Electron-deficient alloxazinium salts: Efficient organocatalysts of mild and chemoselective sulfoxidations with hydrogen peroxide

Article abstract of DOI:10.1002/adsc.201300617

A series of substituted alloxazinium perchlorates has been prepared and tested as catalysts for the oxidation of sulfides to sulfoxides with hydrogen peroxide. The logarithms of the observed rate constants of thioanisole oxidation correlate with the Hammett σ constants of the substituents on the alloxazinium catalysts, as well as with their reduction potentials E ^0′ and their pK_R+ values, representing the alloxazinium salt/pseudobase equilibrium. The stronger the electron-withdrawing substituent, the more efficient is the alloxazinium catalyst. The alloxazinium salts with a cyano or trifluoromethyl group in position 8 proved to be the most efficient, operating at room temperature at small loadings, down to 0.1 mol%, achieving turnover number values of up to 640 and acceleration by a factor of 350 relative to the non-catalyzed oxidation. The 8-cyanoalloxazinium perchlorate was evaluated as the best catalyst; however, due to its relatively good accessibility, the 8-(trifluoromethyl)alloxazinium perchlorate seems to be the catalyst of choice for sulfoxidations with hydrogen peroxide. It was successfully tested for the sulfoxidation of a series of aliphatic and aromatic sulfides on a preparative scale. It produced the corresponding sulfoxides in quantitative conversions and with high isolated yields (87-98%). No over-oxidation to sulfone was ever observed. Copyright

Full text of DOI:10.1002/adsc.201300617

FULL PAPERS
DOI: 10.1002/adsc.201300617
Electron-Deficient Alloxazinium Salts: Efficient Organocatalysts
of Mild and Chemoselective Sulfoxidations with Hydrogen
Peroxide
a
b
c
d
Petra Mꢀnovꢁ, Hana Dvo rˇ ꢁkovꢁ, Vꢁclav Eigner, Ji rˇ ꢂ Ludvꢂk,
a,
and Radek Cibulka *
Department of Organic Chemistry, Institute of Chemical Technology, Prague, Technickꢁ 5, 166 28 Prague 6, Czech
Republic
a
Fax : (+420)-220-444-288; phone: (+420)-220-443-688; e-mail: cibulkar@vscht.cz
Central Laboratories, Institute of Chemical Technology, Prague, Technickꢁ 5, 166 28 Prague 6, Czech Republic
Department of Solid State Chemistry, Institute of Chemical Technology, Prague, Technickꢁ 5, 166 28 Prague 6, Czech
Republic
b
c
d
Department of Molecular Electrochemistry, J. Heyrovsk y´ Institute of Physical Chemistry, AS CR, v.v.i., Dolej sˇ kova 3, 182
2
3 Prague 8, Czech Republic
Received: July 13, 2013; Published online: November 15, 2013
Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/adsc.201300617.
Abstract: A series of substituted alloxazinium per- to 640 and acceleration by a factor of 350 relative to
chlorates has been prepared and tested as catalysts the non-catalyzed oxidation. The 8-cyanoalloxazini-
for the oxidation of sulfides to sulfoxides with hydro- um perchlorate was evaluated as the best catalyst;
gen peroxide. The logarithms of the observed rate however, due to its relatively good accessibility, the
constants of thioanisole oxidation correlate with the 8-(trifluoromethyl)alloxazinium perchlorate seems to
Hammett s constants of the substituents on the be the catalyst of choice for sulfoxidations with hy-
alloxazinium catalysts, as well as with their reduction drogen peroxide. It was successfully tested for the
0’
potentials E and their pK values, representing the sulfoxidation of a series of aliphatic and aromatic
R+
alloxazinium salt/pseudobase equilibrium. The stron- sulfides on a preparative scale. It produced the corre-
ger the electron-withdrawing substituent, the more sponding sulfoxides in quantitative conversions and
efficient is the alloxazinium catalyst. The alloxazini- with high isolated yields (87–98%). No over-oxida-
um salts with a cyano or trifluoromethyl group in po- tion to sulfone was ever observed.
sition 8 proved to be the most efficient, operating at
room temperature at small loadings, down to Keywords: electrochemistry; flavins; green chemis-
0.1 mol%, achieving turnover number values of up try; organic catalysis; oxidation; peroxides
Introduction
method. In recent decades, much attention has been
paid to environmentally friendly sulfoxidations, in
Sulfoxides are important intermediates and building particular those using hydrogen peroxide as a stoichio-
[
1]
[6]
blocks in organic synthesis. In addition, several sulf- metric oxidizing agent. Hydrogen peroxide is cheap,
[
2]
oxides show remarkable biological activity
and easy to handle and produces water as the only side-
some of them have recently been used as pharmaceut- product; however, it needs activation for the majority
ical substances, for example, the proton pump inhibi- of substrates. Numerous catalytic systems based on
[
3]
[7]
tors known as prazols (omeprazol, lansoprazol, etc.)
transition metals have been developed. Surprisingly,
[
4]
or the psychostimulant modafinil. Sulfoxides are organocatalytic sulfoxidations with hydrogen peroxide
[
8,9]
usually obtained by the oxidation of the correspond- are still relatively rare
in spite of the fact that or-
This synthetic approach is straightfor- ganic catalysts are usually characterized by low toxici-
ward but in many cases not selective, producing the ty and wide availability.
[1a,5]
ing sulfides.
corresponding sulfone as a by-product formed by
Flavinium (isoalloxazinium) salts 1 have become es-
over-oxidation. Thus, chemoselectivity is a necessary tablished as one of the most promising organocata-
requirement for
a
new potential sulfoxidation lysts for chemoselective oxidations of sulfides to sulf-
Adv. Synth. Catal. 2013, 355, 3451 – 3462
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3451

Petra Mꢀnovꢁ et al.
FULL PAPERS
Scheme 1. Catalytic cycle for the H O2 oxidation of sub-
2
strate Sred catalyzed by flavinium salt 1.
[
10,11]
oxides with hydrogen peroxide (Scheme 1).
Even
some chiral flavinium catalysts have been prepared
and used for stereoselective sulfoxidations, with mod-
[
12]
erate to high enantioselectivities.
The principle of
the catalysis by flavinium salts 1 is based on the in-
situ formation of flavin-4a-hydroperoxide, 1-OOH,
which oxidizes a substrate more effectively than hy-
Scheme 2. Catalytic cycle for the oxidation of substrate Sred
catalyzed by alloxazinium salt 2 in the presence of a base B.
[
11a]
drogen peroxide itself (Scheme 1).
After oxygen
transfer to the substrate and subsequent formation of
AHCTUNGTRENNUNGo xidations. The current low efficiency of alloxazinium
hydroxyflavin, 1-OH, the flavinium catalyst is regen- catalysts 2a and 2b results from their reduced reactivi-
erated in the rate-determining step of the catalytic ty towards hydrogen peroxide and thus, from their re-
+
cycle. Since a proton (H ) is required for the last step, duced ability to form hydroperoxide 2-OOH. This is
flavinium catalysts are not active under basic condi- caused by their lower electrophilicity compared with
[
13]
tions. As a consequence, they have negligible activity the isoalloxazinium derivatives.
in the N-oxidations of tertiary amines, which proceed can be tuned by substitution, we decided to prepare
by the same mechanism.
As electrophilicity
[
13]
a series of substituted alloxazinium salts 2 and to eval-
Recently, we have shown that structurally related uate the effect of substitution on their electrophilicity
alloxazinium salts 2a and 2b behave complementarily and catalytic activity in sulfoxidation reactions. We
to 1 in oxidations; they catalyze N-oxidations, but do expected that the introduction of an electron-with-
[
13]
not catalyze sulfoxidations.
This is a result of the drawing group might be a suitable way of preparing
change of the rate-determining step which is, for the alloxazinium catalysts with a higher activity in H O2
2
oxidations catalyzed by alloxazinium salts, the forma- sulfoxidations.
tion of the hydroperoxide (Scheme 2). Nevertheless,
there are ways to activate salt 2a for sulfoxidations,
for instance, by adding a catalytic amount of a suitable
base to the catalytic mixture to allow the formation of
the hydroperoxy adduct from the alloxazinium salt, or
by using an alloxazine catalyst in the form of an allox-
[
13]
azine-C4a adduct, for example, 2a-OCH3. Another
[
14]
efficient procedure developed by Bꢄckvall uses the
reduced form, 2a-H , which is activated by oxygen
2
before entering the catalytic cycle (Scheme 2). Such
improved alloxazine-based catalytic systems show Results and Discussion
good efficiency in sulfoxidations. However, they are
still significantly less efficient than the isomeric iso- Synthesis of Alloxazinium Catalysts
[
13]
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
alloxazinium catalysts.
Keeping in mind their better synthetic availability The synthesis of alloxazinium salts 2 began with the
[
13,14]
compared to flavinium salts 1 and rapid regeneration standard procedure,
consisting of the condensa-
in the catalytic cycle, alloxazinium salts still have high tion of substituted diamines 3 with alloxan and the
potential to be efficient and useful catalysts for sulf- subsequent alkylation of the obtained alloxazines 4
3452
asc.wiley-vch.de
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2013, 355, 3451 – 3462

Electron-Deficient Alloxazinium Salts: Efficient Organocatalysts
Scheme 3. Synthesis of alloxazinium salts 2.
with methyl iodide. This produced the 1,3-dimethylal- tive work-up was 5-ethyl-1,3-dimethylalloxazinium
loxazines 5 in good yields (Scheme 3). Trifluorometh- perchlorate (2a). However, optimization of the reac-
yl and cyano derivatives 4 were formed as mixtures of tion conditions (shorter reaction time, lower hydrogen
the 7- and 8-isomers in ratios 3:7 and 1:1, respectively. pressure) eventually led to the desired 7,8-dichloro-5-
These mixtures were directly alkylated, yielding 7- ethyl-1,3-dimethylalloxazinium perchlorate (2d) in
and 8-substituted 3-methylalloxazines in ratios 2:9 75% yield. Milder reaction conditions, that is, aque-
and 1:1, respectively. The regioisomers were separated ous ethanol as a solvent and lower hydrogen pressure,
by crystallization. Only a single crystallization was coupled with a longer reaction time (5 days) was
needed to obtain pure 8-trifluoromethylalloxazine 5e. shown to be more effective for the reductive alkyla-
In contrast, despite numerous attempts, we were not tion of 5e to 5-ethyl-1,3-dimethyl-8-trifluoromethyl-
able to isolate the pure 7-regioisomer. With the cyano
ACHTUNGTRENNUaGN lloxazinium perchlorate (2e), which was obtained in
derivatives, a series of successive crystallizations and moderate yield.
work-ups of the mother liquor were necessary. On the
Reductive alkylation on Pd(C) could not be used
positive side, both isomers 4f and 4g were eventually for the preparation of alloxazinium salts 2f and 2g
obtained as pure compounds. Our results with the tri- with an easily reducible cyano group. Therefore,
fluoromethyl derivatives 4 are in contrast with those a modified procedure for the reductive alkylation
[
14c]
previously published by Bꢄckvall,
who reported using sodium dithionite and sodium cyanoborohy-
obtaining a mixture of 7- and 8-trifluoromethyl deriv- dride in place of hydrogenation was used. Even when
ative 4 in the ratio 84:16 after the condensation step these mild conditions were applied, the desired prod-
and pure 7-trifluoromethyl derivative after the subse- ucts 2f and 2g were obtained in low yields.
quent alkylation of the mixture of regiomers with
ethyl iodide. To be certain, we confirmed the position
of the substituents in 5e and 5f by X-ray structural H O Sulfoxidations Catalyzed by Substituted
2
2
analysis (for details see the Supporting Informa- Alloxazinium Salts
[15]
tion). In addition, the positions of the substituents
in the corresponding alloxazinium salts were also de- The catalytic efficiency of the series of alloxazinium
termined by NOE experiments (see the Supporting catalysts 2 was tested using the oxidation of the
Information).
model substrate thioanisole 6a under the usual condi-
[13,14]
For the introduction of the ethyl group into posi- tions,
that is, in deuterated methanol with 2 mol%
tion 5, leading to the desired 5-ethylalloxazinium salts of the catalyst and 1.5 equiv. of aqueous H O . The
2
2
[
11e,13]
2, we first attempted our usual procedure,
con- observed initial rates vobs of the oxidations and the
sisting of reductive alkylation of dihydroflavin with rate enhancements relative to non-catalyzed reaction
acetaldehyde in hydrochloric or acetic acid in an auto- expressed as v /v , are summarized in Table 1 (en-
obs
0
clave (5 atm, H , Pd/C), followed by oxidation with tries 1–8; for details of the progress of the oxidations,
2
sodium nitrite. This approach was successful only for see the Supporting Information). Data for the iso-
compounds 5a–5c and the synthesis of the remaining
ACHTUNGTRENNUNaG lloxazinium catalyst 1a and for dihydroalloxazine 2a-
derivatives had to be optimized for each individual H are included for comparison (entries 9 and 10).
2
derivative. During the reductive alkylation of di-
The results clearly indicate the positive effect of
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
chloroalloxazine 5d using the above-mentioned condi- electron-withdrawing groups on the catalytic activity
ACHTUNGTRENNUNG
tions, hydrogenolysis of both chlorine atoms occurred of the alloxazinium salts. By comparison, in the pres-
readily and the only product isolated after the oxida- ence of the non-substituted derivative 2a or derivative
Adv. Synth. Catal. 2013, 355, 3451 – 3462
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
3453

Petra Mꢀnovꢁ et al.
FULL PAPERS
[
a]
[b]
Table 1. Observed initial rates and rate enhancements
for the oxidations of thioanisole 6a catalyzed by 2.
contained 66% of 2f and 33% of 2g. The mixture 2h
is less active than the pure 8-cyano derivative 2g for
the oxidation of thioanisole (entry 8), but neverthe-
less, 2h still proved to be a very potent catalyst.
À1
Entry
Catalyst
vobs [mmolh ]
vobs/v0
1
2
3
4
5
6
7
8
9
2a
2b
2c
2d
2e
2f
16
8.3
116
1050
1250
1350
2570
2.1
1.1
16
140
170
180
350
214
The most efficient compounds 2e and 2g were
chosen for testing their ability to catalyze the H O2
2
oxidation of various sulfides 6 to sulfoxides 7 on
a preparative scale (Table 2). Using 2 mol% of the
catalyst 2e (entries 1–6) or 2g (entry 8), yields of 87–
98% were achieved irrespective of the type of sub-
2g[
c]
strate. However, the type of substrate does influence
the reaction time necessary to obtain full conversion
2h
1560[
d]
[d]
1a
2a-H2
1550
210
[
e]
[e]
(
monitored by TLC). The aliphatic substrate 6e was
10
470
64
completely oxidized within 10 min (entry 5). The oxi-
dation of the aromatic sulfide 6f with an electron-do-
nating group was complete within 30 and 20 min for
2e and 2g, respectively (entries 6 and 8). In contrast,
the oxidation of sulfides with the electron-withdraw-
ing chloro (6b) or nitro (6c) groups took 24 h (en-
[
[
a]
Initial rate up to 15% conversion.
Rate enhancement: the ratio of the initial reaction rates
b]
of the catalyzed (vobs) and non-catalyzed (v ) reaction;
0
À1
v =7.3 mmolh .
0
[c]
[d]
[e]
Mixture of 2f and 2g (2:1).
[13]
Ref.
[16]
Ref. ; rate enhancement v /v =74 for p-tolyl methyl tries 2 and 3). It is noteworthy that neither overoxida-
obs
0
[
14a]
sufide oxidation was reported.
tion to sulfone nor the oxidation of the double bond
in 6d was detected.
Having observed promising catalytic activity for
2
b with two methyl groups, the rate of oxidation is both catalysts, we decided to reduce the catalyst load-
almost the same as that of a non-catalyzed reaction ing to 0.2 mol% of 2e (entry 7) and 0.1 mol% of 2g
entries 1 and 2). The difluoroalloxazinium salt 2c ac- (entries 9 and 10). In both cases, quantitative conver-
(
celerates the oxidation significantly (entry 3). Howev- sion of p-tolyl methyl sulfide 6f to the corresponding
er, a really remarkable rate enhancement was ob- sulfoxide was observed and the sulfoxides were isolat-
served with alloxazinium salts bearing two chloro ed in yields higher than 90%. Impressive TONs of
groups, a trifluoromethyl or a cyano group (entries 4– 395 and 640 were achieved for 2e and 2g, respectively.
7), especially using catalyst 2g with the cyano function
Because the synthesis of the cyano salt 2g is rather
in position 8. Salt 2g accelerates the oxidation by tedious and the catalyst is obtained in low yields, the
a factor of 350 relative to the non-catalyzed reaction, readily available and efficient 8-trifluoromethyl deriv-
making it substantially more effective than the most ative 2e becomes the catalyst of choice. The stability
efficient isoalloxazinium salt 1a (cf. entries 7 and 9). and re-usability of 2e was tested by successive addi-
The catalytic system with 2g is also far more efficient tions of the substrate and hydrogen peroxide after the
than the one utilizing dihydroalloxazine 2a-H (cf. en- previous run had been completed. A total of three
2
tries 7 and 10). The achieved TOF in the initial stage runs was conducted and the activity of the catalyst
of the thioanisole oxidation catalyzed by 2g was was found not to have decreased. In fact, the reaction
À1
6
40 h . The observed activating effect of alloxazini- rates were even slightly higher in the second and
um salt by the substitution with cyano group is indeed third runs (Figure 1), which is surprising because the
significant. The oxidation of thioanisole catalyzed by alloxazinium salt should not need activation (see
2
g is 160 times faster comparing with those in the Scheme 2). On the other hand, the rate enhancement
presence of 2a taking in account the initial reaction in the second run is negligible in contrast to H O oxi-
2
2
rate. The analogical activation (by introducing dations catalyzed by 5-alkyl-1,5-dihydroalloxazines 2-
a cyano group) of 1-ethylpyrazinium catalyst leads H . For these, the 5-alkyl-1,5-dihydroalloxazines need
2
only to 30-fold increase in reaction rate of H O sul- activation by oxygen, which slows down the first ki-
2
2
[16]
[14c]
foxidation.
netic run significantly.
The preparation of the cyanoalloxazinium salts 2f
and 2g is unfavourable, due to the difficult separation
of the precursors, 7- and 8-cyano-1,3-dimethylalloxa- Evaluation of Substitution Effect on the Efficiency of
zines 5f and 5g, which required several subsequent Alloxazinium Catalysts
crystallizations. The difficulty of separating the iso-
mers can be overcome by using a defined mixture of The significant positive effect of electron-withdrawing
both salts 2h, resulting from the quaternization of groups on the catalytic activity of alloxazinium salts
a mixture of 7- and 8-cyano-1,3-dimethylalloxazines in sulfoxidations with hydrogen peroxide is clearly
(
5f and 5g) in a 1:1 ratio; the catalyst obtained 2h shown by kinetic data. It is most probable that the
3454
asc.wiley-vch.de
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2013, 355, 3451 – 3462

Electron-Deficient Alloxazinium Salts: Efficient Organocatalysts
[a]
Table 2. Preparative oxidations of sulfides by hydrogen peroxide catalyzed by alloxazinium salts 2e and 2g.
[b]
[c]
[d]
Entry Substrate
Product
Catalyst Catalyst loading
mol%]
Time
[h]
Yield
[%]
Blank conv.
[%]
TON
[
1
2
3
2e
2e
2e
2
2
2
0.5
24
24
87
92
94
2
43
34
28
24
37
4
5
2e
2e
2
2
2.5
97
93
2
48
46
0.16
<1
6
2e
2
0.5
98
3
48
7
8
9
1
6f
6f
6f
6f
7f
7f
7f
7f
2e
2g
2g
2g
0.2
2
0.2
0.1
4
0.33
3
95
94
98
94
16
2
14
30
395
46
420
640
0
7
[
a]
Conditions for the oxidations: m (substrate)=100 mg, 1.5 equiv. H O , 2 mol% of the catalyst, CH OH (2 mL), T=258C,
2
2
3
monitored by TLC (toluene).
Isolated yield.
Conversion of the non-catalyzed reaction after the same reaction time.
[
[
[
b]
c]
d]
Calculated from isolated yield excluding contribution of non-catalyzed reaction.
electron-withdrawing groups make the alloxazinium peroxide (the equilibrium constants for the reaction
nucleus more electrophilic, thus allowing the addition of heteroarenium salt with water and hydrogen perox-
of hydrogen peroxide to form the oxidative species, ide do not differ substantially as evident from the
alloxazine-4a-hydroperoxide 2-OOH. To evaluate the values for 5-ethyl-tetraacetylriboflavinium ion: K
=
H2O
À4
À4
[19]
effect of substituents on the electrophilicity of the al- 1.6ꢅ10 M; K
=2.5ꢅ10 M; ref. ). The obtained
H2O2
0’
loxazinium nucleus, we determined the reduction po- pKR+ and E values for the alloxazinium salts 2, to-
tentials of the series of salts 2 by means of cyclic vol- gether with those for the isoalloxazinium derivative
tammetry. Reductions of all the investigated alloxazi- 1a, are summarized in Table 3.
nium salts consisted of two reversible one-electron
Consistent with our expectations, all the catalysts
processes, leading first to a radical and, subsequently, with electron-withdrawing substituents show higher
to an anion (see scheme in Table 3). This course of electrophilicity (less negative reduction potentials)
the alloxazinium salt reduction is consistent with the and a greater tendency to form adducts (lower pKR+
[
17]
reduction of neutral flavins, as well as of isoalloxa- values) than the non-substituted alloxazine catalyst 2a
[11b]
zinium salts 1.
(Table 3). The effect of electron-donating methyl
The ability to add nucleophiles was quantified by groups in 2b is the opposite. To find the relationship
pKR+ values, representing the alloxazinium salt 2/ between the catalytic efficiency and the electrophilici-
pseudobase 2-OH equilibrium (see scheme in ty of the substituted alloxazinium salts we plotted
X
H
Table 3). Formally, pKR+ is a measure of the ability of logACHTUNGTRENNUNG
a heteroarenium salt to react with water.
[
13,18]
X
theless, it is reasonable to assume that this value also by the substituted salt (k ) relative to the rate of oxi-
reflects the ability of the heteroarenium nucleus to dation catalyzed by the non-substituted derivative
H
add other similar nucleophiles, for example, hydrogen (k ), against pK values and the first reduction po-
R+
Adv. Synth. Catal. 2013, 355, 3451 – 3462
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
3455

Petra Mꢀnovꢁ et al.
FULL PAPERS
0’
the pK and (E ) values, with negative and positive
a
1
slopes, respectively. This means, for example, that the
most efficient catalyst 2g has the lowest pKR+ value
and the most positive value of the reduction potential.
The electronic effect of the substituents on the cata-
lytic efficiency of alloxazinium salts is also evident
X
H
from the good correlation between log ACHUTNGRENNUG( k /k ) and the
Hammett s values (Figure 2c). In this correlation, we
defined the substituents at C-7 as meta and those at
C-8 as para, since the attack of the nucleophile occurs
at position C-4a (Scheme 4). Accordingly, the higher
[20]
s constant for CN (s =0.66, s =0.56; ref. ) ex-
p
p
m
plains higher reactivity of the 8-cyano in contrast to
the 7-cyano derivative. The formation of hydroperoxy
adduct is more facilitated by a cyano group in posi-
tion 8 due to conjugation effect. For the 7,8-disubsti-
tuted flavins, the s and s values of the substituents
p
m
[
21]
were added together.
Interestingly, although 8-trifluoromethyl 2e and
Figure 1. Oxidation of thioanisole catalyzed by 2e in three both cyano derivatives 2f and 2g exhibit higher cata-
st
nd
rd
consecutive runs. * 1 run, & 2 run, I 3 run, ~ blank. lytic activities than the simple isoalloxazinium salt 1a,
st
0’
Conditions for the 1 run: n (thioanisole)=0.2 mmol, n they have higher pKR+ and lower E values. This does
(
(
H O )=0.3 mmol,
n
(catalyst)=0.004 mmol, CD OD
3
1
2
2
not fit the trends observed among the alloxazinium
salts (Figure 2a and b and Table 3). It shows that al-
though both parameters nicely reflect the catalytic be-
haviour in one series of compounds (e.g., alloxazini-
um salts), they cannot be applied across two groups
600 mL), T=258C, monitored by H NMR. Conditions for
the 2 and 3 runs: 0.2 mmol of thioanisole and 0.2 mmol
of H O added to the reaction mixture.
nd
rd
2
2
0’
tentials (E ) . Figure 2a and b demonstrate that the of structurally different catalysts. Similarly, differences
log ACHTUNGTRENNUNG( k /k ) values show good linear correlations with in the catalytic activity of the isoalloxazinium and al-
1
X
H
loxazinium salts in oxidations with oxygen could not
be explained on the basis of their reduction poten-
[
11b]
tials.
Table 3. Comparison of pK values and formal reduction
R+
0’
0
’
Nevertheless, it is undisputed that the pKR+ and E
azinium 1a) salts (Fl =alloxazinium or isoalloxazinium, Fl- values of alloxazinium salts 2e–g with electron-with-
potentials E for flavinium (i.e., alloxazinium 2 and isoallox-
+
OH=4a-hydroxyalloxazine or 4a-hydroxyisoalloxazine).
drawing groups (Table 3) are shifted towards values
for the isoalloxazinium derivative 1a and, consequent-
ly, salts 2e–g resemble the isoalloxazinium salt 1a with
respect to its catalytic efficiency for H O sulfoxida-
2
2
tions (Table 1). We wondered if this trend in catalytic
activity is reflected in other reactions and therefore
we tested salts 2 as catalysts for N-oxidations using
triethylamine as a model substrate (Table 4). As ex-
pected, the electron-deficient alloxazinium salts 2d–2g
showed only negligible catalytic activity, as did the
isoalloxazinium salt 1a. The negative effect of the
electron-withdrawing groups on the reactivity of salts
[
a]
0’
[b]
Flavinium salt
pKR+
E [mV]
0’
0’
A
H
U
G
R
N
U
G
ACHTUNGENTRNUNG( E )
2
1
2
2
2
2
2
2
2
1
a
b
c
d
e
f
7.9
9.2
6.9
5.5
4.8
4.1
3.4[
100
À15
180
235
275
305
320
À710
À783
À620
À570
À575
À555
À530
À350
2
in N-oxidations is also illustrated by the linear
Hammett correlation having a reverse slope com-
pared to the S-oxidations (cf. Figure 2c and d). One
can conclude that strongly electron-withdrawing sub-
stituents on the alloxazinium ring cause a shift in its
catalytic activity from “alloxazinium-like” (catalyzes
N-oxidations and does not catalyze S-oxidations) to
“isoalloxazinium-like” (catalyzes S-oxidations and
does not catalyze N-oxidations).
g
a
c]
[c]
[c]
2.0
350
[
[
a]
+
+
pKR+ =Àlog KR+; KR+ =[FlOH][H ]/
A
H
U
G
R
N
U
G
b]
Formal reduction potentials in acetonitrile vs. ferrocene/
ferrocenium determined by cyclic voltammetry on plati-
num electrode (for details see Experimental Section).
[
c]
[13]
Ref.
3456
asc.wiley-vch.de
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2013, 355, 3451 – 3462

Electron-Deficient Alloxazinium Salts: Efficient Organocatalysts
X
H
Figure 2. Dependence of log ACHTGNUTRNEUNG( k /k ) for sulfoxidations catalyzed by alloxazinium salts on pK values (a) and reduction po-
R+
X H
0
tentials (E ’) (b) of the catalysts, and on Hammett s values of substituents (c). Dependence of log ACHNTUGRNENUG( k /k ) for N-oxidations
1
[20]
on Hammett s values (d). Hammett s values were taken from the reference.
[a]
[b]
Table 4. Observed initial rates and rate enhancements
for the oxidations of triethylamine catalyzed by 2.
À1
Catalyst
vobs [mmolh ]
vobs/v0
2
2b
2c
2
2
2f
a
330
480
110
30
21
23
48
70
15
4.4
3.1
3.3
2.1
Scheme 4. Position of the substituents relative to catalytic
centre in alloxazinium salts.
d
e
Conclusions
2
g
15 [
c]
[c]
A series of substituted alloxazinium perchlorates 2 1a
9.0
1.3
has been prepared as potential sulfoxidation catalysts.
Introduction of electron-withdrawing group has signif-
icantly positive effect on the activity of the alloxazini-
um catalysts. From the prepared series, 7,8-dichloro
[a]
Initial rate up to 15% conversion.
Rate enhancement: the ratio of the initial reaction rates
[b]
of the catalyzed (vobs) and non-catalyzed (v ) reaction;
0
À1
v
Ref.
A
H
U
G
R
N
N
(NEt )=6.9 mmolh .
0
3
[c]
[13]
2d, 8-trifluoromethyl 2e and both 7- and 8-cyanoallox-
Adv. Synth. Catal. 2013, 355, 3451 – 3462
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
3457

Petra Mꢀnovꢁ et al.
FULL PAPERS
azinium salts 2f and 2g proved to be the most effi- 0.063 mm, Merck). Melting points were measured on a Boe-
tius melting point apparatus and are uncorrected.
cient, as reflected by the rate enhancement of the oxi-
dation of thioanisole from 140 to 350 relative to the
non-catalyzed reaction. These rate enhancements and
high TON values achieved rank 2e and mainly 2g
among the most efficient flavin-based catalysts for
H O sulfoxidation.
Starting materials, reagents and substrates were obtained
from commercial suppliers and used without further purifi-
cation. The solvents were purified and dried using standard
[22]
[11b]
[11e]
[13]
[14c]
procedures.
Compounds 1a,
2a,
2b,
5c,
and
[23]
4
d
were prepared according to previously described pro-
2
2
cedures.
Electron-poor alloxazinium salts 2 have several ad-
vantages compared with other flavinium-based cata-
lysts. They are more efficient than isoalloxazinium
perchlorates 1. This higher efficiency can be explained
by a better balance between individual steps of the
catalytic cycle (see Scheme 2). The formation of allox-
azine hydroperoxide is facilitated by the presence of
electron-withdrawing groups and the recovery of al-
loxazinium salt is facilitated by the restoration of aro-
Synthesis of Catalyst 2c
Acetaldehyde (600 mL, 10.73 mmol) and palladium on
carbon (10%, 35 mg) were added to a suspension of 5c
(
100 mg, 0.36 mmol) in acetic acid (6.0 mL) and water
(
0.6 mL). The resulting mixture was stirred for 24 h in an au-
toclave at room temperature under hydrogen atmosphere
0.5 MPa). The catalyst was removed by filtration, the acetic
(
[14a]
maticity.
Compared to the 5-ethyl-1,5-dihydro-
acid was evaporated under reduced pressure, and the re-
maining solid was dried under vacuum. The residue was sus-
ACHTUNGTRENNUNGa lloxazines 2-H previously used for the oxidation re-
2
[14]
action, the alloxazinium salts 2 are easier to handle pended in perchloric acid (2M, 4.0 mL), and sodium per-
and do not need pre-activation by oxygen. Therefore chlorate (522 mg, 4.26 mmol) and sodium nitrite (200 mg,
the salts 2 with electron-withdrawing substituents 2.90 mmol) were added at 08C. The mixture was stirred for
2
h at room temperature. The precipitated solid was collect-
should be suitable as redox-active unit in chiral cata-
lysts for sulfoxidations which are now under prepara-
tion in our laboratory.
ed by filtration, washed successively with 2M perchloric
acid and water, and dried under vacuum to afford 5-ethyl-
7
,8-difluoro-1,3-dimethylalloxazinium perchlorate (2c) as
We have shown that the catalytic activity of the
a yellow powder; yield: 105 mg (72%); mp 195–1988C.
alloxazinium salts correlates with their electrophilicity
1
H NMR (500 MHz, CD CN, 243 K): d=1.72 (t, J=7.1 Hz,
3
0
expressed by reduction potentials (E ’) and pK
+
1
R+
3H, N CH CH ), 3.46 (s, 3H, 3-NCH ), 3.75 (s, 3H, 1-
2 3 3
+
0
values. More positive (E ’) values and lower pK
1
R+
NCH ), 4.90 (bs, 1H, N CH CH ), 6.19 (bs, 1H,
3 2 3
+
values correlate with higher reactivity towards sul- N CH CH ), 8.22 (t, J=8.7 Hz, 1H, 9-ArH), 8.53 (dd, J=
fides. This shows that, based on these parameters, one 11.0, 7.5 Hz, 1H, 6-ArH); C NMR (125 MHz, CD
can estimate the efficiency of a catalyst having the 243 K): d=15.1 (N CH CH ), 30.5 (3-NCH
same heteroaromatic core. Direct comparison of cata-
lysts with a different structure can be carried out only
on a qualitative level. Nevertheless, keeping in mind
this limitation, reduction potentials or pKR+ values
can be used for the prediction of efficiency of new
sulfoxidation catalysts that act according an analogous
mechanism.
2
3
1
3
CN,
), 31.4 (1-
NCH ), 53.7 (N CH CH ), 108.6 (d, J=26 Hz, 6-CH), 116.6
3
+
2 3 3
+
3
2
3
(
d, J=18 Hz, 9-CH), 122.4 (4a-C), 129.3 (d, J=11 Hz, 5a-
C), 146.3 (d, J=14 Hz, 9a-C), 148.6 (10a-C), 149.4 (2-CO),
1
2
55.8 (4-CO), 156.0 (dd, J=263, 14 Hz, CF), 157.1 (dd, J=
66, 16 Hz, CF); F NMR (282 MHz, CD CN): d=À117.57
19
3
(
m, 1F, CF), À116.8 (m, 1F, CF); HR-MS: m/z=307.1011,
+
calcd. for C H N O F [M] : 307.1007; UV-VIS (water):
1
4
13
4
À1
2
2
À1
À1
À1
l=440 (e=5400M cm ), 370 nm (e=5600M cm ).
Synthesis of Catalyst 2d
Experimental Section
7
,8-Dichloro-1,3-dimethylalloxazine (5d): A mixture of 7,8-
NMR spectra were recorded on a Varian Mercury Plus 300
dichloroalloxazine 4d (1.56 g, 5.52 mmol), potassium carbon-
ate (4.22 g, 30.53 mmol), methyl iodide (0.90 mL,
14.46 mmol), and DMF (50 mL) was stirred at room temper-
ature overnight. All the solvent was removed under reduced
pressure. The residue was dissolved in water (130 mL) and
chloroform (180 mL), and the layers were separated. The
aqueous phase was then extracted with chloroform (3ꢅ
50 mL). The combined organic extracts were washed with
water (100 mL), dried over magnesium sulfate and evaporat-
ed to dryness. The crude product was recrystallized from
ethanol to give 5d as a brown powder; yield: 1.19 g (70%);
1
13
(
299.97 MHz for H, 75.44 MHz for C, and 282.23 MHz for
1
1
9
F), Bruker Avance DRX 500 (500.13 MHz for H and
1
3
1
(
25.77 MHz for1
C) and Bruker 600 Avance III
13
600.13 MHz for H and 150.90 MHz for C) at 298 K unless
otherwise indicated. Chemical shifts d are given in ppm,
using residual solvent or tetramethylsilane as an internal
standard. Coupling constants are reported in Hz. UV-VIS
spectra were recorded on a Varian Cary 50 spectrophotome-
ter. Elemental analyses (C, H, N) were performed on
a Perkin–Elmer 240 analyser. High-resolution mass spectra
were obtained on Q-Tof Micro (Waters), equipped with
a quadrupole and TOF analyzers and MCP detector. TLC
analyses were carried out on a DC Alufolien Kieselgel 60
F254 (Merck). Preparative column chromatography separa-
tions were performed on a silica gel Kieselgel 60 (0.040–
[24]
1
mp 253–2548C (lit. 2568C). H NMR (300 MHz, DMSO-
d ): d=3.60 (s, 3H, NCH ), 3.80 (s, 3H, NCH ), 8.18 (s, 1H,
6
3
3
1
3
ArH), 8.44 (s, 1H, ArH); C NMR (75 MHz, CDCl ): d=
3
29.6, 30.0, 128.6, 130.7, 131.1, 134.2, 138.7, 139.3, 142.2,
146.2, 150.6, 159.4.
3458
asc.wiley-vch.de
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2013, 355, 3451 – 3462

Electron-Deficient Alloxazinium Salts: Efficient Organocatalysts
7
,8-Dichloro-5-ethyl-1,3-dimethylalloxazinium perchlorate
2d): Acetaldehyde (270 mL, 4.83 mmol) and palladium on
carbon (10%, 16 mg) were added to a suspension of 5d
fate. The solvent was evaporated to dryness to afford a mix-
ture of 7- and 8-isomers in a 2:9 ratio; yield: 90 mg (82%).
Pure 8-trifluoromethyl isomer 5e was obtained after re-
(
(
(
50 mg, 0.16 mmol) in acetic acid (2.7 mL) and water
270 mL). The resulting mixture was stirred for 6 h in an au-
crystallization from ethanol; yield: 50 mg (46%); mp 268–
1
2708C. H NMR (600 MHz, CDCl ): d=3.64 (s, 3H, 3-
3
toclave at room temperature under hydrogen atmosphere
0.2 MPa). The catalyst was removed by filtration, the acetic
acid was evaporated under reduced pressure, and the re-
maining solid was dried under vacuum. The residue was sus-
pended in perchloric acid (2M, 1.8 mL), and sodium per-
chlorate (235 mg, 1.92 mmol) and sodium nitrite (90 mg,
NCH ), 3.84 (s, 3H, 1-NCH ), 7.92 (dd, J=8.8, 1.8 Hz, 1H,
3
3
(
7-ArH), 8.36 (s, 1H, 9-ArH), 8.47 (d, J=8.8 Hz, 1H, 6-
13
ArH); C NMR (150 MHz, CDCl ): d=29.4 (3-NCH ), 29.8
3 3
(1-NCH ), 123.2 (q, J=270 Hz, CF ), 124.7 (7-CH), 125.8 (9-
3
3
CH), 131.5 (4a-C), 132.0 (6-CH), 135.0 (q, J=33 Hz, 8-C),
140.7 (5a-C), 142.4 (9a-C), 146.2 (10a-C), 150.4 (2-CO),
1
9
1.30 mmol) were added successively at 08C. The mixture
159.2 (4-CO); F NMR (282 MHz, CD CN): d=À63.7 (m, 3
3
was stirred for 2 h at room temperature. The precipitated
solid was collected by filtration, washed with 2M perchloric
acid and water and dried under reduced pressure. The crude
product was dissolved in a minimum volume of acetonitrile
and precipitated with diethyl ether to give 2d as yellow-
F, CF ); anal. calcd: for C H F N O (310.24): C 50.33, H
3
13
9
3
4
2
2.92, N 18.06; found. C 50.51, H 2.47, N 18.14.
5-Ethyl-1,3-dimethyl-8-(trifluoromethyl)alloxazinium per-
chlorate (2e): 1,3-Dimethyl-8-(trifluoromethyl)alloxazine
(5e) (35 mg, 0.11 mmol) and palladium on carbon (10%,
12 mg) were suspended in ethanol (2 mL) and water
(1.5 mL). Concentrated hydrochloric acid (214 mL) and acet-
aldehyde (214 mL, 3.83 mmol) were added to the mixture.
The resulting mixture was stirred for 5 days in an autoclave
at room temperature under hydrogen atmosphere
(0.2 MPa). The catalyst was removed by filtration through
celite, the solvents were evaporated under reduced pressure,
and the remaining solid was dried in vacuo. The residue was
suspended in perchloric acid (2m, 1.2 mL), and sodium per-
chlorate (164 mg, 1.34 mmol) and sodium nitrite (63 mg,
0.91 mmol) were added successively at 08C. The mixture
was stirred for 2 h at room temperature. The precipitated
solid was collected by filtration, washed with 2m perchloric
acid and dried under reduced pressure. The crude product
was dissolved in a minimum volume of acetonitrile and pre-
cipitated with diethyl ether to give 2e as yellow powder,
orange powder; yield: 53 mg (75%); mp 227–2318C.
1
H NMR (500 MHz, CD CN, 253 K): d=1.76 (t, J=7.1 Hz,
3
+
3
H, N CH CH ), 3.50 (s, 3H, 3-NCH ), 3.78 (s, 3H, 1-
2
3
3
+
NCH ), 5.03 (bs, 1H, N CH CH ), 6.20 (bs, 1H,
3
2
3
+
N CH CH ), 8.59 (s, 1H, 9-ArH), 8.84 (s, 1H, 6-ArH);
2
3
13
+
C NMR (125 MHz, CD CN, 253 K): d=14.6 (N CH CH ),
9.9 (3-NCH ), 30.8 (1-NCH ), 52.6 (N CH CH ), 121.1 (6-
3
2
3
+
2
3 3 2 3
CH), 122.0 (4a-C), 129.0 (5a-C), 130.2 (9-CH), 140.5 (7-C),
41.9 (8-C), 145.5 (9a-C), 148.2 (10a-C), 148.6 (2-CO), 155.1
4-CO); HR-MS: m/z=339.0426, calcd. for C H N O Cl
1
(
1
4
13
À1
4
2
À1
2
+
[M] : 339.0416. UV-VIS (water): l=458 (e=9100M cm ),
À1
À1
3
94 nm (e=11500M cm ).
Synthesis of Catalyst 2e
1
yield: 34 mg (68%); mp 233–2358C. H NMR (500 MHz,
[
14c]
7
/8-(Trifluoromethyl)alloxazine
drate (383 mg, 2.39 mmol) and boric acid (154 mg,
.49 mmol) were heated in acetic acid (13 mL) until all the
solid dissolved. 4-Trifluoromethylbenzene-1,2-amine
400 mg, 2.27 mmol) was added to the solution while stir-
(4e): Alloxan monohy-
+
CD CN, 243 K): d=1.77 (t, J=7.1 Hz, 3H, N CH CH ),
3
2
3
3
1
.48 (s, 3H, 3-NCH ), 3.79 (s, 3H, 1-NCH ), 5.12–5.14 (m,
3 3
+ +
2
H, N CH CH ), 6.17–6.19 (m, 1H, N CH CH ), 8.39 (d,
2
3
2
3
J=9.3 Hz, 1H, 7-ArH), 8.70 (s, 1H, 9-ArH), 8.74 (d, J=
(
13
9
.5 Hz, 1H, 6-ArH); C NMR (125 MHz, CD CN, 243 K):
3
+
ring. The reaction mixture was stirred for 3 h at room tem-
perature. A pale yellow-brown precipitate was filtered off
and washed successively with acetic acid and water to give
d=15.3 (N CH CH ), 30.6 (3-NCH ), 31.5 (1-NCH ), 53.5
2
3
3
3
+
(
N CH CH ), 123.0 (6-CH), 123.1 (q, J=270 Hz, CF ), 124.4
2
3
3
(
4a-C), 128.8 (9-CH), 131.0 (7-CH), 131.9 (5a-C), 136.7 (q,
4
e as a mixture of 7- and 8- isomers in a 3:7 ratio; yield:
J=34 Hz, 8-C), 146.8 (9a-C), 149.1 (10a-C), 149.3 (2-CO),
55.7 (4-CO); HR-MS: m/z=339.1060, calcd. for
488 mg (76%).
1
1
Data for the 7-trifluoromethyl isomer: H NMR (300 MHz,
+
C H N O F [M] : 339.1069; UV-VIS (water): l=446 (e=
6
1
5
14
4
À1
2
3
À1
DMSO-d ): d=8.06–8.09 (d, J=8.9 Hz, 1H, 6-ArH), 8.12–
À1
À1
6
500M cm ), 357 nm (e=6400M cm ).
8
1
.15 (dd, J=8.9, 1.5 Hz, 1H, 7-ArH), 8.54 (d, J=1.7 Hz,
H, 9-ArH), 11.90 (bs, 1H, NH), 12.10 (bs, 1H, NH).
1
Data for the 8-trifluoromethyl isomer: H NMR (300 MHz,
DMSO-d ): d=7.97–8.01 (dd, J=8.7, 1.7 Hz, 1H, 8-ArH),
6
Synthesis of Catalysts 2f, 2g, and 2h
8
9
.22 (d, J=1.8 Hz, 1H, 6-ArH), 8.34–8.39 (d, J=8.7 Hz, 1H,
-ArH), 11.90 (bs, 1H, NH), 12.10 (bs, 1H, NH).
7/8-Cyanoalloxazine (4f): Alloxan monohydrate (120 mg,
0.75 mmol) and boric acid (51 mg, 0.83 mmol) were heated
in acetic acid (2.5 mL) until all the solids dissolved. 3,4-Di-
1
,3-Dimethyl-8-(trifluoromethyl)alloxazine (5e): A mix-
ture of 4e (100 mg, 0.35 mmol), potassium carbonate
160 mg, 1.16 mmol), methyl iodide (90 mg, 0.63 mmol), and
(
ACHTUNGTRENNUNaG minobenzonitrile (100 mg, 0.75 mmol) was added to the so-
DMF (15 mL) was stirred at room temperature for 4 h.
After the reaction was complete (TLC monitoring using
ethyl acetate as a mobile phase), all the solvent was re-
moved under reduced pressure. The residue was dissolved in
water (15 mL) and chloroform (15 mL), and the layers were
separated. The aqueous phase was then extracted with
chloroform (5ꢅ10 mL). The combined organic extracts were
washed with brine (15 mL) and dried over magnesium sul-
lution while stirring. A pale yellow precipitate started to
form immediately. The reaction mixture was stirred for
90 min at room temperature. The precipitate was filtered
off, and washed with water and ethanol. Drying gave 4f as
a mixture of 7/8-isomers in a 1:1 ratio; yield: 162 mg (90%);
mp >3508C.
1
Data for the 7-cyano isomer: H NMR (300 MHz, DMSO-
d6): d=7.99 (d, J=9.4 Hz, 1H, 9-ArH), 8.16 (dd, J=8.8,
Adv. Synth. Catal. 2013, 355, 3451 – 3462
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
3459

Petra Mꢀnovꢁ et al.
FULL PAPERS
2.1 Hz, 1H, 8-ArH), 8.78 (d, J=1.5 Hz, 1H, 6-ArH), 11.86
dissolved in a minimum volume of acetonitrile and precipi-
(
bs, 1H, NH), 12.15 (bs, 1H, NH).
Data for the 8-cyano isomer: H NMR (300 MHz, DMSO-
tated with diethyl ether affording 2f as a yellow powder;
1
1
yield: 20 mg (9%); mp 223–2278C. H NMR (500 MHz,
+
d6): d=8.02 (dd, J=8.2, 1.8 Hz, 1H, 7-ArH), 8.30 (d, J=
.1 Hz, 1H, 6-ArH), 8.49 (d, J=1.5 Hz, 1H, 9-ArH), 11.86
bs, 1H, NH), 12.15 (bs, 1H, NH).
-Cyano-1,3-dimethylalloxazine (5f), 8-cyano-1,3-dimeth-
alloxazine (5g): A mixture of 4f (1.215 g, 5.08 mmol), po-
tassium carbonate (3.650 g, 26.41 mmol), methyl iodide
740 mL, 11.89 mmol), and DMF (40 mL) was stirred at
CD
3
CN, 253 K): d=1.77 (t, J=7.2 Hz, 3H, N CH
CH
2
),
3
9
(
3.48 (s, 3H, 3-NCH
N CH CH ), 6.19 (bs, 1H, N CH CH
2 3 2 3
13
3
), 3.78 (s, 3H, 1-NCH
3
), 5.07 (bs, 1H,
), 8.40 (s, 2H, 8,9-
ArH), 9.05 (s, 1H, 6-ArH); C NMR (125 MHz, CD CN,
), 31.6 (1-
), 117.4 (CN), 118.2 (7-C), 124.8
+
+
7
3
+
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
yl
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
253 K): d=15.5 (N CH CH ), 30.6 (3-NCH
2 3 3
+
NCH ), 53.6 (N CH CH
3 2 3
(
(5a-C), 127.4 (6-CH), 129.7 (4a-C), 131.9 (9-CH), 137.7 (8-
room temperature for 18 h. Potassium carbonate was re-
moved by filtration and the solvent was evaporated under
reduced pressure. The residue was dissolved in water
CH), 148.0 (9a-C), 149.2 (10a-C), 149.7 (2-CO), 155.6 (4-
+
CO); HR-MS: m/z=296.1142, calcd. for C15
H
N
O
[M] :
14
5
2
À1
À1
296.1147; UV-VIS (water): l=445 (e=6800M cm ),
À1
À1
3
75 nm (e=6200M cm ).
-Cyano-5-ethyl-1,3-dimethylalloxazinium
2g): Compound 2g was prepared according to the proce-
(
200 mL) and dichloromethane (200 mL), and the layers
8
perchlorate
were separated. The aqueous phase was then extracted with
dichloromethane (10ꢅ50 mL). The combined organic ex-
tracts were washed with water (250 mL), dried over magne-
sium sulfate and evaporated to afford a mixture of 7/8-iso-
mers in a 1:1 ratio; yield: 1.358 g (99%).
Recrystallization from ethanol afforded a crystalline mix-
ture of the 7/8-isomers in a 1:3 ratio and a mother liquor
containing the 7/8-isomers in a 3:1 ratio. A series of succes-
sive crystallizations from ethanol afforded pure 7-cyano de-
rivative 5f (yield: 346 mg, 26%) and pure 8-cyano derivative
(
dure described for 2f, starting from 5g (100 mg, 0.374 mmol)
to furnish a yellow powder; yield: 38.7 mg (26%); mp 213–
1
2
7
1
188C. H NMR (500 MHz, CD CN, 253 K): d=1.80 (t, J=
3
+
.0 Hz, 3H, N CH CH ), 3.53 (s, 3H, 3-NCH ), 3.82 (s, 3H,
2
3
3
+
-NCH ), 5.14 (bs, 1H, N CH CH ), 6.21 (bs, 1H,
3
2
3
+
N CH CH ), 8.41 (dd, J=9.5, 1.7 Hz, 1H, 7-ArH), 8.72 (d,
2
3
J=9.5 Hz, 1H, 6-ArH), 8.82 (d, J=1.5 Hz, 1H, 9-ArH).
1
3
+
C NMR (125 MHz, CD CN, 253 K): d=15.2 (N CH CH ),
3
2
3
+
3
0.7 (3-NCH ), 31.5 (1-NCH ), 53.5 (N CH CH ), 116.9
5
g (yield: 557 mg, 41%).
Data for the 7-cyano isomer: mp 293–2968C. H NMR
600 MHz, CDCl ): d=3.64 (s, 3H, 3-NCH ), 3.86 (s, 3H, 1-
3 3 2 3
1
(CN), 120.1 (8-C), 122.4 (6-CH), 124.7 (4a-C), 132.0 (5a-C),
1
1
36.0 (7-CH), 136.5 (9-CH), 143.6 (10a-C), 146.4 (9a-C),
(
3
3
49.2 (2-CO), 155.6 (4-CO); HR-MS: m/z=296.1151, calcd.
NCH ), 8.03 (dd, J=8.8, 1.8 Hz, 1H, 8-ArH), 8.15 (d, J=
8
calcd. for C H N O (267.25): C 58.43, H 3.39, N 26.21;
found: C 58.89, H 3.57, N 26.15.
3
+
for C H N O [M] : 296.1147; UV-VIS (water): l=451 (e=
1
5
14
À1
5
À1
2
.8 Hz, 1H, 9-ArH), 8.71 (d, J=1.7 Hz, 1H, 6-ArH); anal.
À1
À1
6
400M cm ), 370 nm (e=7400M cm ).
/8-Cyano-5-ethyl-1,3-dimethylalloxazinium
(2h): Catalyst 2h was prepared from a 1:1 mixture of 5f and
g (100 mg, 0.374 mmol) according to the procedure de-
13
9
5
2
7
perchlorate
1
Data for the 8-cyano isomer: mp 265–2708C. H NMR
600 MHz, CDCl ): d=3.63 (s, 3H, 3-NCH ), 3.85 (s, 3H, 1-
5
(
3
3
scribed above for the preparation of 2f and 2g. The obtained
catalyst consisted of 66% 2f and 33% 2g according to
NCH ), 7.90 (dd, J=8.8, 1.5 Hz, 1H, 7-ArH), 8.43 (d, J=
3
1
.2 Hz, 1H, 9-ArH), 8.46 (d, J=8.4 Hz, 1H, 6-ArH);
1
13
H NMR. Yellow powder; yield: 29 mg (19%).
C NMR (150 MHz, CDCl ): d=29.4 (3-NCH ), 29.9 (1-
3
3
NCH ), 117.0 (8-C), 117.5 (CN), 129.5 (7-CH), 131.9 (4a-C),
3
Kinetic Measurements
1
1
32.3 (6-CH), 133.0 (9-CH), 140.9 (5a-C), 142.4 (9a-C),
46.4 (10a-C), 150.3 (2-CO), 159.0 (4-CO); anal. calcd. for
The catalyst (0.004 mmol) and the substrate (0.200 mmol)
were dissolved in MeOH-d4 (600 mL). The solution was
transferred into an NMR tube and thermostatted for 15 min
at 25.0Æ0.58C. The reaction was started by the addition of
hydrogen peroxide (30%, 31 mL, 0.300 mmol). In the blank
experiments, no catalyst was present and the hydrogen per-
C H N O (267.25): C 58.43, H 3.39, N 26.21; found: C
13
9
5
2
5
8.44, H 2.94, N 26.17.
-Cyano-5-ethyl-1,3-dimethylalloxazinium
2f): Sodium dithionite (392 mg, 2.25 mmol), sodium cyano-
borohydride (354 mg, 5.63 mmol) and acetic acid (169 mL,
.95 mmol) were added to a solution of 7-cyano-1,3-dime-
thylalloxazine (5f) (150 mg, 0.56 mmol) in dry DMF
13.2 mL). The flask was sealed with a rubber septum and
flushed with nitrogen. After 10 min, acetaldehyde (3.0 mL,
3.666 mmol) was added, and the inflow of nitrogen was
turned off. The mixture was stirred at 608C for 48 h. After
4 h, a fresh batch of acetaldehyde (1 mL, 17.89 mmol) was
7
perchlorate
(
2
oxide was added to a solution of the substrate in MeOH-d .
4
1
The reaction was monitored by H NMR until 90% conver-
(
sion was achieved. The PhSCH and NCH CH signals were
3
2
3
integrated in the oxidation of thioanisole and triethylamine,
respectively. The initial rate constants vobs were calculated as
5
the slope of linear conversion vs. time dependence until
2
X
1
5% conversion. The pseudo-first order rate constants k
added. Next, the solution was cooled to room temperature
and extracted between water (20 mL) and dichloromethane
H
and k were calculated by non-linear regression of conver-
sion vs. time dependence until 80% conversion.
(
10ꢅ8 mL). The combined organic extracts were washed
with water (20 mL) and the solvents were removed under
reduced pressure. The residue was suspended in perchloric
acid (2m, 6 mL). Sodium perchlorate (780 mg, 6.37 mmol)
and sodium nitrite (300 mg, 4.35 mmol) were added succes-
sively at 08C. The resulting mixture was stirred for 90 min at
room temperature. The precipitated yellow solid was collect-
ed by filtration, washed with 2M perchloric acid and water,
and dried under reduced pressure. The crude product was
Preparative Sulfoxidations
The substrate (100 mg) and the catalyst (0.1–2 mol%) were
dissolved in methanol (2 mL). The solution was thermostat-
ted for 20 min at 25.0Æ0.58C. The reaction was initiated by
the addition of hydrogen peroxide (30% aqueous solution;
1.5 equiv). The reaction mixture was stirred at 25.0Æ0.58C
and the course of the reaction was monitored by TLC using
3460
asc.wiley-vch.de
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2013, 355, 3451 – 3462

Electron-Deficient Alloxazinium Salts: Efficient Organocatalysts
toluene as a mobile phase. After complete conversion was
achieved (or after 24 h in the case of substrates more diffi-
cult to oxidize), sodium sulfite (100 mg) was added to
quench the excess of hydrogen peroxide. Most of the sol-
vents were evaporated under reduced pressure and the resi-
due was dissolved in water (6 mL) and dichloromethane
Hernꢁndez-Torres, M. Ribagorda, A. Urbano, Chem.
Commun. 2009, 6129–6144.
[2] J. Legros, J. R. Dehli, C. Bolm, Adv. Synth. Catal. 2005,
347, 19–31.
[3] a) K. S. Jain, A. K. Shah, J. Bariwal, S. M. Shelke, A. P.
Kale, J. R. Jagtap, A. V. Bhosale, Bioorg. Med. Chem.
2007, 15, 1181–1205; b) C. M. Spencer, D. Faulds,
Drugs 2000, 60, 321–329; c) A. M. Rouhi, Chem. Eng.
News 2004, 82, 47–62.
[4] a) R. Kumar, Drugs 2008, 68, 1803–1839; b) M. Russo,
Clinical Medicine: Therapeutics 2009, 415–432.
[5] a) B. Ferber, H. B. Kagan, Adv. Synth. Catal. 2007, 349,
493–507; b) E. Wojaczynska, J. Wojaczynski, Chem.
Rev. 2010, 110, 4303–4356.
(
6 mL). The phases were separated and the aqueous phase
was extracted with dichloromethane (4ꢅ6 mL). The com-
bined organic phases were washed with brine (10 mL) and
dried over magnesium sulfate. The solvent was evaporated
under reduced pressure and the crude product was purified
by column chromatography. The products were character-
1
ized by H NMR spectra which corresponded to those pub-
[25]
lished in the literature (see the Supporting Information).
[
[
[
6] a) C. L. Hill, Angew. Chem. 2004, 116, 406–408; Angew.
Chem. Int. Ed. 2004, 43, 402–404; b) R. A. Sheldon,
I. W. C. E. Arends, U. Hanefeld, Green Chemistry and
Catalysis, Wiley-VCH, Weinheim, 2007; c) R. Noyori,
M. Aoki, K. Sato, Chem. Commun. 2006, 16, 1977–
Alloxazinium Salt/Pseudobase Equilibrium
The pKR+ values representing alloxazinium salt 2/pseudo-
base 2-OH equilibrium in water were calculated by non-
linear regression analysis of absorbance versus pH data at
the maximum absorption wavelengths of salt 2 and its pseu-
dobase 2-OH. The pKR+ values are the means of the ob-
tained values (for details see the Supporting Information).
The pH of the systems was adjusted by perchloric acid (for
pH <1.6), phosphate buffers (for pH 1.6–7.2), and borate
buffers (for pH> 7.3).
1
986.
7] a) K. Katarzyna, Z. Kolarska, K. Mitka, P. Kowalski,
Tetrahedron 2005, 61, 8315–8327; b) K. P. Bryliakov,
E. P. Talsi, Curr. Org. Chem. 2008, 12, 386–404; c) C. L.
Sun, B. J. Li, Z. J. Shi, Chem. Rev. 2011, 111, 1293–
1
314; d) K. P. Bryliakov, E. P. Talsi, Curr. Org. Chem.
2
012, 16, 1215–1242.
8] For reviews about organocatalytic sulfoxidations, see:
a) J.-E. Bꢄckvall, in Modern Oxidation Methods, (Ed.:
J.-E. Bꢄckvall), Wiley-VCH Verlag, Weinheim, 2010,
pp 277–313; b) K. A. Stingl, S. B. Tsogoeva, Tetrahe-
dron: Asymmetry 2010, 21, 1055–1074; c) A. Russo, C.
De Fusco, A. Lattanzi, ChemCatChem 2012, 4, 901–
Electrochemical Measurements
Electrochemical measurements were performed with a PA4
polarographic analyzer (Laboratornꢂ p rˇ ꢂstroje, Prague) using
a standard three-electrode system in a two-compartment
cell with a stationary platinum disc working electrode
9
16.
(
diameter 1 mm). Saturated calomel reference electrode was
ˇ
´
, Q. Wang, B. List, J. Am. Chem. Soc.
separated from the investigated aprotic solution by a glass
frit and a salt bridge filled with the same electrolyte. A plat-
inum wire served as an auxiliary electrode. The cyclic vol-
tammetry measurements were carried out in acetonitrile
[9] S. Liao, I. Coric
2012, 134, 10765–10768.
[10] For review about flavinium-based catalysis see: a) G. de
Gonzalo, M. W. Fraaije, ChemCatChem 2013, 5, 403–
415; b) F. G. Gelalcha, Chem. Rev. 2007, 107, 3338–
3361; c) Y. Imada, T. Naota, Chem. Rec. 2007, 7, 354–
361.
[11] For application of flavinium catalysts in sulfoxidations,
see: a) S.-I. Murahashi, T. Oda, Y. Masui, J. Am. Chem.
Soc. 1989, 111, 5002–5003; b) Y. Imada, H. Iida, S.
Ono, Y. Masui, S.-I. Murahashi, Chem. Asian J. 2006, 1,
À3
containing a flavinium salt (c=1ꢅ10 M) and tetrabutylam-
À1
monium hexafluorophosphate (c=1ꢅ10 M) as supporting
electrolyte under argon atmosphere The scan rate was
À1
0
1
00 mVs The resulting formal standard potentials E ’ were
calculated as the mean of cathodic and anodic peak poten-
0
c
p
tials relative to ferrocene/ferrocenium couple [E ’=(E +
a
E )/2].
p
1
36–147; c) Y. Imada, T. Ohno, T. Naota, Tetrahedron
Lett. 2007, 48, 937–939; d) L. Baxovꢁ, R. Cibulka, F.
Hampl, J. Mol. Catal. A 2007, 277, 53–60; e) R. Cibul-
ka, L. Baxovꢁ, H. Dvo rˇ ꢁkovꢁ, F. Hampl, P. Mꢀnovꢁ, V.
Mojr, B. Plancq, S. Sayin, Collect. Czech. Chem.
Commun. 2009, 74, 973–993; f) J. Zˇ urek, R. Cibulka, H.
Acknowledgements
The authors wish to thank the Czech Science Foundation
(
Grant No. P207/12/0447) for financial support.
Dvo rˇ ꢁkovꢁ, J. Svoboda, Tetrahedron Lett. 2010, 51,
1
083–1086; g) B. J. Marsh, D. R. Carbery, Tetrahedron
Lett. 2010, 51, 2362–2365; h) Y. Imada, T. Kitagawa,
H.-K. Wang, N. Komiya, T. Naota, Tetrahedron Lett.
References
2
013, 54, 621–624.
[
1] a) I. Fernꢁndez, N. Khiar, Chem. Rev. 2003, 103, 3651–
[12] For application of chiral flavinium salts in enantioselec-
tive sulfoxidations, see: a) S. Shinkai, T. Yamaguchi, O.
Manabe, F. Toda, J. Chem. Soc. Chem. Commun. 1988,
1399–1401; b) S.-I. Murahashi, Angew. Chem. 1995, 107,
2670–2693; Angew. Chem. Int. Ed. Engl. 1995, 34,
2443–2565; c) R. Jurok, R. Cibulka, H. Dvo rˇ ꢁkovꢁ, F.
Hampl, J. Hoda cˇ ovꢁ, Eur. J. Org. Chem. 2010, 5217–
3
705; b) H. Pellissier, Tetrahedron 2006, 62, 5559–5601;
c) J. L. Garcꢂa Ruano, J. Alemꢁn, M. B. Cid, M. A.
Fernꢁndez-IbaÇez, M. C. Maestro, M. R. Martꢂn, A. M.
Martꢂn-Castro, in: Organosulfur Chemistry in Asym-
metric Synthesis, (Eds.: T. Toru, C. Bolm), Wiley-VCH,
Weinheim, 2008, pp 55–159; d) M. C. CarreÇo, G. H.
Adv. Synth. Catal. 2013, 355, 3451 – 3462
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
3461

Petra Mꢀnovꢁ et al.
FULL PAPERS
5
224; d) V. Mojr, V. Herzig, M. Bud eˇ sˇ ꢂnsk y´ , R. Cibulka,
[18] a) J. W. Bunting, Heterocycles 1980, 14, 2015–2045;
b) J. W. Bunting, Tetrahedron 1987, 43, 4277–4286.
[19] G. Merꢀnyi, J. Lind, J. Am. Chem. Soc. 1991, 113,
3146–3153.
[20] C. Hansch, A. Leo, R. W. Taft, Chem. Rev. 1991, 91,
165–195.
T. Kraus, Chem. Commun. 2010, 46, 7599–7601; e) V.
Mojr, M. Bud eˇ sˇ ꢂnsk y´ , R. Cibulka, T. Kraus, Org.
Biomol. Chem. 2011, 9, 7257–7580; f) T. Hartman, V.
Herzig, M. Bud eˇ sˇ ꢂnsk y´ , J. Jind rˇ ich, R. Cibulka, T.
Kraus, Tetrahedron: Asymmetry 2012, 23, 1571–1583;
g) H. Iida, S. Iwahana, T. Mizoguchi, E. Yashima, J.
Am. Chem. Soc. 2012, 134, 15103–15113.
[
21] In principle, a correction for the interaction between
the substituents should be taken into account. Never-
theless, the effect of disubstitution has been found to
be additive within the validity range of the substituent
constants in many cases, for examples, see: a) Y.
Huang, L. Liu, W. Liu, S. Liu, S. Liu, J. Phys. Chem. A
[
13] P. Mꢀnovꢁ, R. Cibulka, J. Mol. Catal. A: Chemical
2
012, 363–364, 362–370.
[
14] For catalysis of H O sulfoxidations with 5-ethyldihy-
2
2
droalloxazines, see: a) A. B. E. Minidis, J.-E. Bꢄckvall,
Chem. Eur. J. 2001, 7, 297–302; b) A. A. Lindꢀn, L.
Krꢆger, J.-E. Bꢄckvall, J. Org. Chem. 2003, 68, 5890–
2
011, 115, 14697–14707; b) O. Pytela, J. Kulhꢁnek, M.
ˇ
Ludwig, V. Rꢂha, Collect. Czech. Chem. Commun. 1994,
5
896; c) A. A. Lindꢀn, N. Hermanns, S. Ott, L. Krꢆger,
5
9, 627–638.
J.-E. Bꢄckvall, Chem. Eur. J. 2005, 11, 112–119;
d) A. A. Lindꢀn, M. Johansson, N. Hermanns, J.-E.
Bꢄckvall, J. Org. Chem. 2006, 71, 3849–3853.
[
22] D. D. Perrin, W. L. F Armarego, Purification of Labora-
th
tory Chemicals, 4 edn., Elsevier Science Ltd., Oxford,
1
996.
[
15] CCDC 887463 (5e) and CCDC 887464 (5f) contain the
supplementary crystallographic data for this paper.
These data can be obtained free of charge from The
[
23] I. R. Ager, A. C. Barnes, G. W. Danswan, P. W. Hair-
sine, D. P. Kay, P. D. Kennewell, S. S. Matharu, P.
Miller, P. Robson, D. A. Rowlands, W. R. Tully, R.
Westwood, J. Med. Chem. 1988, 31, 1098–1115.
Cambridge
www.ccdc.cam.ac.uk/data_request/cif.
16] P. Mꢀnovꢁ, F. Kafka, H. Dvo rˇ ꢁkovꢁ, S. Gunnoo, M.
Crystallographic Data
Centre via
[
[
24] F. Yoneda, S. Matsumoto, Y. Sakuma, S. Fukazawa, J.
[
[
Chem. Soc. Perkin Trans. 1 1975, 19, 1907–1909.
ˇ
Sanda, R. Cibulka, Adv. Synth. Catal. 2011, 353, 865–
70.
17] a) M. Stankovich, in: Chemistry and Biochemistry of
Flavoenzymes, (F. Mꢆller, Ed.), CRC, Boca Raton,
25] a) M. H. Ali, W. C. Stevens, Synthesis 1977, 764; b) B.
Li, A.-H. Liu, L.-N. He, Z.-Z. Yang, J. Gao, K.-H.
Chen, Green Chem. 2012, 14, 130–135; c) M. Kirihara,
J. Yamamoto, T. Noguchi, A. Itou, S. Naito, Y. Hirai,
Tetrahedron 2009, 65, 10477–10484.
8
1
991, pp 401–425; b) C. J. Rizzo, J. J. Hasford, J. Am.
Chem. Soc. 1998, 120, 2251–2255.
3462
asc.wiley-vch.de
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2013, 355, 3451 – 3462