Reductive photocarboxylation of phenantnrene: A mechanistic investigation

Article abstract of DOI:10.1021/jo951702n

Irradiation of a solution of phenanthrene (PHN) in DMSO saturated with CO₂ in the presence of N,N-dimethylaniline (DMA) produced 9,10-dihydrophenanthrene-9-carboxylic acid (1) in 55% yield, trans-9,10-dihydrophenanthrene-9,10-dicarboxylic acid (2) in 11% yield, a trace of phenanthrene-9-carboxylic acid (3), and a trace of 10-[p-(N,N-dimethylamino)phenyl]-9,10-dihydrophenanthrene-9-carboxylic acid (5). Addition of cumene, a hydrogen donor, or water, a proton donor, decreased the yield of 2, while addition of certain salts increased its yield. 9-Carboxy-9,10-dihydrophenanthr-10-yl, generated by irradiation of phenanthrene-9-carboxylic acid in the presence of DMA, is proposed to be an intermediate in the formation of the acids. The quantum yield for the formation of 2 increased to a maximum of 0.13 with increasing light intensity. High CO₂ concentrations in DMSO changed the reaction pathway, greatly reducing the yields of 1 and 2, and phenanthrene-9-carboxylic acid (3) and 10-[p-(N,N-dimethylamino)phenyl]-9,10-dihydrophenanthrene-9-carboxylic acid (5) were formed instead. On the basis of these results reduction of 9-carboxy-9,10-dihydrophenanthr-10-yl with the phenanthrene radical anion is proposed to be a step in the mechanism accounting for trans-9,10-dihydrophenanthrene-9,10-dicarboxylic acid formation. Transient spectroscopic evidence in support of this proposal is presented.

Full text of DOI:10.1021/jo951702n

J . Org. Chem. 1996, 61, 1065-1072
1065
Red u ctive P h otoca r boxyla tion of P h en a n th r en e: A Mech a n istic
In vestiga tion
Alexander V. Nikolaitchik, Michael A. J . Rodgers, and Douglas C. Neckers*
Center for Photochemical Sciences,1 Bowling Green State University, Bowling Green, Ohio 43403
Received September 18, 1995X
Irradiation of a solution of phenanthrene (PHN) in DMSO saturated with CO2 in the presence of
N,N-dimethylaniline (DMA) produced 9,10-dihydrophenanthrene-9-carboxylic acid (1) in 55% yield,
trans-9,10-dihydrophenanthrene-9,10-dicarboxylic acid (2) in 11% yield, a trace of phenanthrene-
9-carboxylic acid (3), and a trace of 10-[p-(N,N-dimethylamino)phenyl]-9,10-dihydrophenanthrene-
9-carboxylic acid (5). Addition of cumene, a hydrogen donor, or water, a proton donor, decreased
the yield of 2, while addition of certain salts increased its yield. 9-Carboxy-9,10-dihydrophenanthr-
10-yl, generated by irradiation of phenanthrene-9-carboxylic acid in the presence of DMA, is proposed
to be an intermediate in the formation of the acids. The quantum yield for the formation of 2
increased to a maximum of 0.13 with increasing light intensity. High CO2 concentrations in DMSO
changed the reaction pathway, greatly reducing the yields of 1 and 2, and phenanthrene-9-carboxylic
acid (3) and 10-[p-(N,N-dimethylamino)phenyl]-9,10-dihydrophenanthrene-9-carboxylic acid (5) were
formed instead. On the basis of these results reduction of 9-carboxy-9,10-dihydrophenanthr-10-yl
with the phenanthrene radical anion is proposed to be a step in the mechanism accounting for
trans-9,10-dihydrophenanthrene-9,10-dicarboxylic acid formation. Transient spectroscopic evidence
in support of this proposal is presented.
In tr od u ction
vs SCE),6 and the oxidation potential of DMA (0.71 V vs
SCE in CH3CN),6 radical ion formation is calculated to
be exergonic by ca. -0.41 eV. Consistent with this value,
a diffusion controlled quenching of 1PHN* by DMA
without exciplex emission is observed in polar solvents.7
This suggests rapid formation of radical ions which then
undergo further reactions with other molecules present
in the reaction mixture.
Recent advances in metal complex catalysis and pho-
toinduced electron-transfer reactions have given rise to
a renewed interest in organic reactions involving carbon
dioxide.2 The carbon atom in CO2 possesses relatively
large partial charge making it susceptible to nucleophilic
attack.
Tazuke et al.3 reported that aromatic hydrocarbons,
such as anthracene, phenanthrene, and pyrene under-
went reductive photocarboxylation when irradiated in
DMSO or DMF in the presence of CO2 and aromatic
amines. The highest product yield observed in the case
of phenanthrene was that of 9,10-dihydrophenanthrene-
9-carboxylic acid which was obtained in 60% yield. The
mechanism suggested for its formation involved photo-
induced electron transfer reaction between the singlet
excited hydrocarbon and the amine, the former being an
electron acceptor and the latter being an electron donor.
In a polar aprotic solvent the electron transfer step is
followed by separation of radical ions with subsequent
trapping of the hydrocarbon radical anion with CO2.
The direction and energetics of the electron transfer
reaction to form a radical ion pair can be estimated from
the Rehm-Weller equation:4
.-
The reduction potential of CO2 to CO2 is reported to
be -2.21V vs SCE in DMF.8 Thus direct reduction of
CO2 is endergonic for any oxidant in PHN/DMA system.
One can find many analogies between the reductive
photocarboxylation of aromatic hydrocarbons and the
addition of CO2 to the adducts of aromatic hydrocarbons
formed by electron transfer from the alkali metals9 since
both are addition reactions of CO2 to carbon-centered
nucleophiles. However, there are also a number of
differences. For example, the addition of CO2 to the
phenanthrene-sodium adduct results in formation of
trans-9,10-dihydrophenanthrene-9,10-dicarboxylic acid as
the only product.10 The mechanism for the formation of
9,10-dihydrophenanthrene-9-carboxylic acid in the pho-
toreaction was recently reported11 while the mechanism
of the former reaction remains unclear.
We have conducted a detailed investigation of the
reductive photocarboxylation of phenanthrene and iso-
lated several previously unreported products. Our mecha-
nistic studies follow.
ox
red
∆G ) Ed - Ea - E00 + 2.6 eV/ꢀ - 0.13 eV
where Edox is an oxidation potential of the donor; Ea is
a reduction potential of the acceptor; E00 is excited state
energy of the reacting state, and ꢀ is dielectric constant
of the solvent. Using values for the singlet energy of
PHN (3.58 eV),5 its reduction potential (-2.44 V in DMF
red
Resu lts a n d Discu ssion
P r od u ct Id en tifica tion . Th e Effect of Ad d itives
on th e Yield s of P r od u cts. Reductive photocarboxy-
(6) Mann, C. K.; Barnes, K. K. Electrochemical Reactions in Non-
aqueous Systems; Marcel Dekker: New York, 1970.
(7) Chen, J .-M.; Ho, T.-I.; Mou, C. Y. J . Phys Chem. 1990, 94, 2889.
(8) Lamy, E.; Nadjo, L.; Saveant, J . M. J . Electroanal. Chem. 1977,
78, 403.
X Abstract published in Advance ACS Abstracts, J anuary 15, 1996.
(1) Contribution No. 245 from the Center for Photochemical Sci-
ences.
(2) Lapidus, A.; Ping, Y. Russ. Chem. Rev. 1981, 50 (1), 63.
(3) Tazuke, S.; Ozawa, H. J . Chem. Soc., Chem. Commun. 1975, 237.
(4) Weller, A. Z. Phys. Chem. (Wiesbaden) 1982, 133, 93.
(5) Murov, S. L. Handbook of Photochemistry; Marcel Dekker: New
York, 1973.
(9) Paul, D.; Lipkin, D.; Weissman, S. J . Am. Chem. Soc. 1956, 78,
116.
(10) J eanes, A.; Adams, R. J . Am. Chem. Soc. 1937, 59, 2608.
(11) Tazuke, S.; Kazama, S.; Kitamura, N. J . J . Org. Chem. 1986,
51 (24), 4548.
0022-3263/96/1961-1065$12.00/0 © 1996 American Chemical Society

1066 J . Org. Chem., Vol. 61, No. 3, 1996
Nikolaitchik et al.
Sch em e 1
Ta ble 1. Effect of Ad d itives on th e Yield of 1 a n d 2
Sch em e 2
time,
PHN
yield of 1, yield of 2,
additive c, M min consumption, %
%
%
none
cumene
water
-
1
1
90
90
90
62
62
65
55
59
68
11
6
0.5
a
Reaction system: [PHN] ) 0.05 M, [DMA] ) 0.5 M, DMSO
saturated with CO2. Determined by HPLC. c Isolated yield is
b
based on consumed PHN. Determined by NMR of the acid
mixture.
Sch em e 3
lation of phenanthrene in DMSO in the presence of N,N-
dimethylaniline (DMA) produces 9,10-dihydrophenan-
threne-9-carboxylic acid (1), trans-9,10-dihydrophen-
anthrene-9,10-dicarboxylic acid (2), a trace of phenan-
threne-9-carboxylic acid (3),12 and a trace of 10-[p-(N,N-
dimethylamino)phenyl]-9,10-dihydrophenanthrene-9-car-
boxylic acid (5), Scheme 1. The yields of 1 and 2 varied
upon inclusion of various additives in the system, Table
1.
Ta ble 2. Qu a n tu m Yield for th e F or m a tion of 2 a s a
F u n ction of Ligh t In ten sitya
quantum yield for
the formation of 2
Io,b Einsteins/s
time,c s
In the absence of additives the yield of 2 was low and,
upon the addition of cumene, a hydrogen donor, or water,
a proton donor, decreased even further while the yield
of 1 increased slightly. Since the observation of an
increased yield of 1 with the addition of hydrogen donors
was consistent with previously reported data,11 we pos-
tulated a common intermediate preceded the formation
of both 1 and 2. Assuming that the addition of CO2 to
the phenanthrene radical anion occurred first, 9-carboxy-
9,10-dihydrophenanthr-10-yl anion (4) was suggested as
an intermediate in the formation of 1. Cumene, an
effective hydrogen donor, trapped 4, forming 1. Water
is an ineffective hydrogen atom donor hence one can
assume that hydrogen abstraction from it is insignificant.
When H2O was replaced with D2O in the experiment
monodeuterated compound 1 was formed as indicated by
1H NMR and GC-MS data. The 1H NMR spectrum
indicated that the site of deuteration was the 10 position
of 1, and monodeuterated 1 contributed 68% to the total
yield of the monocarboxylic acid as calculated from the
mass spectrum of the corresponding methyl ester.13 We
thus concluded that 9-carboxy-9,10-dihydrophenanthr-
10-yl was reduced in the course of the reaction with a
formation of the corresponding anion 6 which was then
trapped with D2O at the 10 position. We estimated the
reduction potential of 4 to be between -1 and -2 V by
analogy with the reduction potential of the R-phenylethyl
radical being -1.6 V vs SCE in CH3CN.14
8.01 × 10-8
5.69 × 10-8
3.61 × 10-8
2.15 × 10-8
5400
6763
12000
20164
0.137
0.115
0.073
0.069
a
Reaction system: [PHN] ) 0.01 M, [DMA] ) 0.2 M in DMSO
b
saturated with CO2. Wavelength of irradiation light is centered
at 348 nm, 8.5 nm bandpass. c The same number of photons is
absorbed by each system.
to dimers of DMA radicals formed as a result of such a
proton transfer.
It appeared that a strong reductant was required to
react with 4. Possible candidates were DMA, excited
phenanthrene, or phenanthrene radical anion. Possible
reactions leading to the formation of 2 are presented on
Scheme 2. All but one of these pathways were eliminated
in the course of our investigation.
We were able to conclude this by employing an alter-
native method for the generation of 9-carboxy-9,10-
dihydrophenanthr-10-yl. Irradiation of 3 in the presence
of DMA in DMSO saturated with CO2 produced 1 in 86%
yield. It is not unreasonable to propose that the mech-
anism of this reaction involves photoinduced electron
transfer between 3 and DMA followed by proton transfer
and hydrogen abstraction as shown on Scheme 3.
Proton transfer in the above reaction produces radical
4. Absence of 9,10-dihydrophenanthrene-9,10-dicarboxy-
lic acid allows us to eliminate two proposed pathways
presented on the left in Scheme 2.
Qu a n tu m Yield s of 2 a s a F u n ction of In cid en t
Ligh t In ten sity. The three pathways presented in a
group of three in Scheme 2 each require two photons for
the formation of one molecule of 9,10-dihydrophenan-
threne-9,10-dicarboxylic acid. Thus the quantum yield
of formation of 2 should increases with light intensity
(increased number of photons absorbed by PHN) as it
does, Table 2.
Protonation of the anions of carboxylic acids can occur
as a result of proton transfer from DMA.+ since proton
loss by amino radical cations has been reported previ-
ously.11 In addition, the GS-MS analysis of the photo-
products which are insoluble in aqueous KOH evidences
the presence of compounds with M+ ) 240 corresponding
(12) Detected by HPLC. Identified by comparison with authentic
sample.
(13) Synthesized with diazomethane.
(14) Wayner, D. D. M.; McPhee, D. J .; Griller D. J . Am. Chem. Soc.
1988, 110, 132.
We filtered UV light through a monochromator (348

Reductive Photocarboxylation of Phenanthrene
J . Org. Chem., Vol. 61, No. 3, 1996 1067
nm with 8.5-nm band pass) to insure all the photons are
absorbed by PHN in the quantum yield measurements.
It is possible to rule out singlet excited PHN as the
reductant of 4 since values close to the diffusion con-
trolled limit have been observed for its quenching rate
constant by DMA in polar solvents.7 Under the condi-
tions of our experiment ([DMA] ) 0.2 M, [PHN] ) 0.01
M) we estimate the lifetime of singlet PHN to be less than
1.5 ns. This value makes unlikely any bimolecular
1
reaction in which PHN* participates.
F igu r e 1. High pressure setup. 1: CO2 cylinder; 2: pressure
valves; 3: water trap; 4: high pressure optical cell; 5: interfer-
ence filter; 6: temperature control unit; 7: pressure sensor;
8: magnetic stirrer; 9: infrared absorbing filter; 10: lenses;
11: 450 W xenon lamp.
We were also able to detect triplet excited phenan-
threne in this reaction using flash-photolysis, and it has
been reported that triplet quenchers did not influence
the yield of 1.11 In our study we found that addition of
2,5-dimethyl-2,4-hexadiene to the reaction did not effect
the yield of 2. In addition, from Rehm-Weller equation
reduction of 4 with 3PHN* is an endergonic process
(Eox(PHN) ) 1.58 V vs SCE, E00t(PHN) ) 2.41 eV).5,6 Thus
Ta ble 3. In flu en ce of [CO2] on th e Rea ction P a th w a ya
pressure of CO2, atm
detected productsb
ratio 1/2c
0.5
1
5
1, 2
1, 2
1, 2, 3
3, 5
3.3
2.5
3.0
-
3
we eliminated PHN* as a participant in the reduction
of 4.
48
These results implicate PHN.-as the reductant of 4.
Apparently, PHN•- is a strong reducing agent owing to
the fact that its oxidation restores conjugation.
a
Reaction system: [PHN] ) 0.01 M, [DMA] ) 0.2 M in DMSO.
Detected by HPLC. c Ratio of HPLC peak areas. UV detection
b
at 280 nm.
The rate constants of CO2 addition to carbon centered
anions often reach the limits of diffusion control.15 Since
the concentration of CO2 in DMSO at atmospheric
pressure reaches 0.21 M one expects rapid quenching of
PHN•-. This process would make the steady-state con-
centration of PHN•- small, thus reducing the possibility
of bimolecular reaction with another radical.
Tazuke et al.11 proposed that equilibria are established
in which there are significant concentrations of the ion
radical involving the phenanthrene radical anion, CO2,
and the CO2 adduct. Under such circumstances the
concentration of the latter species should be controlled
by the concentration (activity) of CO2 in solution. We
thus investigated the influence of [CO2] on reductive
carboxylation of PHN.
We observed changes in the reaction pathway with an
increase in the CO2 partial pressure. At 50 atm CO2
partial pressure, only trace amounts of 1 and 2 were
detected and two other products became dominant. One
was identified as 3 from UV-vis spectrum and GC-MS
spectrum of methyl ester. The second product was
identified as 10-[p-(N,N-dimethylamino)phenyl-9,10-di-
hydrophenanthrene-9-carboxylic acid (5) by comparison
of the HPLC traces of the unknown with an authentic
sample synthesized by irradiation of phenanthrene-9-
carboxylic acid methyl ester in CH3CN in the presence
of DMA followed by hydrolysis. The reactions are sum-
marized in Scheme 4.
If PHN•- is a reductant of 4, and there is an equilib-
rium between PHN•- and the CO2 adduct, then a high
CO2 concentration would increase the concentration of
the carboxylated radical while reducing the steady-state
concentration of PHN•-, thus making reduction of 4 less
likely. This would depress the yield of 2. Indeed, we
observed the absence of 2 among the products of the
reductive photocarboxylation at high CO2 concentration.
In flu en ce of CO2 Con cen tr a tion on th e Rea ction
P a th w a y. CO2 concentration in DMSO is a major factor
in determining the reaction pathway. Though the con-
centration of CO2 in a saturated solution of a number of
solvents at 25 °C was previously reported,11 our measure-
ment, following the procedure similar to that described
by Kraeutler et al.16 confirmed that the concentration is
0.21 M in DMSO at 23 °C and atmospheric pressure.
DMSO is miscible with liquid CO2 so the value of [CO2]
therein at 50 atm CO2 partial pressure can reach 10 M.
Concentrations lower than 0.21 M were attained through
purging DMSO with a mixture of argon and CO2. The
experimental set up assembled to measure the effect of
CO2 concentration in DMSO on the products of the
reductive photocarboxylation is shown on Figure 1.
These results are consistent with our proposal that
PHN•- was a reductant of 4. They also supported the
proposed equilibrium between the phenanthrene radical
anion and the CO2 adduct.
Mech a n ism of Red u ctive P h otoca r boxyla tion of
P HN. The mechanism of reductive photocarboxylation
of phenanthrene is outlined in Scheme 5. Electron
transfer processes between 1PHN* and DMA in polar
solvents yield mainly free radical ions.7 3PHN* is also
formed and this is discussed in the flash photolysis
section. Addition of CO2 to PHN•- is reversible as
indicated by the results under high [CO2].
Combination of a 450 W xenon lamp with 340 nm
interference filter (10 nm bandpass) produced a narrow
wavelength band of UV light of sufficient intensity. 99%
of the filtered UV light is absorbed by PHN. The
irradiation was conducted for a period of time which
corresponded to phenanthrene consumption of less than
10%. All the incident light entering the optical cell was
absorbed by the reaction solution. The results are
summarized in Table 3.
Significant reduction in the yield of 2 upon addition of
H2O suggested that the addition of CO2 to 6 is a much
slower process than the initial addition of CO2 to PHN•-
Since, in DMSO, the reactivity of water is lessened by
H-bonding, the addition of CO2 to PHN•- is faster than
reaction of H2O as indicated by the results in the presence
of D2O. Formation of 1 monodeuterated at the 10
position when D2O was added to the reaction mixture is
(15) Inoue, S.; Yamazaki, N. Organic and Bio-organic Chemistry of
Carbon Dioxide; Kodansha: Tokyo, 1981.
(16) Kraeutler, B.; Bard, A. J . J . Am. Chem. Soc. 1978, 100, 2239.

1068 J . Org. Chem., Vol. 61, No. 3, 1996
Nikolaitchik et al.
Sch em e 4
Sch em e 5
still futher evidence in support of the mechanistic scheme
described above.
Sa lt Effect on th e Yield of 1 a n d 2. Recently a
number of groups have reported salt effects on photoin-
duced electron-transfer reactions.17 Yanagida et al.,18 for
example, reported a 10-fold increase in the yield of
benzilic acid in the photoreduction of CO2 by benzophe-
none in the presence of tetraethylammonium chloride,
while Santamaria observed an increased selectivity in
photochemical N-demethylation of tertiary amines in the
presence of an added salts such as LiClO4 or Mg(ClO4)2.19
In our system, a slight increase in the yield of 1 and a
30% increase in the yield of 2 were observed in the
presence of 0.1 M Mg(ClO4)2. The effects of tetramethyl-
ammonium chloride, tetraethylammonium chloride, tet-
rabutylammonium bromide, and LiClO4 on the yields of
the acids were comparable with experimantal error
(10%). We propose that the presence of salt effected the
reduction of 4 to the corresponding anion. It is possible
that the salt suppressed back electron-transfer thus
increasing the total yield of 2. The specific nature of the
individual salt effects is still to be determined.
F igu r e 2. Transient absorption spectra of Ar-purged DMSO
solution [PHN] ) 0.01 M, [DMA] ) 0.2 M: 1: 10 µs after the
pulse; 2: 11.2 µs; 3: 13.6 µs; 4: 22 µs; 5: 54 µs.
and 490 nm indicated the presence of at least two
transient species. We attribute the major portion of the
absorption to the phenanthrene radical anion (PHN•-
)
F la sh -P h otolysis Exp er im en ts. Flash-photolysis of
PHN-DMA system in DMSO was carried out in the
presence and absence of CO2. Transient absorption
spectra are presented in Figure 2. We observed absorp-
tion from a number of transients in the 400-500 nm
region. A difference in the character of the decay at 430
which has been shown to have a broad absorption with
λ
max at 450 nm tailing up to 700 nm.9 The absorbance at
490 nm corresponds to the PHN triplet. Addition of the
known triplet quenchers, 2,5-dimethyl-2,4-hexadiene or
oxygen, to the solution results in the quenching of this
transient absorbance consistent with PHN triplet behav-
ior. The wavelength of the absorbance is consistent with
that for PHN triplet in polar solvents.20 We did not
observe a distinct contribution from N,N-dimethylaniline
(17) (a) Mizuno, K.; Ichinose, N.; Otsuji, Y. Chem. Lett. 1985, 455.
(b) Mizuno, K.; Ichinose, N.; Tamai, T.; Otsuji, Y. Tetrahedron Lett.
1985, 26 (47), 5823. (c) Goodcon, B.; Shuster, G. B. Tetrahedron Lett.
1986, 27 (27), 3123.
(18) Ogata, T.; Hiranaga, K.; Matsuoka, S.; Wada, Y.; Yanagida, S.
Chem. Lett. 1993, 984.
(19) Santamaria, J . Pure Appl. Chem. 1995, 67 (1), 141.
(20) Chow, Y. L.; Buono-Core, G. E.; Marciniak, B; Beddard, C. Can
J . Chem. 1983, 61, 801.

Reductive Photocarboxylation of Phenanthrene
J . Org. Chem., Vol. 61, No. 3, 1996 1069
F igu r e 4. Transient absorption decays of PHN•- at 430 nm
in DMSO. 1: Purged with Ar; 2: purged with CO2. [PHN] )
0.01 M, [DMA] ) 0.2 M.
F igu r e 3. 1: The spectrum of separated solutions of PHN
(0.01 M) in DMSO and DMA (0.5 M) in DMSO. 2: Difference
absorption spectrum between solution of PHN and DMA in
DMSO ([PHN] ) 0.01 M, [DMA] ) 0.5 M) and the spectrum
of separated solutions of PHN in DMSO and DMA in DMSO.
radical cation (DMA.+) likely because the DMA.+ concen-
tration is too low to permit its detection in DMSO.
For the range of concentrations used in our experi-
ments, it was likely that all the singlet PHN was
quenched by DMA, thus ruling out formation of triplet
PHN via intersystem crossing from singlet PHN.
We have no evidence for complexation of CO2 with any
of these compounds since we found no change in the
absorption spectra of PHN, DMA, and PHN-DMA in
DMSO upon bubbling with CO2.
A number of groups have reported triplet formation
in polar solvents resulting from some combination of the
deactivation of vibrationally excited exciplexes, the re-
combinaton of radical ions, and excited charge-transfer
complex deactivation. Ottolenghi et al. proposed this
mechanism for a number of aromatic hydrocarbon/amine
systems.21
We believe that the deactivation of vibrationally ex-
cited PHN-DMA exciplex and the recombination of the
radical ions is responsible for triplet formation in our
system. The difference absorption spectrum of the solu-
tions of PHN and DMA in DMSO indicates the presence
of a ground state complex between the hydrocarbon and
the amine, Figure 3. For the ratio of concentrations
PHN/DMA used in flash-photolysis experiments this
complex absorbed 41% of the excitation light at 355 nm.
Although it is not confirmed, it is possible that deactiva-
tion of excited PHN-DMA charge-transfer complex is
also contributing to PHN triplet formation. We have no
data on the photochemistry of PHN-DMA charge-
transfer complex.
In Ar-purged DMSO solution of PHN and DMA second
order decay of PHN.- was observed, Figure 4. We
measured k/ꢀ, (k is the rate constant, ꢀ is extinction
coefficient of PHN•- at 430 nm), to be 2.26 ( 0.02 × 106
cm s-1. The observed second order decay of PHN•- can
be explained by assuming the bimolecular reactions of
PHN•- with DMA.+ are taking place. A new fast com-
ponent was introduced into the decay by purging with
CO2, Figure 4. We did not determine the rate constants
F igu r e 5. Transient absorption spectra of Ar-purged CH3-
CN solution [PHN] ) 0.01 M, [DMA] ) 0.2 M. 1: 10.15 µs
after the pulse; 2: 10.45 µs; 3: 11.10 µs; 4: 15 µs; 5: 50 µs.
of the contributing components due to complexity of the
decay in DMSO.
We observed different transient absorption spectra in
CH3CN, Figure 5. An absorbance with λmax ) 465 nm
was observed in the transient spectra at longer time. The
rate of decay of this absorbance was slower when
compared with that of other transients, and it was
possible to observe significant absorbance up to 200 µs
after the laser pulse. This absorbance did not appear in
the transient spectra when tributylamine replaced DMA
as the electron donor. Addition of nitromethane to the
acetonitrile solution produced a transient absorption
spectrum identical to that observed in the earlier experi-
ment at longer times after the laser pulse. We attribute
the absorbance at λmax ) 465 nm to DMA.+ by comparison
of the observed spectrum with that reported by Holcman
et al.22
The kinetics of the decay of PHN•- in Ar-purged CH3-
CN differed from those observed in Ar-purged DMSO,
Figure 6. In CH3CN we observed first-order decay of
PHN•- with a rate constant of 7.5 ( 0.1 × 105 s-1. Upon
purging with CO2 the decay rate increased and the
observed first-order rate constant became 1.28 ( 0.05 ×
106 s-1. Assuming the concentration of CO2 in CH3CN
as 0.17 M11 we calculated the rate constant of the reaction
of PHN•- with CO2 to be 3.14 ( 0.07 × 106 M-1 s-1
.
(21) Orbach, N.; Ottolenghi, M. Intersystem Crossing and Ionic
Recombination Studied by Pulsed Laser Excitation of Charge-Transfer
Systems. In The Exciplex; Gordon, M., Ware, W. R., Eds.; Academic
Press Inc.: New York, 1975.
(22) Holcman, J .; Sehested, K. J . Phys. Chem. 1977, 81 (20), 1963.

1070 J . Org. Chem., Vol. 61, No. 3, 1996
Nikolaitchik et al.
with this solvent as it is shown by flash-photolysis
results.
Exp er im en ta l Section
Ma ter ia ls a n d Equ ip m en t. HPLC grade DMSO, DMF,
and acetonitrile were purchased from Aldrich. DMSO and
DMF were refluxed over calcium hydride for 1 h and fraction-
ally distilled under vacuum. Acetonitrile was refluxed over
phosphorus pentoxide and distilled before use. N,N-Dimethyl-
aniline 99.5% (Aldrich) was dried over potassium hydroxide
overnight and vacuum distilled. Phenanthrene 98% was
purchased from Lancaster and purified according to a proce-
dure previously described.24 CO2 (99.9%) was purchased from
Liquid Carbonic.
NMR spectra were recorded using Varian Gemini-200 and
Unity Plus-400 NMR spectrometers. GC-MS was performed
on a Hewlett-Packard 5987A instrument. HPLC was per-
formed on a Hewlett-Packard 1050 series instrument equipped
with multiple wavelength diode-array and fluorescence detec-
tors. UV spectra were obtained on a Hewlett-Packard 8452A
diode array spectrophotometer, and IR spectra were collected
on a Mattson Instruments 6020 Galaxy Series FT-IR. A Spex
Fluorolog 2 spectrofluorimeter was used for quantum yield
measurements.
F igu r e 6. Decays of PHN•- at 430 nm in (1) Ar-purged DMSO;
(2) Ar-purged CH3CN. [PHN] ) 0.01 M, [DMA] ) 0.2 M.
The observed results suggested PHN.- quenching by
CH3CN which is consistent with the report of Tazuke et
al.11 that PHN did not undergo reductive photocarboxy-
lation in that solvent. We found that 9,10-dihydro-
phananthrene-9-carboxylic acid was produced in 9% yield
and trans-9,10-dihydrophenanthrene-9,10-dicarboxylic
acid was produced in trace amounts upon irradiation
solution of PHN, DMA, and CO2 in CH3CN. Under
similar experimental conditions, the yield of 1 was lower
than that in DMSO. Quenching of PHN•- by CH3CN
explains the lower yield of the acids formed in reductive
photocarboxylation of phenanthrene. Observation of
DMA.+ in CH3CN at longer times after the laser pulse
was consistent with proposed quenching of PHN•- with
CH3CN to produce a less reactive ion radical. In the
absence of such quenching one could expect a bimolecular
coupling reaction of the ion radicals to consume DMA.+.
P h otor ea ction s. La r ge-Sca le. A 100 mL Pyrex internal
irradiation photochemical reactor (Ace Glass) was used.
A
450-W Hanovia high pressure mercury lamp (Ace Glass Inc.)
was used as a light source. The standard sample volume was
100 mL. CO2, passed through drierite drying tube, was
continuously bubbled through sample solution during photo-
reaction. The standard concentrations of PHN and DMA were
0.05 M and 0.5 M.
Sm a ll-Sca le. Small-scale reactions were conducted in a
merry-go-round photochemical reactor using a 450 W high
pressure mercury lamp placed in a Pyrex cooling immersion
well as the irradiation source. Sample solutions were placed
in Pyrex test-tubes (13-120 mm) through which CO2 was
bubbled for 15 min before irradiation. The test-tubes were
sealed and irradiated for 30 min. The resulting solutions were
analyzed by HPLC.
P r od u cts Sep a r a tion a n d Id en tifica tion . After photo-
reaction, the solvent and amine were removed by vacuum
distillation. KOH (40 mL/1 M) was added to the residue. The
aqueous mixture was extracted with two 20 mL portions of
chloroform and then with 20 mL of ether. The chloroform and
ether extracts were combined, dried over anhydrous magne-
sium sulfate, and evaporated. The resulting residue was
analyzed by GC-MS.
The aqueous solution was acidified with 3 M HCl to pH )
4 and extracted with two 20 mL portions of ethyl acetate. The
combined ethyl acetate extracts were dried over magnesium
sulfate, and the solvent was evaporated. Chloroform (20 mL)
was added to the residue and then the mixture was heated.
The boiling chloroform mixture was filtered and, when cooled,
produced a white solid. The solid was chromatographed on
silicic acid, the products being eluted with methylene chloride
and ethyl acetate. The first product eluted was identified as
9,10-dihydrophenanthrene-9-carboxylic acid (1).3 The second
product was trans-9,10-dihydrophenanthrene-9,10-dicarboxylic
acid (2) which was recrystallized from 95% ethanol: mp 234-
240 °C dec (lit.10 235-242 °C); NMR (DMSO-d6) δ 4.3 (2H, s),
7.3 (6H, m), 7.84 (2H, d), 12.55 (2H, s); MS m/ e 268 (M+), 250,
236, 222, 205, 178; HRMS m/ e measured 268.0741 (∆ ) -0.6
mDa); calculated 268.0735; IR (KBr, cm-1) 1702 (nCdO). The
dimethyl ester of 2 was synthesized using diazomethane and
purified by chromatography on silica gel, (yield quantitative,
mp 127 °C (lit.10 128 °C)).
The anion radical of 9,10-dicyanoanthracene is known
to react with CH3CN.23 We think a similar reaction is
occurring with PHN.-. Currently we are investigating
this process.
Con clu sion s
We have shown that reductive photocarboxylation of
phenanthrene in the presence of N,N-dimethylaniline
proceeds via the radical intermediate 9-carboxy-9,10-
dihydrophenanthr-10-yl which, upon hydrogen abstrac-
tion, forms 9,10-dihydrophenanthrene-9-carboxylic acid.
This radical intermediate is also reduced with phenan-
threne anion radical producing the corresponding anion
and phenanthrene. This pathway results in the forma-
tion of trans-9,10-dihydrophenanthrene-9,10-dicarboxylic
acid. We found that addition of CO2 to phenanthrene
anion radical is a reversible process as indicated by the
results at high CO2 concentration in solution. Consider-
ing the similarity between photochemical reductive car-
boxylation of aromatic hydrocarbons and reductive car-
boxylation of these using alkali metals, we propose that
the same mechanism for the formation of dicarboxylic
acids is operating in both cases. This mechanism in-
volves reduction of the radical formed upon primary
addition of CO2 to PHN.- with a second PHN.-.
The chloroform filtrate was concentrated and chromoto-
graphed on silicic acid thus also producing a white solid. This
compound was identified as 9,10-dihydrophenanthrene-9-
carboxylic acid (1).
Significantly reduced yield of the acids in CH3CN can
be explained by quenching of phenanthrene anion radical
(23) Ohashi, M.; Kudo, H.; Yamada, S. J . Am. Chem. Soc. 1979, 101,
2201.
(24) Nikolaitchik, A. Spectrum 1994, 7 (4), 21.

Reductive Photocarboxylation of Phenanthrene
J . Org. Chem., Vol. 61, No. 3, 1996 1071
An a lysis by HP LC. The following method was used for
carboxylic acid analysis. Tetramethylammonium chloride (2
mL of 0.01 M solution) in H2O was added to 25 mL of reaction
solution. The filtered sample (20 mL) was injected into an
HPLC equipped with a Hewlett-Packard 10 µm LiChrosorb
column (200 mm × 4.6 mm). Water-methanol was used as
the mobile phase. The following solvent gradient was used
(time, % methanol): (0 min, 10), (10 min, 70), the flow rate
being 1 mL/min. The elution was monitored using UV-vis
detection at 268 nm. The solid carboxylic acids were analyzed
using a similar procedure, except the sample solution was
prepared by dissolving the solid acids in methanol. For
phenanthrene analysis, 10 mL of reaction solution was injected
into HPLC equipped with 5 µm Hypersil column (200 mm ×
4.6 mm). Hexane was used as the mobile phase, the flow rate
being 0.5 mL/min. Detection was performed with UV-vis
detector at 346 nm as well as with fluorescence detector (350
nm excitation, 400 nm observation).
Qu a n tu m Yield of F or m a tion of 1 a n d 2 a s a F u n ction
of Ligh t In ten sity. The reaction solution (0.01 M PHN, 0.18
M DMA in DMSO, total 2.5 g) was placed in a quartz cuvette,
and CO2 was bubbled through the sample. The irradiation
was performed in a sample compartment of Spex Fluorolog 2
spectrofluorimeter with an excitation monochromator set at
348 nm, 8.5 nm bandpass. Under these conditions all the
incident light is absorbed by PHN in the reaction solution. CO2
was bubbled continuously through the sample during irradia-
tion, and the progress of the reaction was monitored by
measuring the UV absorbance of the reaction solution at 348
nm. Neutral density filters (Oriel Instruments) were used to
control the light intensity. The time of exposure was varied
depending on the intensity of incident light. In all reactions
PHN consumption was less than 5%. After irradiation,
samples were analyzed for phenanthrene and the carboxylic
acids by HPLC as described above. Light intensity at 348 nm
was measured using potassium ferrioxalate actinometry.5
procedure similar to that described by Anzalone.26 A stirred
mixture of 9-cyanophenanthrene (3.75 g, 18 mmol) and NaOH
(1.75 g, 43 mmol) in 60 mL diethylene glycol was refluxed for
8 h. The cooled mixture was diluted with 100 mL of H2O and
acidified to pH ) 1 with 3 M HCl. The white precipitate of
the acid was filtered, dried, and recrystallized from methanol
(yield 86%; mp 252 °C (lit.27 252 °C)). The methyl ester of 3
was synthesized by refluxing the acid in methanol containing
a few drops of concentrated sulfuric acid for 4 h. The product
was recrystallized from methanol (yield 91%; mp 115 °C (lit.28
115 °C).
Syn th esis of 9,10-Dih yd r op h en a n th r en e-9-ca r boxylic
Acid (1). Phenanthrene-9-carboxylic acid (0.35 g; 1.6 mmol)
and DMA (3.81 g; 32 mmol) were dissolved in 50 mL of DMSO
and the solution was placed in the photochemical reactor
described above, purged with Ar, and irradiated for 180 min.
The solvent was evaporated in vacuum, 30 mL of 0.5 M KOH
was added to the residue, and the solution was extracted with
chloroform and ether. The white amorphous precipitate
formed upon acidification of the aqueous solution with 0.5 M
HCl was collected and washed with H2O. Chromatography
on silicic acid with CHCl3 produced 0.29 g (81% yield) of 1.
Mp 111 °C (lit.29 111 °C dec); NMR (CDCl3) δ 3.23 (2H, m),
3.84 (1H, t), 7.35 (6H, m), 7.78 (2H, m).
Syn th esis of tr a n s-9,10-Dih yd r op h en a n th r en e-9,10-
d ica r boxylic Acid (2). Compound 2 was synthesized via
reaction of the phenanthrene sodium adduct with CO2 follow-
ing the procedure of Adams et al.10 Yield 34%. The analytical
characteristics are the same as those for compound 2 prepared
by the photochemical procedure.
Syn th esis of 10-[p-(N,N-Dim eth yla m in o)p h en yl]-9,10-
d ih yd r op h en a n th r en e-9-ca r boxylic Acid (5). Methyl es-
ter of 10-[p-(N,N-dimethylamino)phenyl]-9,10-dihydrophenan-
threne-9-carboxylic acid was synthesized using the following
procedure. Phenanthrene-9-carboxylic acid (0.3 g, 1.27 mmol)
methyl ester and N,N-dimethylaniline (3 g, 25 mmol) were
dissolved in 100 mL of anhydrous acetonitrile and placed in
the photochemical reactor described above; the solution was
purged with Ar and then irradiated with 366 nm light. The
progress of the reaction was monitored by TLC. Upon comple-
tion the solvent was evaporated in vacuum and the products
were chromatographed on silica gel with hexane-ethyl acetate
(10:1). A 13% yield (61 mg) of 10-[p-(N,N-dimethylamino)-
phenyl]-9,10-dihydrophenanthrene-9-carboxylic acid methyl
ester was obtained after chromatography: mp 57 °C; 1H NMR
(CDCl3) δ 2.92 (6H, s), 3.59 (5H, m), 6.67 (2H, d, J ) 15.9 Hz),
7.04 (2H, d, J ) 15.8 Hz), 7.25 (6H, m), 7.74 (2H, dd, J ) 9, 7
Hz). Anal. Calcd for C24H23O2N: C, 80.67; H, 6.44; N, 3.92%.
Found: C, 79.81; H, 6.69; N, 4.06%. MS: 357 (M+), 298, 237,
120.
In flu en ce of CO2 Con cen tr a tion on th e P a th w a y of
Red u ctive P h otoca r boxyla tion . A high pressure optical
cell, equipped with three sapphire windows, was constructed
from 303 stainless steel. Cell design has been described
elsewhere.25 The cell was equipped with a pressure transducer
(Omega Engineering Inc. PX-302-7.5KGV), a temperature
control system (Lauda RM 6), and a pressure regulating valve
(High Pressure equipment, 30-13HF4). The general scheme
of the irradiation set up is presented on Figure 2.
A 500 W Xenon lamp (Oriel Instruments) was used as a light
source. A 10 cm quartz tube with water in combination with
an interference filter (340 nm, 10 nm bandpass, Oriel Instru-
ments) was used to cut out the desired wavelength range.
Experiments were conducted as follows.
The reaction solution (15 mL) was loaded into the optical
cell, purged with CO2 passed through a water trap (molecular
sieves, 3A, Aldrich), and pressurized. The cell was equilibrated
for 30 min prior to measurement. All the measurements were
performed at 25 °C. The solution was continuously stirred
during the experiments, and 1 atm partial pressure CO2 was
continuously bubbled through the solution in the optical cell.
A 1:1 mixture of Ar and CO2 was used to obtain 0.5 atm partial
pressure of CO2.
Mea su r em en t of CO2 Con cen tr a tion in DMSO. DMSO
(30 mL) was placed in a 50 mL graduated cylinder connected
with a gas mixer and a flask containing 200 mL of 0.1 M NaOH
in H2O saturated with Ba(OH)2. The experiment was con-
ducted at 22 °C. CO2 was bubbled through DMSO for 30 min,
after which the cylinder was connected to the flask with Ba-
(OH)2 and purged with Ar for 30 min, the gas being dispersed
in aqueous solution. The white precipitate was filtered under
N2 and weighed. BaCO3 (1.249 g) corresponded to CO2
concentration of 0.21 M.
The hydrolysis of 9,10-dihydrophenanthrene-10-(p-N,N-
dimethylaminophenyl)-9-carboxylic acid methyl ester was
performed following a procedure similar to that described by
Bartlett et al.30 32 mg (0.0089 mmol) of 10-[p-(N,N-dimethyl-
amino)phenyl]-9,10-dihydrophenanthrene-9-carboxylic acid
methyl ester was treated with 6 mL of 0.63 M lithium propyl
mercaptide in HMPA. The solution was stirred for 2 h, diluted
with 50 mL of H2O, extracted with three 20 mL portions of
ether, and acidified with 1 M HCl. The precipitate was washed
with H2O and analyzed by HPLC using the procedure for
carboxylic acids described earlier. Retention time and UV
spectrum of this amino acid were identical with those obtained
for an unknown compound synthesized under high pressure
of CO2.
F la sh -P h otolysis Exp er im en ts. Nanosecond flash-pho-
tolysis experiments were performed on a set up described by
Ford and Rodgers31 using the third harmonic of a Q-switched
(26) Anzalone, L.; Hirsch, J . A. J . Org. Chem. 1985, 50 (12), 2130.
(27) Mossetig, E.; Van De Kamp, J . J . Am. Chem. Soc. 1932, 54,
3328.
(28) Schoppee, C. W. J . Chem. Soc. 1933, 40.
(29) De Koning, H.; Wiedhaup, K.; Pandit, V. K.; Huisman, H. O.
Rec. Trav. Chim. 1964, 83, 364.
Syn t h esis of P h en a n t h r en e-9-ca r b oxylic Acid (3).
Phenanthrene-9-carboxylic acid (3) was synthesized by hy-
drolysis of 9-cyanophenanthrene (Aldrich, 97%) following the
(25) McHugh, M. A.; Krukonis, V. J . Supercritical Fluid Extrac-
tion: Principles and Practice; Butterworths: Boston, 1986.
(30) Bartlett, P. A.; J onson, W. S. Tetrahedron Lett. 1970, 46, 4459.
(31) Ford, W. E.; Rodgers, M. A. J . J . Phys. Chem. 1994, 98, 3822.

1072 J . Org. Chem., Vol. 61, No. 3, 1996
Nikolaitchik et al.
Nd:YAG laser (Continuum, YG660) as an excitation source.
Samples were purged with Ar or CO2 prior to measurements.
dation (NSF : DMR-9013109). The NMR spectrometers
were purchased with funds from grants from the NSF
(CHE9302619), the Ohio Board of Regents, and Bowling
Green State University. We are pleased to acknowledge
their support. One of us (A.V.N.) is grateful to Ciba
Geigy Inc., Polymer Division, and the McMaster En-
dowment of the Center for Photochemical Sciences,
Bowling Green State University, for research
assistantships.
Ack n ow led gm en t. The authors thank Dr. G. S.
Hammond, Dr. T. H. Kinstle for helpful discussions and
valuable suggestions made during preparation of the
manuscript, Dr. Z. Shi for assistance in obtaining high
resolution mass spectra, and Dr. Rong Bao for valuable
comments in the course of the investigation. This work
has been supported by the Office of Naval Research
(N00014-93-1-0772 and the National Science Foun-
J O951702N