62
X. Xue et al. / Catalysis Communications 33 (2013) 61–65
[
33] Briefly, 0.61 g sample of NaVO
er solution of sodium acetate/acetic acid at pH=4.8. Then, 4.0 g of
-Na H[PW 34]·19H O was added to the above solution and stirred
3
was added to 20 ml (1.0 M) buff-
Table 1
a
Ammoximation of 2-butanone with different catalysts.
△
8
9
O
2
Entry
Catalyst
Yield of MEKO
(mol%)
Selectivity to
MEKO (mol%)
at 25 °C for 48 h. The solution was treated with 3 g of solid KCl and
stirred for 30 min. Methanol (50 ml) was added to produce precipi-
tate. The solid was filtered and dried in air to give 4.3 g of tan-
1
Na
2
WO
4
·2H
2
O+Na
2
HPO
4
·12H
2
O+
21
99
NaVO
3
·2H
2
O
−
1
2
3
4
5
6
7
K
K
K
K
4
5
6
6
PW11VO40·2H
PW10 40·4H
HSiW 40·3H
PW 40·4H
2
O
25
60
66
87
30
74
99
99
99
99
99
99
–
orange powder. IR spectrum (KBr, cm ): 1085, 1055, 962, 876,
88, 597, 509, and 481. 3 P MAS NMR: −13.2 ppm. Calcd for K
PW 40]·4H O: K, 8.42; P, 1.11; V, 5.49; W, 59.41. Found: K, 8.49;
P, 1.06; V, 5.38; W, 59.35.
1
V
2
O
2
O
7
[
6
9
V
3
O
2
O
9 3
V O
2
9
V
3
O
2
O
b
K PW V O ·4H O
6
6
9
9
3
3
40
2
2
c
K
PW
V
O
40·4H
O
8
None
Trace
a
3
. Results and discussion
2 2 3 2
Reaction conditions: 1 mmol substrate; 2 mmol H O (30%); 3 mmol NH ·H O(25%);
1
0 μmol catalyst; 2 ml isopropanol as solvent at 25 °C for 10 h; yields and selectivity were
determined by GC using bromobenzene as internal standard and were calculated based on
substrate, H and NH ·H O were added batchwise for four times every 0.5 h.
O as solvent.
2 ml t-BuOH as solvent.
Although ammoximation reaction stoichiometrically requires equal
moles of substrate, NH ·H O and H , as illustrated in Scheme 1,
oximes may undergo acid catalyzed Beckmann rearrangement to
amides. For aldoximes, they could dehydrate to yield nitriles [6]. Also
considering that ammonia is highly volatile, so excess ammonia is
needed to create a alkaline environment. However, too much ammonia
O
2 2
3
2
3
2
O
2 2
b
2
ml H
2
c
Although conversions of aldehydes were higher than those of
accelerates H
unavoidably be decomposed under the reaction environment. So mod-
erate quantities of NH ·H O and H were needed. After optimization
reaction conditions, the ratio of reactants were as follows, substrate:
NH ·H O:H was 1:3:2.
Table 1 showed the ammoximation of 2-butanone with different cat-
alysts. Na HPO ·12H O and NaVO ·2H O showed only 21%yield of oxime
entry 1). Some keggin-type vanadium substituted polyoxometalates
entries 2–5) were used to catalyze the ammoximation and it was
2
O
2
decomposition. Meanwhile, part of H
2
O
2
would
ketones, oxime selectivity was not so high as ketones owing to pro-
ducing a small amount of amides and nitriles (entries 7–12). The
selectivity to oxime catalyzed by K PW V O40 is generally higher
6 9 3
3
2
O
2 2
when compared with Neumann's catalyst [6]. As Table 2 shows,
the conversions of benzaldehyde, 4-methylbenzaldehyde, as well as
4-chlorobenzaldehyde are all more than 90% and the selectivity to oximes
are over 80%, indicating the peculiar selectivity of this tri-vanadium
substituted polyoxometalate (entries 7–9). The oxime yield of
2-chlorobenzaldehyde is lower than that of 4-chlorobenzaldehyde; this
should be ascribed to the steric hindrance effect, which the ortho-
chlorin group obstructs the attack of hydroxylamine to the carbonyl of al-
dehyde (entries 9–10). For the higher activity of 4-nitrobenzaldehyde
compared to 3-nitrobenzaldehyde(entries 11–12), the steric hindrance
effect should also answer for the difference.
3
2
2 2
O
2
4
2
3
2
(
(
found that oxime yield increased gradually with the increase of vanadium
substituted number (entries 2–5).
2
+
Bhaumik et al. [5] reported that VO loaded resin heterogeneously
catalyzed the ammoximation of cyclohexanone and VO2 was the
+
d 6
catalytic site by characterization. The structure of terminal VO in VO
6
−
2+
octahedra of Keggin anion PW
loaded resin. So the VO portion should be the catalytically active site
for the ammoximation reaction catalyzed by the K PW 40·4H O.
Furthermore, the catalytic activity should increase with the increasing
of vanadium atom number. There are three VO octahedra portion in
40 anion, so tri-vanadium substituted tungstophosphate
40·4H O exhibited the highest activity among all the ex-
amined catalysts.
The 2-butanone ammoximation reaction was also carried out homo-
geneously by K PW 40·4H O in water; however, the result was
9 3
V O
40 is very similar to that of VO
The manners of adding NH
oxime yield because NH ·H O promotes H
they were added using a micro-injection pump or added interruptedly to
enhance H efficiency. We can see that when NH ·H O and H were
·H O and H O
3 2 2 2
also have influence on
d
3
2
2 2
O decomposition. Generally,
6
9 3
V O
2
O
2 2
3
2
2 2
O
6
added once before reaction without batching, oxime yield decreased to
76% (entry 13).
Scheme 2 shows the proposed mechanism of ammoximation of
ketones and aldehydes catalyzed by this polyoxometalate. Firstly, the
reaction of ammonia and hydrogen peroxide leads to the formation of
6
−
9 3
the PW V O
of K PW
6
9 3
V O
2
6
9 3
V O
2
6 9 3 2
hydroxylamine catalyzed by K PW V O40·4H O. Since this vanadium-
disappointing with only 30% yield of oxime (entry 6). The solvent of
tert-butanol afforded moderately high yield of oxime (74%) (entry 7).
In comparison, isopropanol showed the best result (87% yield of
containing polyoxometalate has no characteristic of porosity, so we pre-
sume that the catalytic reaction proceed not only on the surface but also
in the bulk phase, that is, the pseudo-liquid phase behavior [34] of
this keggin-type polyoxometalate. Although the polyoxometalate was
insoluble in the solvent of isopropanol, the small polar molecules of
oxime) on ammoximation of 2-butanone catalyzed by K
40·4H O. In the control experiment, almost no oxime was detected
in the absence of catalyst (entry 8).
6 9 3
PW V
O
2
H
2
O
2
and NH
which interact with the active sites of vanadium atoms to catalyze for-
mation of NH OH. In the pseudoliquid catalysis process, the catalyst
3
could enter into the bulk of the solid polyoxometalate,
Having obtained these results, we extended our catalytic system to
the ammoximation of other ketones and aldehydes. Table 2 presents
the results on ammoximation of ketones and aldehydes based
2
appeared as solid but behaved like liquid, which could also explain
the high catalytic activity of this catalytic system. Then the product of
6 9 3 2 2 2
on K PW V O40·4H O with H O and ammonia (see supporting
information). Most of the substrates were successfully converted to cor-
responding oximes in high yields. Acetone, 2-butanone, 2-pentanone,
cyclohexanone and cycloheptanone, all ketones got satisfying oxime
yields of over 80% except for cyclopentanone which is only 44% (entries
2
NH OH came out to the surface and released to the organic phase of
isopropanol. Then, it interacts with ketones or aldehydes to form the
final product of oximes by nucleophilic addition (Scheme 2). In this pro-
6 9 3 2
cess, the solid catalyst of K PW V O40·4H O may also interact with sub-
1–6). The selectivity to oximes for ketones is all up to 99%.
strates via hydrogen bonds to increase the electrophilicity of them.
6 9 3 2
Scheme 1. Schematic representation of the ammoximation of ketones and aldehydes with K [PW V O40]·4H O.