Studies on substituted aromatic diselenides as catalysts for selective alcohol oxidation using tert-butyl hydroperoxide

Article abstract of DOI:10.1002/ejoc.201100487

We have investigated the Ph₂Se₂-catalysed oxidation of alcohols with tert-butyl hydroperoxide (TBHP) by using a variety of spectroscopic techniques and reaction calorimetry. We showed that the active oxidant is benzeneseleninic anhydride (BSA), formed by reaction of Ph ₂Se₂ with TBHP. The influence of aromatic substituents on the activity of a selection of nine aromatic diselenides involved in alcohol oxidations was studied. Calorimetric experiments showed that compounds that are activated more rapidly have higher initial activity in the catalytic oxidation of benzyl alcohol. An exception to this rule was observed when the diselenide contained a dimethylamino group in the ortho position, which can apparently compensate for its slow pre-activation by a special ortho effect. An aliphatic alcohol, 1-decanol was also used as a substrate. Besides diphenyl diselenide only three of the substituted diaryl diselenides studied showed reasonable activity after pre-activation with TBHP. Dimesityl diselenide displayed the highest activity and selectivity for the oxidation of 1-decanol. Specificity for a given aldehyde could be achieved by using an oxidant and substrate feed protocol. The oxidation steps from Ph₂Se₂ to benzeneseleninic anhydride, which is a gooddehydrogenation agent, have been elucidated by a combination of spectroscopic techniques, both in situ and ex situ. The oxidation potentials of derivatives of this anhydride have been investigated in the oxidation of 1-decanol.

Full text of DOI:10.1002/ejoc.201100487

FULL PAPER
DOI: 10.1002/ejoc.201100487
Studies on Substituted Aromatic Diselenides as Catalysts for Selective Alcohol
Oxidation Using tert-Butyl Hydroperoxide
John C. van der Toorn,[a] Gerjan Kemperman,[b] Roger A. Sheldon,[a] and
Isabel W. C. E. Arends*[a]
Keywords: Selenium / Oxidation / Calorimetry / Peroxides / Alcohols
We have investigated the Ph2Se2-catalysed oxidation of
alcohols with tert-butyl hydroperoxide (TBHP) by using a
variety of spectroscopic techniques and reaction calorimetry.
We showed that the active oxidant is benzeneseleninic anhy-
dride (BSA), formed by reaction of Ph2Se2 with TBHP. The
influence of aromatic substituents on the activity of a selec-
tion of nine aromatic diselenides involved in alcohol oxi-
dations was studied. Calorimetric experiments showed that
compounds that are activated more rapidly have higher ini-
tial activity in the catalytic oxidation of benzyl alcohol. An
exception to this rule was observed when the diselenide con-
tained a dimethylamino group in the ortho position, which
can apparently compensate for its slow pre-activation by a
special ortho effect. An aliphatic alcohol, 1-decanol was also
used as a substrate. Besides diphenyl diselenide only three of
the substituted diaryl diselenides studied showed reasonable
activity after pre-activation with TBHP. Dimesityl diselenide
displayed the highest activity and selectivity for the oxidation
of 1-decanol. Specificity for a given aldehyde could be
achieved by using an oxidant and substrate feed protocol.
in the oxidation of the allylic alcohol (E)-2-hexen-1-ol with
TBHP, although the catalyst loading was not reported.[5]
Using N-chlorosulfonamides, such as Chloramine-T so-
dium salt or N,4-dichlorobenzenesulfonamide sodium salt,
Kuwajima and co-workers observed a large improvement
with diphenyl diselenides containing ketone or ester groups
as ortho substituents on the catalysts in the oxidation of
activated alcohols. In a more recent article, bis[2-(2-pyrid-
yl)phenyl] diselenide was shown to be a more efficient cata-
lyst in combination with an N-chlorosulfonamide as the ox-
idant, and catalyst loadings could be reduced to 0.2 mol-
%.[7] These findings demonstrate that aromatic diselenides
still hold great promise for catalytic oxidations of alcohols.
Recently, we revisited the oxidation of alcohols using
Ph2Se2 as the catalyst and TBHP as the terminal oxidant.
Results of a mechanistic study indicated the involvement of
an autocatalytic step with benzeneseleninic anhydride
(BSA) as the key intermediate (Figure 2). Based on these
results we were able to develop a procedure for the catalytic
oxidation of activated alcohols using TBHP and 1 mol-%
of Ph2Se2 as catalyst.[8] The formation of BSA was moni-
tored by using reaction calorimetry in the absence of the
alcohol substrate. Its formation exhibited a distinct exother-
mic profile, which we attributed to some sort of autocataly-
sis. Driven by the potential of diselenides as catalysts, there
is a need to gain deeper insight in the mechanism of the
reaction.
Introduction
The selective oxidation of alcohols is a prominent trans-
formation in organic chemistry, and there is an ongoing
quest for selective and sustainable methods.[1] The use of
aromatic diselenide-based oxidation catalysts was first re-
ported in 1978,[2] and these reagents have subsequently been
shown to be versatile and selective catalysts in a number
of oxidative transformations.[3] Among other reactions, they
catalyze the oxidation of alcohols, albeit with generally high
catalyst loadings (Figure 1). Many groups have reported the
use of different diselenides in alcohol oxidations.[4–7] How-
ever, a systematic investigation of different diselenides in
combination with an easily accessible oxidant has not been
published. One of the best selenium catalysts in terms of
substrate scope is dimesityl diselenide, although the catalyst
loading is usually 50 mol-%.[4] Kuwajima and co-workers
reported the effective use of bis(p-chlorophenyl) diselenide
Figure 1. Catalytic oxidation of alcohols with aromatic diselenides.
[a] Department of Biotechnology, Delft University of Technology,
Julianalaan 136, 2628 BL Delft, The Netherlands
Fax: +31-15-2781415
E-mail: i.w.c.e.arends@tudelft.nl
[b] Department of Process Chemistry, Merck Sharp and Dohme,
Molenstraat 110, 5342 CC Oss, The Netherlands
Our first goal in the present study was to further extend
our knowledge of how Ph2Se2 is activated by TBHP at ele-
vated temperatures. To accomplish this goal a range of spec-
troscopic techniques was applied (Step 1). Our second goal
Eur. J. Org. Chem. 2011, 4345–4352
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4345

I. W. C. E. Arends et al.
FULL PAPER
oxidation of the diselenide. We know that, at this stage, the
BSA is not yet formed since we have observed that the oxi-
dation of benzyl alcohol does not start when added after
10 min.[8] In contrast, if we added the benzyl alcohol after
40 min, an equimolar amount was oxidized (95% yield).
Figure 2. Catalytic cycle of the diselenide-catalyzed oxidation of
alcohols.
Figure 3. Spectral changes of Ph2Se2 upon oxidation with TBHP
and subsequent addition of benzyl alcohol (after 40 min).
was to investigate the influence of various substituents on
the pre-activation of substituted diselenides. The potential
of these compounds as catalysts in the oxidation of
alcohols, including aliphatic ones, was studied with the aim
of expanding the scope of alcohol substrates amenable to
oxidation.
To further investigate the nature of the active species, we
turned to in situ Raman and IR spectroscopy. The IR data
are relevant in view of the Se=O stretch vibration, which
has
a complicated and intense absorption around
850 cm–1.[10] The IR data corroborated previously acquired
UV data as there was again a pronounced spectral change
after 5 min. The time plots of two distinct selenium–oxygen
vibration frequencies (860 and 840 cm–1) are shown in Fig-
ures 4 and 5. From the combination of the UV and IR data
we hypothesize that after 5 min the first Se=O bond is
formed (Figure 4). When this process is almost complete,
the second Se=O stretch appears in the spectrum (Figure 5).
Results and Discussion
Spectroscopic Studies of the Formation of BSA as the
Active Oxidant from Diphenyl Diselenide
In our previous studies of the Ph2Se2-catalyzed oxidation
of benzyl alcohol we observed an induction period during
which no substrate was converted. This suggested that the
activation of Ph2Se2 needed to be complete in order for the
catalytic reaction to start. When testing this hypothesis by
simply mixing Ph2Se2 with TBHP in toluene we observed a
peak in reaction calorimetry after approximately 25 min.
We found that if we added the substrate after this exother-
mic event, conversion of the substrate was almost instan-
taneous. The oxidation of Ph2Se2 to BSA (Figure 2, Step 1)
consists of three two-electron oxidations, and several pos-
sible pathways can be envisaged.
We started our study using UV/Vis spectroscopy, which
is highly suitable for determining the initial oxidation as
oxidized selenium species, such as diphenyl selenoxide, do
not absorb light above 300 nm.[9] Samples were withdrawn
from the reaction mixture and were consecutively diluted,
thus allowing for absorption measurement acquisitions
(Figure 3). After mixing the warm sample with cold tolu-
ene, the signal was constant for several hours. The absorp-
Figure 4. Time profile of IR band at 860 cm–1 during the activation
of Ph2Se2 by TBHP. At t = 39 min, 1 equiv. of benzyl alcohol was
added.
In Raman spectroscopy, the Se–Se stretch vibration has
tion maximum around 330 nm almost immediately de- a distinct frequency, providing us with information on the
creased to one third of its original intensity upon exposure time of the putative Se–Se bond cleavage. Unfortunately,
of 1 equiv. of Ph2Se2 to 1 equiv. of TBHP. After 10 min, no spectra could only be recorded in heptane, which suffered
further changes in absorption could be detected. We assign a major drawback; after 23 min, BSA started to precipitate
the observed change in the absorption signal to the first from the solution. However, the activation profile of Ph2Se2
4346
www.eurjoc.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2011, 4345–4352

Aromatic Diselenides as Catalysts for Selective Alcohol Oxidation
Figure 7. Path of autocatalytic activation of Ph2Se2 by TBHP.
Figure 5. Time profile of IR band at 840 cm–1 during the activation
of Ph2Se2 by TBHP. At t = 39 min, 1 equiv. of benzyl alcohol was
added.
oxyphenylseleninate (5) is presumed to be a more effective
oxygen transfer oxidant than TBHP itself. Using NMR
spectroscopy, we expected to observe other oxidized sele-
nium compounds in addition to BSA at the end of the reac-
tion. We studied the activation of Ph2Se2 by 1 equiv. of
TBHP using NMR spectroscopy. The time course of the
aromatic region of the NMR signals is shown in Figure 8.
by TBHP in reaction calorimetry was almost identical (data
not reported here). Therefore, we feel that the data obtained
in either toluene or heptane are interchangeable. From our
previous experiments we expected a change in composition
between 10 to 15 min after addition of the TBHP. Indeed,
after 15 min a large drop in the Se–Se stretch vibration was
observed (Figure 6). We presume that this event is indicative
of BSA formation, which corroborates the calorimetric pro-
file showing strong exothermicity after 15 min of reaction.
A tentative mechanism can be envisaged for the activation
of Ph2Se2 (Figure 7).
Figure 8. NMR spectra of the activation of Ph2Se2 by TBHP, data
shown for the region of δ = 8.5–7.5 ppm. After t = 30 min, 1 equiv.
of benzyl alcohol was added.
Figure 6. Normalized Raman spectral data; upper line Se–Se
stretch vibration (310 cm–1), lower line C=O stretch vibration
(1710 cm–1). TBHP was added at t = 0 min and benzyl alcohol
(0.9 equiv.) at t = 39 min.
The spectra revealed two new sets of peaks only after
15 min, which increased in intensity up to 25 min. The
The first oxidation (Ox I) of 1 to 2 is consistent with UV newly found signals correspond to BSA, in a quantity of
and IR spectral data and occurs within 5–10 min. After this, 1/3 of the mixture, while the other 2/3 of the mixture was the
the second oxidation of 2 to 3 (Ox II) takes place as evi- remaining Ph2Se2. We were not able to find any distinct
denced by IR spectroscopy. The final oxidation of com- signals that could be attributed to other intermediates on
pound 3 to BSA (4) then occurs between 15 and 20 min, the path from Ph2Se2 to BSA. This is consistent with a lit-
as shown by Raman spectroscopy and calorimetry. In this erature report on the oxidation of bis(p-fluorophenyl) dis-
scenario (Figure 7) we have also included the formation of elenide monitored by 19F NMR spectroscopy.[11] This is a
species 5. In our previous study we postulated that this in- remarkable feature of the oxidation of these two aromatic
termediate is most likely responsible for the autocatalysis diselenides. Apparently, the second and third oxidation of
(vide supra) that occurs when 4 is formed. tert-Butyl per- Ph2Se2 proceeds much more easily than the first as sug-
Eur. J. Org. Chem. 2011, 4345–4352
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4347

I. W. C. E. Arends et al.
FULL PAPER
gested by the formation of a 2:1 mixture of Ph2Se2 and cially not available compounds, except for 9 (which was pre-
BSA, rather than a more complex mixture of selenides in pared by direct lithiation of pentafluorobenzene followed
different oxidation states.
by reaction with selenium[14]), were constructed by using
the corresponding Grignard reagents. During these synthe-
ses, we noticed that the Grignard reactions often led to
large amounts of monoselenide adducts besides the diselen-
ide.[15] The only way these could be separated was by re-
versed-phase silica chromatography. Bis[2-(dimethylamino)-
phenyl] diselenide (8) was successfully prepared by using a
Knochel Mg/Br exchange, after which point, the higher ate
complex was created by adding 1 equiv. of dioxane.[16] This
complex was treated with elemental selenium to yield the
target compound.
Study of Substituted Diselenides as Pre-catalysts in
Oxidations with TBHP
The influence of aromatic substituents, on the potential
of diselenides as pre-catalysts in oxidations with TBHP was
investigated. The goal was to extend the substrate scope of
the current method towards non-activated alcohols. For this
purpose a selection of substituted diaryl diselenides was
tested (Figure 9).
The reaction of compounds 1–9 with TBHP was studied
by using calorimetry. In this way the so-called activation
time – the time needed to form the anhydride as the cata-
lytic species – and activation energy could be measured. The
influence of substituents on these parameters is listed in
Table 1. Notably, the diselenide concentration was lower
(6.3 mm) than in previous experiments, leading to a longer
induction period of 50 min instead of the aforementioned
25 min.
It turns out that Ph2Se2 is one of the slowest compounds
to achieve activation. The fastest of the series was found to
be compound 4, although this compound delivered the low-
est heat of activation among all compounds studied. Com-
pound 8 showed the same profile as 1, but it had the highest
activation energy, more than twice that of most other dis-
elenides. Very remarkable were the data for the series of
compounds 4–6; the dimesityl diselenide was faster to
achieve activation than either the 4- or 3-methyl-substituted
catalysts. Methylation at the 3- or 4-positions significantly
slowed the oxidation.
Figure 9. Selected compounds used to investigate the substituent
effect.
Compound 1 is, by definition, our reference catalyst.
Compound 2 [bis(2-nitrophenyl) diselenide] is a well-known
and often-applied epoxidation and Baeyer–Villiger oxi-
dation catalyst.[12] Compound 3 has been referred to as an
effective catalyst in the oxidation of allylic alcohols.[5] Com-
pound 4 is known to be a very effective pre-catalyst for
alcohol oxidations, although the catalyst loading is often
very high.[5] Compounds 5 and 6 were selected for compari-
son with compound 4. Compounds 7 and 8, in contrast to
Substituted Diaryl Diselenides in the Catalytic Oxidation of
Benzyl Alcohol
Having established the activation behaviour of catalysts
compound 2, possess electron-donating groups in the ortho in the absence of substrate, the overall reaction efficiency
position of the aromatic ring. Bis(pentafluorophenyl) dis- for the catalytic oxidation of benzyl alcohol by using 5 mol-
elenide (9) was selected, because the corresponding seleninic % of compounds 1–8 was studied. The results are presented
acid was previously reported as an active reagent in dehy- in Figure 10 for the compounds with the lowest rates and
drogenations of non-activated alcohols.[13] The commer- in Figure 11 for compounds with the highest rates.
Table 1. Induction period (activation times) with substituted diselenides. In each reaction diselenide (0.25 mmol) was dissolved in toluene
(40 mL), and TBHP (0.5 mmol) was added at 80 °C. The activation time is defined as the time intervening the addition of TBHP and
acquisition of the highest peak in the exothermic profile. The activation energy is calculated by integrating the peak area.
Entry
Diselenide
Substitution pattern
Activation time
[min]
Activation energy
[kJ/mol]
1
2
3
4
5
6
7
8
9
1
2
3
4
5
6
7
8
9
–
50
n.d.[a]
52
22
38
150
n.d.[a]
230
120
150
2-nitro
4-chloro
2,4,6-trimethyl
4-methyl
3-methyl
42
145
2-methoxy
2-(dimethylamino)
pentafluoro
n.d.[a]
49
n.d.[a]
345
52
190
[a] n.d. = not detected.
4348
www.eurjoc.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2011, 4345–4352

Aromatic Diselenides as Catalysts for Selective Alcohol Oxidation
taining substituent on the other hand can possibly enhance
the catalytic mode of action, a phenomenon that has been
found in selenium-based glutathione peroxidase mimics.[17]
Oxidation of an Aliphatic Alcohol: 1-Decanol
Previously we had shown that, using an oxidant and sub-
strate feed in a loop cycle, high selectivity and efficiency
can be attained with diphenyl diselenide as the oxidation
catalyst.[8] To further explore the scope of our substituted
diselenides, a catalytic test was performed with 5 mol-% di-
selenide and 1-decanol as the substrate. Unfortunately, the
maximum yield of decanal attained was 8%, which declined
over time as it was further oxidized to capric acid. One of
the other by-products was the decyl ester of capric acid.
Therefore, we decided to explore the use of an oxidant
and substrate feed technique. First the capacity of different
diselenides in the dehydrogenation of 1-decanol was studied
in a stoichiometric fashion: pre-activation with 1 equiv. of
TBHP was followed by the addition of a stoichiometric
amount of 1-decanol. The conversion of substrate was
monitored over time (Figure 12). In these experiments tolu-
ene was replaced by (trifluoromethyl)benzene to prevent
oxidation of the solvent. Compounds 5 and 6 were not
tested in these experiments because of their lack of activity
in the catalytic oxidation of 1-decanol.
Figure 10. Slowest profiles for the oxidation of benzyl alcohol with
diselenides (5%) at 80 °C and TBHP (1.1 equiv.).
Figure 11. Fastest profiles for the oxidation of benzyl alcohol with
diselenides (5%) at 80 °C and TBHP (1.1 equiv.).
We observed that the slightly electron-donating methyl
groups seem to promote fast initial turnover of the sub-
strate. All alkyl-substituted diaryl diselenides were found to
be among the fastest to convert substrate. Second, the ortho
position seemed to be very important as contradictory re-
sults had been obtained with diselenides substituted at this
position. Compound 7, the original epoxidation catalyst,
was found to possess the lowest activity. When comparing
calorimetry data obtained during pre-activation studies
with activity data, a clear correlation is observed between
Figure 12. Yield of decanal at different time points with pre-acti-
vated diselenides. Diselenide (0.5 mmol) and TBHP (1.1 mmol)
were premixed at 80 °C in (trifluoromethyl)benzene for 2 h, fol-
lowed by the addition of 1-decanol (1 mmol).
Compounds 4, 7 and 9 all completely converted the
the time of activation and the initial catalytic activity of the alcohol to the aldehyde. The speed of oxidation was con-
catalyst. The two compounds that did not show activation siderably slower than in the same experiment with Ph2Se2
in the calorimetric experiments (2 and 7), showed the lowest and benzyl alcohol (the latter gave full conversion to benz-
oxidation activity. The main exception was compound 8, aldehyde within 10 min). In the case of the oxidation of 1-
although we hypothesize that the 2-(dimethylamino) group decanol conversion took at least 120 min. In these experi-
could possibly act as a base in reactions containing this ments the effect of pre-activation was excluded, and what
diselenide. Calorimetry data obtained with 8 suggests that we observed was the primary dehydrogenating activity of
it does not have an effect on activation, but that the dimeth- the corresponding selenenic anhydride with 1-decanol. It
ylamino moiety may play an important role in dictating can therefore be envisaged that Step 3 in Figure 2 is fav-
catalytic activity.
oured by strong inductive and electron-donating substitu-
The above results demonstrate that formation of the ents, such as the o-methoxy group in compound 7 and the
active intermediate is highly influenced by substituents on three methyl substituents in compound 4. Surprisingly, bis-
the aromatic ring. A strong electron-withdrawing or elec- (pentafluorophenyl) diselenide (9), once pre-activated,
tron-donating substituent at the ortho position diminishes showed good activity. These results show that the actual
the activity of the diselenide in catalysis. A nitrogen-con- rate-determining step during oxidation is not well under-
Eur. J. Org. Chem. 2011, 4345–4352
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4349

I. W. C. E. Arends et al.
FULL PAPER
stood, although we presume that a number of factors are 70 °C.[13] This dehydration may be more difficult for substi-
important in the dehydrogenation of alcohols by BSA and tuted diselenides, thus requiring higher temperatures. Al-
its derivatives.
though it is possible to heat up the mixture to higher tem-
We decided to test the most promising catalysts in the peratures, TBHP is unstable at temperatures above 90 °C,
batch-wise oxidation of 1-decanol, using our previously de- so this is not a viable option. However, compared to pre-
scribed protocol. The protocol was changed in order to pro- vious results in the literature, our system has the lowest cat-
long the period between the addition of 1-decanol and a alyst loading (25%) and highest yield (98%) in the oxi-
new amount of TBHP for regeneration of the catalyst. The dation of 1-decanol.
stepwise addition of substrate and oxidant is visualized in
Figures 13 and 14.
Conclusions
We must conclude that, for the mechanism of oxidation
of alcohols by substituted diselenides, there are two impor-
tant parameters. First, there is the mode of activation of
the diselenide by the oxidant (Figure 2, Step 1). Certain dis-
elenides such as dimesityl diselenide (4) are easily activated
by TBHP, whereas others are not activated at all. Secondly,
the oxidation potential of the activated substituted diselen-
ides differs greatly. Not all activated diselenides are capable
of dehydrogenating non-activated alcohols (Figure 2,
Steps 2 and 3). Finally, regeneration of the anhydride of the
Figure 13. Oxidation of 1-decanol with different diselenides
(0.25 mmol) according to the oxidant and substrate feed protocol.
active species (Figure 2, Step 4) is also expected to change
when the substituents are varied. The traditional catalytic
oxidation of 1-decanol with aromatic diselenides was ham-
pered by formation of side-products due to overoxidation
and transesterification. However, when diselenides were
first pre-activated, followed by addition of 1-decanol, com-
pounds 4, 7 and 9 were capable of converting this substrate
selectively to the aldehyde. Obviously, the steric and elec-
tronic demands for the hydrogen abstraction are optimal
for these catalysts, whereas other diselenides were not cap-
able of quantitatively converting substrate. The most prom-
ising diselenides were tested in the catalytic oxidation of
1-decanol by using our previously optimized oxidant and
substrate feed protocol. It was shown that this was indeed
a good way of improving catalyst performance.
Figure 14. Capric acid formation during the oxidant and substrate
feed protocol with different diselenides in the oxidation of 1-dec-
anol. Note that the x-axis starts at t = 2 h.
We have shown that changing the electronic properties of
the diphenyl diselenides provides an entry for improving
their catalytic properties. However, more studies are needed
in order to predict these properties and to design more
active dehydrogenating selenium species.
All of the selected diselenides turned out to be good oxi-
dation pre-catalysts by using the oxidant and substrate feed
protocol. One of the main problems was over-oxidation of
the product leading to capric acid, as shown in Figure 12.
This reaction was also found to be selenium-catalyzed as a
blank run with TBHP and 1-decanal did not show capric
acid formation. The formation of intermediate perester 5 is
Experimental Section
General: All reagents and solvents were used as received from their
most likely one of the main problems as this can oxidize respective vendors. 1H (400 Hz) and 13C (100 Hz) NMR spectra
were recorded with a Bruker AC 400 spectrometer by using tet-
ramethylsilane as an external standard. UV spectra were recorded
with a UV/Vis Hewlett Packard 8452A Diode Array Spectropho-
tometer. Gas chromatography analysis was carried out with a Shi-
madzu GC2010 instrument equipped with a non-polar CP Sil-5 CB
50 mϫ0.53 mm colomn by using an FID detector. Mass spectrom-
etry data were collected with a Shimadzu GC–MS-QP2010 Ultra,
equipped with a non-polar CP Sil-5 CB 50 mϫ0.53 mm column.
Elemental analyses were determined with a Carlo Erba Ea 1108
instrument. Melting points were recorded with a Büchi melting-
aldehydes to the corresponding acids. In the presence of the
seleninic acid, perester formation is also possible, and thus
the formation of by-product cannot be suppressed. One of
the main problems in this protocol is most probably the
reformation of the intermediate anhydride, the active spe-
cies in our reaction. It can be reasonably assumed that oxi-
dation of benzeneselenol (the compound formed in Step 3,
Figure 2) to the benzeneseleninic acid occurs rapidly. It is
known that benzeneseleninic acid dehydrates to the anhy-
dride at relatively low temperatures, typically around point apparatus B-545.
4350
www.eurjoc.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2011, 4345–4352

Aromatic Diselenides as Catalysts for Selective Alcohol Oxidation
Spectroscopy: Reaction calorimetry was performed with
a
The colour of the solution changed during this time from light-
brown to bright green. After cooling the mixture to room temp., it
was poured onto a mixture of ice and saturated aqueous NH4Cl.
This was stirred for 10 min, and the water layer was extracted 5
times with Et2O. The combined organic layers were dried with
MgSO4, and the solvent was evaporated in vacuo. The residue was
taken up in EtOH, and KOH (100 mg) was added. A stream of air
was passed over the solution for 45 min after which the mixture
was stirred in an open flask overnight. A white compound had
crystallized, which was filtered off (dimer), and the mother liquor
was subjected to preparative RP18 silica gel chromatography, which
could separate the diselenide from the monoselenide [eluent:
MeCN/H2O (8:2), Rf = 0.32], which yielded a light-yellow crystal-
line compound. M.p. 48 °C. Yield: 879 mg, 2.57 mmol (17%). 1H
NMR (400 MHz, CDCl3, 25 °C): δ = 7.48 (d, J = 6.4 Hz, 2 H),
7.07 (d, J = 7.6 Hz, 2 H), 2.34 (s, 3 H) ppm. 13C NMR (100 MHz,
CDCl3, 25 °C): δ = 138.4 (Cq), 132.7, 130.3, 128.1 (Cq), 21.5 ppm.
MS: m/z = 342. C14H14Se2 (341.94): calcd. C 49.43, H 4.15; found
C 50.1, H 4.22.
Multimax apparatus: Programmable 4 parallel reactor box, reac-
tion volume from 25 to 70 mL with overhead stirring, temperature
range from –25 to 150 °C, reflux cooler and inter-gas purging. Each
reactor can be set individually for temperature and stirring. Tem-
perature control modes: Jacket and reactor contents. Reaction calo-
rimetry was done by adding a known amount of heat to the reac-
tion mixture using a calibration probe, followed by integration of
the signals obtained (150 Ω, 24 V). To a solution of a specific con-
centration of the diselenide at 80 °C was added the appropriate
amount of TBHP in decane after which the temperature difference
between the internal sensor (Tr; temperature of the reaction) and
the external sensor (Tj; temperature of the jacket) was monitored.
UV/Vis: To a heated mixture of 0.1 m Ph2Se2 in toluene (10 mL)
was added at 80 °C an equimolar amount of TBHP in decane. Af-
ter 40 min, an equimolar amount of benzyl alcohol was added to
the mixture. At selected time intervals, aliquots (50 μL) were with-
drawn from the reaction mixture and diluted with toluene
(2.95 mL) at room temp. The UV spectrum was recorded instantly
with a UV/Vis Hewlett Packard 8452A Diode Array Spectropho-
tometer. In situ IR spectroscopy: The IR data were recorded by
using a Bruker Matrix-MF FT-IR instrument in combination with
a fiber-optic diamond ATR probe. A 0.1 m solution of Ph2Se2 in
toluene was heated to 80 °C, and TBHP (2 equiv.) in decane was
added. After 40 min, benzyl alcohol (2 equiv.) was added. In situ
Raman spectroscopy: The Raman data were recorded by using a
Raman RXN1 analyser from Kaiser Optical systems with a stan-
dard immersion probe. A solution of Ph2Se2 in heptane (0.1 m) was
heated to 80 °C, and TBHP (2 equiv.) in decane was added. After
40 min, benzyl alcohol (2 equiv.) was added. NMR analysis of acti-
vation of Ph2Se2 by TBHP: kinetic reactions were performed the
same way as the UV/Vis analysis (vide supra), but the solvent for
the reactions and for the dilutions was deuterated toluene.
Bis(3-methylphenyl) Diselenide (6): To a 1 m solution of m-tolylmag-
nesium bromide (10 mL), Se powder was added portion-wise, and
after complete addition the mixture was stirred for another 3 h dur-
ing which time a viscous slurry formed. The reaction mixture was
poured onto a mixture of ice and saturated aqueous NH4Cl. Dur-
ing this workup, large amounts of red selenium formed. This mix-
ture was stirred for 10 min, and the water layer was extracted 5
times with Et2O. The combined organic layers were dried with
MgSO4, and the solvent was evaporated in vacuo. The residue was
subjected to RP18 silica gel chromatography which could separate
the diselenide from the monoselenide [eluent: MeCN/H2O (8:2), Rf
= 0.35], which yielded a light-yellow crystalline compound. M.p.
53 °C. Yield: 366 mg, 1.07 mmol (21%). 1H NMR (400 MHz,
CDCl3, 25 °C): δ = 7.34 (m, 1 H), 7.25 (m, 3 H), 2.31 (s, 3 H) ppm.
13C NMR (100 MHz, CDCl3, 25 °C): δ = 138.9 (Cq), 133.0, 129.3,
129.0, 128.7, 128.1, 21.3 ppm. MS: m/z = 342. C14H14Se2 (341.94):
calcd. C 49.43, H 4.15; found C 49.7, H 4.28.
Catalytic Oxidations: To a stirred solution of substrate (2 mmol),
1,2-dimethoxybenzene (0.5 mmol, 69 mg, internal standard) and
the selected diselenide (0.1 mmol) in solvent (10 mL) was added at
80 °C TBHP (2.2 mmol). At several intervals, aliquots (50 μL) were
withdrawn, quenched with Na2SO3 [100 mg in EtOAc (1.5 mL)],
and the solids were filtered off. The mixture was subsequently ana-
lyzed by GC.
Bis(2-methoxyphenyl) Diselenide (7): To a 1 m solution in THF of
the Grignard reagent (40 mL), Se powder was added portion-wise
in such a rate that the temperature did not exceed 35 °C in the
reaction; 10 min after the complete addition of the Se, the mixture
started to polymerize, and thus the mixture was hydrolyzed with
saturated aqueous NH4Cl. The mixture was extracted 3 times with
EtOAc, the organic layers were combined, dried with MgSO4, and
subsequently the solids were filtered off. All volatiles were removed
in vacuo, and the compound was purified on SiO2 (EtOAc/hexanes,
95:5). The orange liquid was subjected to kugelrohr distillation,
and two fractions were obtained, one boiling at 170 °C (0.1 mbar)
and the other at 210 °C (0.1 mbar). The diselenide was recrys-
tallized from EtOH, which yielded bright orange crystals. M.p.
79 °C Yield: 3.587 g, 9.59 mmol (48%). 1H NMR (400 MHz,
CDCl3, 25 °C): δ = 7.54 (d, J = 7.6 Hz, 1 H), 7.21 (t, J = 8.0 Hz,
1 H), 6.87 (t, J = 7.2 Hz, 1 H), 6.82 (d, J = 8.0 Hz, 1 H), 3.91 (s,
3 H) ppm. 13C NMR (100 MHz, CDCl3, 25 °C): δ = 157.3 (Cq),
131.0, 128.6, 122.3, 119.0 (Cq), 110.6, 56.4 ppm. MS: m/z = 374.
C14H14O2Se2 (373.93): calcd. C 45.18, H 3.79; found C 45.08, H
3.70.
Stoichiometric Oxidations: To a stirred solution of the diselenide
(0.5 mmol), dodecane (0.25 mmol, internal standard) in (trifluoro-
methyl)benzene (10 mL) at 80 °C was added TBHP (1 mmol from
a 5.5 m solution in decane), and after the appropriate amount of
time 1-decanol (0.5 mmol) was added. At several intervals, aliquots
(50 μL) were withdrawn, quenched with Na2SO3 [100 mg in EtOAc
(1.5 mL)], and the solids were filtered off. The mixture was sub-
sequently analyzed by GC.
Oxidant and Substrate Feed Reactions: To a stirred solution of the
diselenide (0.25 mmol), dodecane (0.25 mmol, internal standard) in
(trifluoromethyl)benzene (5 mL) at 80 °C was added TBHP
(0.5 mmol from a 5.5 m solution in decane), and after the appropri-
ate amount of time 1-decanol (0.25 mmol) was added. After 2 h, 1-
decanol (0.25 mmol) was added. After an additional 1 h, TBHP
(0.25 mmol) was added. These steps were repeated four times. At
several intervals, aliquots (50 μL) were withdrawn, quenched with
Na2SO3 [100 mg in EtOAc (1.5 mL)], and the solids were filtered
off. The mixture was subsequently analyzed by GC.
Bis[2-(dimethylamino)phenyl] Diselenide (8): To the starting com-
pound (5 mmol) in THF (10 mL) was added iPrMgCl·LiCl com-
plex (1.1 equiv.), and after full addition dioxane (1 equiv.) was
added. A precipitate was observed, and conversion of the starting
material was monitored by taking samples and analysing them by
GC–MS. After stirring overnight, conversion was complete, and Se
Syntheses
Bis(4-methylphenyl) Diselenide (5): To a 1 m solution of p-tolylmag-
nesium bromide (30 mL), Se powder was added portion-wise, and
after complete addition the mixture was heated to 40 °C for 30 min.
Eur. J. Org. Chem. 2011, 4345–4352
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4351

I. W. C. E. Arends et al.
FULL PAPER
[4] M. Shimizu, I. Kuwajima, Tetrahedron Lett. 1979, 20, 2801–
2804.
powder was added gradually. The mixture was stirred for another
2 h and then hydrolyzed with aqueous NH4Cl. The mixture was
purified by SiO2 chromatography (heptane) to yield a light-yellow
crystalline compound. M.p. 41 °C Yield: 438 mg, 1.10 mmol (44%).
1H NMR (400 MHz, CDCl3, 25 °C): δ = 7.52 (d, J = 8.1 Hz, 1 H),
7.15 (m, 2 H), 7.00 (t, J = 6.4 Hz, 1 H), 2.80 (s, 6 H) ppm. 13C
NMR (100 MHz, CDCl3, 25 °C): δ = 152.8 (Cq), 129.3, 127.6,
126.3, 120.8, 45.8 ppm. MS: m/z = 199 [M2+]. C16H20N2Se2
(400.00): calcd. C 48.25, H 5.06, N 7.03; found C 48.02, H 5.11, N
7.10.
[5] I. Kuwajima, M. Shimizu, H. Urabe, J. Org. Chem. 1982, 47,
837–842.
[6] T. Onami, M. Ikeda, S. S. Woodard, Bull. Chem. Soc. Jpn.
1996, 69, 3601–3605.
[7] H. Ehara, M. Noguchi, S. Sayama, T. Onami, J. Chem. Soc.
Perkin Trans. 1 2000, 9, 1429–1431.
[8]
J. C. van der Toorn, G. Kemperman, R. A. Sheldon, I. Arends,
J. Org. Chem. 2009, 74, 3085–3089.
[9]
R. D. McCulla, W. S. Jenks, J. Am. Chem. Soc. 2004, 126,
16058–16065.
[10]
G. Socrates, Infrared and Raman Characteristic Group Fre-
quencies, 3rd ed., Wiley, Chichester, West Sussex, 2004.
Acknowledgments
[11] R. A. Gancarz, J. L. Kice, Tetrahedron Lett. 1981, 22, 1661–
These investigations were supported (in part) by the Netherlands
Organisation for Scientific Research (NWO) in the NWO-ACTS
research programme B-Basic. The authors would like to thank
Peter Hilberink with help in using the Multimax equipment for the
reaction calorimetry.
1662.
[12] J. Mlochowski, M. Brzaszcz, M. Giurg, J. Palus, H. Wojtowicz,
Eur. J. Org. Chem. 2003, 4329–4339.
[13] D. H. R. Barton, T. L. Wang, Tetrahedron Lett. 1994, 35, 5149–
5152.
[14] T. M. Klapötke, B. Krumm, K. Polborn, Eur. J. Inorg. Chem.
1999, 1359–1366.
[1] R. A. Sheldon, I. Arends, A. Dijksman, Catal. Today 2000, 57,
157–166.
[2] D. H. R. Barton, A. G. Brewster, A. H. F. Hui, D. J. Lester,
S. V. Ley, T. G. Back, J. Chem. Soc., Chem. Commun. 1978,
952–954.
[15] J. Muller, A. Terfort, Inorg. Chim. Acta 2006, 359, 4821–4827.
[16] A. Krasovskiy, B. F. Straub, P. Knochel, Angew. Chem. 2006,
118, 165; Angew. Chem. Int. Ed. 2006, 45, 159–162.
[17] G. Mugesh, W. W. du Mont, H. Sies, Chem. Rev. 2001, 101,
2125–2179.
[3] M. Giurg, L. Syper, Phosphorus, Sulfur, Silicon Relat. Elem.
2008, 183, 970–985.
Received: April 6, 2011
Published Online: June 10, 2011
4352
www.eurjoc.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2011, 4345–4352