Unconventional titania photocatalysis: Direct deployment of carboxylic acids in alkylations and annulations

Article abstract of DOI:10.1021/ja306168h

Under dry, anaerobic conditions, TiO₂ photocatalysis of carboxylic acid precursors resulted in carbon-carbon bond-forming processes. High yields of dimers were obtained from TiO₂ treatment of carboxylic acids alone. On inclusion of electron-deficient alkenes, efficient alkylations were achieved with methoxymethyl and phenoxymethyl radicals. In reactions with maleic anhydride or maleimides, phenoxyacetic acid produced chromenedione derivatives in addition to adducts. These photocatalytic reactions are simple and cheap to perform, and the TiO₂ is easily removed by filtration. The anaerobic photocatalysis strategy offers a range of synthetic possibilities.

Full text of DOI:10.1021/ja306168h

Communication
pubs.acs.org/JACS
Unconventional Titania Photocatalysis: Direct Deployment of
Carboxylic Acids in Alkylations and Annulations
David W. Manley,† Roy T. McBurney,† Phillip Miller,† Russell F. Howe,‡ Shona Rhydderch,‡
and John C. Walton*,†
†School of Chemistry, University of St. Andrews, EaStChem, St. Andrews, Fife KY16 9ST, U.K.
‡Chemistry Department, University of Aberdeen, Aberdeen AB24 3UE, U.K.
S
* Supporting Information
through photolyses with TiO2. We used Degussa (Evonik)
ABSTRACT: Under dry, anaerobic conditions, TiO2
photocatalysis of carboxylic acid precursors resulted in
carbon−carbon bond-forming processes. High yields of
dimers were obtained from TiO2 treatment of carboxylic
acids alone. On inclusion of electron-deficient alkenes,
efficient alkylations were achieved with methoxymethyl
and phenoxymethyl radicals. In reactions with maleic
anhydride or maleimides, phenoxyacetic acid produced
chromenedione derivatives in addition to adducts. These
photocatalytic reactions are simple and cheap to perform,
and the TiO2 is easily removed by filtration. The anaerobic
photocatalysis strategy offers a range of synthetic
possibilities.
P25 TiO2, consisting of a 3:1 anatase/rutile mixture (particle
size ∼21 nm, surface area 50 m2 g−1, anatase excitation
wavelength 385 nm, band gap 3.2 eV)13 as 1−5 mg mL−1
dispersions in acetonitrile. Experiments with P25 platinized
with 0.1% Pt (Pt−P25)14 were also carried out. Oxidative/
degradative processes were precluded by purging the
dispersions with Ar and drying the MeCN over CaH2. In a
typical setup, the dispersion was stirred in a Pyrex tube clamped
between two face-to-face sunlamp arrays (UVA).15 After
irradiation, the photocatalyst was simply removed by filtration.
We found that a good range of carboxylic acids did indeed react
under these conditions. Good yields of dimeric C−C-coupled
products could be isolated in the absence of coreactants. On
inclusion of electron-deficient alkenes, alkylations proceeded
effectively; for aryloxyacetic acids, addition−cyclization cas-
cades supervened.
emiconductor photocatalysis (SCPC) has achieved interna-
Photocatalyses of individual carboxylic acids were first
investigated. Photolyses of single acids with P25 were
inefficient, but the use of Pt−P25 with acids 1a−e gave very
pleasing yields of the homodimerization products 2 derived
S
tional prominence for its role in the destructive oxidation
of organic molecules during environmental remediation.1 The
electron/hole pair generated when a SC powder is photo-
irradiated2 interacts with water and oxygen to produce reactive
oxygen species (hydroxyl radicals, superoxide).3 These readily
convert organic matter to CO2, water, and mineral acids.4
Numerous applications of SCPC have been found with
sterilizing, deodorizing, and antifouling materials as well as in
self-cleaning glass and antifogging coatings. Alternatively, under
anaerobic conditions, the TiO2 electron/hole pair can generate
C-centered radicals from certain substrates for use in
constructive chemical transformations. However, only a modest
number of such molecular assembly applications have been
reported.5
•
from C-centered radicals (XCH2 ; Scheme 1). Arylacetic acids
Scheme 1. TiO2-Promoted Dimerizations of Carboxylic
Acids
afforded good to excellent dimer yields (Table 1, entries 1−4).
Phenoxyacetic acid (entry 5) and benzyloxyacetic acid (entry 6)
also produced useful yields of the corresponding C−C-coupled
products. Only in the case of 2-naphthylacetic acid was a
significant amount of the reduced byproduct 2-methylnaph-
thalene observed (entry 2).
To expand the synthetic scope of our TiO2 chemistry, we
examined radical alkylation of alkenes. Since most of the acid-
derived radicals had nucleophilic character, alkenes containing
electron-withdrawing substituents were chosen.16 When maleic
anhydride was included in the dispersions, photolyses with P25
led to the formation of adducts 3 incorporating the carboxylic
Two notable examples are the exploratory study of additions
of enol ethers to various acceptors6 and research on the
addition of tertiary amines to electron-deficient alkenes.7
Carboxylic acids also undergo hole oxidation, generating C-
centered radicals following decarboxylation. The simplest
́
example is the photo-Kolbe reaction, first studied in the late
1970s.8 This is potentially an attractive method of radical
generation because of the availability of many natural and
synthetic carboxylic acids that could then be used without
functionalization. Most conventional methods of radical
generation from carboxylic acids rely on preparing unappealing
precursors such as diacyl peroxides, peresters,9 Barton esters,10
or Hunsdieker salts11 or involve hypervalent iodine.12
By analogy with the photo-Kolbe
that carboxylic acids might be reductively decarboxylated
́
reaction, we anticipated
Received: June 25, 2012
Published: August 7, 2012
© 2012 American Chemical Society
13580
dx.doi.org/10.1021/ja306168h | J. Am. Chem. Soc. 2012, 134, 13580−13583

Journal of the American Chemical Society
Communication
several N-substituted maleimides. Good to excellent overall
yields were obtained under all conditions (Table 2). In each
case, the adduct 7 and cycloadduct 8 were accompanied by
moderate amounts (11−43%) of the corresponding reduced
alkene (succinic anhydride or succinimide).
Table 1. Yields of Dimers from SCPC of Carboxylic Acids
yield (%)
entry
acid: X
1a: Ph
irrad. time (h)
2
XCH3
0
1
2
3
4
5
6
19
52
20
20
21
21
81
56
83
78
52
38
a
a
1b: 2-naphthyl
1c: 2-thiophene
1d: 3-thiophene
1e: OPh
40
0
0
0
0
Table 2. Product Yields from Reactions of Phenoxyacetic
a
Acid with Maleic Anhydride and Maleimides
time
(h)
yield of
1f: OBn
entry
X
conditions
P25
7 + 8 (%)
7:8 ratio
a
NMR yield.
1
2
O
26
18
45
17
20
45
12
11
16
12
16
60
1.3:1.0
6.0:1.0
0.3:1.0
0.42:1.0
0.42:1.0
0.10:1.0
1.0:1.0
1.0:1.0
0.12:1.0
1.1:1.0
0.5:1.0
b
O
0.1% Pt−P25
56
acid moiety R and an additional H atom (Scheme 2). In each
case, significant amounts of succinic anhydride (9−16%)
c
b
3
O
P25
63
4
NH
P25
P25
78
78
5
NMe
Scheme 2. TiO2-Promoted Additions of Carboxylic Acids to
Alkenes
c
b
6
NMe
P25
75
7
NBu-t
NPh
P25
67
79
8
P25
b
9
NPh
0.1% Pt−P25
P25
77
10
11
NCO2Me
NCO2Me
68
b
0.1% Pt−P25
74
a
In all reactions, 1e (1 equiv), 6 (2 equiv), and TiO2 (1.5 equiv) were
used. Unless otherwise stated, the TiO2/MeCN dispersion was 1 mg
b
c
mL−1 and isolated yields are reported. NMR yield. TiO2/MeCN
dispersion was 5 mg mL−1.
Major advantages of our photocatalytic methodology are the
inherent opportunities for adjusting and adapting it. The
synthetic value of the process in Scheme 3 was enhanced
because the selectivity could be tuned in favor of either the
furochromene derivative 8 (X = O) or the adduct 7 (X = O).
The P25 photocatalyst in a 1 mg mL−1 dispersion gave a
modest selectivity of 7:8 = 1.3:1.0 (Table 2, entry 1). Use of
platinized Pt−P25 greatly favored the adduct (7:8 = 6.0:1.0;
entry 2), whereas a denser dispersion of P25 inverted the
selectivity in favor of the cycloadduct 8 (7:8 = 0.3:1.0; entry 3).
The selectivity for the pyrrolochromene derivatives 8 (X = NR)
varied to a minor extent depending on the N substituent
(compare entries 4, 5, 7, 8, and 10). N-Methylmaleimide gave
the best selectivity for the cycloadduct 8 (X = NMe), and a
denser dispersion of P25 improved this still further (entry 6).
In the reaction with N-phenylmaleimide, Pt−P25 was selective
for the adduct (entry 9). A modest selectivity in favor of the
cycloadduct was achieved with Pt−P25 in the case of the N-
methoxycarbonyl substituent (compare entries 10 and 11).
Platinization provides an electron sink and thus increases the
availability of holes by hindering recombination. This modified
photocatalyst would therefore be expected to favor the
cycloadducts 8 because hole oxidation of the ring-closed radical
is a key step (see the mechanism in Scheme 4 below). The
increased selectivities for 8 shown in entries 9 and 11 are thus
explained. Unexpectedly, Pt−P25 was selective for adduct 7 in
the case of maleic anhydride (entry 2). The reason for this
inverted selectivity is unknown. However, the same results were
obtained when the photolyses were repeated. With the denser
P25 dispersion, less light could enter the system, and hence,
longer photolysis times were needed. However, the higher
reactant concentrations favored encounters between the ring-
closed radicals and the P25. Hence, these conditions were also
expected to increase the hole availability and favor cycloadduct
formation. Entries 3 and 6 do indeed show this.
accompanied products 3. The yields of the alkyl anhydrides 3
were low for aliphatic carboxylic acids (3a) but increased
substantially for methoxyacetic acid (3b) (Scheme 2). Good
yields of the analogous adducts 4 were obtained when
acrylamide was used as the substrate, and substantial yields of
adducts 5 were isolated from the P25-photocatalyzed reactions
of methoxyacetic acid with N-phenyl- and N-methylmaleimide
(Scheme 2). As expected,17 minor amounts of alkene [2 + 2]
photodimers were observed with the maleimides.
The reaction of phenoxyacetic acid (1e) with maleic
anhydride opened a divergent reaction channel in that adduct
7 was accompanied by a significant amount of dihydrofur-
ochromenedione derivative 8 (X = O) (Scheme 3). This
appeared to have formed when the initial adduct radical
underwent an intramolecular closure onto the phenyl ring
followed by rearomatization. Similar cycloadducts 8 (X = NR)
were obtained from TiO2-photocatalyzed reactions of 1e with
Scheme 3. Dihydrochromene Derivative Formation
13581
dx.doi.org/10.1021/ja306168h | J. Am. Chem. Soc. 2012, 134, 13580−13583

Journal of the American Chemical Society
By analogy with the photo-Kolbe reaction, the first step was
Communication
́
Scheme 4. Proposed Mechanism for P25-Photocatalyzed
Reactions of Carboxylic Acids with Maleic Anhydride and
Maleimides
expected to be hole capture by the carboxylic acid at the TiO2
surface to generate the corresponding radical cation.8 Loss of a
proton would then generate an acyloxyl radical, RCH2C(O)O•,
which would rapidly eject CO2 to produce an alkyl radical,
18
•
RCH2 .
The continuous-wave X-band EPR spectrum obtained during
UV irradiation (400 nm) of a frozen suspension of P25 with t-
BuCO2H in CH3CN (Figure 1a) showed the 10-line spectrum
Figure 1. EPR spectra during UV irradiation of TiO2 with carboxylic
acids. (a) EPR spectrum of t-Bu• during UV irradiation of t-BuCO2H
and P25 in frozen CH3CN at 80 K. (b) Isotropic spectrum of t-Bu•
(central 4 lines only) during UV irradiation of t-BuCO2H with PC-500
•
at 300 K in PhH. (c) Isotropic spectrum of PhOCH2 during UV
irradiation of 1e and PC-500 in fluid PhH at 300 K.
of the t-Bu• radical superimposed on high-field signals due to
trapped electrons (Ti3+) in the titania. Additionally, the same
radical was observed as a transient during UV irradiation of the
acid with PC-500 in fluid PhH at 300 K (Figure 1b; only the
central four lines are shown). UV irradiation of a fluid
or succinimide obtained in our reactions. However, this
electron transfer route implies that the yield of succinic
anhydride or succinimide should equal the yield of the
accompanying chromene derivative 8.24 In fact, equal yields
of 8 and anhydride or succinimide were not observed in any of
our reactions. Furthermore, we found that irradiations of TiO2
dispersions of maleic anhydride and the maleimides in the
absence of the carboxylic acids produced anhydride or
succinimides in significant yields.25 We conclude that direct
hole capture by the cyclohexadienyl radicals is probably the
main process, although of course the alkenes retain their role as
electron acceptors.
19
dispersion of TiO2 with acid 1e in PhH at 300 K resulted in
an isotropic spectrum (Figure 1c). The EPR parameters [g =
2.0024, a(2H) = 17.4 G] were in good agreement with the
literature for the PhOCH2 radical.20 Thus, the EPR spectra
•
provided effective support for the first steps of the mechanism.
The lack of anisotropy in the solution spectra (Figure 1b,c)
demonstrated that at least some of the t-Bu• and PhOCH2
•
radicals were freely tumbling in solution and not bound to the
TiO2 surface.
Additions of the weakly nucleophilic RCH2 radicals to the
In conclusion, we have found that UVA irradiations of
dispersions of TiO2 and carboxylic acids under dry anaerobic
conditions lead to decarboxylation and free radical generation.
In the absence of substrates, the corresponding dimers were
prepared in good to excellent yields. On inclusion of electron-
deficient alkenes, alkylations occurred quite efficiently. The
reaction of phenoxyacetic acid and a maleimide resulted in a
cascade process in which a pyrrolochromene derivative
accompanied the alkylated succinimide. The process could be
tuned by catalyst modification to afford either the alkyl adduct
or the pyrrolechromene. These photocatalyses are easily and
cheaply carried out in the laboratory, and the heterogeneous
catalyst is simply filtered off during workup. Mechanistic
aspects and applications of these methods in the synthesis of
complex molecules are currently under investigation.
•
electron-deficient double bonds of the alkenes are expected to
be fast.21 The resulting adduct radicals might abstract H atoms
from an endogenous donor. It seems more likely, however, that
they pick up electrons from the TiO2 particles, thus generating
the corresponding enolate ions, which are easily protonated22
to afford adducts 2, 7, etc. (Scheme 4). For the electrophilic
anhydrido- or imidoalkyl radicals produced on addition of
•
PhOCH2 radicals, a competition exists between reduction to
adducts 7 or homolytic closure onto the phenyl ring. In the
latter case, resonance-stabilized cyclohexadienyl radicals are
obtained, which rearomatize to yield the functionalized
chromenes 8. This rearomatization could be the result of
hole capture from TiO2 by the cyclohexadienyl radicals and
subsequent proton loss, as shown in Scheme 4. Alternatively,
electron transfer to more anhydride or maleimide might take
place, as suggested in related work by Hoffmann.7,23
Protonation of the resulting anhydride or maleimide radical
anions, followed by further electron capture and protonation
steps, would explain the significant yields of succinic anhydride
ASSOCIATED CONTENT
■
S
* Supporting Information
Experimental procedures, details of EPR spectroscopic experi-
ments, and NMR spectra for new compounds. This material is
available free of charge via the Internet at http://pubs.acs.org.
13582
dx.doi.org/10.1021/ja306168h | J. Am. Chem. Soc. 2012, 134, 13580−13583

Journal of the American Chemical Society
Communication
(17) (a) Boule, P.; Lemaire, J. J. Chim. Phys. Phys.-Chim. Biol. 1980,
77, 161−165. (b) Cantín, A.; Corma, A.; Leiva, S.; Rey, F.; Rius, J.;
Valencia, S. J. Am. Chem. Soc. 2005, 127, 11560−11561.
(18) (a) Janzen, E. G.; Evans, C. A.; Nishi, Y. J. Am. Chem. Soc. 1972,
94, 8236−8238. (b) Chateauneuf, J.; Lusztyk, J.; Ingold, K. U. J. Am.
Chem. Soc. 1988, 110, 2877−2885. (c) Chateauneuf, J.; Lusztyk, J.;
Ingold, K. U. J. Am. Chem. Soc. 1988, 110, 2886−2893. (d) Fraind, A.;
Turncliff, R.; Fox, T.; Sodano, J.; Ryzhkov, L. R. J. Phys. Org. Chem.
2011, 24, 809−820.
AUTHOR INFORMATION
■
Corresponding Author
jcw@st-and.ac.uk
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
(19) Millenium PC-500 TiO2 was used in this case because its finer
mesh enabled it to stay in suspension without stirring during EPR
spectral scans.
We thank the EPSRC for funding (Grants EP/I003479/1 and
EP/I00372X/1) and the EPSRC National Mass Spectrometry
Service, Swansea.
(20) Hudson, A.; Root, K. D. J. J. Chem. Soc. B 1970, 656−660.
(21) (a) Tedder, J. M.; Walton, J. C. Acc. Chem. Res. 1976, 9, 183−
191. (b) Fischer, H.; Radom, L. Angew. Chem., Int. Ed. 2001, 40,
1340−1371. (c) Poutsma, M. L. J. Phys. Org. Chem. 2008, 21, 758−
782.
REFERENCES
■
(1) (a) Hoffmann, M. R.; Martin, S. T.; Choi, W.; Bahnemann, D. W.
Chem. Rev. 1995, 95, 69−96. (b) Vorontsov, A. V.; Kozlova, E. A.;
Besov, A. S.; Kozlov, D. V.; Kiselev, S. A.; Safatov, A. S. Kinet. Catal.
2010, 51, 801−808. (c) Ravelli, D.; Fagnoni, M.; Dondi, D.; Albini, A.
J. Adv. Oxid. Technol. 2011, 14, 40−46.
(2) (a) Mills, A.; Le Hunte, S. J. Photochem. Photobiol., A 1997, 108,
1−35. (b) Kumar, S. G.; Devi, L. G. J. Phys. Chem. A 2011, 115,
13211−13241.
(3) (a) Ravelli, D.; Dondi, D.; Fagnoni, M.; Albini, A. Chem. Soc. Rev.
2009, 38, 1999−2011. (b) Li, W.; Li, D.; Lin, Y.; Wang, P.; Chen, W.;
Fu, X.; Shao, Y. J. Phys. Chem. C 2012, 116, 3552−3560. (c) Lipovsky,
A.; Ievitski, L.; Tzitrinovich, Z.; Gedanken, A.; Lubart, R. Photochem.
Photobiol. 2012, 88, 14−20.
(4) (a) Linsebigler, A. L.; Lu, G.; Yates, J. T., Jr. Chem. Rev. 1995, 95,
735−758. (b) Choy, W. K.; Chu, W. Ind. Eng. Chem. Res. 2007, 46,
4740−4746.
(22) Kraeutler, B.; Bard, A. J. J. Am. Chem. Soc. 1978, 100, 5985−
5992.
(23) Bertrand, S.; Hoffman, N.; Humbei, S.; Pete, J. P. J. Org. Chem.
2000, 65, 8690−8703.
(24) Provided that the anhydride- or maleimide-derived radicals and
ions are not diverted down alternative reaction pathways.
(25) On inclusion of MeOH to mop up holes, the yields of the
succinic derivatives increased to 77−91%.
(5) (a) Shiraishi, Y.; Saito, N.; Hirai, T. J. Am. Chem. Soc. 2005, 127,
12820−12822. (b) Palmisano, G.; Augugliaro, V.; Pagliaro, M.;
Palmisano, L. Chem. Commun. 2007, 3425−3437.
(6) (a) Kunneth, R.; Feldmer, C.; Knoch, F.; Kisch, H. Chem.Eur.
̈
J. 1995, 1, 441−448. (b) Keck, H.; Schindler, W.; Knoch, F.; Kisch, H.
Chem.Eur. J. 1997, 3, 1638−1645.
(7) (a) Marinkovic, S.; Hoffmann, N. Chem. Commun. 2001, 1576−
1578. (b) Marinkovic, S.; Hoffmann, N. Int. J. Photoenergy 2003, 5,
175−182. (c) Marinkovic,
́
S.; Hoffmann, N. Eur. J. Org. Chem. 2004,
3102−3107.
(8) (a) Kraeutler, B.; Bard, A. J. J. Am. Chem. Soc. 1978, 100, 2239−
2240. (b) Kraeutler, B.; Jaeger, C. D.; Bard, A. J. J. Am. Chem. Soc.
1978, 100, 4903−4905.
(9) For a review, see: Nonhebel, D. C.; Walton, J. C. Free Radical
Chemistry; Cambridge University Press: Cambridge, U.K., 1974;
Chapter 5, pp 99−127.
(10) (a) Barton, D. H. R.; Crich, D.; Motherwell, W. B. J. Chem. Soc.,
Chem. Commun. 1983, 939−941. (b) Motherwell, W. B.; Crich, D. Free
Radical Chain Reactions in Organic Synthesis; Academic Press: London,
1992.
(11) (a) Johnson, R. G.; Ingham, R. K. Chem. Rev. 1956, 56, 219−
269. (b) Baker, F. W.; Holtz, H. D.; Stock, L. M. J. Org. Chem. 1963,
28, 514−516. (c) Yin, F.; Wang, Z.; Li, Z.; Li, C. J. Am. Chem. Soc.
2012, 134, 10401−10404. (d) Wang, Z.; Zhu, L.; Yin, F.; Su, Z.; Li, Z.;
Li, C. J. Am. Chem. Soc. 2012, 134, 4258−4263.
(12) Muraki, T.; Togo, H.; Yokoyama, M. Rev. Heteroat. Chem. 1997,
17, 213−243.
(13) (a) Ohno, T.; Sarukawa, K.; Tokieda, K.; Matsumura, M. J.
Catal. 2001, 203, 82−86. (b) Hurum, D. C.; Agrios, A. G.; Gray, K. A.;
Rajh, T.; Thurnauer, M. C. J. Phys. Chem. B 2003, 107, 4545−4549.
(14) Mills, A. J. Chem. Soc., Chem. Commun. 1982, 367−368.
(15) Two Phillips commercial face-tanning units containing four 15
W Cleo tubes (λ ≈ 350 nm) were used in most of the experiments.
(16) For a discussion of polar effects in free radical reactions, see:
(a) Roberts, B. P. Chem. Soc. Rev. 1999, 28, 25−35. (b) Crich, D.;
Grant, D.; Krishnamurthy, V.; Patel, M. Acc. Chem. Res. 2007, 40,
453−463.
13583
dx.doi.org/10.1021/ja306168h | J. Am. Chem. Soc. 2012, 134, 13580−13583