acetonitrile/anisole exclusively
radical. This radical would also be the source of benzene yields the methoxybiphenyls 2o,m,p in a ratio of 81:2:17. In
(1h) produced by hydrogen atom abstraction from the sol- relation to the o/m/p ratio observed for 1 (Table 2, Entry 1)
vent. A possible route for the production of the phenyl radi- the relative amount of meta product formed is remarkably
cal is photoinduced electron transfer from anisole to the low. The very limited yield of the meta-biphenyl product is
·
diazonium salt, which should yield PhN2 and next Ph·.[31,32] reminiscent of the o/m/p ratios obtained in the reactions of
This process would be expected to be most efficient upon the triplet species 2,6-dichloro-oxocyclohexadienyl carbene
irradiation in the charge-transfer bands of the anisole-dia- (73:0:27)[34] and tosylnitrene (71:2:27)[35] with anisole. The
zonium salt complex. To investigate the importance of a triplet carbene-type regioselectivity is in agreement with the
radical-mediated route towards methoxybiphenyl product proposal that the photolysis of 2 produces a phenyl cation
formation, a) the reactivity of an independently prepared I3 of triplet nature.[7c,7d,36]
5366
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 5364–5375

Fingerprints of Singlet and Triplet Phenyl Cations
Table 1. Products of the photolysis of compounds 1–7 in acetonitrile/anisole.[a]
[a] Yields are calculated as percentages of total product formation upon complete conversion after 90 min of photolysis. [b] Only 4% of
the precursor was converted after 90 min of photolysis. [c] Reaction carried out in trifluoroethanol/anisole (1:1). [d] The lower yield of
methoxybiphenyls is partly caused by the in-cage trapping of the leaving group iodobenzene, yielding iodobiphenyls. Also, only 70% of
the salt was converted.
Triplet phenyl cations possess an open-shell diradical Thus, attack of the triplet cation at the meta position of
structure, rather reminiscent of a triplet carbene. Triplet anisole rarely yields the meta biphenyl product. The chemo-
carbenes add to C=C double bonds in two radical-mediated selectivity observed with 2 is discussed later together with
steps to form cyclopropyl rings. We propose that the first those observed with 3, 4 and 5 (vide infra).
step of the biphenyl formation is the addition of a radical
to anisole (Scheme 4). This step produces a triplet (Wigner’s
rules[37]) distonic diradical cation. After spin-inversion, the
second step comprises closure towards a phenonium ion,
4-Chlorophenol (3)
which can take place in two directions. Formation of σ
The photolytic behaviour of 3 is solvent-dependent: the
complexes and proton loss lead to the biphenyls. Their iso- reaction occurs much more readily in trifluoroethanol than
mer pattern is thus determined in three steps. The first step in acetonitrile.[38,39] Therefore, 3 was not only irradiated in
takes place with the same regioselectivity as the attack of a acetonitrile/anisole, but also in trifluoroethanol/anisole. In
phenyl radical on anisole (vide supra). In the second step, both solvent systems the only products are 3o,m,p in a ratio
the direction of closure to the phenonium ions will presum- of 80:1:19. These results agree with the chemo- and regiose-
ably be preferentially towards the more stable species. Ac- lectivity of product formation observed in the photolysis of
cording to PM3 calculations the 1,2 complex is less stable 2 and are therefore due to the triplet phenyl cation mecha-
(by 5 kcalmol–1) than the 2,3 and 3,4 complexes, which are nism of Scheme 4. The product-forming intermediate is
degenerate in energy. The third step, opening of the phen- either the triplet phenyl cation or the deprotonated form
onium ions, should preferentially lead to the more stable, thereof: the triplet 4-oxocyclohexa-3,5-dienylidene. Inde-
ortho or para methoxy-substituted cyclohexadienyl cation. pendent generation of that species by photolysis of 4-diazo-
Eur. J. Org. Chem. 2007, 5364–5375
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5367

M. Slegt, H. S. Overkleeft, G. Lodder
FULL PAPER
Table 2. ortho/meta/para ratios of the methoxybiphenyl photoproducts.[a]
[a] Percentages are averages of the yields in three separate experiments; the maximum experimental error is Ϯ2%. [b] Photolysis in
acetonitrile/anisole (1:1). [c] Photolysis in trifluoroethanol/anisole (1:1).
quinone under the same reaction conditions as used for 3 the ratio reported for its thermal solvolysis in a mixture of
yields 3o,m,p in a 72:0:28 ratio.[40a] Therefore, 4-oxocy- acetonitrile, benzene and anisole, which yields 5o,m,p in a
clohexadienylidene is unlikely to be the product-forming in- ratio of 77.7:9.9:12.4.[41] This difference may be due to the
termediate. Presumably, unlike in water,[7e,38,39] the rate of occurrence of a pure SN1 mechanism in the photolysis and
the reaction of the triplet cation with the nucleophile ani- of an SN1/SN2 borderline mechanism in the thermolysis. No
sole exceeds its rate of deprotonation.
thermal solvolysis data for 4 are known.
The observed o/m/p ratios of the biphenyl photoproducts
of 4 and 5 are similar to the o/m/p ratio of the products of
the 4-nitrophenyl radical with anisole, which is known from
three separate sources to be 69:15:16, 68:16:16 and
4-Acetylphenyldiazonium (4) and 4-Nitrophenyldiazonium
(5) Tetrafluoroborates
The photolyses of 4 and 5 in acetonitrile/anisole yield the 72:12:16.[42–44] This suggests that the attack on anisole is
methoxybiphenyls 4o,m,p and 5o,m,p in ratios of 75:12:13 governed by the radical qualities of an intermediate. How-
and 75:13:12, respectively (Table 2, Entries 4 and 5). As in ever, this species is not the 4-acetylphenyl or the 4-ni-
the case of 2 and 3, no ether or Ritter products are formed trophenyl radical, formed by photoinduced electron trans-
(Table 1). Also no Schiemann product is formed. The o/m/ fer between the diazonium salts 4 and 5 and anisole. This
p ratio observed in the case of 5 is slightly different from was shown by performing the photoreactions at λexc = 300
5368
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 5364–5375

Fingerprints of Singlet and Triplet Phenyl Cations
Scheme 3. Mechanism of formation of photoproducts from 1.
Table 3. Reported o/m/p ratios in the reactions between the phenyl and 5 produces triplet phenyl cations[7a,7b] and that of 1 a
singlet phenyl cation.[6] Both the nitro and the acetyl sub-
cations from various precursors and anisole.
stituents are known to enhance intersystem crossing when
attached to an aromatic moiety. Upon excitation of a salt
to a singlet excited state, ISC to a triplet excited state takes
place (Scheme 1). The subsequent fragmentation will yield
a triplet phenyl cation. This phenyl cation is trapped before
it can spin-invert to a singlet state.
Reported ratio Precursor
Ref.
phenyldiazonium BF4 (hν)[a]
1,4-ditritiobenzene (β-decay)
this work
–
68:13:19
65:13:22
[26]
–
[23]
[23]
[27]
[28]
[29]
[30]
56.0:12.5:31.5
59:10:31
55.2:13.4:31.6
64.0:8.5:27.5
57.6:10.9:31.5
61.8:9.7:28.7
phenyldiazonium BF4 (∆)[b]
–
phenyldiazonium BF4 (∆)[c]
phenyldiazonium BF4 (∆)[d]
–
phenyldiazonium BF4 (∆)[e]
–
phenylazotriphenylmethane/TFA (∆)[f]
phenyldiazonium triflate (∆)[g]
Remarkably, the o/m/p ratios observed with 4 and 5 are
different from the o/m/p ratios observed with 2 and 3, even
though both series involve triplet phenyl cations. Scheme 5
explains why the triplet phenyl cations derived from 4 and
5 produce biphenyls with a different regioselectivity than
the triplet phenyl cations derived from 2 and 3. Other than
in the mechanism proposed for the photolysis of 2 and 3
in Scheme 4, the closure of the diradical cation toward the
phenonium ion here does not take place (Scheme 5). The
para EWG substituents are not able to harbour the positive
charge and thereby reduce the carbenoid character of the
reactive species involved. The methoxy biphenyls 4o,m,p
and 5o,m,p are formed after intramolecular electron trans-
fer and loss of a proton. This means that the o/m/p ratio of
the photoproducts is solely governed by the radical addition
step to anisole. This is also the case in the reaction of the
4-NO2 phenyl radical with anisole.[42–44] This explains why
the o/m/p biphenyl product ratio observed for 5 with anisole
(75:12:13) agrees with the ratio observed for the radical re-
action (69:15:16, 68:16:16 and 72:12:16). It also explains
[a] Molar ratio diazonium salt to anisole 1:460. [b] Molar ratio
diazonium salt to anisole 1:80. [c] In pure anisole. [d] Molar ratio
diazonium salt to anisole 1:14. [e] Molar ratio diazonium salt to
anisole 1:50. [f] Molar ratio starting material to anisole 1:25.
[g] Molar ratio diazonium salt to anisole 1:38.
and 350 nm, wavelengths closer to the charge-transfer ab-
sorption bands than λexc = 254 nm.[31] No variations in o/m/
p ratios and only slight increases in the amount of the radi-
cal-derived arenes 4h and 5h were found (4h/4o,m,p = 1:7.8
at λexc = 254 nm, 1:7.6 at λexc = 300 nm and 1:7.6 at λexc
350 nm) (5h/5o,m,p = 1:24 at λexc = 254 nm, 1:22 at λexc
300 nm and 1:19 at λexc = 350 nm).
The difference between the o/m/p ratios observed in the
photolysis of 4 and 5 and that seen in the photolysis of 1,
as well as the differences in product profiles (Table 1, En-
tries 4 and 5 vs. 1), reflect the difference in the natures of
=
=
the reactive intermediates responsible for product forma- why that ratio, as well as that observed for 4 (75:13:12), is
tion, in agreement with the proposal that irradiation of 4 the same as the ratio observed for the reaction of the parent
Eur. J. Org. Chem. 2007, 5364–5375
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5369

M. Slegt, H. S. Overkleeft, G. Lodder
FULL PAPER
Scheme 4. Mechanism of product formation upon attack of an EDG-substituted triplet phenyl cation (I3) on anisole (S = NH2, OH).
phenyl radical (produced by photolysis of iodobenzene, vide 4-(Diethylamino)phenyldiazonium Tetrafluoroborate (6)
supra) with anisole (75:13:12).
The EDG- and EWG-substituted triplet phenyl cations
The photolysis of 6 yields the methoxybiphenyls 6o,m,p
produced from 2, 3, 4 and 5 display remarkable chemoselec- in a ratio of 64:3:33 (Table 2, Entry 6). Also, O- (6e) and F-
tivity. The n-electrons of anisole, acetonitrile and BF4– have (6s) -alkylation products are formed (Table 1). Neither the
no appeal for the electron-deficient intermediate. Only reac- o/m/p ratio nor the product profile match that of the parent
tion with π-electrons occurs. An explanation for this behav- diazonium salt 1, and neither do they match the ratios and
iour is that the dispersed charge and electron density of the profiles observed in the photolysis of 2, 3, 4 and 5. In
triplet phenyl cations makes them soft electrophiles (Lewis particular, the twofold higher yield of Friedel–Crafts prod-
acids), which preferentially react with soft nucleophiles uct para 6p and the considerable yield of the O-alkylation
(Lewis bases), according to the HSAB principle.
product 6e stand out. Thus, the product-forming intermedi-
5370
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 5364–5375

Fingerprints of Singlet and Triplet Phenyl Cations
Scheme 5. ortho attack of an EWG-substituted triplet phenyl cation (I3) on anisole [S=C(=O)CH3, NO2].
ate is neither a closed-shell singlet phenyl cation nor a trip- closed-shell singlet (I1-H) and the triplet phenyl cation (I3-
let phenyl cation.
H).[13d] This is not the case for the 4-aminophenyl cation.
Presumably, diazonium salt 6 fragments in the singlet ex- According to CASSCF(8,8)/6-31G(d)//CASCF(8,8)/6-
cited state and gives the closed-shell singlet phenyl cation 31G(d) calculations, (nonplanar) I1-NH2 is only
I1, which equilibrates with the singlet open-shell phenyl cat- 0.4 kcalmol–1 more stable than I2-NH2. Triplet I3-NH2 is
ion I2 (Scheme 6). The singlet open-shell isomer of the par- about 16 kcalmol–1 more stable than I1-NH2 and I2-
ent phenyl cation (I2-H) has been studied computation- NH2.[40b] The polarity of the solvent system would be ex-
ally,[13b,13c] but was not considered a viable reactive interme- pected to govern the equilibrium between I1 and I2. In the
diate, because it lies much higher in energy than both the relatively apolar solvent system used, the open-shell species,
Scheme 6. Mechanism of the photolysis of 4-(diethylamino)phenyldiazonium salt 6.
Scheme 7. Reactions between the open-shell singlet phenyl cation I2 and anisole [S = N(Et)2]. The depicted 2,3 mode of attack serves as
an example; the 1,2 and 3,4 modes of attack also take place.
Eur. J. Org. Chem. 2007, 5364–5375
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5371

M. Slegt, H. S. Overkleeft, G. Lodder
FULL PAPER
with its more dispersed positive charge, may increase in im- Diphenyliodonium Tetrafluoroborate (7)
portance over the closed-shell species. The singlet open-
shell isomer I2 reacts with anisole with singlet carbene-like
Diphenyliodonium salts are often used in photolithogra-
behaviour before it can spin-convert into its more stable phy and in the photographic industry.[45] When subjected to
triplet manifold I3 (Scheme 6). That reaction yields phe- UV light, they are reported to react through phenyl cations,
nonium ions in one step (Scheme 7). After opening to the stemming from the cleavage of a C–I bond in the singlet
more stable σ complex and loss of a proton, the methoxybi- excited state, or through phenyl radicals, thought to origi-
phenyls 6o,m,p are produced. Thus, unlike the biphenyl for- nate from a C–I bond fission in the triplet excited state.[46]
mation from 2 and 3, and also unlike that from 4 and 5, The photolysis of 7 in acetonitrile/anisole yields the meth-
the regioselectivity of biphenyl formation from 6 is deter- oxybiphenyls 7o,m,p (= 1o,m,p) in a 68:12:19 ratio. Within
mined in two steps: a) the selectivity of the initial singlet experimental error this ratio is the same as the ratio seen in
carbene-like addition to anisole, and b) the formation of the the photolysis of phenyldiazonium tetrafluoroborate (1). In
more stable σ complex.
addition, the O- and N-alkylation products diphenyl ether
The excess of diphenyl ether 6e produced in the photore- (7e = 1e) and acetanilide (7r = 1r) are also formed, again
action of 6 can be interpreted by assuming a singlet car- as in the irradiation of 1.
bene-type insertion by the open-shell phenyl cation I2 in
the Cmethyl–O bond of anisole (Scheme 7). No Cmethyl–H as a function of the wavelength of excitation shows that at
insertion product is found.
λexc = 300 nm,[47] products 7o,m,p are formed in the o/m/p
Study of the photoreaction of 7 in the presence of anisole
The conversion of the closed-shell singlet into the open- ratio of 70:12:18, equal within experimental error to the
shell triplet phenyl cation (I1 Ǟ I3) is generally thought to ratio observed in the experiment at λexc = 254 nm
consist of one single step. The trapping of the open-shell sing- (69:12:19). Further, the relative amounts of phenyl radical-
let 4-(diethylamino)phenyl cation I2 disturbs this picture. The derived benzene [7h = (1h)] are the same at both wave-
cationic molecule undergoes considerable structural changes lengths (7h/7o,m,p = 1:4.1) and much less than the amount
in its conversion from the closed-shell singlet to an open-shell of benzene produced upon independent generation of the
triplet. Therefore, the displacement of the nuclei and the spin phenyl radical by photolysis of iodobenzene under the exact
inversion process may be separate reaction steps.
same reaction conditions (7h/7o,m,p = 1:2.1; vide supra).
Scheme 8. Proposed mechanism of the photolysis of diphenyliodonium tetrafluoroborate (7).
5372
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 5364–5375

Fingerprints of Singlet and Triplet Phenyl Cations
This indicates that a phenyl radical is not a major product- shell, a singlet open shell, an EDG-substituted triplet cat-
forming intermediate in the photolysis of 7.
ion, and an EWG-substituted triplet phenyl cation. The dif-
Both the methoxybiphenyl isomer ratio and the product ferent natures of the various phenyl cation isomers also
profile in the photolysis of 7 indicate that product forma- cause the differences in chemoselectivity. This spin-selective
tion largely takes place through a closed-shell singlet phenyl chemistry provides a tool to fingerprint the singlet/triplet
cation. This is quite surprising because the iodine atom nature of any phenyl cation.
would be expected to enhance the rate of ISC and produce
a triplet excited state and next a triplet cation after C–I tolysis of diphenyliodonium tetrafluoroborate indicates that
bond fission. the product-forming intermediate is a singlet species, de-
The fingerprint of the phenyl cation produced in the pho-
Several transient absorption studies of the photolysis of spite the presence of the heavy atom iodine. An alternative
7 on the nano- and picosecond timescales reveal the radical for the widely accepted mechanism of phenyliodonium salt
cation of iodobenzene as an intermediate.[48,49] Its presence photolysis (often applied in photolithography) is proposed.
has been used as support for the longstanding homolytic
mechanism for triplet state dissociation, yielding the iodo-
Experimental Section
benzene radical cation and a (not detected) phenyl radi-
cal.[46,50] The iodobenzene radical cation, however, may be a
secondary intermediate rather than a primary one. A triplet
phenyl cation/iodobenzene pair produced on a femtosecond
timescale may, by electron transfer, yield the iodobenzene
Materials: Starting materials 2, 3 and iodobenzene are commer-
cially available and were used as received. Diazonium salt 6 is com-
mercially available and was dissolved in acetonitrile and brought
to crystallisation by addition of n-pentane prior to use. Compounds
radical cation observed on the nano- and picosecond times- 1,[51] 4,[52] 5[51] and 7[53] were synthesised by the literature pro-
cedures. Anisole was distilled under argon to obtain GC purity.
Acetonitrile, trifluoroethanol and anisole were argon-purged prior
to their use in the photolysis experiments.
cales. The triplet phenyl cation also spin-inverts to the
lower-energy singlet closed-shell phenyl cation, the product-
forming intermediate in the photolysis of 7 (Scheme 8). It
is proposed that C–I+ bond cleavage primarily takes place Photochemistry: The photochemical reactions were carried out un-
der argon in quartz tubes with rubber seals. The starting materials
were dissolved at 0.05  in acetonitrile/anisole (1:1, v/v, 10 mL) or
trifluoroethanol/anisole (1:1, v/v, 10 mL). n-Decane was used as in-
ternal standard. The tubes were placed in a Rayonet Reactor
(RPR200) fitted with seven 254 nm, 300 nm or 350 nm lamps and
equipped with a magnetic stirrer. The photolyses of the salts 1, 4,
5, 6 and 7 were followed as a function of time by taking aliquots
(0.050 mL samples) and adding them to water (0.5 mL) + diethyl
ether (0.050 mL). The organic layers were analysed by GC and GC-
MS and the assignment of the structures was confirmed by coinjec-
in the triplet excited state.
The mechanism proposed in Scheme 8 is supported by
results for diphenyliodonium, diphenylbromonium and di-
phenylchloronium hexafluorophosphate photochemistry re-
ported in the literature.[46b] The o/m/p ratios of the halobi-
phenyls produced by the in-cage phenylation of halobenz-
enes are fingerprints for the nature of the reactive interme-
diates. The photolysis of the diphenylchloronium salt in
acetonitrile yields chlorobiphenyls in an o/m/p ratio of
48:31:21. Practically the same ratio is found for the di- tion of commercially available or independently prepared products.
After completion of the irradiations, the reaction mixtures were
poured into water (10 mL) and extracted twice with diethyl ether
(5 mL). The combined ether fractions were analysed by GC and
GC-MS. The photolyses of 2, 3 and iodobenzene were followed,
and the products analysed, by direct injection of aliquots on GC
and GC-MS.
phenylbromonium salt: 51:30:19. These ratios indicate a
closed-shell singlet phenyl cation as reactive intermedi-
ate.[26] However, for the diphenyliodonium salt the o/m/p
ratio is 73:13:15. Diphenyliodonium salt photolysis carried
out in the high-energy triplet-sensitising solvent acetone
produces an even more different o/m/p ratio of
70:5:24.[46a,46b] The low yield of meta product observed in
the case of the iodonium salt is reminiscent of the photoly-
ses of 2 and 3, which involve triplet phenyl cations. Even
without the aid of a sensitiser, diphenyliodonium salts pro-
duce triplet excited states with considerable efficiency.
Cleavage of a C–I bond in a triplet state produces a triplet
phenyl cation that is either trapped by in-cage iodobenzene
or spin-inverts to its singlet ground state.
Photoproducts: The products 1h (= 7h), 4h, 5h, 6h, 1s (= 7s), 1r
(= 7r), 1e (= 7e), 1o, and 1p were identified by means of GC, GC-
MS and coinjection with the aid of commercially obtained refer-
ence samples. Of all o/m/p biphenyl mixtures 2–6o,m,p the para and
ortho isomers were independently synthesised, from commercially
available starting materials, by parallel Suzuki cross-coupling.[54] p-
Iodoaniline, p-iodophenol, p-iodoacetophenone, p-iodonitroben-
zene and 1-(diethylamino)-4-iodobenzene (1 mmol each) were dis-
solved together under argon in THF (50 mL). As catalyst,
Pd(PPh3)4 was added (5ϫ10–2 mmol, 1 mol-% with respect to total
amount of iodo compounds). Cesium carbonate was used as base
in equimolar amount with respect to the total amount of iodo com-
pounds (5 mmol). After addition of 4-methoxyphenylboronic acid
(6 mmol), the temperature of the solution was raised to reflux and
kept at reflux overnight. The reaction mixture was filtered through
Celite and dried with calcium chloride. After evaporation of the
THF, the slurry was dissolved in dry diethyl ether (20 mL), filtered
and analysed by GC and GC-MS to confirm the identity of the
product. The five products were formed in yields varying from 20
Conclusions
The photolysis of seven photolabile compounds, known
to produce phenyl cations, in acetonitrile in the presence of
anisole yields methoxybiphenyls in four distinct o/m/p ra-
tios. The differences in regioselectivity are due to the dif-
ferent natures of the product-forming intermediates, re-
sulting in different mechanisms of reaction: a singlet closed to 90%.
Eur. J. Org. Chem. 2007, 5364–5375
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5373

M. Slegt, H. S. Overkleeft, G. Lodder
FULL PAPER
Boule, B. Szczepanik, K. Rotkiewicz, G. Grabner, J. Phys.
Chem. A 2000, 104, 9525–9534.
[9] a) E. Fasani, M. Mella, D. Caccia, S. Tassi, M. Fagnoni, A.
Albini, Chem. Commun. 1997, 1329–1330; b) E. Fasani, A. Pro-
fumo, A. Albini, Photochem. Photobiol. 1998, 68, 666–674; c)
E. Fasani, F. F. Barberis Negra, M. Mella, S. Monti, A. Albini,
J. Org. Chem. 1999, 64, 5388–5395.
The parallel Suzuki cross-coupling reaction was repeated with (2-
methoxyphenyl)boronic acid, which gave the five desired ortho iso-
mers in 20% to 90% yields. Schiemann product 6s was prepared
from 6 by the literature method.[55] Diphenyl ether 6e was prepared
(among other products) by heating a mixture of diazotised 4-(dieth-
ylamino)aniline (0.02 mol) and phenol (0.02 mol) in acetonitrile
(2 mL) to 50 °C for 1 h. Of the iodobiphenyls produced in trace
amounts in the irradiation of 7, the ortho and para products were
synthesised separately by diazotisation of 2- and 4-aminobiphenyl
in 50% water/concentrated aqueous HCl at melting ice tempera-
ture, treatment with KI followed by extraction with diethyl ether,
washing of the organic layer with a Na2S2O3 solution (5%) and
brine, and evaporation of the solvent.
[10] S. Torrisi, S. Sortino, J. Agric. Food Chem. 2004, 52, 5943–
5949.
[11] Dicoordinated Carbocations (Eds.: Z. Rappoport, P. Stang),
Wiley, Chichester, UK, 1997, chapter 2, pp. 10–97.
[12] a) R. Gleiter, R. Hoffmann, W. D. Stohrer, Chem. Ber. 1972,
105, 8–24; b) H. H. Jaffe, K. G. Koser, J. Org. Chem. 1975, 40,
3082–3084; c) J. D. Dill, P. von R. Schleyer, J. S. Binkley, R.
Seeger, J. A. Pople, E. Haselbach, J. Am. Chem. Soc. 1976, 98,
5428–5431.
Equipment: UV spectra were recorded at room temperature with a
[13] a) M. Kraus, S. Roszak, J. Mol. Struct. (THEOCHEM) 1994,
310, 155–160; b) J. Hrusak, D. Schroder, S. Iwata, J. Chem.
Phys. 1997, 106, 7541–7549; c) A. Nicolaides, D. M. Smith, F.
Jensen, L. Radom, J. Am. Chem. Soc. 1997,119, 8083–8088; d)
M. Aschi, J. N. Harvey, J. Chem. Soc. Perkin Trans. 2 1999,
1059–1062; e) J. N. Harvey, M. Aschi, H. Schwarz, W. Koch,
Theor. Chem. Acc. 1998, 99, 95–99.
double-beam Varian DMS 200 spectrophotometer, if applicable
with pure solvent in the reference cell. H NMR spectra were re-
1
corded on a JEOL 200 instrument, in CDCl3 as solvent. For analyt-
ical GC a Hewlett–Packard 6890 model was used, equipped with
a automatic injector, fitted with a CP-Sil5-CB column (25 m, ø =
0.25 mm, 1.2 µm) and with hydrogen as carrier gas. The Flame-
Ionisation Detector (FID) was calibrated with commercially avail-
able reference chemicals. HP Chemstation was used for the analysis
of the analytical data.
[14]
a) M. Winkler, W. Sander, Angew. Chem. Int. Ed. 2000, 39,
2014–2016; b) M. Winkler, W. Sander, J. Org. Chem. 2006, 71,
6357–6367.
[15] M. Speranza, Chem. Rev. 1993, 93, 2933–2980.
[16] A. Filippi, G. Lilla, G. Occhiucci, C. Sparapani, O. Ursini, M.
Speranza, J. Org. Chem. 1995, 60, 1250–1264.
Mass spectra were measured on a GC-MS set-up consisting of a
Hewlett–Packard 5890 series 2 model GC, equipped with a auto-
matic injector, fitted with a AT-5MS column (30 m, ø = 0.25 mm,
0.25 µm) with helium as carrier gas. The GC was coupled to a
Finnigan Mat SSQ 710 mass spectrometer, with employment of
electron impact as the ionisation method. The GC-MS data were
analysed with Xcaliber.
[17] B. R. Ussing, D. A. Singleton, J. Am. Chem. Soc. 2005, 127,
2888–2899.
[18] a) P. S. J. Canning, H. Maskill, K. McCrudden, B. Sexton, Bull.
Chem. Soc. Jpn. 2002, 75, 789–800; b) D. Smith, Z. B. Maksic,
H. Maskill, J. Chem. Soc. Perkin Trans. 2 2002, 906–913.
[19] M. Klessinger, J. Michl, Excited States and Photochemistry of
Organic Molecules, Wiley, Chichester, UK, 1995.
[20] a) J. D. Dill, P. von R. Schleyer, J. A. Pople, J. Am. Chem. Soc.
1977, 99, 1–8; b) K. K. Laali, G. Rasul, G. K. S. Prakash, G. A.
Olah, J. Org. Chem. 2002, 67, 2913–2918.
[21] M. Slegt, F. Minne, H. Zuilhof, H. S. Overkleeft, G. Lodder,
Eur. J. Org. Chem. 2007, 5353–5363; preceding paper.
[22] G. J. Kavarnos, Fundamentals of Photoinduced Electron Trans-
fer, VCH Publishers, New York, 1993.
[1] A. G. Griesbeck, M. Abe, S. Bondock, Acc. Chem. Res. 2004,
37, 919–928.
[2] a) M. Jones Jr, R. A. Moss, in Reactive Intermediate Chemistry
(Eds.: R. A. Moss, M. S. Platz, M. Jones Jr), John Wiley &
Sons, Hoboken, 2004, pp. 273–328; b) H. Tomioka, in Reactive
Intermediate Chemistry (Eds.: R. A. Moss, M. S. Platz, M.
Jones Jr), John Wiley & Sons, Hoboken, 2004, pp. 375–461; c)
B. B. Wright, M. S. Platz, J. Am. Chem. Soc. 1984, 106, 4175–
4180; d) P. B. Grasse, B.-E. Brauer, J. J. Zupancic, K. J. Kauf-
mann, G. B. Schuster, J. Am. Chem. Soc. 1983, 105, 6833–6845.
[3] a) M. S. Platz, in Reactive Intermediate Chemistry (Eds.: R. A.
Moss, M. S. Platz, M. Jones Jr), John Wiley & Sons, Hoboken,
2004, pp. 501–560; b) A. Albini, G. Bettinetti, G. Minoli, J.
Am. Chem. Soc. 1997, 119, 7308–7315.
[4] D. E. Falvey, in Reactive Intermediate Chemistry (Eds.: R. A.
Moss, M. S. Platz, M. Jones Jr), John Wiley & Sons, Hoboken,
2004, pp. 593–650.
[5] S. M. Gasper, C. Devadoss, G. B. Schuster, J. Am. Chem. Soc.
1995, 117, 5206–5211.
[6] a) W. Ando, in The Chemistry of Diazonium and Diazo Groups
(Ed.: S. Patai), Wiley, Chichester, 1978, part 1, pp. 341–487; b)
H. Zollinger, in The Chemistry of the Triple-bonded Functional
Groups (Eds.: S. Patai, Z. Rappoport), Wiley, Chichester, 1983,
part 1, pp. 603–669.
[7] a) F. Milanesi, M. Fagnoni, A. Albini, Chem. Commun. 2003,
216–217; b) S. Milanesi, M. Fagnoni, A. Albini, J. Org. Chem.
2005, 70, 603–610; c) M. Fagnoni, M. Mella, A. Albini, Org.
Lett. 1999, 1, 1299–1301; d) B. Guizzardi, M. Mella, M. Fag-
noni, M. Freccero, A. Albini, J. Org. Chem. 2001, 66, 6353–
6363; e) S. Protti, M. Fagnoni, M. Mella, A. Albini, J. Org.
Chem. 2004, 69, 3465–3473; f) M. Fagnoni, Lett. Org. Chem.
2006, 3, 253–259.
[23]
H. Eustathopoulos, J. Court, J. M. Bonnier, J. Chem. Soc. Per-
kin Trans. 2 1983, 803–807.
[24]
[25]
H. Suschitzky, Adv. Fluorine Chem. 1965, 4, 1–30.
J. J. Ritter, P. P. Minieri, J. Am. Chem. Soc. 1948, 70, 4045–
4048.
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
G. Angelini, Y. Keheyan, M. Speranza, Helv. Chim. Acta 1988,
71, 107–119.
R. A. Abramovitch, F. F. Gadallah, J. Chem. Soc. B, Phys. Org.
1968, 497–500.
M. Kobayashi, H. Minato, E. Yamada, N. Kobori, Bull. Chem.
Soc. Jpn. 1970, 43, 215–219.
N. Kamigata, M. Kobayashi, H. Minato, Bull. Chem. Soc. Jpn.
1972, 45, 1231–1235.
N. Kamigata, R. Hisida, H. Minato, M. Kobayashi, Bull.
Chem. Soc. Jpn. 1973, 46, 1016–1018.
D. Kosynkin, T. M. Bockman, J. K. Kochi, J. Am. Chem. Soc.
1997, 119, 4846–4855.
T. M. Bockman, D. Kosynkin, J. K. Kochi, J. Org. Chem. 1997,
62, 5811–5820.
a) D. R. Augood, J. I. G. Cadogan, D. H. Hey, G. H. Williams,
J. Chem. Soc. 1953, 3412–3417; b) R. L. Dannley, E. C. Gregg,
R. E. Phelps, C. B. Coleman, J. Am. Chem. Soc. 1954, 76, 445–
448.
[34]
[35]
M. J. S. Dewar, K. Narayanaswami, J. Am. Chem. Soc. 1964,
86, 2422–2427.
[8] a) P. Boule, K. Othmen, C. Richard, B. Szczepanik, G.
Grabner, Int. J. Photoenergy 1999, 1, 49; b) K. Othmen, P.
R. A. Abramovitch, J. Roy, V. Uma, Can. J. Chem. 1965, 43,
3407–3418.
5374
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 5364–5375

Fingerprints of Singlet and Triplet Phenyl Cations
[36] M. Hara, T. Sunagawa, Y. Oseki, T. Majima, J. Photochem.
Photobiol. A: Chem. 2004, 163, 153–158.
[47] The photoreaction of 7 at 350 nm, in our set-up, did not yield
reproducible product formation, probably because it is very in-
efficient at this excitation wavelength due to the lack of absorp-
tion.
[48] R. J. Devoe, M. R. V. Sahyun, N. Serpone, D. K. Sharma, Can.
J. Chem. 1987, 65, 2342–2349.
[49] a) E. Klemm, E. Riesenberg, A. Graness, Z. Chem. 1983, 23,
222; b) S. P. Pappas, J. H. Jilek, Photogr. Sci. Eng. 1979, 23,
140; c) S. P. Pappas, B. C. Pappas, L. R. Gatechair, W. Schna-
bel, J. Polym. Sci., Polym. Chem. Ed. 1984, 22, 69–76; d) S. P.
Pappas, L. R. Gatechair, J. H. Jilek, J. Polym. Sci., Polym.
Chem. Ed. 1984, 22, 77–84; e) S. P. Pappas, B. C. Pappas, L. R.
Gatechair, J. H. Jilek, W. Schnabel, Polym. Photochem. 1984, 5,
1–22.
[37] E. Wigner, Z. Phys. 1927, 40, 492–500; E. Wigner, Z. Phys.
1927, 40, 883–892; E. Wigner, Z. Phys. 1927, 43, 624–652.
[38] a) G. Grabner, C. Richard, G. Köhler, J. Am. Chem. Soc. 1994,
116, 11470–11480; b) A. P. Durand, R. Brown, D. Worall, F.
Wilkinson, J. Chem. Soc. Perkin Trans. 2 1998, 365–370.
[39] I. Manet, S. Monti, M. Fagnoni, S. Protti, A. Albini, Chem.
Eur. J. 2005, 11, 140–151.
[40] a) M. Slegt, Selective Alkylation by Photogenerated Aryl and
Vinyl Cations, Thesis, Leiden University, 2006, p. 63; b) See ref.
40a, p. 55.
[41] N. Kamigata, M. Kobayashi, H. Minato, Bull. Chem. Soc. Jpn.
1972, 45, 2047–2050.
[42] R. A. Abramovitch, O. A. Koleoso, J. Chem. Soc. B 1969, 779–
[50] J. W. Knapczyk, J. J. Lubinkowski, W. E. McEwen, Tetrahedron
Lett. 1972, 35, 3739–3742.
785.
[43] N. Arai, K. Narasaka, Bull. Chem. Soc. Jpn. 1995, 68, 1707–
1714.
[44] R. Itô, T. Migita, N. Morikawa, O. Simamura, Tetrahedron
1965, 21, 955–961.
[51] P. Hanson, J. R. Jones, A. B. Taylor, P. H. Walton, A. W.
Timms, J. Chem. Soc. Perkin Trans. 2 2002, 1135–1150.
[52] P. S. J. Canning, K. McCrudden, H. Maskill, B. Sexton, J.
Chem. Soc. Perkin Trans. 2 1999, 2735–2740.
[45] J. V. Crivello, in Radiation Curing In Polymer Science and Tech-
nology, J. P. Fouassier, J. F. Rabek, Chapman & Hall, London,
1993, Vol. 2, pp. 435–471.
[46] a) J. L. Dektar, N. P. Hacker, J. Org. Chem. 1990, 55, 639–647;
b) J. L. Dektar, N. P. Hacker, J. Org. Chem. 1991, 56, 1838–
1844; c) N. P. Hacker, D. V. Leff, J. L. Dektar, J. Org. Chem.
1991, 56, 2280–2282.
[53] M. Ochiai, M. Toyonari, T. Nagoaka, D.-W. Chen, M. Kida,
Tetrahedron Lett. 1997, 38, 6709–6712.
[54] N. Miyaura, A. Suzuki, Chem. Rev. 1995, 95, 2457–2483.
[55] G. Schiemann, W. Winkelmüller, Ber. Dtsch. Chem. Ges. 1933,
66, 727–730.
Received: April 17, 2007
Published Online: September 14, 2007
Eur. J. Org. Chem. 2007, 5364–5375
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5375