Mild and Efficient Flavin-Catalyzed H2O2 Oxidation of Tertiar Amines to Amine N-Oxides

Article abstract of DOI:10.1021/jo980926d

A mild and highly effective H₂O₂ oxidation of tertiary amines has been developed by the use of flavin catalysis. Eight aliphatic amines were oxidized to their corresponding N-oxides in fast and selective reactions. For all substrates a considerable rate enhancement was observed compared to the noncatalyzed reactions. The product N-oxides were isolated in good yields using this mild oxidation system based on the environmentally attractive oxidant H₂O₂. As the catalyst, an N¹N⁵- dialkylated flavin was used as an analogue of the biologically important flavin redox cofactor. The catalytic cycle proposed for the flavin catalysis accounts for the observation that, in addition to the hydrogen peroxide oxidant, molecular oxygen is required for the initiation of the process.

Full text of DOI:10.1021/jo980926d

6
650
J . Org. Chem. 1998, 63, 6650-6655
2 2
Mild a n d Efficien t F la vin -Ca ta lyzed H O Oxid a tion of Ter tia r y
Am in es to Am in e N-Oxid es
†
,†,‡
Katarina Bergstad and J an-E. B a¨ ckvall*
Department of Organic Chemistry, University of Uppsala, Box 531, SE-751 21 Uppsala, Sweden,
and Department of Organic Chemistry, Arrhenius Laboratory, Stockholm University,
SE-106 91 Stockholm, Sweden
Received May 15, 1998
2 2
A mild and highly effective H O oxidation of tertiary amines has been developed by the use of
flavin catalysis. Eight aliphatic amines were oxidized to their corresponding N-oxides in fast and
selective reactions. For all substrates a considerable rate enhancement was observed compared to
the noncatalyzed reactions. The product N-oxides were isolated in good yields using this mild
1
5
2 2
oxidation system based on the environmentally attractive oxidant H O . As the catalyst, an N ,N -
dialkylated flavin was used as an analogue of the biologically important flavin redox cofactor. The
catalytic cycle proposed for the flavin catalysis accounts for the observation that, in addition to the
hydrogen peroxide oxidant, molecular oxygen is required for the initiation of the process.
In tr od u ction
Oxidation reactions with environmentally acceptable
recent years. These oxidants are highly attractive since
they are cheap and produce no toxic waste products in
contrast to many commonly employed inorganic electron-
accepting agents. Aqueous hydrogen peroxide is easy to
handle and has a high percentage of available oxygen
compared to most other oxidizing compounds except
dioxygen.
1
-19
oxidants such as molecular oxygen
and hydrogen
peroxide1
,7,8,15,20-37
have been intensively studied during
†
University of Uppsala.
University of Stockholm. Correspondence to this address. E-mail:
‡
jeb@organ.su.se.
Although hydrogen peroxide has a high oxidation
potential (1.76 V), there is generally a high activation
(
(
1) Murahashi, S.-I. Angew. Chem., Int. Ed. Engl. 1995, 34, 2443.
2) Hosokawa, T.; Nakahira, T.; Takano, M.; Murahashi, S.-I. J . Mol.
3
8
barrier for its reaction with organic substrates. As a
2 2
consequence, a number of catalytic procedures for H O
Catal. 1992, 74, 489.
3) Sim a´ ndi, L. I. Catalytic Activation of Dioxygen by Metal Com-
plexes; Kluwer: Dordrecht, The Netherlands, 1992.
4) B a¨ ckvall, J .-E.; Hopkins, R. B.; Grennberg, H.; Mader, M. M.;
Awasthi, A. K. J . Am. Chem. Soc. 1990, 112, 5160.
5) Bergstad, K.; Grennberg, H.; B a¨ ckvall, J .-E. Organometallics
998, 17, 45.
6) Bystr o¨ m, S. E.; Larsson, E. M.; Åkermark, B. J . Org. Chem. 1990,
5, 5674.
(
oxidation have been developed in which the activation
barrier is significantly lowered. Examples of catalytic
(
2
2-27
(
reactions are transition metal-catalyzed
and poly-
1
15,30-34
oxometalate-catalyzed
ganic substrates by H
oxidations of different or-
. Only to a limited extent have
(
O
2
5
2
(
(
7) Meunier, B. Chem. Rev. 1992, 92, 1411.
8) Arts, S. J . H. F.; Mombarg, E. J . M.; van Bekkum, H.; Sheldon,
bio- or biomimetic catalysts been employed in such
processes.
Amine N-oxides are widely employed oxidants which
7,20,28,29,35-37
R. A. Synthesis 1997, 597.
9) Mark o´ , I. E.; Giles, P. R.; Tsukazaki, M.; Chell e´ -Regnaut, I.;
Urch, C. J .; Brown, S. M. J . Am. Chem. Soc. 1997, 119, 12661.
(
can be prepared by H
2
O
2
oxidation of tertiary amines in
(
(
(
(
(
10) Lenz, R.; Ley, S. V. J . Chem. Soc., Perkin Trans. 1 1997, 3291.
11) Neumann, R. J . Org. Chem. 1989, 54, 4607.
12) Neumann, R.; Levin, M. J . Org. Chem. 1991, 56, 5707.
13) Neumann, R.; Dahan, M. Nature 1997, 388, 353.
14) Grate, J . H.; Hamm, D. R.; Mahajan, S. In Polyoxometalates:
39,40
a slow reaction.
Recently, catalytic H oxidations
2 2
O
(27) Sato, K.; Aoki, M.; Takagi, J .; Noyori, R. J . Am. Chem. Soc.
1997, 119, 12386.
From Platonic Solids to Anti-Retroviral Activity; Pope, M. T., M u¨ ller,
A., Eds.; Kluwer: Dordrecht, The Netherlands, 1994; p 281.
(28) Thellend, A.; Battioni, P.; Sanderson, W.; Mansuy, D. Synthesis
1997, 1387.
(
15) Hill, C. L.; Prosser-McCartha, C. M. Coord. Chem. Rev. 1995,
(29) Andersson, M.; Willetts, A.; Allenmark, S. J . Org. Chem. 1997,
62, 8455.
1
43, 407.
(
(
16) Larock, R. C.; Hightower, T. R. J . Org. Chem. 1993, 58, 5298.
17) van Benthem, R. A. T. M.; Hiemstra, H.; Michels, J . J .;
(30) Khenkin, A. M.; Hill, C. L. J . Am. Chem. Soc. 1993, 115, 8178.
(31) Duncan, D. C.; Chambers, R. C.; Hecht, E.; Hill, C. L. J . Am.
Chem. Soc. 1995, 117, 681.
(32) Neumann, R.; Gara, M. J . Am. Chem. Soc. 1994, 116, 5509.
(33) Neumann, R.; Gara, M. J . Am. Chem. Soc. 1995, 117, 5066.
(34) Ishii, Y.; Ogawa, M. In Reviews on Heteroatom Chemistry; Oae,
S., Ed.; MYU: Tokyo, 1990; Vol. 3, p 121.
Speckamp, W. N. J . Chem. Soc., Chem. Commun. 1994, 357.
18) R o¨ nn, M.; Andersson, P. G.; B a¨ ckvall, J .-E. Tetrahedron Lett.
997, 38, 3603.
19) Yoshino, Y.; Hayashi, Y.; Iwahama, T.; Sakaguchi, S.; Ishii, Y.
J . Org. Chem. 1997, 62, 6810.
(
1
(
(
(
20) Meunier, B.; Sorokin, A. Acc. Chem. Res. 1997, 30, 470.
21) Strukul, G. Catalytic Oxidations with Hydrogen Peroxide as
(35) Shinkai, S.; Yamaguchi, T.; Manabe, O.; Toda, F. J . Chem. Soc.,
Chem. Commun. 1988, 1399.
Oxidant; Kluwer: Dordrecht, The Netherlands, 1992.
22) Herrmann, W. A.; Fischer, R. W.; Marz, D. W. Angew. Chem.,
Int. Ed. Engl. 1991, 30, 1638.
23) Rom a˜ o, C. C.; K u¨ hn, F. E.; Herrmann, W. A. Chem. Rev. 1997,
7, 3197.
(36) Mazzini, C.; Lebreton, J .; Furstoss, R. J . Org. Chem. 1996, 61,
8.
(37) Murahashi, S.-I.; Oda, T.; Masui, Y. J . Am. Chem. Soc. 1989,
111, 5002.
(38) Curci, R.; Edwards, J . O. In Catalytic Oxidations with Hydrogen
Peroxide as Oxidant; Strukul, G., Ed.; Kluwer: Dordrecht, The
Netherlands, 1992; p 45.
(
(
9
(24) Al-Ajlouni, A. M.; Espenson, J . H. J . Am. Chem. Soc. 1995, 117,
9
243.
(
(
25) Zauche, T. H.; Espenson, J . H. Inorg. Chem. 1997, 36, 5257.
26) Rudolph, J .; Reddy, K. L.; Chiang, J . P.; Sharpless, K. B. J .
(39) Cope, A. C.; Ciganek, E. Org. Synth. Coll. Vol. 1963, IV, 612.
(40) VanRheenen, V.; Cha, D. Y.; Hartley, W. M. Org. Synth. Coll.
Vol. 1988, VI, 342.
Am. Chem. Soc. 1997, 119, 6189.
S0022-3263(98)00926-8 CCC: $15.00 © 1998 American Chemical Society
Published on Web 08/26/1998

Oxidation of Tertiary Amines to Amine N-Oxides
J . Org. Chem., Vol. 63, No. 19, 1998 6651
of aromatic N-heterocyclic compounds to their corre-
sponding N-oxides employing a biomimetic manganese
porphyrin28 or methyltrioxorhenium(VII)
41,42
as catalyst
were reported. Other oxidants employed in the oxidation
4
3
of tertiary amines include peracids, magnesium monop-
4
4
45
erphthalate, 2-sulfonyloxaziridines, R-azohydroper-
oxides46 and dioxiranes. Amine N-oxides are used as
stoichiometric oxidants, for example in osmium-catalyzed
47
biologically important flavin redox cofactor.3
5-37
These
cationic isoalloxazines are known to give an intermediate
flavin hydroperoxide like 1, an active oxygen donor, upon
4
0,48,49
dihydroxylation of olefins,
in ruthenium-catalyzed
oxidation of alcohols5
0,51
and in a mild procedure for
conversion of halides to aldehydes.5
2-54
61
reaction with hydrogen peroxide.
1
5
In connection with a project on the development of new
Interestingly, N ,N -dialkylated 5,10-dihydroallox-
oxidations of organic substrates,4
,5,55-57
one
O
2
or H
2
O
2
azines such as 3 are suggested to give a hydroperoxide
6
2
objective was to use an amine N-oxide as an in situ
generated oxidant. We therefore needed to develop a
mild and efficient oxidation of tertiary amines, prefer-
entially by the use of an environmentally friendly oxidant
such as hydrogen peroxide. Obviously, a direct oxidation
similar to 1 upon reaction with molecular oxygen. We
considered these alloxazine derivatives a highly attrac-
tive group to study as models for flavins in catalytic
oxidations. In this paper we report on the first, to the
1
5
best of our knowledge, efficient catalysis by an N ,N -
dialkyl blocked flavin analogue 3 applied on the oxidation
of tertiary amines by H O . Turnover numbers up to 182
2 2
of the amine by H O is too slow, and a catalytic
N-oxidation was therefore called for. Although a fast
oxidation of tertiary amines to amine N-oxides by the
stoichiometric flavin hydroperoxide 1 has been de-
2
2
were obtained with this flavin analogue as a biomimetic
catalyst (eq 1).
Resu lts a n d Discu ssion
Syn th esis of th e F la vin Ca ta lyst 3. The catalyst 3
was synthesized in a four-step procedure according to
Scheme 1. The tricyclic alloxazine skeleton was built up
via oxidation of barbituric acid, followed by condensation
between the alloxan (4) thus formed and o-phenylenedi-
scribed,58,59 no catalytic reaction employing a flavin as a
catalyst has been realized. In fact, only a few examples
of flavin-catalyzed H
2
O
2
oxidations of organic substrates
In one of these reactions secondary
amines were oxidized to imine oxides, however at a
are known.3
5-37,60
3
amine. After methylation of the two sp -nitrogens in 5,
the desired 5,10-dihydroalloxazine 3 was obtained by a
one-pot 1,4-hydrogenation/reductive alkylation proce-
dure.62 A hard electrophile like acetaldehyde preferen-
3
7
moderate rate and with less than 10 turnovers.
The few previously described catalytic flavin systems
for oxidation of organic substrates employ a perchlorate
salt of an isoalloxazine (e.g. 2) as a model of the
5
tially adds to the N -carbon in a reductive alkylation of
the 5,10-dihydroalloxazine, as has been described by
Hemmerich et al previously.63 Reduction by sodium
dithionite, a widely employed method for the reduction
of isoalloxazines,64 was tried in the presence of sodium
(
41) Cop e´ ret, C.; Adolfsson, H.; Khuong, T.-A. V.; Yudin, A. K.;
Sharpless, K. B. J . Org. Chem. 1998, 63, 1740.
42) The MTO catalyst has also been used for H
tertiary amines: Zhu, Z.; Espenson, J . H. J . Org. Chem. 1995, 60, 1326.
43) Mosher, H. S.; Turner, L.; Carlsmith, A. Org. Synth. Coll. Vol.
963, IV, 828.
44) Brougham, P.; Cooper, M. S.; Cummerson, D. A.; Heaney, H.;
Thompson, N. Synthesis 1987, 1015.
45) Zajac, W. W.; Walters, T. R.; Darcy, M. G. J . Org. Chem. 1988,
3, 5856.
(
2 2
O oxidation of
65
cyanoborohydride but resulted in an overreduced prod-
(
1
uct. The catalyst 3 is like most models of dihydroflavins
very sensitive toward molecular oxygen.6
2,64
Conse-
(
quently, it was found to be crucial to carry out the workup
under a strictly inert atmosphere for a successful prepa-
ration. The flavin analogue 3 was obtained as a yellow
solid, which was characterized by spectroscopic means.
In contrast, workup in air resulted in a brown powder
with different spectroscopic properties.
(
5
1
5
(46) Baumstark, A. L.; Dotrong, M.; Vasquez, P. C. Tetrahedron Lett.
987, 28, 1963.
(47) Ferrer, M.; S a´ nchez-Baeza, F.; Messeguer, A. Tetrahedron 1997,
3, 15877.
(
(
(
48) Schr o¨ der, M. Chem. Rev. 1980, 80, 187.
49) Ahrgren, L.; Sutin, L. Org. Proc. Res. Dev. 1997, 1, 425.
50) Griffith, W. P.; Ley, S. V.; Whitcombe, G. P.; White, A. D. J .
F la vin -Ca t a lyzed
2 2
H O Oxid a t ion of Ter t ia r y
Chem. Soc., Chem. Commun. 1987, 1625.
51) Ley, S. V.; Norman, J .; Griffith, W. P.; Marsden, S. P. Synthesis
994, 639.
Am in es. Reaction of N-methylmorpholine (7) with H O
2 2
(
in the presence of air and 2.5 mol % of the flavin 3
resulted in a rapid conversion of the amine to its
corresponding N-oxide 8. After about 1 h, more than 80%
conversion had been obtained (Table 1, entry 1). Several
1
(
52) Franzen, V.; Otto, S. Chem. Ber. 1961, 94, 1360.
(53) Suzuki, S.; Onishi, T.; Fujita, Y.; Misawa, H.; Otera, J . Bull.
Chem. Soc. J pn. 1986, 59, 3287.
(
(
54) Godfrey, A. G.; Ganem, B. Tetrahedron Lett. 1990, 31, 4825.
55) B a¨ ckvall, J .-E.; Chowdhury, R. L.; Karlsson, U. J . Chem. Soc.,
Chem. Commun. 1991, 473.
56) Grennberg, H.; Faizon, S.; B a¨ ckvall, J .-E. Angew. Chem., Int.
Ed. Engl. 1993, 32, 263.
57) Wang, G.-Z.; Andreasson, U.; B a¨ ckvall, J .-E. J . Chem. Soc.,
Chem. Commun. 1994, 1037.
(61) Bruice, T. C. Acc. Chem. Res. 1980, 13, 256.
(62) Mager, H. I. X.; Berends, W. Tetrahedron 1976, 32, 2303.
(63) Ghisla, S.; Hemmerich, P. J . Chem. Soc., Perkin Trans. 1 1972,
1564.
(64) Ghisla, S.; Hartmann, U.; Hemmerich, P.; M u¨ ller, F. J ustus
Liebigs Ann. Chem. 1973, 1388.
(65) Hutchins, R. O.; Hutchins, M. K. In Comprehensive Organic
Synthesis; Trost, B. M., Fleming, I., Eds.; Pergamon Press: Oxford,
U.K., 1991; Vol. 8, p 25.
(
(
(58) Ball, S.; Bruice, T. C. J . Am. Chem. Soc. 1979, 101, 4017.
(59) Ball, S.; Bruice, T. C. J . Am. Chem. Soc. 1980, 102, 6498.
(60) In one case the flavin system has been used to induce a radical
chain reaction: Mager, H. I. X.; Tu, S.-C. Tetrahedron 1994, 50, 6759.

6
652 J . Org. Chem., Vol. 63, No. 19, 1998
Bergstad and B a¨ ckvall
Sch em e 1. Syn th esis of th e F la vin Ca ta lyst 3a
F igu r e 1. Rate of conversion for catalyzed and noncatalyzed
2 2
H O oxidation of N,N-dimethyldodecylamine (9) to its corre-
sponding N-oxide.
a(a) CrO3, HOAc/H2O; (b) o-phenylenediamine, H3BO3, HOAc;
c) MeI, K2CO3, DMF; (d) H2, Pd/C, CH3COH, HCl, EtOH/H2O.
(
Ta ble 1. Oxid a tion of Ter tia r y Am in es Em p loyin g th e
oxide (8) is a commonly employed oxidant for the reoxi-
dation of transition metals such as osmium and ruthe-
nium (vide supra). Larger N-oxides such as 10 are often
F la vin -Hyd r ogen P er oxid e System a s Oxid a n ta
prepared by H
are useful surfactants.
2 2
O oxidation on an industrial scale as they
6
8
However, these large-scale
direct oxidations by hydrogen peroxide may require
elevated temperatures and also prolonged reaction times.
These disadvantages could possibly be overcome by the
use of a catalytic system such as the one described here.
N,N-Dimethylamine N-oxides, e.g. 14, are important
intermediates in the synthetically useful preparation of
3
9,69
olefins via the well-known Cope elimination.
Inter-
estingly, the catalytic oxidation of amine 13 occurs with
a rate comparable to most other substrates (Table 1). The
noncatalyzed reaction was found to be particularly slow
and required 2 days for completion, which has also been
observed by Cope.39 In the Meisenheimer rearrange-
ment, allylic or benzylic amine N-oxides serve as sub-
strates in a transformation into O-alkylated N,N-
disubstituted hydroxylamines.70 The N-oxide 16 would
be a suitable substrate for this reaction.
Heteroaromatic compounds are normally not oxidized
by hydrogen peroxide67 but require stronger oxidants
such as peracids. The flavin hydroperoxide 1 is described
2 2
to be between H O and mCPBA in reactivity as oxygen
donor.71 An interesting substrate to try to oxidize with
the catalytic system employing a flavin analogue would
therefore be pyridine. However, attempted oxidation of
pyridine with our system failed, and apparently the
oxidizing power of the intermediate hydroperoxide is too
low for this heteroaromatic substrate.
To estimate the rate enhancement for the catalytic
reaction, the noncatalyzed oxidation of tertiary amines
with hydrogen peroxide was studied in control experi-
ments. For most substrates less than 20% conversion
had been obtained after 2-7 h reaction time. This is to
be compared with more than 85% conversion within 1 h
for all reactions performed in the presence of the flavin
catalyst (Table 1). In Figure 1, the conversion versus
reaction time is plotted for catalyzed and noncatalyzed
oxidation of amine 9 as an example of the observed
difference in rate. For all aliphatic amines studied, a
a
The reactions were performed using 2.5 mol % flavin in MeOH-
b
d4 as described in the Experimental Section. Calculated by
division of the initial rates for catalyzed and noncatalyzed reac-
tions, respectively. 80% conversion. Estimated ratio of the
reactivities of catalytic flavin hydroperoxide and H2O2 (see text).
c
d
e
2
.8 mol % flavin was used.
other amines having different kinds of R groups attached
to the tertiary nitrogen atom were also oxidized in fast
reactions by the catalytic flavin-hydrogen peroxide sys-
tem (Table 1). Interestingly, although H
to be the active oxidant (vide infra), molecular oxygen
(66) When the reaction was performed under argon, the solubility
of the catalyst was low, and no conversion could be seen. However,
upon inlet of molecular oxygen the catalyst immediately dissolved, and
a rapid oxidation of the amine was observed.
2 2
O is thought
6
6
(
air) was necessary for generation of an active catalyst.
The product amine N-oxides are in many cases syntheti-
(67) Albini, A. Synthesis 1993, 263.
(
(
(
68) Sauer, J . D. Surfactant Sci. Ser. 1990, 34, 275.
69) Cope, A. C.; Trumbull, E. R. Org. React. 1960, 11, 317.
70) Meisenheimer, J . Ber. Dtsch. Chem. Ges. 1919, 52B, 1667.
cally useful compounds or important intermediates for
further transformations.6
7,68
N-Methylmorpholine N-
(71) Bruice, T. C. Isr. J . Chem. 1984, 24, 54.

Oxidation of Tertiary Amines to Amine N-Oxides
J . Org. Chem., Vol. 63, No. 19, 1998 6653
substantial rate enhancement was obtained in the pres-
ence of catalyst 3 (Table 1).
In the experiments reported in Table 1 an excess of
hydrogen peroxide (2.6 equiv) was employed to get a
better accuracy in the determination of the initial rate
of the noncatalyzed reactions, which are known to be
slow.3
9,40
The rate of the noncatalyzed reaction is pro-
concentration. The flavin-catalyzed
reaction, on the other hand, is not dependent on the H
2 2
portional to the H O
2
O
2
concentration (except at very low concentrations), which
is consistent with a mechanism involving the catalyst in
the rate-limiting step of the catalytic cycle (vide infra).72
Therefore, the rate enhancement factors in Table 1
F igu r e 2. Rate of the reaction when N,N-dimethyldodecy-
lamine (9) as substrate is added in several portions.
should roughly be 2.6 times larger when 1 equiv of H
is used. To demonstrate this point we have run one
experiment with 0.5 equiv of H . The rate enhance-
2 2
O
2
O
2
system. Up to 85% isolated yield and turnover numbers
of 73-85 on the flavin could easily be achieved (Table
2). By addition of hydrogen peroxide in portions to a
reaction employing only 0.4 mol % catalyst, the TON
could be raised to 182 (Table 2, entry 5). This, but also
the TON of 73-85 (Table 2, entries 1-4), is much higher
than what has generally been observed in the previously
described catalytic flavin systems, where less than 20
ment for oxidation of amine 13 could thus be increased
from 83 (Table 1, entry 4) to as much as 355 simply by
the use of a lower peroxide concentration.73 The possibil-
ity of using low concentrations of the oxidant (e.g. via
slow addition) makes our mild oxidation system highly
attractive for oxidation of tertiary amines in the presence
of other functionalities that may be sensitive to hydrogen
peroxide.
3
5-37,75
TON have been reported for the majorityofsubstrates.
To get an even better insight into how good catalyst
the flavin is, the ratio of the reactivity of flavin hydro-
peroxide compared to hydrogen peroxide can be calcu-
lated. A rate enhancement of 61 (Table 1, entry 1) can
be corrected for the fact that only 2.5 mol % flavin was
used by division of 61 with 0.025 giving 2440. However,
Su ggested Ca ta lytic Cycle F or th e F la vin An a -
logu e 3. In the catalytic oxidations of amines the 5,10-
dihydroalloxazine 3 most likely acts as a precursor of the
active catalyst (Scheme 2). By reaction with molecular
oxygen (step i) an intermediate flavin hydroperoxide 23
62,76
is formed.
Most likely the presence of an N-ethyl
as an excess of H
2 2
O was used (2.6 equiv), the number
substituent in the 5-position of 23 is of importance for
the success of the catalytic reaction. Earlier studies have
shown that in the absence of an alkyl group in the
2
440 should be multiplied by 2.6 to get an appropriate
comparison of the reactivities of the two hydroperoxides.
Thus, under the reaction conditions used in this study
the flavin hydroperoxide is more than 6000 times more
reactive than H O toward amine 7. In these calculations
2 2
it has been assumed that there is a fast oxidation of flavin
to its corresponding hydroperoxide, which is reasonable
10
corresponding position in N -alkylated flavins, a facile
71
2 2
elimination of H O occurs. Moreover, previously de-
scribed catalytic systems where this substituent is lack-
36,37
ing have turned out to be inactive.
The flavin hydroperoxide 23 would be able to transfer
its electrophilic oxygen to an amine, in close analogy with
what has been described by Bruice et al for the stoichio-
2 2
under conditions using a large excess of H O compared
to flavin. However, Murahashi has described formation
of a cationic flavin complex as the rate-determining step
in a catalytic cycle for a flavin such as 2.72 If this would
be valid also for flavin 3 under the reaction conditions
described here the calculated ratios between the reac-
tivities of the two different hydroperoxides should be even
higher than those reported in Table 1.74
metric oxidation of tertiary amines by hydroperoxide 1
58,59
(steps ii and iii).
Elimination of a hydroxide ion from
2
4 would give the cationic alloxazine 25 (step iv), which
on reaction with hydrogen peroxide can regenerate the
37,61,77
hydroperoxide 23.
In contrast to systems employing
isoalloxazines (e.g. 2) as flavin models, most likely there
The stability of a catalyst is of importance for the
development of a good catalytic system. This feature was
tested by addition of the substrate in several portions to
the reaction mixture. The catalyst was found to be active
even after a complete consumption of substrate, corre-
sponding to 42 turnover numbers (TON) for the flavin
catalyst. During the second addition a somewhat slower
oxidation was observed (Figure 2) but the rate was still
significantly higher than for the noncatalyzed reaction.
The amine N-oxides were also synthesized on a pre-
parative scale by employing this efficient flavin catalytic
is no stable 4a-hydroxy intermediate 24 formed with this
62,63
alloxazine system (Scheme 2).
In the present study
of flavin-catalyzed amine oxidations both molecular
oxygen and hydrogen peroxide have been shown to be
essential for an oxidation to occur. This observation is
explained by the catalytic cycle proposed (Scheme 2).
Con clu sion
Flavins are important cofactors in many biologically
important redox reactions. Their redox chemistry can
be taken advantage of in synthetic organic chemistry, as
have been exemplified by the mild and highly efficient
(72) Murahashi et al. describe the formation of a cationic flavin
complex as the rate-determining step in a detailed mechanistic study
for the flavin-catalyzed oxidation of methyl phenyl sulfide. See ref 37.
(
73) The reactions were performed using the same conditions as
described for the kinetic study in the Experimental Section, except
for the use of a lower amount of H (14 µL, 0.12 mmol). A larger
volume of MeOH-d (0.66 mL) was used to give the same total volume.
74) Flavin hydroperoxides are described to be 10 times more
reactive than H as oxygen donors: Bach, R. D.; Su, M.-D. J . Am.
Chem. Soc. 1994, 116, 5392.
(75) In the flavin-catalyzed oxidation of dibutyl sulfide a TON of 99
has been reported. See ref 37.
2
O
2
(76) This is in analogy with what has been reported for the reaction
3
4
of 1,5-dihydroflavins with
O
2
: Kemal, C.; Bruice, T. C. Proc. Nat. Acad.
4
(
Sci. U.S.A. 1976, 73, 995.
(77) Kemal, C.; Chan, T. W.; Bruice, T. C. J . Am. Chem. Soc. 1977,
99, 7273.
2
O
2

6
654 J . Org. Chem., Vol. 63, No. 19, 1998
Ta ble 2. P r ep a r a tion of Am in e N-Oxid es fr om Ter tia r y Am in es
Bergstad and B a¨ ckvall
entry
substrate
mol % of catal
reacn time (h)
yield (%)a
TONb
1
2
3
4
N-methylmorpholine (7)
1
1
1
1
8 h
2 h
8 h
2 h
4 h
75
85
73
82
73
75
85
73
82
182
N,N-dimethyldodecylamine (9)
N,N-dimethyl(cyclohexylmethyl)amine (13)
N,N-dimethylbenzylamine (15)
N,N-dimethyldodecylamine (9)
c
5
0.4
a
b
Isolated yields after purification. A minor pathway for formation of amine N-oxide via a direct oxidation by H2O2 cannot be completely
c
ruled out. H2O2 (3.3 mmol) was added in three portions with 40 min intervals to a solution of substrate (3 mmol) and flavin 3 (3.2 mg)
in 7.7 mL of MeOH, and the mixture was stirred for another 2 h 40 min.
Sch em e 2. Th e ca ta lytic cycle for fla vin m ed ia ted
oxid a tion s of a m in es.
were purchased from Lancaster except for N-methylmorpho-
line (7) (Fluka), dodecyl bromide (Fluka), N,N-dimethyl-
(cyclohexylmethyl)amine (13) (Aldrich) and dimethylamine
40% aqueous solution, Merck). MeOH-d was from Dr.
(
4
Glaser, AG Basel. Merckoquant peroxide test strips were
bought from Merck. Alloxan monohydrate (4) was prepared
by oxidation of barbituric acid.78 N,N-Dimethyldodecylamine
(
9), N,N-dimethyl-2-octylamine (11), N,N-dimethylbenzy-
lamine (15), and N,N-dimethylcycloheptylamine (17) were
prepared by reacting the corresponding bromides with N,N-
dimethylamine (see Supporting Information for details). DMF
2
was dried over CaH and distilled at reduced pressure before
use. Acetaldehyde was distilled over catalytic amounts of
p-toluenesulfonic acid. Other commercial chemicals were used
as received without further purification. The amine N-oxides
were purified by chromatography on basic aluminum oxide
(Aldrich, 150 mesh), while other compounds were purified on
silica gel (Merck silica gel 60, 230-400 mesh). Progress of
the amine oxidation reactions was followed by TLC on Merck
aluminum oxide 60 F254 plates (neutral) or Merck silica gel 60
F
254 plates. For hydrogenation reactions a Parr pressure
reaction apparatus was used.
Alloxa zin e (5). o-Phenylenediamine (6.75 g, 62 mmol) was
dissolved in HOAc (100 mL). To this solution was added a
mixture of 4 (10.0 g, 62 mmol) and H
in hot HOAc (400 mL). The reaction mixture was stirred for
5 min at room temperature. The precipitate formed during
3 3
BO (4.25 g, 69 mmol)
7
the reaction was filtered off and washed with HOAc and then
with ether. Drying gave 5 as a green solid (9.34 g, 70%). The
1
13
H and C NMR spectra of the product were identical to those
of commercial 5.
,3-Dim eth yla lloxa zin e (6). Alloxazine (5) (3.99 g, 18.6
mmol) and K CO (13.35 g, 96.6 mmol) were added to dry DMF
150 mL) (all of the 5 did not dissolve). The atmosphere was
changed to N , and MeI (2.7 mL, 43.4 mmol) was added. The
1
2
3
(
2
reaction mixture was stirred at room temperature for 2 h. The
solvent was removed at reduced pressure, and the remaining
H
2
O
2
oxidation of tertiary amines described in this paper.
The present study represents the first example where an
solid was dissolved in CH
separated, and the aqueous phase was extracted with 7 × CH
Cl . The combined organic phases were washed with saturated
2 2 2
Cl and H O. The phases were
1
5
N ,N -dialkyl blocked flavin is used in catalysis. The
flavin catalyst has been shown to have good efficiency,
as reflected by the high TON obtained when only 0.4 mol
2
-
2
aqueous NaCl and dried over MgSO
solvent gave 6 as a yellow powder. Purification of the crude
product by column chromatography (95:5 CH Cl /EtOAc) gave
as a yellow powder (3.99 g, 88%). NMR data were in
4
. Evaporation of the
%
catalyst was used. Moreover, this is the first catalytic
2
2
system where a fully reduced alloxazine acts as a
precursor of the active flavin catalyst, which is easily
generated in the presence of molecular oxygen (air). In
the oxidation of a number of tertiary amines a consider-
able rate enhancement was observed compared to the
noncatalyzed reactions. The product amine N-oxides
have been isolated in good yields using 0.4-1 mol %
flavin catalyst under these mild conditions.
Further studies are in progress on the application of
the flavin-based N-oxidation system on different catalytic
systems where an in situ generation of an amine N-oxide
would be desirable.
6
7
9,80
accordance with those previously reported in the literature.
6
2
1
,3-Dim eth yl-5-eth yl-5,10-dih ydr oalloxazin e (3). Com-
pound 6 (200 mg, 0.83 mmol) and Pd/C (10%, 81 mg, “preac-
tivated” under reduced pressure for 30 min) were suspended
in degassed EtOH (17 mL) and H O (16 mL) under Ar
2
bubbling. Concentrated HCl (1.4 mL) and freshly distilled
acetaldehyde (1.4 mL, 25 mmol) were added to the mixture.
The mixture was hydrogenated at room temperature and 1.1
atm pressure for 26 h. The workup was done under inert
atmosphere using conventional Schlenk techniques. The mix-
ture was filtrated through Celite giving a yellow solution. The
pH was adjusted to 7-8 by addition of concentrated NH
3
(5
Exp er im en ta l Section
(78) Holmgren, A. V.; Wenner, W. Org. Synth. Coll. Vol. 1963, IV,
Gen er a l P r oced u r es. 1H and 13C NMR spectra were
recorded at 400 and 100.6 MHz or 300 and 75.4 MHz,
respectively. Chemical shifts (δ) are reported in ppm, using
23.
(
79) Grande, H. J .; van Schagen, C. G.; J arbandhan, T.; M u¨ ller, F.
Helv. Chim. Acta 1977, 60, 348.
80) Grande, H. J .; Gast, R.; van Schagen, C. G.; van Berkel, W. J .
H.; M u¨ ller, F. Helv. Chim. Acta 1977, 60, 367.
(
4
residual solvent or Me Si as internal standard. Most reagents

Oxidation of Tertiary Amines to Amine N-Oxides
J . Org. Chem., Vol. 63, No. 19, 1998 6655
8
3
mL). At this pH the product seemed to be extremely air
sensitive. As soon as the color of the mixture started to change
from yellow to orange-brown, small amounts of aqueous
Na S O were added which gave the yellow color back again.
2 2 4
Most of the solvents were evaporated. The solution was cooled
N,N-dimethylbenzylamine N-oxide (16), N-methylpiperidine
N-oxide (20),8
authentic samples prepared by H
according to literature procedures (N,N-dimethyl-2-octylamine
4,85
and triethylamine N-oxide (22) ) or to
oxidation of the amines
86
2
O
2
8
7
N-oxide (12), N,N-dimethyl(cyclohexylmethyl)amine N-oxide
(14),39 and N,N-dimethylcycloheptylamine N-oxide (18) ).
B. With ou t F la vin Ca ta lyst. N-Meth ylm or p h olin e
N-Oxid e (8). To a solution of N-methylmorpholine (27 µL,
88
in ice-water bath, and the solid was collected by filtration.
The solid was suspended in degassed H
small amounts of dissolved Na S O
2 2 4
2
O (1.4 mL) containing
and stirred for 20 min.
Filtration gave a yellow powder, which was washed several
times with degassed H O. The solid was dried overnight at
0.25 mmol) in MeOH-d
(75 µL of a 27% aqueous solution, 0.655 mmol). The time
measurement was started, and the reaction was followed by
4
(0.6 mL) in an NMR tube was added
2
2 2
H O
reduced pressure to give 3 as a yellow powder (166 mg, 74%).
When 3 was stored at -30 °C under argon, its catalytic activity
remained for several months. Since 3 is very sensitive toward
1
H NMR for 10 h (28% conversion). The 1H and C NMR
13
spectra of the product were identical to those of commercial
8.
Oxid a tion of Oth er Ter tia r y Am in es. The reactions
were followed for 7-12 h. The products were characterized
as described above.
Gen er a l P r oced u r e for P r ep a r a t ion of Am in e N-
Oxides. N,N-Dim eth yldodecylam in e N-Oxide (10). Amine
9 (214 mg, 1.0 mmol) was stirred in MeOH (2.5 mL). To this
mixture the flavin catalyst 3 (3.1 mg, 0.011 mmol, 1.1 mol %)
molecular oxygen, the NMR sample was prepared in CDCl
having a layer of aqueous (D O) Na on top. The sample
was prepared under argon. H NMR (CDCl , 300 MHz): δ
.86 (m, 2H), 6.78 (dt, 1H, J ) 2.0, 7.4 Hz), 6.60 (dd, 1H, J )
.2, 7.7 Hz), 4.8 (br s, 1H), 3.47 (s, 3H), 3.44 (q, 2H, J ) 7.1
3
2
2
S
2
O
4
1
3
6
1
1
3
Hz), 3.34 (s, 3H), 1.13 (t, 3H, J ) 7.0 Hz). C NMR (CDCl
7
1
3
,
5.4 MHz): δ 158.0, 150.3, 146.3, 136.2, 134.9, 125.0, 123.5,
22.5, 114.7, 99.6, 50.6, 28.7, 28.3, 11.4.81
Gen er a l P r oced u r e for th e Kin etic Stu d y. The conver-
and H
added. After 2 h of stirring at room temperature, excess H
was destroyed by addition of solid MnO (93 mg). Filtration
through Celite, followed by evaporation of the solvents under
reduced pressure, gave the crude product, which was purified
2 2
O (125 µL of a 27% aqueous solution, 1.1 mmol) were
sion rates for reactions performed with flavin catalysis and
for those run in the absence of the catalyst were determined
2 2
O
2
1
by integration of the corresponding H signals in amine and
the product amine N-oxide. For most compounds the CHNMe
signals were integrated; however, in a few cases the CHNMe
2 3 2 2 2 2
on basic Al O (gradient 100% CH Cl to 95:5 CH Cl /MeOH).
or CH
2
NMe signals were integrated due to better separation
Immediate evaporation of the solvents afforded 195 mg (85%)
of 10 as a white solid. The NMR spectra of the product were
in accordance with literature data.82
from other peaks. The rate enhancement was calculated by
division of the rate at low conversion (10%) for the catalyzed
and noncatalyzed reactions, respectively.
Ackn owledgm en t. Financial support from the Swed-
ish Research Council for Engineering Sciences and the
Swedish Natural Science Research Council is gratefully
acknowledged.
A. With F la vin Ca ta lyst. N-Meth ylm or p h olin e N-
Oxid e (8). Flavin 3 (1.7 mg, 0.006 mmol) was added to a
solution of N-methylmorpholine (27 µL, 0.25 mmol) in MeOH-
4 2 2
d (0.6 mL) in an NMR tube. To this mixture H O (75 µL of
a 27% aqueous solution, 0.655 mmol) was added, and the time
measurement was started. After a few minutes almost all 3
had dissolved to give a clear and yellow solution. During the
course of the reaction small amounts of a white precipitate
Su p p or t in g In for m a t ion Ava ila b le: 1H and 13C NMR
spectra of flavin 3, experimental procedure for the preparation
of amines 9, 11, 15, and 17, spectral data for amine N-oxides
12, 14, and 18, and experimental details for amine oxidations
as described in the kinetic study (8 pages). This material is
contained in libraries on microfiche, immediately follows this
article in the microfilm version of the journal, and can be
ordered from the ACS; see any masthead page for ordering
information.
1
were formed. The reaction was monitored by H NMR (5 min
intervals for the first 35 min, thereafter with 10 min intervals
until the reaction time was 65 min (80% conversion)). The
sample was manually shaken between the NMR measure-
1
13
ments. The H and C NMR spectra of the product were
identical to those of commercial 8.
Oxid a tion of Oth er Ter tia r y Am in es. The reactions
J O980926D
1
were followed by H NMR for 35-65 min (5 min intervals for
the first 35 min, thereafter with 10 min intervals until >85%
(83) Lorand, J . P.; Anderson, J . L. J .; Shafer, B. P.; Verral, D. L. J .
conversion). The NMR spectra of the products were compared
Org. Chem. 1993, 58, 1560.
(84) Chastanet, J .; Roussi, G. J . Org. Chem. 1985, 50, 2911.
to literature data (N,N-dimethyldodecylamine N-oxide (10),82
(85) Al-Rawi, J . M. A.; Flamerz, S.; Khuthier, A. H. Spectrochim.
Acta A. 1985, 41, 1391.
(
81) The 13C NMR shifts of the N -ethyl group were found to be
5
(86) Bravo, R.; Laurent, J .-P. J . Chem. Res., Miniprint 1983, 710.
(87) Z a´ vada, J .; P a´ nkov a´ , M.; Svoboda, M. Collect. Czech. Chem.
Commun. 1973, 38, 2102.
(88) Cope, A. C.; Pike, R. A.; Spencer, C. F. J . Am. Chem. Soc. 1953,
75, 3212.
almost identical to what has been reported for the similar compound
-ethyl-3,7,8,10-tetramethyl-1,5-dihydroisoalloxazine: van Schagen, C.
G.; M u¨ ller, F. Helv. Chim. Acta 1980, 63, 2187.
82) Brycki, B.; Szafran, M. Magn. Reson. Chem. 1992, 30, 535.
5
(