Cyclohexane Functionalization Catalyzed by Octahedral Molecular Sieve (OMS-1) Materials

Article abstract of DOI:10.1006/jcat.1998.2049

Both the abundance of alkanes and their extremely low activity have greatly interested several researchers. In this paper, different metal substituted 3×3 octahedral molecular sieves (OMS-1) materials were used to catalyze the functionalization of cyclohexane by using tert-butyl hydroperoxide as oxidant and tert-butyl alcohol as solvent at different temperatures (60, 80, and 100°C). [Fe]-OMS-1 at 80°C exhibits the best activity and selectivity. The solvent t-butyl alcohol (the reduced state of t-butyl hydroperoxide) was first introduced to the reaction system which makes the system simple to study. The effects of catalyst amount and ratio of tert-butyl alcohol to cyclohexane were examined. Variable speed stirring (200-800 rpm) experiments suggest that under conditions reported here that diffusion is not a problem. Studies of the liquid phase after separation from the solid OMS-1 catalysts have shown that metal does not leach into the solution and that heterogeneous catalysis occurs. At 80°C, the conversion of cyclohexane or the total yield of products can reach 13.1% in 40 h. The yields of cyclohexanone, cyclohexanol, and cyclohexyl hydroperoxide were 6.57,2.83, and 1.38%, respectively, and t-butyl cyclohexyl perether was 2.36%. The reaction conditions are mild, and the catalysts retain their crystallinity after reaction. Moreover, the catalyst can be easily separated from the reaction mixture and used catalysts retain similar catalytic activity over a 40-h time period.

Full text of DOI:10.1006/jcat.1998.2049

JOURNAL OF CATALYSIS 176, 275–284 (1998)
ARTICLE NO. CA982049
Cyclohexane Functionalization Catalyzed by Octahedral Molecular
Sieve (OMS-1) Materials
,
, ,1
ꢀ
ꢀ
ꢀ
Jin-Yun Wang, Guan-Guang Xia,‡ Yuan-Gen Yin, Steven L. Suib, † ‡ and C. L. O’Young§
ꢀ
Department of Chemistry, †Department of Chemical Engineering, ‡Institute of Materials Science, University of Connecticut, Storrs,
Connecticut 06269-4060; and §Texaco Research Center, Texaco, Inc., P.O. Box 509, Beacon, New York 12508
Received February 18, 1997; revised December 29, 1997; accepted February 23, 1998
fied TiO2 (17, 18) show good oxidation results, but with the
formation of significant amounts of COx (15–18).
Both the abundance of alkanes and their extremely low activity
have greatly interested several researchers. In this paper, different
metal substituted 3 ꢁ 3 octahedral molecular sieves (OMS-1) mate-
rials were used to catalyze the functionalization of cyclohexane by
using tert-butyl hydroperoxide as oxidant and tert-butyl alcohol as
solvent at different temperatures (60, 80, and 100ꢂC). [Fe]-OMS-1
at 80ꢂC exhibits the best activity and selectivity. The solvent t-butyl
alcohol (the reduced state of t-butyl hydroperoxide) was first in-
troduced to the reaction system which makes the system simple to
study. The effects of catalyst amount and ratio of tert-butyl alco-
hol to cyclohexane were examined. Variable speed stirring (200–
800 rpm) experiments suggest that under conditions reported here
that diffusion is not a problem. Studies of the liquid phase after sep-
aration from the solid OMS-1 catalysts have shown that metal does
not leach into the solution and that heterogeneous catalysis occurs.
At 80ꢂC, the conversion of cyclohexane or the total yield of prod-
ucts can reach 13.1% in 40 h. The yields of cyclohexanone, cyclo-
hexanol, and cyclohexyl hydroperoxide were 6.57, 2.83, and 1.38%,
respectively, and t-butyl cyclohexyl perether was 2.36%. The reac-
tion conditions are mild, and the catalysts retain their crystallinity
after reaction. Moreover, the catalyst can be easily separated from
the reaction mixture and used catalysts retain similar catalytic ac-
c
Hydrogen peroxide and tert-butyl hydroperoxide
(TBHP) have been used widely for selective oxidation.
Systems based on functionalization of cyclohexane via ox-
idation with H2O2 catalyzed by organometallic complexes
have attained high conversions and selectivities (10, 19–24).
Fewer reports have been published using heterogeneous
catalysts. Titanium silicate (25, 26), vanadium silicate (27),
chromium silicate (28), and Mn2+-exchanged clay (29), as
well as vanadium-substituted MCM-41 zeolite (30), were
employed as catalysts and all these systems show interest-
ing oxidation reactions. However, the presence of large
amounts of water usually culminates in a two-phase system
with inherent separation problems. Moreover, metal (Ti,
V, and Cr) silicates show good activities for the oxidation
of linear alkanes (25, 31–33), but much lower activities for
the oxidation of cycloalkanes, like cyclohexane (33, 34).
We have used TBHP (70% aqueous solution) as the ox-
idant, with t-butyl alcohol (TBA) as the solvent to avoid
phase separation problems. TBA is also the decomposition
product of TBHP, so that complete miscibility of the reac-
tion mixture can be ensured. Several papers were published
(1, 28, 30, 33–37) with TBHP as the oxidant, but mostly us-
ing complicated organometallic complexes as catalysts. In
pursuing our interest in the catalytic behavior of manganese
oxides, we have used manganese oxide catalysts that have
the todorokite structure (hereafter designated as OMS-1
(38, 39); see Fig. 1).
tivity over a 40-h time period.
ꢃ 1998 Academic Press
1. INTRODUCTION
The abundance of saturated hydrocarbons has led to
the interest of several researchers since the early 1980s
to try to activate these materials (1–3). However, their
very low reactivity poses a severe challenge toward ex-
ploration and potential commercial use. A lot of work in-
volving organometallic complexes that imitate the activ-
ity of enzymes such as cytochrome p-450 monooxygenase
has been done (4–9). Usually dioxygen is unable to oxidize
these inert hydrocarbons even in the presence of active
organometallic complex catalysts (10–14). Photochemical
reactions with UV radiation catalyzed by TiO2/TS-1 (15),
by metal–porphyrin complexes (16) or by porphyrin modi-
Octahedral molecular sieves used in this research are mi-
croporous tunnel-structure materials. They are mixed va-
lent systems that are good semiconductors. These synthetic
materials have been modeled after natural manganese nod-
ules. The synthetic versions are much purer than natural
materials and can be systematically doped with various
transition metals in the framework of MnO6 octahedra. Ex-
changeable cations in the tunnels of these materials can be
replaced via ion-exchange methods.
OMS-1 possesses a 3 ꢁ 3 tunnel structure with a tunnel
1 To whom correspondence should be addressed.
˚
opening of about 6.9 A. The basic building blocks of this
275
0021-9517/98 $25.00
c
Copyright ꢃ 1998 by Academic Press
All rights of reproduction in any form reserved.

276
WANG ET AL.
Otherwise the composition of the mixture will be speci-
fied. Acetophenone was added as an internal standard. The
reaction products were qualitatively analyzed by GC-MS
and GC-IR and quantitatively analyzed by GC (HP 5890
series II) with a Supelcowax-10 column.
All the catalysts were sieved to a size of 50 mesh or
˚
smaller with crystallite sizes of about 200 A as determined
by transmission electron microscopy.
III. RESULTS
2+
FIG. 1. Structure of OMS-1, synthetic todorokite, Mg
0.98ꢄ1.35
A. Effects of Reaction Temperatures for Different Catalysts
2+
4+
Mn
Mn
O12 ꢅ 4.47 ꢄ 4.55 H2O.
1.89ꢄ1.94
4.38ꢄ4.54
The different substituted OMS-1 and commercial MnO2
(Chemetals Division, Diamond Shamrock Chemicals) were
tested at 60, 80, and 100ꢂC, respectively. Chemical analyses
of the OMS-1 materials show that their compositions are:
material are MnO6 octahedra that are interconnected by
edges and corners. When doped with a lower valent transi-
tion metal, these transition metal cations (M2+/3+) can re-
place Mn cations in the framework to form MO6 units or
ions that can occupy tunnel positions. Catalytic studies of
OMS-1and metaldoped OMS-1materialsin the functional-
ization of cyclohexane with TBHP as oxidant are presented
here.
Mg1.10Mn6.41O12ꢅ4.31H2O (OMS-1);Mg1.12Fe0.20Mn6.21O12
ꢅ
4.21H2O ([Fe]-OMS-1); Mg1.21Co0.24Mn5.98O12 ꢅ 3.58H2O
([Co]-OMS-1); Mg1.08Cu0.21Mn6.18O12 ꢅ 3.84H2O ([Cu]-OM
S-1); and Mg1.15Cr0.19Mn6.25O12 ꢅ 3.66H2O ([Cr]-OMS-1).
Table 1 shows results of reaction at 60ꢂC. OMS-1 mate-
rials show significant catalytic activity as compared to com-
mercial MnO2. Metal substituted OMS-1 materials have
better catalytic activity than OMS-1 itself and [Fe]-OMS-1
has the best activity among the catalysts tested (see Fig. 2).
However, cyclohexyl hydroperoxide (CHP) is the main
product for allthe catalystsat 60ꢂC. The conversion ofcyclo-
hexane and the total yield of all products (no other prod-
ucts were detected) decrease in the order [Fe]-OMS-1 >
[Co]-OMS-1 > [Cu]-OMS-1 > [Cr]-OMS-1 > OMS-1 ꢆ
MnO2.
II. EXPERIMENTS
A. Synthesis of Doped OMS-1 (38, 39)
The preparation of catalysts was done in the follow-
ing way: a 60 mL volume of a 5.0 M NaOH solution was
added dropwise to a vigorously stirred 40 mL mixture of
0.5 M MnCl2, 0.1 M MgCl2 to form a suspension in a plas-
tic flask at room temperature. After that, 40 mL of 0.2 M
KMnO4 solution was added dropwise to a vigorously stirred
suspension, and stirring was stopped upon completion of
this addition. The plastic flask was sealed and aged at
35ꢂC for 2 days before filtration and the solid powder was
washed with distilled deionized water (DDW). The filtered
residue was ion-exchanged by using 200 mL of 1.0 M MgCl2
overnight at room temperature, and the exchanged material
was filtered and washed with DDW. The mixture was trans-
ferred to an autoclave at 150ꢂC for 2 days. The final product
was filtered and washed, and then dried at 110ꢂC overnight.
For the synthesis of framework substituted OMS-1 mate-
rials, the only difference is that 8 mL of 0.1 M transition
metal salt was added to the mixture of MnCl2 and MgCl2
(40).
The data of Table 2 suggest that when the reaction tem-
perature is raised to 80ꢂC that the product distribution
changes significantly. The conversion of cyclohexane at
80ꢂC increases considerably as compared to that at 60ꢂC
for all catalysts. The yields of cyclohexanol (A) and cyclo-
hexanone (K) increase greatly at the expense of CHP. The
yield of t-butyl cyclohexyl perether (P) is more than twice
that at 60ꢂC. In addition, the activity order of the catalysts
at 80ꢂC is different from that at 60ꢂC. [Cu]- and [Cr]-OMS-1
TABLE 1
Cyclohexane Oxidation at 60ꢂC for 40 h
Type of catalysts
% conversion
% A
% K
% CHP
% P
B. Catalytic Reactions
[Fe]-OMS-1
[Co]-OMS-1
[Cu]-OMS-1
[Cr]-OMS-1
6.79
4.60
4.59
4.42
3.98
0.11
1.78
0.99
0.88
0.98
0.84
0
2.45
1.13
1.15
1.13
0.90
0
1.81
2.12
2.02
1.99
1.95
0.11
0.75
0.36
0.54
0.32
0.29
0
Stirred reaction mixtures in sealed glass vials were heated
at reaction temperatures of 80ꢂC or lower. The reactions at
100ꢂC were done in stirred autoclaves which had Teflon lin- OMS-1
MnO2 commercial
ers. All the reaction vessels were placed in a temperature-
controlled oil bath. The reaction mixture is generally com-
posed of: OMS-1, 67 mg; cyclohexane, 1.8 g; TPHP, 2.0 g
Note. A = cyclohexanol, K = cyclohexanone, CHP = cyclohexyl hy-
droperoxide, P = t-butyl cyclohexyl perether, catalyst: 67 mg, cyclohexane:
(70% aqueous solution); t-butyl alcohol, 4.2 g as solvent. 1.8 g, TBHP (70% aqueous solution): 2.0 g, t-butyl alcohol: 4.2 g.

CYCLOHEXANE FUNCTIONALIZATION
277
TABLE 3
Cyclohexane Oxidation at 100ꢂC for 40 h
Type of catalysts
% conversion
% A
% K
% CHP
% P
[Fe]-OMS-1
[Co]-OMS-1
[Cu]-OMS-1
[Cr]-OMS-1
OMS-1
15.4
14.9
15.1
15.4
15.0
4.61
4.21
4.33
4.10
4.00
7.49
6.83
6.85
7.01
6.92
0
0
0
0
0
3.32
3.83
4.84
4.28
4.12
Note. A = cyclohexanol, K = cyclohexanone, CHP = cyclohexyl hy-
droperoxide, P = t-butyl cyclohexyl perether; catalyst: 67 mg, cyclohex-
ane: 1.8 g, TBHP (70% aqueous solution): 2.0 g, t-butyl alcohol: 4.2 g.
B. Effects of Stirring Speed
The effects of stirring speed were conducted to exam-
ine mass transfer effects (41). Stirring speeds from 200 to
800 rpm, as well as no stirring, have been used and the re-
sults are listed in Table 4.
FIG. 2. The XRD patterns of [Fe]-OMS-1: (a) fresh; (b) used catalyst
after one run, 40 h; (c) used catalyst after two runs; total time = 80 h.
Table 4 shows that cyclohexane conversions and yields of
different products are different for stirred and nonstirred
reactions. There is a big difference in terms of initial re-
action rates for stirred and nonstirred experiments, but
there is only a small difference for reaction rates with stir-
ring speeds from 200 to 800 rpm. These data suggest that
mass transfer effects can be minimized if a stirring speed of
200 rpm or higher is applied. A stirring speed of 400 rpm was
generally used to test for effects of the amount of TBHP,
the amount of solvent, and other factors in these reactions.
showed higher activity than [Co]-OMS-1 at 80ꢂC. Commer-
cial MnO2 shows the least activity with only P formed at 60
and 80ꢂC.
Two things are very obvious for reactions at 100ꢂC (see
Table 3): first, no CHP exists in the reaction products after
reaction for 40 h; second, all the catalysts (both substituted
and unsubstituted OMS-1 materials) have similar activity.
Generally, ketone and t-butyl cyclohexyl perether occur in
higher yields than at lower temperatures, and the ratios of
K/A increase along with an increase in reaction tempera-
ture. One exception is [Fe]-OMS-1, where the ratio of K/A
is smaller at 100ꢂC than that at 80ꢂC. At 60–100ꢂC, cyclo-
hexane conversion and yields of products increase with the
increase of reaction temperature, and no deep oxidation
products have been detected. However, the increase of con-
version or total yields as the reaction temperature increases
from 80 to 100ꢂC is rather insignificant as compared to the
increase from 60 to 80ꢂC. Therefore, a reaction temperature
of 80ꢂC was chosen for further studies.
C. Effects of Amount of Tert-Butyl Hydrogen Peroxide
The original amount of TBHP was 2.0 g (70% aqueous
solution) in the reaction mixture. In order to study the effect
of the amount of TBHP, all factors and components other
than TBHP were kept the same. The amounts of TBHP
chosen are 1.0, 2.0, 3.0, and 4.0 g, respectively.
Table 5 shows that cyclohexane conversion is affected
by the amount of TBHP. It is clear that yield of every sin-
gle product increases as the amount of TBHP increases.
The efficiency of TBHP decreases as the amount of TBHP
increases. An amount of 2.0 g of TBHP was used in the re-
action mixture for all of the experiments described below
since the cyclohexane conversion showed a large increase
when TBHP was changed from 1.0 to 2.0 g.
TABLE 2
Cyclohexane Oxidation at 80ꢂC for 40 h
Type of catalysts
% conversion
% A
% K
% CHP
% P
D. Effects of Catalyst Amount
[Fe]-OMS-1
[Co]-OMS-1
[Cu]-OMS-1
[Cr]-OMS-1
OMS-1
11.6
9.44
10.3
10.2
9.4
3.45
2.95
3.41
3.18
3.03
0
5.54
4.17
4.59
4.61
4.19
0
0.56
1.13
1.01
1.11
1.08
1.01
2.11
1.09
1.05
1.25
1.08
0
Catalyst amounts were changed from 0 to 84 mg for re-
actions carried out at 80ꢂC. All the results are shown in
Table 6. The data of Table 6 demonstrate that the presence
of very small amounts of catalyst (4 mg) lead to large in-
creases in cyclohexane conversion. However, cyclohexane
conversions and yields of cyclohexanone become flat when
17 mg catalyst are used, although there is a general trend
of higher conversion and higher total yields with higher
MnO2
1.01
Note. A = cyclohexanol, K = cyclohexanone, CHP = cyclohexyl hy-
droperoxide, P = t-butyl cyclohexyl perether; catalyst: 67 mg, cyclohex-
ane: 1.8 g, TBHP (70% aqueous solution): 2.0 g, t-butyl alcohol: 4.2 g.

278
WANG ET AL.
TABLE 4
Effects of Stirring Speed for Cyclohexane Oxidation at 80ꢂC for 40 h
Rate ꢁ 104 a
% selectivity
of TBHP
Stirring speed
% conversion
(mol ꢅ gꢄ1 ꢅ sꢄ1
)
% A
% K
% CHP
% P
0
9.45
11.3
11.6
11.7
11.5
0.59
1.93
1.99
2.12
2.03
2.38
3.31
3.45
3.34
3.36
4.67
5.47
5.54
5.62
5.51
0.39
0.58
0.56
0.59
0.54
2.01
2.17
2.11
2.15
2.09
30.7
32.9
33.5
34.1
33.2
200 rpm
400 rpm
600 rpm
800 rpm
Note. A = cyclohexanol, K = cyclohexanone, CHP = cyclohexyl hydroperoxide, P = t-butyl cyclohexyl perether; catalyst: 67 mg, cyclohexane: 1.8 g,
TBHP (70% aqueous solution): 2.0 g, t-butyl alcohol: 4.2 g.
a Initial rates of reaction.
catalyst amounts. The yield of cyclohexanol rises to a maxi- ratio of TBA to cyclohexane was used in all experiments
mum at 67 mg catalyst, and decreases with further increase discussed below.
of catalyst amount. On the other hand, the yields of cyclo-
hexanoland P increase continuallywith the increase ofcata-
lyst amount. The continuous and rather drastic decrease in
CHP formation with the increase of catalyst amount should
be noted. CHP disappears when 84 mg catalyst was used.
The normalized initial reaction rates which were ob-
tained from differentiated data of conversions with respect
to time are similar, even though the conversions and yields
of different products are not similar.
F. Effects of Reaction Time
Figure 3 shows that the conversion of cyclohexane at
60ꢂC changes over a 25-h period when 67 mg [Fe]-OMS-
1 is used. It appears that the conversion of cyclohexane and
total yields of products increase continuously as time in-
creases. Figure 4 shows the change of product distribution
with time at 60ꢂC within a 25-h reaction period. The yield of
each product increases with reaction time as shown by the
changes in slopes of curves of products in Fig. 4. The forma-
tion of CHP increases steeply during an initial 6-h period.
CHP formation then slows down with further increase of
reaction time and seems to level off near a reaction time of
25 h. The formation of alcohol also increases steeply during
the initial 6 h, but thereafter it slows down very quickly. At
the same time, the concentration of TBHP in the reaction
mixture at the end of 25 h of reaction is about 63% of its
original value.
E. Effects of Solvent Amount
The weight ratio of TBA to cyclohexane was 2.50 in all
previously mentioned experiments and this is the ratio that
can ensure complete miscibility of reactants at room tem-
perature. This ratio was varied to study solvent effects over
a certain range (1.40–2.50). Table 5 shows that the conver-
sion changes as the amount of solvent changes. The ratio of
t-butyl alcohol to cyclohexane of 1.8 is the best for the con-
version of cyclohexane and for the yields of total products.
As the ratio of tert-butyl alcohol to cyclohexane increases,
the yield of CHP goes down, and at the same time, the yield
of cyclohexanol and P go up. The conversion of cyclohex-
ane and total yield of oxidation products reach a maximum
when the weight ratio of TBA to cyclohexane is 1.8. This
Figure 5 shows data for the reaction at 80ꢂC. The conver-
sion of cyclohexane increases dramatically during the first
TABLE 5
Effects of Amount of TBHP on Cyclohexane Oxidation
at 80ꢂC for 40 h
TBHP
%
% selectivity
of TBHP
(70% ) (g) conversion % A % K % CHP % P
1.0
2.0
3.0
4.0
9.37
11.6
13.0
13.8
3.63
3.45
3.71
3.74
4.32
5.54
6.34
6.97
0.21
0.56
0.70
0.75
1.21
2.11
2.25
2.34
40.6
33.5
31.4
29.6
Note. A = cyclohexanol, K = cyclohexanone, CHP = cyclohexyl hy-
droperoxide, P = t-butyl cyclohexyl perether; catalyst: 67 mg; TBA: 4.2 g;
cyclohexane: 1.8 g; reaction time: 40 h.
FIG. 3. Conversion of cyclohexane as a function of time at 60ꢂC with
[Fe]-OMS-1 catalyst: ᭹, conversion of cyclohexane.

CYCLOHEXANE FUNCTIONALIZATION
279
TABLE 6
Effects of Catalyst Amount for Cyclohexane Oxidation at 80ꢂC
Catalyst amount
(mg)
Rate ꢁ 104 a
% selectivity
of TBHP
(mol ꢅ gꢄ1 ꢅ sꢄ1
)
% conversion
% A
% K
% CHP
% P
0
4
8
12
17
49
67
84
0
0
0
0
0
0
—
2.23
2.15
1.93
2.32
2.12
1.99
1.86
7.41
8.90
9.52
10.2
11.3
11.6
12.2
1.04
1.82
2.11
2.58
3.40
3.45
3.34
1.25
2.79
3.64
4.24
5.31
5.54
6.20
4.16
2.97
2.13
1.78
0.90
0.56
0
0.96
1.32
1.63
1.59
1.66
2.10
2.64
51.4
44.7
40.2
37.3
35.7
33.1
29.5
Note. A = cyclohexanol, K = cyclohexanone, CHP = cyclohexyl hydroperoxide, P = t-butyl cyclohexyl perether, cyclohexane: 1.8 g; TBHP (70%
aqueous solution): 2.0 g; TBA: 4.2 g; reaction time: 40 h.
a Normalized initial rate (mole ꢅ catalystꢄ1 ꢅ timeꢄ1).
10 h; thereafter it levels off quickly to 13.1% conversion G. The Reusability of Catalysts and Heterogeneity
at a reaction time of 40 h. This change in conversion ver-
sus time is clearly quite different from data for the reaction
at 60ꢂC (see Fig. 4). Figure 6 shows the change in product
distribution for reactions at 80ꢂC. The initial rates of for-
mation of each product are all faster than at 60ꢂC. For the
formation of alcohol, the data at 80ꢂC are similar to those
at 60ꢂC. The yield of alcohol keeps growing during the first
few hours and it does not increase much after that, and even
drops slightly after 25 h. The trend of formation of P at 80ꢂC
is not very different from at 60ꢂC. The rate of formation of
ketone at 80ꢂC shows a continuous decline in slope with
reaction time, and levels off to 7% yield at 25 h. The con-
version of cyclohexane is 12.7% at 80ꢂC as compared to
6.9% at 60ꢂC. Considerable amounts of TBHP (35.1% ) re-
mained after 25 h of reaction. The rate of formation of CHP
at 80ꢂC shows an obvious maximum at a reaction time of
9 h.
of Reaction
The used [Fe]-OMS-1 materials were filtered, washed
with water, dried at 110ꢂC overnight in air, and then used
as catalysts. A weight ratio of t-butyl alcohol to cyclohex-
ane of 1.8 was chosen. All the other factors and reaction
conditions were the same as described in Section III.F, ex-
cept for time. The results for used catalysts are shown in the
bottom line of Table 6. The conversion of cyclohexane or
total yields of product is 12.8% for the used catalyst. The
yields of cyclohexanol and cyclohexanone of the used cata-
lyst were slightly higher than that of a new catalyst, but the
yields of CHP and P were slightly lower than that of new
catalysts. The powder X-ray diffraction (XRD) data were
collected for the catalysts (fresh, after one run, after two
runs), and all XRD patterns are shown in Fig. 2. It is clear
that crystallinity was retained after two runs and that the
intensities did not show any obvious change. Crystallinity
FIG. 4. Yields of products as a function of time at 60ꢂC with [Fe]-
OMS-1 as catalyst: ᭿, cyclohexanone; ✚, cyclohexanol; , t-butyl cyclo-
hexyl perether; ᭡, cyclohexyl hydroperoxide.
FIG. 5. Conversion of cyclohexane catalyzed with [Fe]-OMS-1 versus
time at 80ꢂC: ᭹, conversion of cyclohexane.

280
WANG ET AL.
increased. Data of Figs. 4 and 6 also suggest that P is stable
as evidenced by the changing yields with reaction time.
The formation of CHP has not yet been reported in liquid
phase reaction systems. The selectivity of CHP decreases
as the reaction temperature increases from 60 to 80ꢂC and
becomes 0 at 100ꢂC, suggesting that CHP is an unstable
intermediate. The fact that CHP has a maximum yield at
9 h, as shown in Fig. 6, clearly indicates that CHP is an
unstable intermediate and that it transforms to other stable
compounds such as ketones.
The efficiency of TBHP decreases as reaction tempera-
ture is increased due to the decomposition of peroxide. The
decomposition of peroxide is faster at higher temperature.
B. Effects of Stirring Speed
FIG. 6. Yields of products catalyzed with [Fe]-OMS-1 versus time at
Even though there are differences in terms of conver-
sions and total yields of different products for stirred and
nonstirred systems, the largest difference is the initial rate
of reaction. The initial reaction rate for the reaction with-
out stirring is less than one-third that with stirring. This indi-
cates that diffusion limitations exist for the reaction without
stirring. However, the initial rates for the reaction with stir-
ring speeds of 200 to 800 rpm are more than 3 times greater
than the rate for reaction without stirring and only small dif-
ferencesbetween the reactionswith stirringratesof200rpm
or higher were observed. No obvious diffusion problem ex-
ists if the stirring speed is greater than 200 rpm as shown
by the small differences in conversions, yields of different
products, and selectivities of TBHP. Micropore diffusion is
unlikely a problem because the particle sizes are small. To
ensure that diffusion problems were eliminated, a stirring
rate of about 400 rpm was selected.
80ꢂC: ᭿, cyclohexanone; ✚, cyclohexanol; , t-butyl cyclohexyl perether;
᭡, cyclohexyl hydroperoxide.
was retained for all metal doped catalysts after reaction as
shown by XRD data.
To study whether the catalytic reaction is heterogeneous
or homogeneous, solutions of reaction mixtures after the
solid was filtered were further mixed with fresh reactants
(cyclohexane and TBHP) and solvent (TBA). No cata-
lyst was added. The mixture was split into two parts. One
part was stored at room temperature without stirring, and
another was stirred and heated at 80ꢂC for 40 h. Both mix-
tures were analyzed by GC methods and no reactivity was
observed for either part. The filtered solutions were also
analyzed by inductively coupled plasma (ICP) analyses. No
transition metal ions were detected in solution.
C. Effects of Amount of Tert-Butyl Hydrogen Peroxide
IV. DISCUSSION
The conversion of cyclohexane increases as the amount
of TBHP used in the reaction increases as shown in
Table 5. More TBHP reactant drives the reaction and fa-
vors the formation of products. The amounts of P and CHP
also increase as TBHP increases. The yields of cyclohex-
anol and cyclohexanone are similar when 1.0 g of TBHP is
added to the reaction system. The ratio of ketone to alcohol
increases when more TBHP is present in the reaction mix-
ture. More TBHP causes more deep oxidation and favors
the production of ketone and minimizes alcohol formation.
A. Effects of Temperature
Data of Table 1 show that the conversion of cyclohexane
and the total yields of oxidation products increase as the
reaction temperature is increased. The increases in conver-
sion and yield from 80 to 100ꢂC are smaller than that from
60 to 80ꢂC. This may be because there is a smaller activation
energy at higher temperature than that at lower tempera-
ture. The ratio of K/A goes up when temperature increases
from 60 to 100ꢂC, and this implies that A is continually
oxidized at a higher temperature.
D. Effects of Catalyst Amount
These observations are different from the data reported
The results of the effect of catalyst amount shown in
by Menage et al. (36) and Leising et al. (1). They used Table 4 demonstrate clearly that [Fe]-OMS-1 is very ac-
TBHP to oxidize cyclohexane and found that the K/A ra- tive in this reaction system, and that even tiny amounts
tio reached almost one when (mu-oxo) diferric complexes of catalyst (4 mg) can lead to significant conversion. The
were used as catalysts. They suggested a mechanism where continuous decrease of CHP formation and formation of
A, K, and P all form concurrently and are all stable. What more stable products like K and P with increasing catalyst
they suggested for P seems to be true in our reaction sys- amount show similar trends when reaction temperature is
tem because the yield of P increases as the temperature is increased. Even though the normalized initial reaction rates

CYCLOHEXANE FUNCTIONALIZATION
281
(disappearance of cyclohexane) are very similar for catalyst
amounts of 4 to 84 mg (initial rates are proportional to the
catalyst amount), the conversions are not similar. The rate
of reaction decreased much faster with a large amount of
catalyst than with a small amount of catalyst. Analysis of the
reaction mixture at various reaction times and at different
temperatures is also needed in order to better understand
the reaction (see Section III.F). In addition, [Fe]-OMS-1
catalysts may lead to a better understanding of the role of
Fe in the natural enzyme cytochrome p-450 which catalyzes
the monoxygenation of methane (2).
E. Effects of Solvent Amount
The t-butyl alcohol solvent plays a very important role in
terms of cyclohexane conversion and product distribution
as shown in Table 5. One interesting trend is that as the ratio
of tert-butyl alcohol to cyclohexane increases, the yield of
CHP decreases. At the same time, the yieldsofcyclohexanol
and P increase. These trends may be due to the reaction
equilibrium being favorable for the formation of P and to
the decrease of CHP when the amount of t-butyl alcohol
increases.
The yield of cyclohexanol increases as the amount of sol-
vent increases, perhaps because as more t-butyl alcohol was
added the physical properties (polarity, etc.) of the reaction
mixture are more favorable for the formation of cyclohex-
anol. The decrease in conversion of cyclohexane on increas-
ing the solvent to cyclohexane ratio from 1.8 to 2.5 implies
that a lowering of reactant concentration is detrimental for
the oxidation reaction. The reason that the conversion of
cyclohexane also decreases when the solvent to cyclohex-
ane ratio changes from 1.8 to 1.4 is that less solvent results
in incomplete dissolution and makes the reaction more dif-
ficult to proceed.
FIG. 7. Reaction rates versus time at different temperatures: a. 80ꢂC;
b. 60ꢂC.
It seems obvious from Fig. 6 that the increase of K is
at the expense of CHP after 9 h. In other words, CHP is
gradually converted to K via dehydration. In a similar way,
A is slowly changed to P by intermolecular dehydration of
A and TBHP.
Conversion data of Figs. 3 and 5 were fit with nonlinear
regression analysis to obtain an equation which was differ-
entiated with respect to time. These differentiated data are
shown in Fig. 7 which is a plot of rate versus time for re-
actions done at 60 and 80ꢂC. The initial rate of reaction at
80ꢂC is about 4 times the rate at 60ꢂC; however, the total
conversion and total yields of different products at 80ꢂC are
less than 2 times that at 60ꢂC. These data are explained by
the faster rate of decomposition of TBHP at the higher tem-
perature. Further evidence for the decomposition of TBHP
is that the efficiency of peroxide was much smaller at 80ꢂC
compared to that at 60ꢂC during the reaction period due to
faster decomposition of TBHP at higher temperature.
The reason why the initial rates (at t = 0 min) are propor-
tional to catalyst amount, whereas total conversion over
the total period of reaction is not is explained as follows.
The catalyst may serve as a radical initiator. Large amounts
of catalyst should produce more radicals with a similar re-
action mixture. A higher percentage of radicals may com-
bine with each other to terminate the radical chain process.
More radicals may also decompose peroxide than at lower
radical concentration when hydrocarbon concentrations
are the same. In other words, the efficiency of radicals to-
wards oxidation may be lower at higher concentrations of
radicals than at lower concentrations.
F. Effects of Reaction Time
Figure 3 shows that the conversion of cyclohexane in-
creases continuously up to 6.9% after 25 h of reaction. At
the same time, 63.0% of the original amount of TBHP is still
present. This indicates that the reaction is far from comple-
tion. A comparison of Figs. 4 and 6 shows that the trends
in formation of K and P (increasing amounts) are similar at
60 and 80ꢂC. However, the changing trends for A and CHP
are quite different at these two temperatures. A eventually
levels off at 80ꢂC, whereas CHP goes through a maximum at
80ꢂC. At 80ꢂC, cyclohexanol levels off after 9 h and starts to
decrease after 30 h. All these data suggest that the alcohol
is continually oxidized.
The rate of formation of CHP versus time is very inter-
esting. This is a clear indication that CHP acts as an inter-
mediate in the reaction. As the reaction proceeds, ketone
is the major products with P being formed to a much lower
G. Heterogeneity of the Reaction and the Reusability
of Catalysts
The liquid phase after being removed from the solid cata-
extent. As the formation of alcohol stops, further reaction lyst was mixed with fresh reactants and tested under simi-
of CHP may not yield alcohol. lar catalytic conditions. No reaction was observed for this

282
WANG ET AL.
ane. Perhaps this is the reason why perether and cyclohex-
TABLE 7
anol do not function as intermediates since the ratio of al-
cohol/ketone reached almost one when (mu-oxo) diferric
complexes were used as catalysts (36).
Effects of Solvent Amount on Cyclohexane Oxidation
at 80ꢂC for 40 h
%
% selectivity
of TBHP
Tateiwa et al. (29) also studied the oxygenation of cyclo-
hexane via TBHP. The catalyst employed in their study is re-
ally a heterogeneous solid. They used Mn2+ ion-exchanged
clay as catalysts and obtained a 7.3% yield. They added
an excess amount of TBHP (molar ratio of TBHP to cy-
clohexane of 4.5) and no t-butyl cyclohexyl perether or
other products except cyclohexanol and cyclohexanone
were reported. They tested commercial MnO2, and no ob-
vious catalytic activity was found as has also been reported
here.
a
TBA/C6H12 conversion % A % K % CHP % P
1.40
1.60
1.80
2.00
2.20
2.50
1.80b
9.18
10.4
1.72 4.28
2.20 4.53
2.83 6.57
2.92 6.84
3.07 6.14
3.45 5.54
3.03 6.86
1.66
1.66
1.38
0.94
0.93
0.56
0.79
1.52
2.02
2.36
2.31
2.52
2.10
2.10
32.5
33.1
33.5
32.9
34.1
33.8
13.1
13.0
12.7
11.6
12.8
Note. A = cyclohexanol, K = cyclohexanone, CHP = cyclohexyl hy-
droperoxide, P = t-butyl cyclohexyl perether, catalyst: 67 mg; cyclohex-
ane: 1.8 g; TBHP (70% aqueous solution): 2.0 g; reaction time: 40 h.
a Weight ratio of tert-butyl alcohol vs cyclohexane.
H. Overall Reaction Pathways
b Reaction catalyzed by used catalyst.
The overall reaction pathway is believed to initially in-
volve two parallel reactions of the peroxide reagent as
shown in Scheme 1. One reaction (1a) involves the de-
composition of the peroxide into alcohol (solvent) and O2.
Evidence for direct decomposition of the peroxide comes
from the efficiency of TBHP for oxidation of cyclohexane.
supernant liquid phase. This is strong evidence that the re-
action is a heterogeneous reaction.
One of the advantages of this system is the reusability of
the catalysts. Similar activities were observed for the spent
catalysts as shown in Table 7. The X-ray diffraction patterns
of Fig. 2 show that the structure was retained after two cata-
lytic cycles. It is clear that the structure of the catalysts is
not changing. The heterogeneous reaction scheme reported
here has an advantage that the catalyst can be easily sepa-
rated and reused over homogeneous systems catalyzed by
organometallic complexes. The heterogeneity of the reac-
tion, the stability, and the reusability of the catalysts make
this system a strong candidate for environmentally friendly
practical applications.
Bressen et al. (37) studied the same reaction of cyclohex-
ane oxidation by TBHP, with either phosphotungstic com-
plexes of Ru(II) or Ru salts of tetrafluoroborate as cata-
lysts in dimethylsulfoxide. The amount of TBHP used was
a large excess as compared to substrate, and the molar ra-
tio of TBHP to cyclohexane was as large as 10. These Ru
catalysts are very active in the decomposition of TBHP.
Although OMS-1 materials also exhibit some activity for
TBHP decomposition, they seem to have a much lower rate
than the Ru systems.
Menage et al. (36) reported a study of the oxidation of cy-
clohexane via TBHP, but with (mu-oxo) diferric complexes
of bipyridine as catalysts. Since the perether was very stable
and reacted only very slowly to give ketone, they believed
that the perether was not a precursor to alcohol or ketone.
Our system is similar to theirs (36) in this regard. However,
they suggest that cyclohexanol was not a precursor to ke-
tone which is different from our case because our yield of
alcohol decreases at long reaction times (see Fig. 6).
More interestingly, the (mu-oxo) diferric catalyst systems
can decompose TBHP to dioxygen gas very quickly, which
is a reaction that competes with the oxidation of cyclohex-
SCHEME 1

CYCLOHEXANE FUNCTIONALIZATION
VI. ACKNOWLEDGMENTS
283
For example, based on total selectivities for the oxidation
products, only 30 to 50% of the consumed TBHP is needed.
These data suggest that there is a dual pathway involved
where decomposition of TBHP is not useful for oxidation of
cyclohexane. As more catalyst is added, there is a tendency
to favor the decomposition reaction pathway. In a similar
way, efficiency of TBHP decreases from around 40% to less
than 31% as the reaction temperature increases from 60 to
100ꢂC for all the doped catalysts.
We thank the Department of Energy, Office of Basic Energy Science,
Division of Chemical Science, and Texaco, Inc. for support of this research.
We thank the editor, Professor W. N. Delgass, for several helpful sugges-
tions regarding experiments that were carried out and interpretation of
the data.
REFERENCES
The other initial pathway may involve radicals such as
formation of hydroxyl and RO (reaction 1b) by the inter-
action of TBHP with catalyst. The reaction is believed to
occur on the basis of product selectivities and the nature of
peroxide.
The initial activation of cyclohexane is proposed to oc-
cur in two separate pathways as shown in reaction 2a of
Scheme 1 which involves formation of a cyclohexyl radical
and H2O. The other activation as shown in reaction 2b of
the Scheme 1 involves formation of cyclohexyl radical and
alcohol.
Product formation may be governed by the combination
of radicals as shown in reaction 3a and 3b of Scheme 1.
The two products (A and CHP) may serve as intermediates
to form other products such as those shown in reaction 3c
(K) and 3d (P). The data of Fig. 6 show a disappearance
of CHP with concomitant formation of K. These data sug-
1. Leising, R. A., Kim, J., Perez, M. A., and Que, L., J. Am. Chem. Soc.
115, 9524 (1993).
2. Hill, C. L., “Activation and Functionalization of Alkanes” (C. L. Hill,
Ed.), p. 243. Wiley, New York, 1989.
3. Sheldon, R. A., and Kochi, J. R., “Metal-Catalyzed Oxidation
of Organic Compounds,” Chap. 11. Academic Press, New York,
1981.
4. Neumann, R., and Gara, M., J. Am. Chem. Soc. 116, 5509 (1994).
5. Sawyer, D. T., Kang, C., and Liobet, A., J. Am Chem. Soc. 115(13),
5817 (1993).
6. Inchley, P., Smith, J. R., and Lower, R. J., New J. Chem. 13(10–11), 669
(1989).
7. White, R. E., and Coon, M. J., Ann. Rev. Biochem. 49, 315 (1980).
8. Herron, N., Tolman, C. A., and Stucky, G. D., J. Chem. Soc., Chem.
Commun., 1521 (1986).
9. Herron, N., Inorg. Chem. 25, 4714 (1986).
10. Cofre, P., Richert, S. A., Sobkowiak, A., and Sawyer, D. T., Inorg.
Chem. 29(14), 2645 (1990).
11. Sheu, C., Sobkowiak, A., Jeon, S., and Sawyer, D. T., J. Am. Chem.
Soc. 112(2), 879 (1990).
gest that CHP is an intermediate in the formation of K. 12. Mahroof-Tahir, M., and Karlin, K. D., J. Am. Chem. Soc. 114, 7599
(1992).
In a similar way, alcohol (A) may form radicals after loss
of H from hydroxyl groups and then may combine with
(CH3)3CO radical to form P (reaction 4 of Scheme 1). The
data of Fig. 6 show the slow disappearance of A after 30 h of
13. Barton, D. H. R., Beviere, S. D., Chavasiri, W., Doller, D., and Hu, B.,
Tetrahedron Lett. 34(12), 1871 (1993).
14. Barton, D. H. R., Csuhai, E., and Doller, D., Tetrahedron 48(42), 9195
(1992).
reaction and the slow increase of P over the total reaction 15. Lu, G.-X., Gao, H.-X., Suo, J.-S., and Li, S.-B., J. Chem. Soc., Chem.
Commun., 2423 (1994).
16. Maldotti, A., Bartocci, C., Amadelli, R., Polo, E., and Mansuy, D.,
period.
J. Chem. Soc., Chem. Commun., 1487 (1991).
17. Polo, E., Amadelli, R, Carassiti, V., and Maldotti, A., “Stud. Surf.
V. CONCLUSIONS
Sci. Catal.: Heterogeneous Catalysis & Fine Chemicals III,” p. 409.
Elsevier, Amsterdam, 1993.
18. Amadelli, R., Breqola, M., Polo, E., Carassiti, V., and Maldotti, A.,
OMS-1 materials are very active in catalyzing cyclohex-
ane oxidation with high yields of useful functional organic
J. Chem. Soc., Chem. Commun., 1355 (1992).
compoundsby using TPHP asan oxidant. Framework metal 19. Sheu, C., Richert, S. A., Cofre, P., Ross, B., Jr., Sobkowiak, A.,
Sawyer, D. T., and Kanofsky, J. R., J. Am. Chem. Soc. 112(5), 1936
(1990).
20. Schuchardt, U., Krahembuhl, C. E. Z., and Carvalho, W. A., New J.
substituted OMS-1 materials are more active than OMS-1
itself. [Fe]-OMS-1 is the best framework substituted cata-
lyst among the catalysts tested, and conversion of cyclohex-
Chem. 15(12), 955 (1991).
ane or total yield of the useful products is 13.1% (mole) at
80ꢂC for 40 h. All the OMS-1 materials retain their crys-
tallinity after reaction. These conversions and yields are
greater than other heterogeneous catalytic cyclohexane ox-
idations studied under similar reaction conditions (28), and
are competitive with the best homogeneous metal complex
21. Banfi, S., Maiocchi, A., Moqqi, A., Montanari, F., and Quici, S.,
J. Chem. Soc., Chem. Commun., 1794 (1990).
22. Barton, D. H. R., and Doller, D., Acc. Chem. Res. 25, 504 (1992).
23. Balavoine, G., Barton, D. H. R., Boivin, J., and Gref, A., Tetrahedron
Lett. 31(5), 659 (1990).
24. Haber, J., Iwanejko, R., and Mlodnicka, T., J. Mol. Catal. 55(1–3), 268
(1989).
catalysts (35). However, the OMS-1 materials have reac- 25. Camblor, M. A., Corma, A., Martinez, A., Perez-Pariente, J., and
Primo, J., “Stud. Surf. Sci. & Catal.: Heterogeneous Catalysis and Fine
Chemicals III,” p. 393. Elsevier, Amsterdam, 1993.
26. Sudhakar Reddy, J., Sivasanker, S., and Ratnasamy, P., J. Mol. Catal.
tion rates that are not as fast as the reactions catalyzed by
some homogenous metal complex catalysts. These OMS-1
heterogenous catalysts are promising due to the high yields
70, 335 (1991).
and ease of separation of the catalyst from the reaction
product mixture.
27. Rao, P. R. H. P., Ramaswamy, A. V., and Ratnasamy, P., J. Catal. 141,
604 (1993).

284
WANG ET AL.
28. Siagh, A. P., and Selvom, T., J. Mol. Catal. A 113, 489 (1996).
29. Tateiwa, J-I., Horiuchi, H., and Uemura, S., J. Chem. Soc., Chem. Com-
mun. 22, 2567 (1994).
36. Menege, S., Vincent, J. M., Lambeaux, G., Chottard, G., Grand, A.,
and Fontecave, M., Inorg. Chem. 32, 4766 (1993).
37. Bressan, M., Morvillo, A., and Romanello, G., J. Mol. Catal. 77, 283
(1992).
30. Neumann, R., and Khenkin, M., Chem. Commun., 2643 (1996).
31. Sudhakar Reddy, J., Sivasanker, S., and Ratnasamy, P., J. Mol. Catal. 38. Shen, Y. F., Zerger, P. R., DeGuzman, R. N., Suib, S. L., McCurdy, L.,
70, 335 (1991). Potter, D. I., and O’Young, C. L., Science 260, 511 (1993).
32. Rao, P. R. H. P., Ramesh Reddy, K., Ramaswamy, A. V., and Rat- 39. Shen, Y. F., Suib, S. L., and O’Young, C. L., J. Am. Chem. Soc. 116(24),
nasamy, P., “Stud. Surf. Sci. Catal.: Heterogeneous Catalysis and Fine
Chemicals III,” p. 385. Elsevier, Amsterdam, 1993.
33. Corma, A., Comblor, M. A., Esteve, P., Martinez, A., and Perez Pari-
ente, J., J. Catal. 145, 151 (1994).
11020 (1994).
40. (a) Suib, S. L., in “Recent Advances and New Horizons in Zeolite
Science and Technology,” Stud. Surf. Sci. Catal. Vol. 102, p. 47. Elsevier,
Amsterdam, 1996.
34. Rao, P. R. H. P., Ramaswamy, A. V., and Ratnasamy, P., J. Catal. 141,
604 (1993).
40. (b) Tian, Z. R., Yin, Y. G., Suib, S. L., and O’Young, C. L., Chem.
Mater. 9, 1126 (1997).
35. Barton, D. H. R., Beviere, S. D., and Hill, D. R., Tetrahedron 50(9),
2665 (1994).
41. Madon, R. J., and Boudart, M., Ind. Eng. Chem. Fund. 21(4), 438
(1982).