Robust Photocatalytic Method Using Ethylene-Bridged Flavinium Salts for the Aerobic Oxidation of Unactivated Benzylic Substrates

Article abstract of DOI:10.1002/adsc.202100024

7,8-Dimethoxy-3-methyl-1,10-ethylenealloxazinium chloride (1a) was found to be a superior photooxidation catalyst among substituted ethylene-bridged flavinium salts (R=7,8-diMeO, 7,8-OCH₂O-, 7,8-diMe, H, 7,8-diCl, 7-CF₃ and 8-CF₃). Selection was carried out based on structure vs catalytic activity and properties relationship investigations. Flavinium salt 1a proved to be robust enough for practical applications in benzylic oxidations/oxygenations, which was demonstrated using a series of substrates with high oxidation potential, i. e., 1-phenylethanol, ethylbenzene, diphenylmethane and diphenylmethanol derivatives substituted with electron-withdrawing groups (Cl or CF₃). The unique capabilities of 1a can be attributed to its high photostability and participation via a relatively long-lived singlet excited state, which was confirmed using spectroscopic studies, electrochemical measurements and TD-DFT calculations. This allows the maximum use of the oxidation power of 1a, which is given by its singlet excited state reduction potential of +2.4 V. 7,8-Dichloro-3-methyl-1,10-ethylenealloxazinium chloride (1 h) can be used as an alternative photocatalyst for even more difficult substrates. (Figure presented.).

Full text of DOI:10.1002/adsc.202100024

FULL PAPER
DOI: 10.1002/adsc.202100024
Very Important Publication
Robust Photocatalytic Method Using Ethylene-Bridged Flavinium
Salts for the Aerobic Oxidation of Unactivated Benzylic
Substrates
Adam Pokluda,a Zubair Anwar,b Veronika Boguschová,a Iwona Anusiewicz,c
Piotr Skurski,c Marek Sikorski,b,* and Radek Cibulkaa,*
a
Department of Organic Chemistry, University of Chemistry and Technology, Prague, Technická 5, 166 28 Prague 6, Czech
Republic
Fax: +420 220 444 288
Tel.: +420 220 444 182
E-mail: cibulkar@vscht.cz
b
Faculty of Chemistry, Adam Mickiewicz University in Poznań, Uniwersytetu Poznańskiego 8, 61–614 Poznań, Poland
Fax: +48 61 8291555
Tel.: +48 61 8291593
E-mail: sikorski@amu.edu.pl
Faculty of Chemistry, University of Gdańsk, Wita Stwosza 63, 80–308 Gdańsk, Poland
c
Manuscript received: January 7, 2021; Revised manuscript received: February 21, 2021;
Version of record online: ■■■, ■■■■
Supporting information for this article is available on the WWW under https://doi.org/10.1002/adsc.202100024
Abstract: 7,8-Dimethoxy-3-methyl-1,10-ethylenealloxazinium chloride (1a) was found to be a superior
photooxidation catalyst among substituted ethylene-bridged flavinium salts (R=7,8-diMeO, 7,8-OCH2O-, 7,8-
diMe, H, 7,8-diCl, 7-CF3 and 8-CF3). Selection was carried out based on structure vs catalytic activity and
properties relationship investigations. Flavinium salt 1a proved to be robust enough for practical applications
in benzylic oxidations/oxygenations, which was demonstrated using a series of substrates with high oxidation
potential, i.e., 1-phenylethanol, ethylbenzene, diphenylmethane and diphenylmethanol derivatives substituted
with electron-withdrawing groups (Cl or CF3). The unique capabilities of 1a can be attributed to its high
photostability and participation via a relatively long-lived singlet excited state, which was confirmed using
spectroscopic studies, electrochemical measurements and TD-DFT calculations. This allows the maximum use
of the oxidation power of 1a, which is given by its singlet excited state reduction potential of +2.4 V. 7,8-
Dichloro-3-methyl-1,10-ethylenealloxazinium chloride (1 h) can be used as an alternative photocatalyst for
even more difficult substrates.
Keywords: Aerobic oxidation; Electron transfer; Photochemistry; Photoredox catalysis; Spectroscopy
Introduction
1.67 V vs SCE).[13] The oxidative power of RFTA can
be enhanced upon complexation to Sc3+[14] or derivati-
During the recent rapid development of visible-light zation of the flavin structure via the introduction of an
photocatalytic oxidative transformations of benzylic electron-withdrawing group(s) or positive charge.[15–17]
C(sp3)-H group,[1,2] flavins[3] as well as acridinium Another improvement of the aerobic flavin photo-
salts,[4] eosin Y,[5] and other organic dyes[6–8] have been oxidative procedure is the addition of iron complexes
shown as catalysts with the potential to be sustainable to decompose the H2O2 by-product,[18] optimizing the
alternatives to more expensive iridium,[9,10] and flow arrangement with the participation of singlet
ruthenium[11,12] complexes. The usability of the most oxygen,[19] use of the micelle-effect,[20] and the involve-
common flavin derivative, riboflavin tetraacetate ment of electrochemistry.[21]
(RFTA), in photooxidative systems has limitations in
Aerobic benzylic oxidations promoted by ethylene-
terms of its excited state reduction potential (E*red = bridged flavinium salt 1g have been recently described
Adv. Synth. Catal. 2021, 363, 1–10
1
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

FULL PAPER
asc.wiley-vch.de
by our group,[16] which has introduced this type of mainly observed in a case of difficult electron-poor
flavin derivative in photoredox catalysis (Scheme 1). substrates that require long reaction times.
The procedure is based on the very positive reduction
To optimize the structure of ethylene bridged
potential of excited 1g* (E*red =2.67 V vs SCE in flavinium catalysts, we decided to synthesise a series
maximum[22]), which is able to activate a benzylic of derivatives 1 bearing various substituents at the 7-
substrate via single electron transfer (SET). The and 8-position to study the structure vs properties
resulting radical reacts with molecular oxygen to form relationship with a focus on the stability, photochem-
the appropriate carbonyl compound. Subsequently, 1g ical and electrochemical properties and reactivity
is regenerated by molecular oxygen. Hydrogen towards model substrates. This approach led us to
peroxide is decomposed by molecular sieves.
discover the most suitable derivatives of 1 for the
This method utilising 1g has been shown to be design of robust aerobic photocatalytic oxidation
useful for a large range of benzylic substrates, i.e., systems. The practical application of these new flavin
substituted toluene derivatives, alkylbenzenes and photocatalysts was verified in terms of the scope of
benzyl alcohols. However, the practical use of 1g has electron-poor benzylic substrates.
some limitations due its low photostability, which was
Results and Discussion
Synthesis
The synthesis of flavinium salts 1 (Scheme 2, see the
Supporting Information S1 for details) was started by
the condensation of an appropriate o-nitroaniline (2)
and alloxan in acetic acid in the presence of boric acid
to give the corresponding isoalloxazine 3. We at-
tempted to incorporate an octyl group at the N(3)-
position of the isoalloxazine via alkylation using octyl
iodide in a presence of potassium carbonate in DMF.
However, while the alkylation was successful for
isoalloxazines 3e, 3f and 3g, the procedure led to the
Scheme 1. Simplified mechanism of benzylic photooxidation decomposition of the isoalloxazine structure when
promoted by flavinium salt 1g.
using 7,8-dimethoxy and 7,8-dichloro derivatives 3a
Scheme 2. Synthesis of ethylene-bridged flavinium salts 1.
Adv. Synth. Catal. 2021, 363, 1–10
2
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

FULL PAPER
asc.wiley-vch.de
and 3h, respectively. Therefore, we prepared a series
The stability of the flavinium salt is an important
of 3-methyl analogues 4 via the condensation of 2a, issue for its application in photocatalytic systems.
2h and 2c with N-methylalloxan (alkyl group was Therefore, we compared the stability of salts 1 under
introduced to avoid problems with solubility of the conditions similar to a photocatalytic procedure re-
catalyst).[16] The synthesis of salts 1 was accomplished cently described:[16] Irradiation near the absorption
via quaternization of the alkylated isoalloxazines 4 in maxima in the presence of trifluoroacetic acid in
SOCl2. We also synthesised non-alkylated 7,8-dimeth- acetonitrile under an oxygen atmosphere. Trifluoro-
oxy derivative 1b to compare its properties with acetic acid was added to avoid the formation of 10a-
alkylated salt 1a. 7,8-Dimethyl derivative 1d was adducts with nucleophiles, which (i) do not absorb in
prepared from (À )-riboflavin according to a literature the visible light region and (ii) accelerate decomposi-
procedure.[23]
tion or allow other (undesired) oxidation processes to
occur.[16,25] In Table 2, the stability was expressed using
the half-life of the flavinium salts (see the Supporting
Information S4 for details).
Spectral and Electrochemical Properties
Electrochemical and spectral properties of the flavi-
Dimethoxy flavinium salts 1a and 1b were found
nium salts 1 were measured to estimate their behaviour to be extremely stable when compared to other
in photocatalytic systems and to obtain the optimal derivatives, whereas methylendioxy derivative 1c was
conditions for their application as a photocatalyst unstable due to the oxidative cleavage of the methylene
(Table 1 and the Supporting Information S2–S4). A group, which was confirmed using mass spectrometry.
bathochromic shift of the absorption maxima identical Derivative 1d with methyl groups had also enhanced
to the fluorescence maxima occurs for derivatives stability compared to non-substituted salt 1e. 7,8-
1a–d bearing electron-donating methyl or alkoxy Dichloro derivative 1h exhibited similar stability to 1e
groups. Moreover, alkoxy groups with a positive and both were more stable than trifluoromethyl
resonance effect cause a significant hyperchromic shift derivatives 1f and 1g.
in the spectra obtained for 1a–c.
The ground state reduction potential was measured
Catalytic Activity in the Model Photooxidation
using cyclic voltammetry and was in good agreement
Reaction
with the Hammett constants (see the Supporting
Information S4). The reduction potential in the singlet The catalytic efficiency of the flavinium salts in a
excited state E*red(S1), the value expressing the photooxidative system were compared using the
maximum ability of 1 to transfer an electron from the benzylic oxidation of model substrates bearing an
substrate molecule, was calculated upon the addition of electron-withdrawing group, 1-chloro-4-ethylbenzene
the ground state reduction potential Ered and first (5-Cl) and 1-ethyl-4-(trifluoromethyl)benzene (5-CF3),
singlet excited state energy estimated from the inter- to give their corresponding acetophenones 6 under the
section between normalized absorbance and emission previously described optimized conditions.[16] The
°
spectra.[24]
temperature was set at 40 C which was found as
optimal from the point of view of reaction conversions
(see the Supporting Information S6). The conversion
of both substrates to 6 after 8 h of irradiation using a
400 or 450 nm diode (close to their absorption
Table 1. Spectral and electrochemical properties of flavinium
salts 1.
Flavinium
salt
λmax
[nm]
λF
ɛ
Ered
E*red
[nm] [LmolÀ 1 [V][a]
cmÀ 1]
[V][b]
Table 2. Stability of flavinium salts 1 under irradiation.[a]
Flavinium salt
Irradiation
wavelength [nm]
Half-life
[min]
1a
1b
1c
1d
1e
449
449
450
449
408
406
401
420
520
520
500
500
490
490
480
500
17500 À 0.38
20500 À 0.30
15000 À 0.36
+2.2
+2.3
+2.3
+2.5
+2.6
+2.7
1a
1b
1c
1d
1e
1f
450
450
450
450
400
400
400
400
1100
900
25
300
65
19
7
85
9000
9000
5500
8000
9000
À 0.24
À 0.21
À 0.03
1f
1g[c]
1h
+0.02 +2.7
+0.12 +2.8
1g
1h
*
[a] Ground state potential of the first reversible wave E(1/1 ) vs
SCE measured by cyclic voltammetry.
[b] Calculated as E*red(S1)=Ered +E0–0(S1); E0–0(S1) taken from
intersection point (see the Supporting Information S2).
[c] Ref. [16].
[a] 1 (0.025 mmol), CF3COOD (1 mmol), CD3CN (1 mL).
°
Irradiated under O2 atmosphere at 40 C and monitored by
1H NMR.
Adv. Synth. Catal. 2021, 363, 1–10
3
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

FULL PAPER
asc.wiley-vch.de
maxima) are displayed in Table 3 (for more data, see substrates. Furthermore, we demonstrated the useful-
the Supporting Information S6). ness of these two catalysts in preparative experiments
Borderline catalytic activity can be observed in the performed on a 1-2 mmol scale (Table 4).
ability to mediate the oxidation of highly electron-
The conditions used for the preparative experiments
deficient 5-CF3. While catalysts bearing electron- were analogous to those used on an analytical scale.
donating substituents exhibit low efficacy, the non- We used a large dilution as it was found to be
substituted catalyst 1e and those bearing electron- beneficial for the reaction rate (see the Supporting
withdrawing groups show significant conversions with Information S6). In terms of the substrate scope, we
the best performance observed with 1h. Chlorinated mainly focused on electron-deficient compounds bear-
substrate 5-Cl was well oxidized in the presence of all ing chloro or trifluoromethyl groups. We studied the
the catalysts studied. One should bear in mind that oxidation of the benzylic CH2 group as well as the
catalysts 1a and 1b are still present and active after corresponding secondary benzylic hydroxyl function.
8 h of reaction according to lifetimes measurements The reaction mixtures were irradiated until the sub-
1
and according to the linear course of conversion vs strate/product ratio changed (monitored by H NMR
time dependence for 5-CF3 oxidation mediated by 1a. spectroscopy) or catalyst decomposed (loss of colour
Indeed, a 4-fold extension of the reaction time led to 4- and fluorescence).
fold increase in the reaction conversion with 1a. On
To our delight, 1a was shown to oxidize almost all
the other hand, oxidation of 5-CF3 by 1g decelerated of the substrates studied on preparative scale: Non-
during time because of catalysts decomposition (see substituted or chloro substituted ethylbenzene deriva-
the Supporting Information S7 for reaction profiles). tives 5-H and 5-Cl (Entries 1 and 3), 1-(4-chlorophen-
Interestingly, 1g showed some activity even after yl)ethanol (7-Cl) (Entry 7), diphenylmethanes 8 (En-
couple of hours irradiation which can be explained tries 10–13) and carbinols 9 (Entries 14–16), with high
either by positive effect of molecular sieves on stability conversion and yield to their corresponding ketones 6
(molecular sieves were not present in photostability and 10, even when using 1.5 mol% catalyst (Entry 12).
measurements) or by non-zero activity of decomposi- With 0.5 mol% of 1a, moderate conversion was
tion products.
achieved (Entry 13). Oxidation of electron rich meth-
oxyderivative 5-OMe (Entry 2) provided ketone 6-
OMe containing small amount (5%) of benzoic acid
(product of overoxidation) and traces of alcohol
Substrate Scope – Preparative Experiments
The high stability of 1a combined with the high intermediate. Similarly, traces of the corresponding
reduction potential and sufficient stability of 1h makes benzoic acid were observed in oxidations of 5-Cl and
these two catalysts promising candidates to replace 1g 7-Cl (Entries 3 and 7). Flavin 1a even oxidized very
in the photooxidative protocol used for electron-poor difficult substrate, alcohol 7-CF3, with high yield
(Entry 8), albeit for the longer reaction time and
catalyst loading required. Conversion of 5-CF3 to 6-
CF3 with 1a was rather small on preparative scale
(Entry 4) in contrast to analytical scale (see Table 3,
Entry 1).
Table 3. Photocatalytic activity of flavinium salts 1.[a]
For substrates 5-CF3 and 7-CF3 with high oxidation
potential, 1h was applied as an alternative and
potentially stronger oxidation catalyst considering its
excited state reduction potential. The oxidations using
1h occurred with good conversions, which were higher
than those obtained using the original catalyst 1g (cf
Entries 5 and 6).
In general, comparing photocatalysts 1a and 1h,
1a is superior. The high stability of catalyst 1a allows
long-term oxidation to be carried out. Moreover, the
reaction using 1a is very simple with no need of
temperature control (it should be noted that the
Catalyst
Irradiation
λ [nm]
Conversion to 6[b,c] [%]
5-Cl
5-CF3
1a
1b
1c
1d
1e
1f
450
450
450
450
400
400
400
400
84
86
39
88
89
78
74
74
15 (61[d])
10
2
8
43
47
65
80
1g
1h
°
reaction mixture reaches up to 45 C during irradi-
[a] 5 (0.1 mmol), 1 (0.005 mmol; 5 mol. %), MS 4 A (15 mg)
and CF3COOH (0.2 mmol) dissolved in CH3CN (0.2 mL).
ation). To demonstrate usefulness of 1a on large scale,
we performed oxidations starting from 2.02 g of 8-Cl
and 2.19 g of 9-Cl (10 mmol) obtaining product 10-Cl
in preparative yields 74 and 85%, respectively, after
12 h irradiation.
°
Irradiated under O2 atmosphere at 40 C for 8 h.
[b] Determined by 1H NMR using dimethylsulfone as a standard.
[c] Traces of benzoic acid were observed.
[d] 32 h.
Adv. Synth. Catal. 2021, 363, 1–10
4
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

FULL PAPER
asc.wiley-vch.de
In contrast to 1a, there is no potential to enhance
Table 4. Preparative experiments on photooxidations.
the yield of the photooxidation reactions using 1h
upon prolonging the reaction time because of relatively
fast bleaching. The reaction temperature used in the
large-scale experiments using 1h had to be controlled
°
at 20 C to decelerate the decomposition of the catalyst.
On the other hand, as a powerful alternative to 1a, 1h
seems to be more useful than original derivative 1g.
Mechanistic Investigations
Our previous mechanistic studies using mass spectrom-
etry and EPR on benzylic oxidations using 1g
confirmed that the reaction occurs via a benzylic
*
+
radical and that protonated flavin radical FlH was a
product of electron and proton transfer from the
substrate.[16,17] Herein, we also found that radical chain
mechanism is not involved as evident from the
measured quantum yields Φ=0.0047 and 0.055 of 5-
Cl photooxidation mediated by 1a and 1h, respec-
tively (see the Supporting Information S13). Never-
theless, a question appeared in regard the excited state
of the flavinium species and whether singlet oxygen
was involved in the reaction. Therefore, we performed
a series of photophysical measurements on 1a or 1h
using a chlorotoluene (CT) model substrate utilising
steady-state and transient absorption spectroscopy as
well as TD-DFT calculations.
Singlet and triplet excitation energies were calcu-
lated employing the TD-DFT/6-31+ +G(d,p) level of
theory (see Experimental and the Supporting Informa-
tion S14 and S17). Figure 1 shows the singlet
excitation energies compared with the experimental
absorption spectra. Table 5 summarises the theoretical
Table 5. Important data for electron transfer with 1a and 1h.
Data for 1a
Data for 1h
E0–0(S1) [eV]
E0–0(T1) [eV]
E*red(S1) [V]
E*red(T1) [V]
τF [ns]
2.80[a] (2.6[b])
1.79[a]
2.84[a] (2.7[b])
2.10[a]
2.42[a] (2.2[b])
1.41[a]
2.96[a] (2.8[b])
2.22[a]
[a] Substrate (2 mmol), catalyst 1 (5 mol. %), MS 4 A (300 mg)
and CF3COOH (2 eq.) dissolved in CH3CN (16 mL).
Irradiated under O2 atmosphere. No temperature control.
[b] Substrate (1 mmol), catalyst 1 (5 mol. %), MS 4 A (150 mg)
and CF3COOH (2 eq.) dissolved in CH3CN (12 mL).
6.10
0.31
0.69
0.46
0.079
0.82
ΦF
ΦΔ
singlet excited state quenching under argon
k(CT) [MÀ 1 sÀ 1]
3.1×1010
0.9×1010
1.5×1010
14×1010
3.4×1010
1.9×1010
k(MB) [MÀ 1 sÀ 1]
k(DMB) [MÀ 1 sÀ 1]
triplet state qÀu1enching
k(CT) [MÀ 1 s ]
°
Irradiated under O2 atmosphere at 20 C.
[c] 10 mol. % of 1a.
[d] 1.5 mol. % of 1a.
[e] 0.5 mol. % of 1a.
[f] NMR conversion.
[g] Preparative yield.
[h] Benzoic acid (�5%) formed.
[i] Not determined.
1.8×108
1.9×1010
4.1×1010
n.d.
k(EDTA) [MÀ 1 sÀ 1]
[a] Value from calculated data.
[b] Value based on spectral measurements.
τF: fluorescence lifetime; ΦF: fluorescence quantum yield; ΦΔ:
singlet oxygen quantum yield.
Adv. Synth. Catal. 2021, 363, 1–10
5
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

FULL PAPER
asc.wiley-vch.de
using laser-flash photolysis (see Table 5 and the
Supporting Information S9 and S11). Interestingly, the
fluorescence quenching rate constants for CT are
higher when compared to the diffusion limit of
acetonitrile and the quenching rate constant of meth-
oxybenzene (MB, Eox =1.73 V vs SCE) and 1,3-
dimethoxybenzene (DMB, Eox =1.50 V vs SCE),
which are better electron donors. This observation can
be explained by the formation of a complex between 1
and CT accelerating the electron transfer process. The
interaction between 1a and CT was confirmed using a
1
titration study monitored using UV-VIS or H NMR
spectroscopy (see the Supporting Information S9).
Analogous interaction between 1h and MB is pro-
posed which is manifested by unexpected higher
quenching rate constant of 1h fluorescence with MB
compared to DMB.
The triplet state quenching rate constant for 1a
with CT was lower by two orders of magnitude when
compared to that obtained for the singlet excited state
and the value is also significantly lower than that with
EDTA, which is known as a good quencher of flavins
(see the Supporting Information S11). The data
obtained for the quenching rates together with the
thermodynamic values for the photoinduced electron
transfer and relatively long lifetime of the singlet
excited state of 1a (6.1 ns in the presence of oxygen)
support the idea that the oxygenation reaction occurs
via the singlet excited state of 1a. On the other hand,
the participation of 1h triplet excited state during the
photooxidation process is expected because of the very
short lifetime of its singlet excited state.
Figure 1. The predicted lowest-energy singlet-singlet π,π*
transitions for 1a (A) and 1h (B) – upper panels – compared to
the experimental absorption spectra in acetonitrile – bottom
panels. Red triangles represent the weak n,π* transitions. The
experimental fluorescence spectra in acetonitrile are also shown
(in blue).
The presence of chlorotoluene (CT) significantly
decreases the efficiency of singlet oxygen sensitization
by both 1a and 1h as evident from the singlet oxygen
quantum yield vs CT concentration relationships (see
values obtained for E0–0(S1) and E0–0(T1), which give the Supporting Information S12). While singlet oxygen
the most information in regard electron transfer in the quantum yield ΦΔ is 0.69 and 0.82 for 1a and 1h,
photooxidation reactions. The E0–0(S1) values calcu- respectively in the absence of CT, it decreases to 0.18
lated from the experimental data based on UV-VIS- and 0.15 at a concentration of 20 mM for this substrate.
fluorescence spectra intersection point are slightly Keeping in mind that the concentration of the substrate
lower (see values in parenthesis). From the values is even higher (>0.1 M) in standard photooxidations,
obtained for the excitation energies and ground state singlet oxygen participation for this reaction is
reduction potentials, the corresponding excited state unlikely.
reduction potentials, E*red(S1) and E*red(T1), were
calculated. Upon comparing the excited state reduction
potentials of 1a and 1h with the ground state oxidation
Conclusions
potential of chlorotoluene (Eox =2.21 V vs SCE), We have described a new photocatalytic procedure for
electron transfer from the triplet excited state seems to the aerobic oxidation/oxygenation of benzylic sub-
be endergonic for 1a and with a ΔG value close to strates mediated by ethylene-bridged flavinium salt 1a
zero for 1h. On the other hand, electron transfer is and blue light. Use of dimethoxyflavinium salt 1a
markedly exergonic for both catalysts in their singlet instead of trifluoromethyl derivative 1g, mentioned in
excited states.
a preliminary report[16] is advantageous because of its
Measuring the bimolecular quenching rate constant substantially higher photostability allowing a simple
of 1a and 1h in the singlet excited state with CT was reaction arrangement without temperature control. The
performed using static and dynamic fluorescence method seems to be unique among flavin-based photo-
measurements. Triplet state quenching was studied oxidation systems as it is both metal-free and robust
Adv. Synth. Catal. 2021, 363, 1–10
6
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

FULL PAPER
asc.wiley-vch.de
and allows the oxidation of highly electron-deficient
substrates. In this aspect, the method exceeds other
photooxidation methodologies performed using organic
dyes and it is comparable with a method using a sub-
General Procedures for Photocatalytical Oxidations
Experiments on analytical scale. 5-Cl or 5-CF3 (0.1 mmol)
and flavinium salt 1 (0.005 mmol, 5 mol. %) were dissolved in
acetonitrile (200 uL) in a vial. CF3COOH (15.3 μL, 2 eq.) and
molecular sieves 4 (15 mg) were added. The reaction mixture
was bubbled with oxygen from balloon through a septum for 2
minutes and then, stirred and irradiated (Luxeon STAR/0,
4x1030 mW@700 mA, 450 or 400 nm) for 8 h under oxygen
stoichiometric
amount
of
sodium
trifluoromethanesulfinate.[26] Our mechanistic investi-
gations show that the reaction starts via electron
transfer from the benzylic substrate to the singlet
excited state of 1a, which offers the maximum
oxidizing power. Alternatively, 1h can be used instead
of 1a for substrates with extremely high oxidation
potential.
°
(balloon) at 40 C. After irradiation, 800 μL of deuterated
DMSO with dimethylsulfone as an inert standard was added.
Mixture was filtrated by syringe frit into a cuvette and 1H NMR
spectrum was measured.
Some ethylene-bridged flavinium salts 1 have been
Experiments on preparative scale using 1a. 1a (36 mg;
0.1 mmol; 5 mol. %) or (11 mg; 0.03 mmol; 1.5 mol. %) or
(3.5 mg; 0.01 mmol; 0.5 mol. %), substrate (2 mmol) and
trifluoroacetic acid (306 μL; 4 mmol) were dissolved in
anhydrous acetonitrile (16 mL) in 50 mL Erlenmeyer flask
equipped with balloon filled with oxygen. Molecular sieves 4 A
(300 mg) were added. The reaction mixture was bubbled with
oxygen from balloon through a septum for 2 minutes and then,
previously
studied
in
organocatalysis
or
biocatalysis.[25b,27] Attempts to achieve the high effi-
ciency of 1 in oxidation reactions have often led to
derivatives bearing electron-withdrawing group(s). In
photocatalysis, too electron-deficient flavinium salts
seem to be disadvantageous. However, methoxy
functionalities have a positive effect on the properties
of the flavin photocatalyst causing its high stability,
while maintaining its highly positive excited state
reduction potential. A similar effect has already been
described among the neutral alloxazines used in energy
transfer cycloadditions,[28] 5-ethylalloxazinium salts
used in [2+2] cycloelimination,[29] and acridinium
salts used in various photooxidative processes.[30] We
plan to further investigate this effect among flavin
derivatives and to evaluate the performance of stable
and powerful salt 1a in other photooxidative proc-
esses.
stirred
and
irradiated
(4x
Luxeon
STAR/0,
4x1030 mW@700 mA, 450 nm) under oxygen atmosphere.
Temperature was not adjusted (orientation measurements
°
showed temperature around 40 C under abovementioned
conditions). Reaction was monitored by H NMR. Silica gel (~
3 g) was added and solvent evaporated. Crude reaction mixture
was purified by flash chromatography (hexane:EtAc). See the
Supporting Information S8 for details.
1
Experiments on preparative scale using 1h or 1g. 1h
(18 mg; 0.05 mmol; 5 mol. %) or 1g (23 mg; 0.05 mmol; 5 mol.
%), substrate (1 mmol) and trifluoroacetic acid (0.153 mL;
2 mmol) were dissolved in anhydrous acetonitrile (12 mL) in
20 mL double-walled reactor equipped with balloon filled with
oxygen. Molecular sieves 4 A (150 mg) were added. The
reaction mixture was bubbled by oxygen from balloon through
a septum for 2 minutes and then, the reaction mixture was
Experimental Section
General Comments to the Starting Material and
Synthesis
stirred
and
irradiated
(6x
Luxeon
STAR/0,
4x1030 mW@700 mA, 400 nm) under oxygen atmosphere.
Starting materials and reagents were obtained from commercial
suppliers and used without further purification. The solvents
were purified and dried using standard procedures. Flavinium
salts 1, were prepared and characterized as described in the the
Supporting Information S1, S15 and S16. NMR spectra were
°
Temperature (20 C) was controlled by thermostat. Reaction
1
was monitored by H NMR. Silica gel (~3 g) was added and
solvent evaporated. Crude reaction mixture was purified by
flash chromatography (hexane:EtAc). See the Supporting
Information S8 for details.
1
recorded on a Varian Mercury Plus 300 (299.97 MHz for H,
75.44 MHz for 13C, and 282.23 MHz for 19F) or Agilent 400-
Experiments on gram-scale. 1a (90 mg; 0.25 mmol; 2.5 mol.
%), substrate (10 mmol) and trifluoroacetic acid (1.5 mL;
20 mmol) were dissolved in anhydrous acetonitrile (160 mL) in
photoreactor equipped by water-cooled 460 nm light source
(novaLIGHT TLED100, 100 W). Molecular sieves 4 A (1.5 g)
were added. The reaction mixture was bubbled with oxygen via
cannula through a septum, stirred and irradiated. Temperature
1
MR DDR2 (399.94 MHz for H and 100.58 MHz for 13C) at
298 K unless otherwise indicated. Chemical shifts δ are given in
ppm, using residual solvent or tetramethylsilane as an internal
standard. Coupling constants J are reported in Hz. High-
resolution mass spectra were obtained on Q-Tof Micro
(Waters), equipped with a quadrupole and time-of-flight (TOF)
analyser and subsequent a multichannel plate (MCP) detector.
Thin layer chromatography (TLC) analyses were carried out on
a DC Alufolien Kieselgel 60 F254 (Merck). Preparative column
chromatography separations were performed on a silica gel
Kieselgel 60 0.040–0.063 mm (Merck). Flash chromatography
was performed at Büchi Pure C-810 at Silica 40 μm irregular
column. Melting points were measured on a Boetius melting
point apparatus and are uncorrected.
°
was adjusted by lamp-cooling system to approx. 30 C. After
6 h of irradiation, the second dose of 1a (90 mg; 0.25 mmol;
2.5 mol. %) was added and the reaction mixture was irradiated
1
for another 6 h (monitoring by H NMR). Silica gel (~6 g) was
added and solvent was evaporated. Crude reaction mixture was
purified by flash chromatography (hexane:EtAc). See the
Supporting Information S8 for details.
Adv. Synth. Catal. 2021, 363, 1–10
7
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

FULL PAPER
asc.wiley-vch.de
equilibrium with ambient air. Quantum yield of singlet oxygen
Electrochemical Measurements
and lifetime of singlet oxygen measurements for 1 was
determined by preparing the series of dilutions ranging from
0.02 to 0.10 nm and excited at their respective absorption
maxima with detection at 1270 nm using. Perinepthenone
(PNF) was used as standard to determine the quantum yield of
singlet oxygen by exciting it at 356 nm (see the Supporting
Information S12 for details).
Cyclic voltammograms were recorded using Metrohm Autolab
PGSTAT128 N, in a 25 mL cell of Metrohm type, in dry
acetonitrile, sample concentration was 5×10À 3 mol/L. Tetrabu-
tylammonium hexafluorophosphate, [Bu4N][PF6], puriss., elec-
trochemical grade, Fluka, was used as supporting electrolyte in
concentration 0.1 mol/L. The sample solution was prepared
directly in the cell, it was deareated by a stream of argon before
measurement. A glassy carbon working electrode was used,
together with Pt wire auxiliary electrode and Ag wire pseudo-
reference electrode. Scan rate was 50, 100, 200, 300 mV/s. Data
were referenced to SCE (ferrocene was used as inner standard).
Ered values were calculated as an average of upper and bottom
wave and as an average of results from measurement by scan
rate 50, 100, 200, 300 mV/s. See the Supporting Information S4
for details.
DFT Calculations
The equilibrium structures and energies of the positively
charged cations from 1a and 1h were determined by employing
the Density Functional Theory (DFT) with Becke’s Three
Parameter Hybrid Method with the LYP (Lee-Yang-Parr)
correlation functional (B3LYP)[31] together with the 6-31+ +
G(d,p)[32] basis sets. Electronic excitation energies for singlet
and triplet states of these systems and the corresponding
oscillator strengths were obtained from the TD-DFT
technique[33] using the same B3LYP functional and 6-31+ +
G(d,p) basis set and at the equilibrium geometry of the
absorbing species. The harmonic vibrational frequencies (un-
scaled) and the energies of the triplet states of 1a and 1h
cations at their equilibrium structures were obtained at the same
TD-DFT 6-31+ +G(d,p) theory level to assure consistency. All
of the calculations were performed with the Gaussian16
program suite.[34] See the Supporting Information S14 and S17
for details.
Photophysical Measurements
Absorption spectra for flavinium salts 1 were performed on
Shimadzu UV-vis spectrophotometer (UV-2550) using quartz
cells of 1 cm. The emission spectra were measured using
Horiba Jobin Yvon Fluorolog-3 spectrofluorometer and the
optical density in each case was less than 0.1 nm. The excitation
wavelengths were 449, 449, 450, 449, 408, 406 and 420 nm for
1a, 1b, 1c, 1d, 1e, 1f and 1h, respectively. See the Supporting
Information S2 for details
Fluorescence quantum yields of flavinium salts 1 were
determined by preparing series of dilution ranging in absorption
from 0.02 to 0.10 and excited at their respective absorption Acknowledgements
maxima. Lumichrome (LC) has been used as a standard. See
the Supporting Information S3 for details.
This project was supported by the Czech Science Foundation
(Grant No. 18-15175S). M.S. and Z.A. would like to thank the
National Science Centre, Poland (NCN), for the support
(research grant 2017/27/B/ST4/02494).
The fluorescence quenching experiments for 1a and 1h were
carried out using chlorotoluene (CT), methoxybenzene (MB)
and dimethoxybenzene (DMB) in the concentration range of
0.5–20 mM using FluoTime 300 Fluorescence Lifetime Spec-
trometer (PicoQuant). The excitation/emission wavelength for
1a and 1h were 440/520 nm and 440/500 nm, respectively. See
the Supporting Information S9 for details
References
[1] X. Zhang, K. P. Rakesh, L. Ravindar, H.-L. Qin, Green
Chem. 2018, 20, 4790–4833.
Transient absorption spectra of 1a and 1h were measured
using laser flash photolysis system build in our lab. All the
measurements have been performed using 1×0.4 cm quartz
cells in the presence and absence of oxygen and quenchers (CT,
EDTA). The 3ω 2ω Q-smart 450 Quantel laser of 355 nm was
used with the energy of 2 mJ/pulse and the signals were
detected through PDA detector. The optical density for each
sample was around 0.50 at 355 nm and the 1 cm face of cuvette
is facing toward laser and 0.4 cm towards lamp. See the
Supporting Information S11 for details
[2] a) K. Chen, P. Zhang, Y. Wang, H. Li, Green Chem.
2014, 16, 2344–2374; b) L. Revathi, L. Ravindar, W.-Y.
Fang, K. P. Rakesh, H.-L. Qin, Adv. Synth. Catal. 2018,
360, 4652–4698.
[3] B. König, S. Kümmel, E. Svobodová, R. Cibulka, in
Physical Sciences Reviews, Vol. 3, 2018, DOI: 10.1515/
psr-2017–0168.
[4] a) K. Ohkubo, S. Fukuzumi, Org. Lett. 2000, 2, 3647–
3650; b) H. Yi, C. Bian, X. Hu, L. Niu, A. Lei, Chem.
Commun. 2015, 51, 14046–14049; c) H. Liu, L. Ma, R.
Zhou, X. Chen, W. Fang, J. Wu, ACS Catal. 2018, 8,
6224–6229.
[5] A. K. Yadav, L. S. Yadav, Tetrahedron Lett. 2015, 56,
6696–6699.
[6] X. Liu, L. Lin, X. Ye, C.-H. Tan, Z. Jiang, Asian J. Org.
Chem. 2017, 6, 422–425.
[7] F. Rusch, J.-C. Schober, M. Brasholz, ChemCatChem
2016, 8, 2881–2884.
Singlet oxygen generation measurements. The characteristic
steady-state emission spectra of singlet oxygen were recorded
on the Jobin Yvon-Spex Fluorolog 3–221 spectrofluorometer
with an H10330B-75 NIR-PMT module, sensitive in the 950 to
1700 nm NIR range. Xenon lamp with a monochromator was
used as an excitation source for the steady-state spectra.
Phosphorescence decays were measured with SpectraLed diode
(λexc =371 nm; λem =1270 nm) on the same spectrofluorometer.
The solutions were studied in 1×1 cm2 quartz cuvette in
Adv. Synth. Catal. 2021, 363, 1–10
8
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

FULL PAPER
asc.wiley-vch.de
[8] W. Schilling, D. Riemer, Y. Zhang, N. Hatami, S. Das, [26] X. Zhu, Y. Liu, C. Liu, H. Yang, H. Fu, Green Chem.
ACS Catal. 2018, 8, 5425–5430. 2020, 22, 4357–4363.
[9] a) G. Pandey, R. Laha, D. Singh, J. Org. Chem. 2016, [27] a) P. Thapa, S. Hazoor, B. Chouhan, T. T. Vuong, F. W.
81, 7161–7171; b) J. W. Tucker, J. M. R. Narayanam,
P. S. Shah, C. R. J. Stephenson, Chem. Commun. 2011,
47, 5040–5042; c) P. Lu, T. Hou, X. Gu, P. Li, Org. Lett.
2015, 17, 1954–1957; d) T. Hou, P. Lu, P. Li, Tetrahe-
dron Lett. 2016, 57, 2273–2276.
Foss, J. Org. Chem. 2020, 85, 9096–9105; b) S. Chen,
M. S. Hossain, F. W. Foss, ACS Sustainable Chem. Eng.
2013, 1, 1045–1051; c) S. Chen, F. W. Foss, Org. Lett.
2012, 14, 5150–5153; d) A. T. Murray, M. J. Dowley, F.
Pradaux-Caggiano, A. Baldansuren, A. J. Fielding, F.
Tuna, C. H. Hendon, A. Walsh, G. C. Lloyd-Jones, M. P.
John, D. R. Carbery, Angew. Chem. Int. Ed. 2015, 54,
8997–9000; Angew. Chem. 2015, 127, 9125–9128;
e) A. T. Murray, P. Matton, N. W. G. Fairhurst, M. P.
John, D. R. Carbery, Org. Lett. 2012, 14, 3656–3659.
[10] a) D. Hager, D. W. C. MacMillan, J. Am. Chem. Soc.
2014, 136, 16986–16989; b) J. Jin, D. W. C. MacMillan,
Angew. Chem. Int. Ed. 2015, 54, 1565–1569; Angew.
Chem. 2015, 127, 1585–1589; c) J. D. Cuthbertson,
D. W. MacMillan, Nature 2015, 519, 74–77; d) K.
Qvortrup, D. A. Rankic, D. W. C. MacMillan, J. Am. [28] V. Mojr, G. Pitrová, K. Straková, D. Prukała, S. Brazevic,
Chem. Soc. 2014, 136, 626–629.
[11] a) P. Kohls, D. Jadhav, G. Pandey, O. Reiser, Org. Lett.
2012, 14, 672–675; b) L. Ruiz Espelt, I. S. McPherson,
E. Svobodová, I. Hoskovcová, G. Burdziński, T. Slanina,
M. Sikorski, R. Cibulka, ChemCatChem 2018, 10, 849–
858.
E. M. Wiensch, T. P. Yoon, J. Am. Chem. Soc. 2015, 137, [29] T. R. M. Hartman, J. Chudoba, E. Svobodová, N.
2452–2455.
Archipowa, J. R. Kutta, R. Cibulka, ChemPlusChem
2021, 86, 373–386.
[12] a) W. Li, X. Zhu, H. Mao, Z. Tang, Y. Cheng, C. Zhu,
Chem. Commun. 2014, 50, 7521–7523; b) W. Li, Y. Zhu, [30] A. Joshi-Pangu, F. Lévesque, H. G. Roth, S. F. Oliver,
Y. Duan, M. Zhang, C. Zhu, Adv. Synth. Catal. 2015,
357, 1277–1282.
L.-C. Campeau, D. Nicewicz, D. A. DiRocco, J. Org.
Chem. 2016, 81, 7244–7249.
[13] S. Fukuzumi, K. Yasui, T. Suenobu, K. Ohkubo, M. [31] a) C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37,
Fujitsuka, O. Ito, J. Phys. Chem. A 2001, 105, 10501–
785–789; b) A. D. Becke, Phys. Rev. A 1988, 38, 3098–
10510.
3100.
[14] B. Mühldorf, R. Wolf, Chem. Commun. 2015, 51, 8425– [32] a) R. Krishnan, J. S. Binkley, R. Seeger, J. A. Pople, J.
8428.
Chem. Phys. 1980, 72, 650–654; b) A. D. McLean, G. S.
Chandler, J. Chem. Phys. 1980, 72, 5639–5648.
[33] a) R. Bauernschmitt, R. Ahlrichs, Chem. Phys. Lett.
1996, 256, 454–464; b) M. E. Casida, C. Jamorski, K. C.
Casida, D. R. Salahub, J. Chem. Phys. 1998, 108, 4439–
4449; c) G. Scalmani, M. J. Frisch, B. Mennucci, J.
Tomasi, R. Cammi, V. Barone, J. Chem. Phys. 2006, 124,
094107.
[15] A. H. Tolba, F. Vávra, J. Chudoba, R. Cibulka, Eur. J.
Org. Chem. 2020, 1579–1585.
[16] J. Zelenka, E. Svobodová, J. Tarábek, I. Hoskovcová, V.
Boguschová, S. Bailly, M. Sikorski, J. Roithová, R.
Cibulka, Org. Lett. 2019, 21, 114–119.
[17] J. Zelenka, R. Cibulka, J. Roithová, Angew. Chem. Int.
Ed. 2019, 58, 15412–15420.
[18] B. Mühldorf, R. Wolf, Angew. Chem. Int. Ed. 2016, 55, [34] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E.
427–430; Angew. Chem. 2016, 128, 437–441.
[19] M. Lesieur, C. Genicot, P. Pasau, Org. Lett. 2018, 20,
1987–1990.
[20] Y. Masahide, N. Takuya, K. Yasumasa, S. Tsutomu, Bull.
Chem. Soc. Jpn. 2003, 76, 601–605.
[21] W. Zhang, K. L. Carpenter, S. Lin, Angew. Chem. Int.
Ed. 2020, 59, 409–417.
[22] The value corresponds to singlet excited state reduction
potential of 1g calculated from ground state reduction
potential and E0–0(S1) estimated from spectral data, see
ref. [16].
[23] T. Sakai, T. Kumoi, T. Ishikawa, T. Nitta, H. Iida, Org.
Biomol. Chem. 2018, 16, 3999–4007.
[24] N. A. Romero, D. A. Nicewicz, Chem. Rev. 2016, 116,
10075–10166.
[25] a) J. Žurek, E. Svobodová, J. Šturala, H. Dvořáková, J.
Svoboda, R. Cibulka, Tetrahedron: Asymmetry 2017, 28,
1780–1791; b) S. Chen, M. S. Hossain, F. W. Foss, Org.
Lett. 2012, 14, 2806–2809.
Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V.
Barone, G. A. Petersson, H. Nakatsuji, X. Li, M.
Caricato, A. V. Marenich, J. Bloino, B. G. Janesko, R.
Gomperts, B. Mennucci, H. P. Hratchian, J. V. Ortiz,
A. F. Izmaylov, J. L. Sonnenberg, Williams, F. Ding, F.
Lipparini, F. Egidi, J. Goings, B. Peng, A. Petrone, T.
Henderson, D. Ranasinghe, V. G. Zakrzewski, J. Gao, N.
Rega, G. Zheng, W. Liang, M. Hada, M. Ehara, K.
Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima,
Y. Honda, O. Kitao, H. Nakai, T. Vreven, K. Throssell,
J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro, M. J.
Bearpark, J. J. Heyd, E. N. Brothers, K. N. Kudin, V. N.
Staroverov, T. A. Keith, R. Kobayashi, J. Normand, K.
Raghavachari, A. P. Rendell, J. C. Burant, S. S. Iyengar,
J. Tomasi, M. Cossi, J. M. Millam, M. Klene, C. Adamo,
R. Cammi, J. W. Ochterski, R. L. Martin, K. Morokuma,
O. Farkas, J. B. Foresman, D. J. Fox, Wallingford, CT,
2016.
Adv. Synth. Catal. 2021, 363, 1–10
9
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

FULL PAPER
Robust Photocatalytic Method Using Ethylene-Bridged
Flavinium Salts for the Aerobic Oxidation of Unactivated
Benzylic Substrates
Adv. Synth. Catal. 2021, 363, 1–10
A. Pokluda, Dr. Z. Anwar, V. Boguschová, Dr. I. Anusiewicz,
Prof. P. Skurski, Prof. M. Sikorski*, Prof. Dr. R. Cibulka*