Co-TUD-1: A ketone-selective catalyst for cyclohexane oxidation

Article abstract of DOI:10.1002/chem.200500166

Cobalt-containing mesoporous TUD-1 with different Si/Co ratios (100, 50, 20, and 10) was synthesized by the direct hydrothermal treatment method (DHT) using triethanolamine (TEA) as a template. The prepared samples (denoted as Co-TUD-1) were characterized by XRD, UV/Vis spectroscopy, elemental analysis, N₂ sorption measurements, ²⁹Si magic-angle spinning (MAS)-NMR, Raman spectroscopy, and HRTEM. The catalytic performance of Co-TUD-1 samples was tested in the liquid-phase oxidation of cyclohexane with tert-butylhydroperoxide (TBHP) as oxidant under solventless conditions. Higher catalytic efficiency was observed in samples with low Co-loading and thus, most likely isolated Co¹¹ active sites, than in samples with high Co-loading and nanoparticles or bulk Co₃O₄ embedded in the TUD-1 silica.

Full text of DOI:10.1002/chem.200500166

DOI: 10.1002/chem.200500166
Co-TUD-1: A Ketone-Selective Catalyst for Cyclohexane Oxidation
[
a]
[b]
[c]
Mohamed S. Hamdy, Anand Ramanathan, Thomas Maschmeyer,
[
b]
[a]
UlfHane ef ld,*
and Jacobus C. Jansen*
Abstract: Cobalt-containing mesopo-
rous TUD-1 with different Si/Co ratios
spinning (MAS)-NMR, Raman spec-
troscopy, and HRTEM. The catalytic
performance of Co-TUD-1 samples
was tested in the liquid-phase oxidation
of cyclohexane with tert-butylhydroper-
oxide (TBHP) as oxidant under sol-
ventless conditions. Higher catalytic ef-
ficiency was observed in samples with
low Co-loading and thus, most likely
(
100, 50, 20, and 10) was synthesized
by the direct hydrothermal treatment
method (DHT) using triethanolamine
II
isolated Co active sites, than in sam-
(
TEA) as a template. The prepared
ples with high Co-loading and nanopar-
ticles or bulk Co O embedded in the
TUD-1 silica.
samples (denoted as Co-TUD-1) were
characterized by XRD, UV/Vis spec-
troscopy, elemental analysis, N sorp-
3
4
Keywords:
Co-TUD-1 · mesoporous materials ·
nanoparticles · oxidation
cobalt
oxide
·
2
2
9
tion measurements,
Si magic-angle
Introduction
catalysis. Industrially, this reaction is carried out homogene-
ously at temperatures above 1508C. Aselectivity of approxi-
mately 85% for mono-oxygenated products (a mixture of K,
A, and the intermediate cyclohexylhydroperoxide (CHHP))
is achieved. The CHHP is decomposed either directly or in
The selective and efficient activation of paraffins and their
conversion into valuable building blocks is a major scientific
challenge. The successful, controlled oxidation of these un-
reactive substances to mono- or difunctionalized compounds
is also of great economic significance. The conversion of cy-
clohexane to cyclohexanone [K] and cyclohexanol [A] with
high selectivity is of importance as these compounds are uti-
[3,4]
a separate step to yield additional ketone and alcohol.
In
a follow-up step, the cyclohexanol produced has to be con-
verted into the desired product, cyclohexanone. Conse-
quently, a heterogeneous system might offer several advan-
tages, such as ease of separation, recycling of the catalyst,
and solvent-free reaction conditions. Cobalt-containing mo-
lecular sieves have been extensively studied for their appli-
[
1,2]
lized in the manufacture of nylon-6 and nylon-6,6.
Cur-
rently, conversions have to be kept well below 10% to ach-
ieve reasonable selectivities and to avoid overoxidation. For
this selective oxidation of cyclohexane, cobalt is used as the
principal metal in both homogeneous and heterogeneous
[5–14]
cation in heterogeneous cyclohexane oxidation.
Howev-
er, the K/Aratios reported in these studies were only
around 1–2. It has been proven that the framework-substi-
[10]
tuted cobalt is responsible for the catalytic activity, and
that there is a correlation between the degree of activity
[
a] Dr. M. S. Hamdy, Prof. Dr. J. C. Jansen
Ceramic Membrane Centre, The Pore, DelftChemTech
Technische Universiteit Delft, Julianalaan 136, 2628 BL
Delft (The Netherlands)
[10]
and the amount of oxidizable cobalt in the framework.
Cyclohexane oxidation was also reported recently in the ap-
plication of nanostructured amorphous iron and cobalt cata-
lysts under an oxygen atmosphere. However, in these re-
ports, leaching of the metal and recycling of the catalyst
Fax : (+31)15-278-4289
E-mail: J.C.Jansen@tnw.tudelft.nl
[
b] Dr. A. Ramanathan, Dr. U. Hanefeld
Gebouw voor Scheikunde, Technische Universiteit Delft
Julianalaan 136, 2628 BL, Delft (The Netherlands)
Fax : (+31)15-278-4289
[15,16]
[17]
were not addressed.
In an earlier study, we demon-
strated that the K/Aratio could be improved by immobiliz-
ing well-defined cobalt acetate oligomers inside the pores of
MCM-41, although leaching of the cobalt due to the forma-
tion of acidic side products could not be avoided. Others
E-mail: u.hanefeld@tudelft.nl
[
c] Prof. Dr. T. Maschmeyer
Laboratory of Advanced Catalysis for Sustainability
School of Chemistry, The University of Sydney
NSW 2006 (Australia)
showed that Co could be incorporated successfully into the
II[18,19]
framework either as isolated Co
or as nanoparticles of
1782
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 1782 – 1789

FULL PAPER
[
20,21]
Co oxide.
The structural limitation (one-dimensional
During the drying step, a substantial loss of volatile com-
ponents (water and ethanol, up to 70 wt%) takes place. At
elevated temperature (hydrothermal treatment step), the
silica framework will form through extensive condensation
reactions between silica oligomers, and TEAand its Co
complexes are expelled from the inorganic framework to
form mesosized, organic-dominated aggregates.
structure) of MCM-41, which affects the catalytic activity,
prompted us to investigate a different heterogenization for
cobalt; the incorporation into a three-dimensional, mesopo-
rous silicate, TUD-1. In this manner, we aim to combine the
improved K/Aratios of our previous studies with the advan-
tages of framework-fixed cobalt, while avoiding the diffu-
sion limitations that might occur in MCM-41. TUD-1 has a
spongelike structure with tunable pore sizes, resulting in
high substrate accessibility. In addition, the synthesis of
TUD-1 is cost-effective and environmentally friendly as it is
surfactant-free. In an initial study, we prepared Co-TUD-
in which the Co atoms were highly dispersed and incor-
porated tetrahedrally into the TUD-1 framework. With this
Co-TUD-1, encouraging results were obtained in the cyclo-
hexane oxidation. Even at 5.5% cyclohexane conversion,
the selectivity for mono-oxygenated products was 89% and
the K/Aratio was 4.35. Moreover, no leaching and activity
loss was observed. This indicated that our hypothesis was
promising. Therefore, we extended our study to two types of
Co incorporation into TUD-1: cobalt was incorporated into
the framework or integrated as
The formation of Co–TEAcomplexes will lead to en-
riched Co species in the TEA-dominated organic phase.
After calcination, Co atoms were deposited onto the inter-
nal mesoporous surface (in situ grafting). Moreover, at
higher Co-loading the chance for Co O formation increases,
[
22]
3
4
II
1
nanosized scale crystals will grow inside the pores, and/or
bulky Co O will be formed as extra framework crystals.
3
4
[23]
Co-TUD-1 as a mesoporous material: Figure 1a illustrates
XRD patterns for the prepared Co-TUD-1 samples com-
pared with the pattern for Co O . All Co-TUD-1 samples
3
4
show a single intensive peak at low angle, indicating that
Co-TUD-1 is a mesostructural material; the peak intensity
decreases slightly as Co-loading increases, indicating the in-
cobalt oxide particles.
Results and Discussion
Synthesis mechanism ofCo-
TUD-1: During the synthesis of
TUD-1, triethanolamine (TEA)
plays an essential role. Initially,
tetraethyl orthosilicate (TEOS)
was hydrolysed to produce sila-
nol species [Eq. (1)], which
then undergo partial condensa-
tion with each other and with
some of the TEA[Eq. (2)] to
form mixtures of mono- and
3 4
Figure 1. a) XRD patterns for Co-TUD-1 samples compared with the pattern for Co O . b) Low-angle XRD
oligomeric-silatrane complexes pattern of Co-TUD-1.
of various silica species.
SiðOC H Þ þ H O ! ðC H OÞ SiOH þ C H OH
ð1Þ
ð2Þ
fluence of Co-loading on the integrity of the mesoporous
structure. Ashift to large angles at higher Co-loading sup-
ports this observation (Figure 1b). In the spectra for Co-5
2
5
4
2
2
5
3
2
5
ðC H OÞ SiOH þ NðCH CH OHÞ !
2
5
3
2
2
3
NðCH CH OÞ SiOH þ 3C H OH
and Co-10, the peaks characteristic for crystalline Co O4
2
2
3
2
5
3
become visible, indicating the presence of Co O particles.
3
4
[24]
Shan et al proposed that TEAhas a second role in the
formation of Ti-TUD-1; to stabilize the titanium alkoxide
through complexation. This is based on the observation that
the three hydroxyl groups in TEAcan easily replace butanol
groups of Ti butoxide. Moreover, the lone-pair electrons of
the nitrogen atom in TEAcan be donated to the empty or-
bital in titanium to form stable complexes. The formation of
a Co–TEAcomplex under the same conditions has been
The elemental analysis and the porosity measurements
obtained from N adsorption at 77 K are listed in Table 1.
2
Elemental analysis (Figure 2) showed that the Si/Co ratio
obtained after calcination corresponds with that present in
the synthesis gel, which indicates that most of the Co cations
are incorporated into the final solid product. Moreover, it
demonstrates the high predictability of the synthesis
method.
Figure 3a shows the N2 sorption isotherms of the Co-
TUD-1 samples. All the samples show type IV adsorption
isotherms, according to IUPAC classification, indicating
[25]
proven earlier. The synthesis mixture thereby becomes an
organic–inorganic hybrid, in which TEAand its Co complex
are homogeneously dispersed in a three-dimensional silica gel.
Chem. Eur. J. 2006, 12, 1782 – 1789
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1783

U. Hanefeld, J. C. Jansen et al.
Table 1. Characterization of Co-TUD-1 samples with different Co-load-
ing.
pore volumes and pore diameters decrease as the Co-load-
ing increases. This result is a further indication that at low
Co-loading the Co is integrated into the framework of
TUD-1.
The Si magic-angle spinning (MAS)-NMR spectrum of
calcined siliceous TUD-1 samples is shown in Figure 4a.
After the deconvolution of the spectrum, three main peaks
[
a]
[b]
[c]
Sample
Si/Co ratio
S
[
BET
2
V
meso
3
D
[nm]
meso
Color
ꢀ1
ꢀ1
m g
]
[cm g
]
synthesis after
mixture calcination
2
9
Co-1
Co-2
100
108
619
605
0.73
0.65
4
grayish
violet
grayish
violet
gray
50
47.8
3.9
Co-5
Co-10
20
10
18.6
9.95
614
684
0.69
0.58
3.6
3.1
gray
[
a] Specific surface area. [b] Mesopore volume. [c] Mesopore diameter.
Figure 2. The elemental composition of Co-TUD-1 in the synthesis gel
plotted against the Si/Co ratio obtained in the final product.
2
9
Figure 4. Si MAS-NMR spectra for a) Si-TUD-1 compared with b) the
Co-5 sample.
were detected. The first at around d=ꢀ110 ppm can be as-
Figure 3. a) Nitrogen sorption isotherms of Co-TUD-1 samples. b) The
pore-size distribution of Co-TUD-1 samples.
signed to (-O-) Si with no OH group attached to the silicon
4
atom (Q ). The second peak at d=ꢀ102 ppm is assigned to
4
(
-O-) Si(OH) with one OH group (Q ). The last peak at d=
3 3
ꢀ
90 ppm, is assigned to (-O-) Si(OH) with two OH groups
2 2
[
26]
[27]
their mesostructured character with narrow pore-size dis-
tribution (Figure 3b). All hysteresis loops exhibit the same
H3-type behavior, that is, they show one well-defined step
at high partial pressure (0.45<P/Po<8), which should be
(Q ).
In contrast, in the spectrum of Co-5 (Figure 4b),
2
three peaks were detected that show a strong paramagnetic
shift due to the cobalt oxide particles/atoms incorporated
into the framework. The peak at d=ꢀ113 ppm is assigned
due to the capillary condensation of N inside the meso-
to Q that was shifted due to the strong paramagnetic field
2
4
pores. Moreover, the increase in Co-loading did not seem to
affect the hysteresis type. This is an indication that the bulk
crystals of Co O , which formed in samples with a higher
Co-loading, were formed outside the silica framework. This
is also supported by the observation that, although the sur-
face areas did not change significantly (see Table 1), the
of Co-oxide particles nearby, and the signal at d=ꢀ102 ppm
is assigned to the Q . More importantly, a new signal at d=
3
ꢀ109 ppm could be detected. This might be attributed to a
3
4
Si atom adjacent to a Co atom incorporated into the TUD-1
2
9
framework as [(-O-) Si-O-Co]. The Si MAS-NMR spectra
3
for all the samples were virtually identical: clearly the maxi-
1784
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 1782 – 1789

Cyclohexane Oxidation Catalysed by Co-TUD-1
FULL PAPER
mum of Co incorporated into the silica matrix is achieved at
1
–2%, and consequently, the signal at d=ꢀ109 ppm has a
similar intensity for all the samples. These conclusions are in
agreement with a recent study on Co-MCM-41-like materi-
al.
The UV/Vis spectra of different Co-TUD-1 samples and
Co O are presented in Figure 5. Spectra of prepared sam-
[
28]
3
4
Figure 6. Laser Raman spectra of Co-TUD-1 samples compared with that
3 4
of Co O . The spectrum of Co-2 is not shown as it is very similar to that
of Co-1.
Figure 5. Diffuse reflectance UV/Vis spectra of Co-TUD-1 samples com-
pared with that for Co
3
4
O .
ples exhibit two peaks at around 525 and 654 nm. Both ab-
4
4
sorption bands can be assigned to the A (F)! T (P) transi-
2
1
2
+
tion of divalent cobalt ions (Co ) in tetrahedral coordina-
[29,30]
tion.
indicates the absence of Co
The absence of peaks at around 480 and 506 nm
in an octahedral environ-
and the absence of a peak at 410 nm indicates the
2
+
[28,31]
ment,
absence of framework Co
3
+ [32]
.
Apeak at around 356 nm
Figure 7. HRTEM image representative for the Co-TUD-1 samples. The
wormlike structure of the mesoporous materials described in detail else-
where is clearly shown.
was detected in Co-5 and Co-10 only. This strong and broad
band in the lower wavelength region will be due to the
charge-transfer bands associated with the non-framework
[38]
3
+
Co species. This is strong evidence for the presence of a
distinct Co O phase, as it is the main peak found in the
[38]
materials. This is a strong indication for the incorporation
3
4
[33]
2+
Co O4 spectrum;
this is also consistent with the XRD
of Co ions into the framework. Given the low Co concen-
3
spectra.
trations and the UV/Vis data, these Co atoms should be iso-
lated in the framework. The images of the Co-5 sample
showed also a spongelike mesoporous matrix beside the
Co O bulky crystals and, more importantly, by using higher
Laser Raman spectra of the Co-TUD-1 samples and
Co O are compared in Figure 6. All spectra showed a band
3
4
ꢀ
1
[34]
at around 998 cm that was assigned to the SiO matrix.
2
3
4
The increase of the Co-loading from Co-5 to Co-10 gives
magnification, the diffraction fringes of the embedded
Co O nanoparticles could be observed.
Table 2 summarizes the characteristics of the prepared
Co-TUD-1 samples as obtained from the different character-
ization techniques applied.
ꢀ
1
rise to three bands at around 690, 485, and 525 cm , as-
signed to the A , E , and F active Raman modes, respec-
3
4
1
g
g
2g
[35,36]
tively, of the direct spinel Co O ,
which is consistent
3
4
with XRD and UV/Vis spectra. The spectra for Co-1 and
Co-2 were virtually identical. The broad signals at approxi-
ꢀ
1
mately 500 and 820 cm were also described for Co-MCM-
Table 2. Appearance of Co in Co-TUD-1 samples.
4
1 samples with isolated and framework-incorporated Co
Sample Co Oxida-
Isolated
Co
Co
3
O
4
Extra framework
bulky Co O
3 4
[37]
species at a very low Co-loading.
II[a]
[a]
[a]
tion state
nanoparticles
The prepared Co-TUD-1 samples were investigated by
transmission electron microscopy; almost 25 images were
taken per sample. All the images of Co-1 (Figure 7) and Co-
II
Co-1
Co-2
Co-5
Co
Co
++
++
+
–
–
+
–
–
+
II
II
III
III
Co +Co
II
Co-10 Co +Co
+
–
++
2
(not presented here) showed only the spongelike, three-di-
mensional structure characteristic of TUD-1 mesoporous
[a] (++) abundantly present, (+) weakly or poorly present, (–) absent.
Chem. Eur. J. 2006, 12, 1782 – 1789
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1785

U. Hanefeld, J. C. Jansen et al.
Catalysis: The major products of cyclohexane oxidation with
tert-butylhydroperoxide (TBHP) over the Co-TUD-1 sam-
ples described above were cyclohexanone [K], cyclohexanol
[A], and cyclohexylhydroperoxide (CHHP). However, trace
amounts of cyclohexyl-tert-butylperether and cyclohexylfor-
mate were also observed as mono-oxygenated products. All
the cobalt catalysts showed an excellent selectivity to mono-
oxygenated products at a remarkable conversion of 5%. By
comparing the results for the cyclohexane oxidation with
these catalysts (Figure 8a–e), it became evident that isolated
cobalt species in TUD-1 are very active in the oxidation of
cyclohexane with TBHP. At the beginning of the reaction,
the K/Aratios for all the cobalt catalysts studied were close
to 1 or even lower (Figure 8d), and this ratio increased as
conversion increased, showing that cyclohexanol [A] is the
primary product, which is later oxidized to cyclohexanone
[
K]. Atypical product distribution over Co-1 after 1 h is
mol%) K=30.5%, A=23.3%, CHHP=39.0%. After 18 h,
the distribution changed to K=62.8%, A=9.4%, CHHP=
3.7%. These values strongly indicate the oxidation of cy-
(
1
clohexanol to cyclohexanone with TBHP. As CHHP is one
of the products formed, it might be suggested that it too can
act as an oxidizing agent. However, as the conversion of cy-
clohexane increased from 1.2 to 10% (over Co-1), the yield
of CHHP also increased continuously. This indicates that
CHHP does not act as an oxidizing agent. On the other
hand, it cannot be excluded that a thermal reaction, such as
[17]
the mixed Russell termination
(Scheme 1), occurs and
contributes to cyclohexanone formation. This selectivity to-
wards cyclohexanone decreased as the loading of cobalt in
the catalyst increased and is especially noticeable in Co-10,
in which the cobalt is present as cobalt oxide clusters (Fig-
ure 8c and d). The loss of selectivity is accompanied by a
loss of activity (Figure 8a and b). Thus, the conversion of
both cyclohexane and cyclohexanol decreases as the cobalt
concentration in TUD-1 increases. This might be attributed
to the increasing aggregation of Co or Co oxide at higher
cobalt concentrations (see above). The agglomeration of Co
reduces the accessibility of the individual cobalt atoms and,
therefore, of the number of active sites. At lower conversion
levels, all these catalysts behaved in a similar manner, as is
evident from the conversion vs selectivity graph (Figure 8e).
However, at a moderate conversion of 5%, the selectivities
of the catalysts show significant differences, with Co-1 being
clearly the most selective catalyst (Co-1>Co-2>Co-5>Co-
10). Under identical experimental conditions, the cobalt
oxide behaved similarly to Co-10, however, the conversions
were relatively low, with CHHP as the major product
(
Table 3). These studies demonstrate that isolated cobalt
Figure 8. Catalytic performance of Co-TUD-1 samples in cyclohexane ox-
idation with TBHP at 708C.
species are more efficient than Co-oxide clusters in the oxi-
dation of cyclohexane and cyclohexanol with TBHP.
[17]
In an earlier study, we improved the K/Aratio signifi-
cantly. However, overoxidation led to leaching of active
cobalt species. Therefore, the aim of this study was to sup-
press overoxidation and the leaching of cobalt that it indu-
ces. Here, analysis for the possible overoxidation products
revealed that dicarboxylic acids (adipic, succinic), monocar-
boxylic acids (valeric, caproic), along with very low concen-
trations of hydroxyacids, such as 5-hydroxyvaleric acid and
6-hydroxycaproic acid, were barely formed at low conver-
sions. Only at higher conversions, as the selectivity of the
catalysts dropped to below 93% (Figure 8e), could acids be
1786
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 1782 – 1789

Cyclohexane Oxidation Catalysed by Co-TUD-1
FULL PAPER
Conclusion
We have demonstrated the application of cobalt-containing,
three-dimensional mesoporous silicate, Co-TUD-1, in the
oxidation of cyclohexane. Improvements in the selectivity
for mono-oxygenated products, as well as in the K/Aratio,
were achieved. Cobalt was successfully incorporated into
the three-dimensional silica matrix of TUD-1 either as iso-
lated Co atoms or as bulky cobalt oxide particles. At lower
concentrations, the cobalt is incorporated into the frame-
work of TUD-1 as an isolated atom, whereas at higher load-
ing of cobalt, oxide clusters
Scheme 1. Mixed Russell termination step between tBuOOC and cyclo-
C H11OOC.
6
along with a very small amount
of cobalt oxide nanoparticles
Table 3. Cyclohexane oxidation over various Co-TUD-1 samples.
[
a]
Catalyst
Time
Conv. cyclohexane
[mol%]
Conv. TBHP
[mol%]
K
[%]
A
[%]
CHHP
[%]
Sel. mono.
[mol%]
K/A
ratio
are observed in the TUD-1.
This new catalyst system is het-
erogeneous, as no catalytic ac-
tivity was detected after hot fil-
tration.
[h]
Co-1
1
5
1.2
4.09
10.3
1.4
3.40
0.6
3.17
8.8
1.39
2.90
0.7
2.22
5.3
1.33
2.69
0.5
2.54
5.8
1.52
2.97
0.34
2.0
3.8
22.73
49.5
7.54
18.76
3.5
16.13
41.1
7.26
17.71
3.1
10.99
21.5
5.46
9.16
1.3
14.39
25.9
8.31
13.0
30.5
45.46
62.8
29.85
38.78
20.2
31.93
54.8
31.15
35.5
17.2
30.30
45.8
38.41
33.49
18.3
23.3
19.36
9.4
13.09
14
22.4
10.59
9.7
10.32
10.08
24.1
9.02
9.5
39
96.58
92.62
90.52
95.18
94.40
97.22
93.95
88.58
89.42
85.71
97.5
96.03
90.98
93.36
84.9
97.46
90.08
88.65
88.5
1.31
2.35
6.72
2.28
2.77
0.9
3.01
5.76
3.02
3.5
0.71
3.36
4.85
3.15
3.74
0.7
4.24
5.05
3.01
4.78
–
26.59
13.1
49.99
37.22
54.7
46.6
17.6
44.02
36.15
56.2
56.70
30.6
41.12
42.3
50.1
44.04
27.1
39.13
45.24
76
1
8
3
4
1
7
8
4
1
1
8
8
4
4
1
8
4
4
4
1
8
[
[
[
b]
b]
b]
.5
2
Co-2
Co-5
Co-10
.5
.5
Experimental Section
1
2
Materials synthesis: Co-TUD-1 sam-
ples with different Co content (Si/Co
ratio=100, 50, 20, and 10 denoted as
Co-1, Co-2, Co-5, and Co-10, respec-
tively) were synthesized by means of
direct hydrothermal treatment (DHT)
according to the molar oxide ratio
1
2
.5
12.18
4.67
26.2
8.33
9.8
35.32
49. 3
35.14
27.75
2 2
SiO :xCoO:0.5TEAOH:1TEA:11H O
(TEAOH=tetraethyl ammonium hy-
2
2
1
[
b]
11.69
5.8
droxide). In a typical synthesis (for
[
23]
83.8
–
–
Co-1),
a mixture of 12.6 g TEA
Co
3
O
4
(97%, ACROS) and deionized water
(3.5 mL) was added drop-wise to a
mixture of 17.4 g tetraethyl orthosili-
cate (TEOS, >98%, ACROS) and
69
–
[
a] Total reaction time. [b] Analysis from hot-filtration studies.
0
.23 g cobalt(ii)sulfate heptahydrate
(
CoSO ·7H O, Aldrich) dissolved in
4
2
observed. To confirm that the catalysts are indeed heteroge-
neous and that no activity leaches, hot-filtration studies
were performed. After 1 h of reaction, the hot (above 508C)
reaction mixture was filtered, and the reaction was contin-
ued with the filtrate. The results (Table 3) indicate that,
after a period of 2–5 h, there is still a low level of activity in
the filtrate, due to the slow termination of residual radical
chains. However, after longer runs (24 h), there is little im-
provement in the conversion of cyclohexane. The heteroge-
neity of this reaction was further reflected in the drop in se-
lectivity for mono-oxygenated products at comparable con-
version levels. Analysis of the Co-TUD-1 samples for their
Co content after filtration revealed no loss of metal. This
once again suggests that Co is incorporated into the TUD-1
framework.
deionized water (5 mL) under vigorous stirring. After stirring for a fur-
ther 30 min, 10.3 g of TEAOH (35%, Aldrich) was added. The mixture
was aged at RT for 24 h, dried at 1008C for 24 h, and then hydrothermal-
ly treated in teflon-lined stainless steel autoclaves at 1808C for 4 h. Final-
ly, the solid samples were calcined at 6008C for 10 h at a ramp rate of
ꢀ
1
18Cmin in air.
Materials characterization: Powder X-ray diffraction (XRD) patterns
were measured by using a Philips PW 1840 diffractometer equipped with
a graphite monochromator using CuKa radiation (l=0.1541 nm). The
samples were scanned over a range of 0.1–808 2q with steps of 0.028. Ni-
trogen adsorption and desorption isotherms were recorded by using a
QuantaChrome Autosorb-6B at 77 K. The pore-size distribution was cal-
culated from the adsorption branch by using the Barret–Joyner–Halenda
[
39]
(
BJH) model. Samples were previously evacuated at 623 K for 16 h.
The BET method was used to calculate the surface area (SBET) of the
samples, and the mesopore volume (Vmeso) was determined by using the t-
[40]
plot method according to Lippens and de Boer. Instrumental neutron
[
41]
activation analysis (INAA)
elemental analysis) was performed by using the “Hoger Onderwijs Re-
actor” nuclear reactor with a thermal power of 2 MW and maximum neu-
for chemical composition determination
(
1
4
ꢀ2 ꢀ1
tron reflux of 2x10
m s . This method can be applied to solid samples
and was used because of difficulties in dissolving the samples. The
method proceeds in three steps: irradiation of the elements with neutrons
in the nuclear reactor, a period of decay, and finally, measurement of the
Chem. Eur. J. 2006, 12, 1782 – 1789
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1787

U. Hanefeld, J. C. Jansen et al.
radioactivity resulting from irradiation. The energy of the radiation and
the half-life period of the radioactivity enable a highly accurate quantita-
[1] W. B. Fisher, J. F. Van Peppen in Kirk Othmer Encyclopedia of
Chemical Technology, 4th ed., Vol. 7 (Ed.: M. Howe-Grant), Wiley,
New York, 1996, pp. 859–871.
[2] M. T. Musser in Ullmannꢀs Encyclopedia of Industrial Organic
Chemicals, Vol. 3, VCH-Wiley, Weinheim, 1999, pp. 1807–1821.
[3] P. A. MacFaul, I. W. C. E. Arends, K. U. Ingold, D. D. M. Wayner, J.
Chem. Soc. Perkin Trans 2 1997, 135–145.
[
41] 29
tive analysis.
Si MAS-NMR experiments were performed at a mag-
netic field of 9.4 T by using a Varian VXR-400 S spectrometer operating
at 104.2 MHz with pulse width of 3.2 msec. Zirconia rotors of 4 mm were
used with a spinning speed set to 8 kHz. The chemical shift was measured
with respect to tetraethylsilane (TMS) as an external standard at 0 ppm.
1
000 scans were collected by using a sweep width of 20000 Hz and an
aquisition delay of 20 sec.
[4] R. A. Sheldon, J. K. Kochi, Metal-Catalyzed Oxidation of Organic
Compounds, Academic Press, New York, 1981, Chapter 11.
[5] P. Tian, Z. Liu, Z. Wu, L. Xu, Y. He, Catal. Today 2004, 93–95, 735–
742.
[6] S. K. Mohapatra, P. Selvam, Top. Catal. 2003, 22, 17–22.
[7] A. F. Masters, J. K. Beattie, A. L. Roa, Catal. Lett. 2001, 75, 159–
162.
[8] R. Raja, G. Sankar, J. M. Thomas, J. Am. Chem. Soc. 1999, 121,
11926–11927.
[9] D. L. Vanoppen, P. A. Jacobs, Catal. Today 1999, 49, 177–183.
[10] G. Sankar, R. Raja, J. M. Thomas, Catal. Lett. 1998, 55, 15–23.
[11] I. Belkhir, A. Germain, F. Fajula, E. Fache, J. Chem. Soc. Faraday
Trans. 1998, 94, 1761–1764.
UV/Vis spectra were collected at ambient temperature by using a Cary-
Win 300 spectrometer with BaSO as a reference. Samples were ground
4
carefully, heated overnight at 1808C, and then scanned from 190–800 nm.
The UV/Vis absorption data were converted to Kubelka–Munk units.
The in situ laser Raman spectra were obtained by using a Renishaw
Raman Imaging Microscope, system 2000. The green (l=514 nm) polar-
ized radiation of an argon-ion laser beam of 20 mW was used for excita-
tion. ALeica DMLM optical microscope with a Leica PL floutar L500/5
objective lens was used to focus the beam onto the sample. The Rama-
scope was calibrated by using a silicon wafer. Samples were dehydrated
ꢀ
1
in situ in an air flow of 100 mLmin by using a temperature-program-
med heated cell (Linkam TS1500). The spectra were collected in the
ꢀ
1
range of 180–1600 cm . HRTEM was carried out by using a Philips
CM30UT electron microscope with a field-emission gun as the source of
electrons operated at 300 kV. Samples were mounted on a copper-sup-
ported carbon-polymer grid by placing a few droplets of a suspension of
ground sample in ethanol onto the grid, followed by drying under ambi-
ent conditions. Electron dispersive X-ray (EDX) elemental analysis was
performed by using a LINK EDX system. Atomic absorption spectrosco-
py (AAS) analysis for leached metal ions in the filtered product mixtures
of cyclohexane oxidation experiments was performed by using a Perkin–
Elmer 4100ZL instrument.
[12] F. J. Luna, S. E. Ukawa, M. Wallau, U. Schuchardt, J. Mol. Catal. A
1
997, 117, 405–411.
[
13] D. L. Vanoppen, D. E. De Vos, M. J. Genet, P. G. Rouxhet, P. A.
Jacobs, Angew. Chem. 1995, 107, 637–639; Angew. Chem. Int. Ed.
Engl. 1995, 34, 560–563.
14] S.-S. Lin, H.-S. Weng, Appl. Catal. A 1993, 105, 289–308.
15] N. Perkas, Y. Koltypin, O. Palchik, A. Gedanken, S. Chandrasekar-
an, Appl. Catal. A 2001, 209, 125–130.
16] V. Kesavan, P. S. Sivanand, S. Chandrasekaran, Y. Koltypin, A. Ge-
danken, Angew. Chem. 1999, 111, 3729–3730; Angew. Chem. Int.
Ed. 1999, 38, 3521–3523.
17] M. Nowotny, L. N. Pedersen, U. Hanefeld, T. Maschmeyer, Chem.
Eur. J. 2002, 8, 3724–3731.
[
[
[
[
Catalytic performance: Astock solution of TBHP in cyclohexane was
prepared by extraction of commercial TBHP (Aldrich, 70% in water)
into an equal volume of cyclohexane. Phase separation was promoted by
saturation of the aqueous layer with NaCl. The organic layer was dried
over MgSO , filtered, and stored at 48C. TBHP content was determined
4
by GC analysis with chlorobenzene as an internal standard.
[
[
18] V. Parvulescu, B. L. Su, Catal. Today 2001, 69, 315–322.
19] S. Lim, D. Ciuparu, C. Pak, F. Dobek, Y. Chen, D. Harding, L. Pfef-
ferle, G. Haller, J. Phys. Chem. A 2003, 107, 11048–11056.
Oxidation of cyclohexane was carried out with 20 mL of a mixture of cy-
clohexane (65 mol%) and TBHP (35 mol%). Chlorobenzene (1 g) was
added as the internal standard. In all the experiments, an equimolar con-
centration of cobalt was used (0.1 mmol). The round-bottomed glass
flask with the reaction mixture containing the catalyst was then immersed
in a thermostated oil bath. The gas phase above the reaction mixture was
filled with nitrogen and a gas burette was attached. The course of the re-
action was followed by analyzing the liquid samples with a gas chromato-
graph (Agilent 6890) equipped with a split inlet (2008C, split ratio 10.0)
and a Sil 5 CB capillary column (ID, 50 m”0.53 mm; constant flow of
[20] D. Dhar, Y. Koltypin, A. Gedanken, S. Chandrasekaran, Catal. Lett.
2003, 86, 197–200.
[
21] J. Haskouri, S. Cabrera, C. Guillem, J. Latorre, A. Beltran, M. D.
Marcos, C. J. Gomes-Garcia, D. Beltran, P. Amoros, Eur. J. Inorg.
Chem. 2004, 9, 1799–1803.
[
[
[
22] J. C. Jansen, Z. Shan, L. Marchese, W. Zhou, N. van der Puil, T.
Maschmeyer, Chem. Commun. 2001, 713–714.
23] R. Anand, M. S. Hamdy, U. Hanefeld, T. Maschmeyer, Catal. Lett.
2
004, 95, 113–117.
24] Z. Shan, E. Gianotti, J. C. Jansen, J. A. Peters, L. Marchese, T.
Maschmeyer, Chem. Eur. J. 2001, 7, 1437–1443.
ꢀ1
2
carrier gas N , 4.0 mLmin ) coupled to a FID detector. The concentra-
tion of carboxylic acid side products was determined by performing GC
analysis from separate samples after conversion into the respective
[
[
25] S. Teller, Y. Marcus, Y. TurꢁYan, Polyhedron 1997, 16, 1047–1056.
26] K. S. W. Sing, D. H. Everett, R. A. W. Haul, L. Moscou, R. A. Pier-
otti, J. Rouquerol, T. Siemieniewska, Pure Appl. Chem. 1985, 57,
[
17,23]
methyl esters.
Identification of the products was achieved by con-
ducting GC–MS. The evolution of molecular oxygen and its consumption
was monitored volumetrically by using the attached gas burette. The con-
version of cyclohexane is defined in mol%, that is, the number of moles
of products formed divided by the initial number of moles of cyclohexane
multiplied by 100. All mass balances were >92%.
6
03–619.
[
[
27] K. J. Mackenzie, M. E. Smith in Multinuclear Solid-State NMR of In-
organic Materials, Pergamon, 2002.
28] J. Haskouri, S. Cabrera, C. J. Gomes-Garcia, C. Guillem, J. Latorre,
A. Beltran, D. Beltran, M. D. Marcos, P. Amoros, Chem. Mater.
2
004, 16, 2805–2813.
[
[
[
29] J. Sponer, J. Cejka, J. Dedecek, B. Wichterlova, Microporous Meso-
porous Mater. 2000, 37, 117–127.
30] R. S. da Cruz, A. J. S. Mascarenhas, H. M. C. Andrade, Appl. Catal.
B 1998, 18, 223–231.
31] J. Y. Yan, M. C. Kung, W. M. Sachtler, H. H. Kung, J. Catal. 1997,
172, 178–186.
Acknowledgements
U.H. thanks the Royal Netherlands Academy of Arts and Sciences
(
KNAW) for a fellowship. M.S.H. thanks Helwan University and the
[32] W. A. Carvalho, P. B. Varaldo, M. Wallau, U. Schuchardt, Zeolite
1997, 18, 408–416.
[33] E. Finocchio, T. Montanari, C. Resini, G. Busca, J. Mol. Catal. A
2003, 204–205, 535–544.
Ministry of Higher Education of the Egyptian Government for a fellow-
ship. The authors are thankful to Delia van Rij for carrying out INAA
analysis, Sander Brouwer for N
nashvili for NMR analysis.
2
sorption experiments, and Kristina Dja-
[34] C. Wang, G. Deo, I. E. Wachs, J. Catal. 1998, 178, 640–648.
1788
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 1782 – 1789

Cyclohexane Oxidation Catalysed by Co-TUD-1
FULL PAPER
[
35] M. A. Stanick, M Hoalla, D. M. Hercules, J. Catal. 1987, 106, 362–
62.
36] H. Jeziorowski, H. Knoezinger, P. Grange, P. Gajardo, J. Phys.
Chem. 1980, 84, 1825–1829.
[40] B. C. Lippens, J. H. de Boer, J. Catal. 1965, 4, 319–323.
3
[41] F. Boynton in Handbook on the Physics and Chemistry of Rare
Earths, Vol. 4 (Eds.: K. Gschneidner, Jr., L. Eyring), North-Holland
Publishing Company, Amsterdam, 1979, pp. 457–470.
[
[
[
37] Q. Tang, Q. Zhang, H. Wu, Y. Wang, J. Catal. 2005, 230, 384–397.
38] P. Waller, Z. Shan, L. Marchese, G. Tartaglione, W. Zhou, J. C.
Jansen, T. Maschmeyer, Chem. Eur. J. 2004, 10, 4970–4976.
39] E. P. Barrett, L. G. Joyner, P. P. Halenda, J. Am. Chem. Soc. 1951,
Received: February 15, 2005
Revised: August 8, 2005
[
7
3, 373–380.
Published online: December 6, 2005
Chem. Eur. J. 2006, 12, 1782 – 1789
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1789