Mechanistic insights and kinetic analysis for the oxidative hydroxylation of arylboronic acids by visible light photoredox catalysis: A metal-free alternative

Article abstract of DOI:10.1021/ja406311g

The photocatalytic hydroxylation of boronic acids with methylene blue as photosensitizer proceeds with high efficiency. Detailed time-resolved studies of the relevant rate constants provide a clear mechanistic understanding of excited-state processes and guided the selection of the photocatalyst and the optimization of experimental conditions.

Full text of DOI:10.1021/ja406311g

Communication
pubs.acs.org/JACS
Mechanistic Insights and Kinetic Analysis for the Oxidative
Hydroxylation of Arylboronic Acids by Visible Light Photoredox
Catalysis: A Metal-Free Alternative
Spencer P. Pitre, Christopher D. McTiernan, Hossein Ismaili, and Juan C. Scaiano*
Department of Chemistry and Centre for Catalysis Research and Innovation, University of Ottawa, 10 Marie Curie, Ottawa K1N
6N5, Canada
S
* Supporting Information
We demonstrated, using two different sources of α-aminoalkyl
ABSTRACT: The photocatalytic hydroxylation of bor-
onic acids with methylene blue as photosensitizer proceeds
with high efficiency. Detailed time-resolved studies of the
relevant rate constants provide a clear mechanistic
understanding of excited-state processes and guided the
selection of the photocatalyst and the optimization of
experimental conditions.
radicals, that the reductive cyclization of an aryl enone,
demonstrated originally by Yoon and co-workers,4 can be
readily achieved in good yield simply by using organic sources
of strongly reducing amine-derived radicals. We wondered if it
would be possible to create a novel catalytic system that would
utilize the excellent reducing capabilities of α-aminoalkyl
radicals and be a viable replacement for traditional transition-
metal catalysts. Herein, we report the use of MB, a member of
the thiazine dye family, for the catalytic formation of α-
aminoalkyl radicals for use in photoredox processes. Impor-
tantly, our selection of this system is based on a detailed
understanding of the excited-state kinetics that govern the
dynamics and efficiency of the photocatalytic process.
MB is an organic photosensitizer that is widely used in both
biology and chemistry, most notably for its ability to generate
singlet oxygen.5 These sensitized reactions almost ubiquitously
occur by way of the triplet state, where the triplet sensitizer and
the substrate interact to form radicals or, in the case of oxygen,
to generate singlet oxygen.6 The two-electron photoreduction
of MB is known to result in the colorless form, leuco-methylene
blue (LMB in Scheme 2), which can be readily oxidized by
oxygen to return to its original form.7 Flash photolysis studies
have indicated that the photoreduction occurs by electron- or
hydrogen-transfer steps, producing a semireduced form of the
methylene blue (MB•) as an intermediate (Scheme 2). This
species can also act as a reducing agent.8,9
n the past decade, photoredox catalysis has attracted
increasing attention,1 as new light-initiated catalytic reactions
I
continue to emerge. Remarkably, this new chemistry is usually
obtained in the absence of any new understanding of the
excited-state kinetics and mechanisms that underlie the new
discoveries. In this contribution, we examine the excited-state
kinetics relevant to the recently reported oxidative hydrox-
ylation of arylboronic acids to yield the corresponding phenols,
Scheme 1.2 We compare the use of Ru(bpy)3Cl2 with
Scheme 1. Photocatalytic Conversion of Arylboronic Acids
a
to Phenols
Scheme 2. Photoinduced Redox Cycling of Methylene Blue
a
9
,
under Aerobic Conditions
a
As proposed by Xiao and co-workers.2 PC represents the photo-
catalyst, Ru(bpy)3Cl2 or MB, the former being used in Xiao’s report,
along with Et3N as the donor.
methylene blue (MB) as a photocatalyst. Through determi-
nation of all the relevant rate constants using laser flash
photolysis techniques, we have discovered a better, less
expensive, and more environmentally benign catalytic system
that also leads to improved reaction yields.
In the vast majority of cases, photocatalysis is performed with
either a ruthenium or iridium photocatalyst using an amine,
typically diisopropylethylamine (iPr2NEt), as a sacrificial
electron donor to complete the catalytic cycle. Recently, we
investigated the role of the amines and the corresponding α-
aminoalkyl radicals produced in the photoredox catalytic cycle.3
a
Note that MB• can also be trapped by oxygen (vide infra).
Received: June 28, 2013
Published: August 16, 2013
© 2013 American Chemical Society
13286
dx.doi.org/10.1021/ja406311g | J. Am. Chem. Soc. 2013, 135, 13286−13289

Journal of the American Chemical Society
Communication
In 1976, Kayser and Young investigated the photoreduction
of MB by both alkyl- and arylamines.8,10 The authors showed
that the reactive state of the dye, the triplet state, can be
quenched by an amine to form a charge-transfer intermediate
that leads to the semireduced form of the dye and an amine
radical cation. It is well known that this amine radical cation
deprotonates to form the corresponding α-aminoalkyl radical;
therefore, MB can be used as a catalytic source of these radicals
for use in photoredox processes. The semireduced form (MB•)
has a modest reactivity toward oxygen (kox = 5 × 105 M−1 s−1)
but is known to be responsible for dye regeneration in other
catalytic processes.9
There has been interest in replacing ruthenium and iridium
with organic dye sensitizers to serve as organo-photocatalysts.
Several recent reports11,12 have employed Eosin Y for direct
C−H arylation of heteroarenes with aryl diazonium salts and
for C−C and C−P bond formation, while Rose Bengal13 has
led to oxyamination, coupling, and acylnitroso ene reactions.
Eosin Y was also compared with other organic dyes.11 A recent
paper by Xiao et al. (vide infra) also mentioned briefly the
potential use of Eosin Y for the hydroxylation of boronic acids.2
We decided to use Xiao’s oxidative hydroxylation of
arylboronic acids to phenols as a simple example to test MB
as the basis for a metal-free photocatalytic system; the system is
ideal as a mechanistic and kinetic tool. Quantitative conversion
of the parent phenylboronic acid to its corresponding phenol
was observed within 7 h using visible light irradiation (see
reaction 1 and Figure 1). Also, MB showed higher catalytic
Table 1. Direct Comparison of Methylene Blue vs
Ru(bpy)3Cl2 for the Oxidative Hydroxylation of
Phenylboronic Acid in 4:1 Acetonitrile:Water (v/v; 7 h
Reaction) and Rate Constants Derived from Laser Flash
Photolysis Studies
photocatalyst
methylene blue
Ru(bpy)3Cl2
63%
58%
conversion
100%
94%
isolated yield
triplet quenching: kq (M−1 s−1
)
O2
2.46 × 109
1.34 × 108
5.81 × 105
2.44 × 108
4.68 × 107
2.97 × 109
b
b
phenylboronic acid
phenol
see Figure S5
see Figure S5
6.21 × 106
3.9 × 106
iPr2NEt
a
Et3N
a
b
MB value in methanol from the literature.8 Supporting Information.
Chart 1. Percent Conversions and Isolated Yields (in
Parentheses) for Various Phenols Derived from Boronic
a
Acids
a
Conditions: hν for 7 h, 1 mol% MB, 5 equiv of iPr2NEt in
MeCN:H2O (4:1 v/v) and under O2.
Intrigued by the high reaction efficiency shown in Table 1,
we explored a number of related boronic acids using MB as
photocatalyst. The striking results are shown in Chart 1. The
reactions were performed in 4:1 acetonitrile:water (v/v), which
proved to be an excellent solvent for the reaction mixture.
Preliminary experiments with DMF as a solvent (used in
literature reports2) gave as much as 40% yield of phenol even
when no amine was added, probably due to the presence of
amines as common impurities.14,15
Figure 1. Conversion vs time plot for reaction 1 using methylene blue
as the photocatalyst.
In order to explore the reasons for the consistent high
efficiency in the case of MB photocatalysis, we determined
relevant rate constants using laser flash photolysis techniques
with 308, 355, and 650 nm laser excitation (see Table 1 and
Supporting Information). The pseudo-first-order rate constant
for excited-state decay was monitored using transient
absorbance or luminescence for MB and Ru(bpy)3Cl2,
respectively; the second-order rate constant is obtained from
the slope of a plot of this decay against the concentration of
quencher. A key rate constant involves the reaction of the
triplet state with the sacrificial electron donor, iPr2NEt in this
case. The corresponding plot shows that MB is quenched 40
times faster than Ru(bpy)3Cl2 (see Table 1, lower section),
most likely one of the main reasons for the system’s remarkable
efficiency, Figure 2.
activity when directly compared to Ru(bpy)3Cl2, and also
provides the advantages of having lower cost and being
completely transition-metal-free. Beyond establishing high
conversions for a series of representative reactions, we have
studied the kinetics of the key processes using laser flash
photolysis techniques. While this Communication centers on
MB, we have also carried out comparative measurements using
Ru(bpy)3Cl2, so as to determine the origin of differences in
reactivity.
The system of reaction 1 was used as a test system in order
to compare the performance of MB with Ru(bpy)3Cl2; the term
“photocatalyst” refers to 1 mol% of one of these molecules. See
the upper section in Table 1.
13287
dx.doi.org/10.1021/ja406311g | J. Am. Chem. Soc. 2013, 135, 13286−13289

Journal of the American Chemical Society
Communication
Scheme 3. Mechanism of Methylene Blue Photocatalysis,
with Box A Adopted from Scheme 1 and the Reported
Mechanism2
Figure 2. Rate of triplet quenching as a function of [iPr2NEt] for
Ru(bpy)3Cl2 (red) and for methylene blue (blue). See Table 1 for
bimolecular rate constants and SI for a more detailed look at
Ru(bpy)3Cl2.
quencher, and typical quenching plots (similar to those in
Figure 2) are curved, reaching a plateau by the time the
concentration is ∼0.1 M (see Supporting Information). The
well-known monomer−dimer equilibrium18 is most likely
responsible for this kinetic anomaly.
We now return to Figure 1, which at low to moderate
conversions shows upward curvature; this observation puzzled
us initially, as normally reactions show negative curvature as the
reagents are consumed. However, data in Figure 1 and eq 2a
provide a clear explanation. Since kboronic ≫ kphenol, fewer
triplets are quenched by the product relative to the boronic acid
reagent, and thus more MB triplets survive to be quenched by
the amine as the reaction proceeds, due to the lowering of the
boronic acid concentration as it is converted to the
corresponding phenol, causing a slight positive curvature early
on. This is yet another example where kinetic data allowed for
detailed mechanistic interpretation.
Both MB and Ru(bpy)3Cl2 are readily quenched by oxygen
to yield singlet oxygen; however, at the concentrations used in
this work, the dominant pathway will be quenching by the
amine. In any event, we examined the quenching of singlet
oxygen by phenylboronic acid and iPr2NEt (see SI) and
obtained rate constants of 1.2 × 104 and 2.3 × 107 M−1 s−1,
respectively, indicating that singlet oxygen is 5000 times more
likely to be quenched by iPr2NEt over phenylboronic acid, and
thus that singlet oxygen does not play a significant role in
product formation. Singlet oxygen generation from MB or
Ru(bpy)3Cl2 was examined and gave consistent data, as
expected.
The fraction of triplets quenched by the amine can be readily
calculated using eq 2; note that while reagents are typically
specified in terms of “equivalents” in synthetic organic
chemistry, these calculations require actual concentrations.
The reported mechanism involves oxidative attack by
superoxide or HOO• formed as part of the catalytic cycle,
+
(% amine quenching) =
typically assigned to electron transfer from Ru(bpy)3 , which
100 × kqamine[amine]
was in turn formed by electron transfer from the sacrificial
electron donor. In our case, both the amine-derived α-amino
radical and the semireduced methylene blue (MB•) are capable
of transferring an electron to oxygen, with the rest of the
reaction with the boronic acid following a previously proposed
mechanism.2 In principle, both radicals, MB• and iPr2NEt+•, are
capable (although unlikely under oxygen) of participating in
disproportionation reactions.2 We have adopted box “A” from
the literature,2 even if the proposed disproportionation of
amine-derived radicals in the presence of oxygen seems
unusual.19 In any event, it is clear that entry into box A in
Schemes 1 and 3 leads exclusively to phenols.
In conclusion, our study demonstrates that in some cases the
use of organic dyes as sensitizers for photocatalysis can lead to
remarkably high yields using low-cost materials that are
ultimately environmentally benign in comparison to their
transition-metal counterparts. Understanding the increased
efficiency and the optimal reaction conditions can frequently
be done best on the basis of a detailed understanding of the
reaction kinetics, which in many cases can be readily
determined using laser flash photolysis techniques.
(τ0−1 + kqamine[amine] + kqboronic[boronic] + kqO2[O2] + kqphenol[phenol])
(2a)
(% amine quenching) ≈
100 × kqamine[amine]
(τ0−1 + kqamine[amine] + kqboronic[boronic] + kqO [O2])
2
(2b)
where the various kq terms refer to the rate constants of Table
1. Equation 2b refers to the initial condition in which phenol is
absent. From the conditions of reaction 1, we calculate that
71% of the MB triplets are quenched by the amine. For
comparison, for Ru(bpy)3Cl2 only 6% of the triplets will be
quenched by the amine. The latter calculation requires an
approximation for the boronic acid term, as the quenching
process is not linear (vide infra); the necessary data are
included in the Supporting Information. In fact, it is well known
that while aromatic amines are very good quenchers of excited
Ru(bpy)3Cl2,16 their aliphatic counterparts are rather poor
quenchers.16,17 In contrast, aliphatic amines are excellent
quenchers of the MB triplet.8 Final yields (as in Table 1 and
Chart 1) will also be influenced by the dynamics of all steps in
Scheme 3, such as the generation of superoxide and the catalyst
cycling; the spectral distribution of lamps and their overlap with
the photocatalyst absorption are also important and, while not
optimized in this work, deserve further examination.
ASSOCIATED CONTENT
* Supporting Information
■
S
Details on reaction conditions, product spectral data,
conversion vs time plots, laser flash photolysis data, quenching
plots, and NMR spectra. This material is available free of charge
via the Internet at http://pubs.acs.org.
Phenylboronic acids are excellent quenchers of excited states,
but it would appear that only the monomeric form is a good
13288
dx.doi.org/10.1021/ja406311g | J. Am. Chem. Soc. 2013, 135, 13286−13289

Journal of the American Chemical Society
Communication
AUTHOR INFORMATION
Corresponding Author
scaiano@photo.chem.uottawa.ca
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by the Natural Sciences and
Engineering Research Council of Canada, the Canadian
Foundation for Innovation, and the Canada Research Chairs
program.
REFERENCES
■
(1) Prier, C. K.; Rankic, D. A.; Macmillan, D. W. Chem. Rev. 2013,
113, 5322.
(2) Zou, Y. Q.; Chen, J. R.; Liu, X. P.; Lu, L. Q.; Davis, R. L.;
Jorgensen, K. A.; Xiao, W. J. Angew. Chem., Int. Ed. 2012, 51, 784.
(3) Ismaili, H.; Pitre, S. P.; Scaiano, J. C. Catal. Sci. Tech. 2013, 3,
935.
(4) Yoon, T. P.; Ischay, M. A.; Du, J. Nature Chem. 2010, 2, 527.
(5) Wilkinson, F.; Helman, W. P.; Ross, A. B. J. Phys. Chem. Ref. Data
1995, 24, 663. Tanielian, C.; Wolff, C. J. Phys. Chem. 1995, 99, 9831.
(6) Turro, N. J.; Ramamurthy, V.; Scaiano, J. C. Modern Molecular
Photochemistry of Organic Molecules; University Science Publishers:
New York, 2010.
(7) Obata, H.; Kogasaka, K.; Koizumi, M. Bull. Chem. Soc. Jpn. 1959,
32, 125.
(8) Kayser, R. H.; Young, R. H. Photochem. Photobiol. 1976, 24, 395.
(9) Klimov, A. D.; Lebedkin, S. F.; Emokhonov, V. N. J. Photochem.
Photobiol.: A 1992, 68, 191.
(10) Kayser, R. H.; Young, R. H. Photochem. Photobiol. 1976, 24, 403.
(11) Neumann, M.; Fuldner, S.; Konig, B.; Zeitler, K. Angew. Chem.,
Int. Ed. 2011, 50, 951.
(12) Hari, D. P.; Schroll, P.; Konig, B. J. Am. Chem. Soc. 2012, 134,
2958.
(13) Liu, H.; Feng, W.; Kee, C. W.; Zhao, Y.; Leow, D.; Pan, Y.; Tan,
C.-H. Green Chem. 2010, 12, 953. Pan, Y.; Kee, C. W.; Chen, L.; Tan,
C.-H. Green Chem. 2011, 13, 2682. Teo, Y. C.; Pan, Y.; Tan, C. H.
ChemCatChem 2013, 5, 235.
(14) Juillard, J. Pure Appl. Chem. 1977, 49, 885.
(15) Considerable formation of semireduced methylene blue was also
observed in its laser flash photolysis in DMF. Interestingly, it was the
time-resolved work that alerted us about possible problems with the
solvent.
(16) Sandrini, D.; Maestri, M.; Belser, P.; Von Zelewsky, A.; Balzani,
V. J. Phys. Chem. 1985, 89, 3675.
(17) Ballardini, R.; Varani, G.; Indelli, M. T.; Scandola, F.; Balzani, V.
J. Am .Chem. Soc. 1978, 100, 7219. Rivarola, C. R.; Bertolotti, S. G.;
Previtali, C. M. Photochem. Photobiol. 2006, 82, 213.
(18) Larkin, J. D.; Bhat, K. L.; Markham, G. D.; Brooks, B. R.;
Schaefer, H. F.; Bock, C. W. J. Phys. Chem. A 2006, 110, 10633.
(19) Most participating radicals will be rapidly scavenged by oxygen,
and thus unlikely to participate in radical−radical reactions. The
exceptions are superoxide and the hydroxyperoxyl radical, which are
unreactive toward oxygen.
13289
dx.doi.org/10.1021/ja406311g | J. Am. Chem. Soc. 2013, 135, 13286−13289