Hydrodechlorination of para-chloroacetophenone in water/ethanol mixtures using organosilane-grafted Rh/SiO2catalysts

Article abstract of DOI:10.1246/cl.140539

Hydrodechlorination was performed in water/ethanol mixtures over organosilane-grafted Rh/SiO₂catalysts. All catalysts in which organosilanes were covalently connected onto silica supports promoted hydrodechlorination, which may be attributed to

Full text of DOI:10.1246/cl.140539

CL-140539
Received: May 28, 2014 | Accepted: June 24, 2014 | Web Released: July 1, 2014
Hydrodechlorination of para-Chloroacetophenone in Water/Ethanol Mixtures
Using Organosilane-grafted Rh/SiO Catalysts
2
1
2
1
2
Tetsuya Yoneda,* Tadashi Aoyama, Koshiro Koizumi, and Toshio Takido
Department of Liberal Arts and Science, College of Science and Technology, Nihon University,
-24-1 Narashinodai, Funabashi, Chiba 274-8501
Department of Applied and Material Chemistry, College of Science and Technology, Nihon University,
-8-14 Kanda-surugadai, Chiyoda-ku, Tokyo 101-8308
1
7
2
1
(
E-mail: yoneda.tetsuya@nihon-u.ac.jp)
Hydrodechlorination was performed in water/ethanol mix-
tures over organosilane-grafted Rh/SiO2 catalysts. All catalysts
in which organosilanes were covalently connected onto silica
supports promoted hydrodechlorination, which may be attrib-
uted to the presence of a hydrophobic zone on the catalyst
surfaces. Additional aromatic π­π interactions between the 4-
tolylsilyl substituent in the catalyst and the aromatic ring of
p-chloroacetophenone may be important for the activity of
the most active Rh/SiO2­TS catalyst that was produced via
silylation using trichloro-4-tolylsilane.
previously,12 and reduced under
a
hydrogen flow of
¹
1
100 mL min
under 0.5 MPa for 60 min at 473 K. Water
(35 mL) was injected into the test tube at 373 K, and CLAP
(2 mmol) in ethanol (5 mL) was introduced using 0.8 MPa
hydrogen. The pressure was adjusted to 1.0 MPa, allowing the
reaction to start. The products were periodically extracted from
the reaction mixture through a pressure-resistant valve and
quantitatively analyzed using a gas chromatograph (Shimadzu
Co., GC-14A) equipped with a flame ionization detector.
Toluene was used as an internal standard. The products were
identified by gas chromatography­mass spectroscopy (Shimadzu
Co., GCMS5050QA). The Brunauer­Emmett­Teller (BET)
specific and metal surface areas were determined using a
BELCAT-B instrument (Bell Japan Inc.). Elemental analyses
were carried out using a JM10 apparatus (J-SCIENCE LAB Co.,
Ltd.). Thermogravimetry­differential thermal analysis (TG-
DTA) profiles were acquired using a TG-DTA 2020s apparatus
(MAC Science Co., Ltd.). Metal loading was measured by
inductively coupled plasma mass spectrometry (Shimadzu Co.,
ICPS-8100) performed at Sumika Chemical Analysis, Ltd.
Except for the BET specific surface area, all measurements were
Catalytic hydrodechlorination (HDC) is expected to solve
problems related to harmful organic chlorinated compounds by
limiting the degradation costs and preventing the production of
additional toxic by-products such as dioxins.1 In particular,
HDC systems using water-based solvents are now desirable to
reduce the use of large amounts of organic solvents that release
volatile or persistent organic substances into the environment.
HDC attempts at dechlorinating chlorinated aromatics in water
have involved directly modifying the surface functionalities of
the catalyst support to increase the reactivity. These attempts
include (i) using silicone polymers to coat Pd/Al2O3 catalysts
­3
11
conducted under the same conditions as previously reported.
Figure 1 shows the conversion and product yields for the
HDC of CLAP, which was performed in the presence of the bare
Rh/SiO2 catalyst in 7:1 (v/v) water/ethanol for 60 min at 373 K.
The detected products consisted of α-methylbenzyl alcohol
(MBA, 13% yield), ethylbenzene (EB, 9% yield), 4-chloro-α-
methylbenzyl alcohol (CMBA, 3% yield), acetophenone (AP,
2% yield), 1-cyclohexylethanol (CHEL, 2% yield), acetylcyclo-
hexane (ACH, 1% yield), ethylcyclohexane (ECH, 1% yield),
4­6
and/or fill their pores and (ii) immobilizing Pd nanoparticles
in amphiphilic copolymers.7
­10
In addition, when grafted to
certain organosilyl derivatives, Pt/SiO2 was found to preferably
catalyze the HDC of para-chloroacetophenone (CLAP) in a
water/ethanol mixture.11 Although all grafted Pt/SiO catalysts
2
showed the excellent HDC performance of CLAP, one drawback
was found, i.e., the production of unpreferable by-product ethyl
α-methylbenzyl ether, which was caused by dehydration between
ethanol in the solvent mixture and α-methylbenzyl alcohol as
one of the HDC products. To prevent the undesirable by-product
formation, we developed new silica-supported rhodium catalysts.
Here, the modified Rh/SiO2 catalysts were found to promote
this reaction in a water/ethanol mixed solvent, shedding light on
the effect of organosilyl grafting reagents on silica supports.
The bare Rh/SiO2 catalyst was prepared on silica by an ion-
exchange reaction (Supporting Information). Next, its surface
was subjected to silylation using five types of organosilane
reagents: butylchlorodimethylsilane (BDMS), chlorodimethyl-
octylsilane (DMOS), chlorodimethyloctadecylsilane (DMODS),
trichloro-4-tolylsilane (TS), and chloro(3,3,4,4,5,5,6,6,7,7,8,8,8-
tridecafluorooctyl)dimethylsilane (TDFDMS). Silylation pro-
duced Rh/SiO2­BDMS, Rh/SiO2­DMOS, Rh/SiO2­DMODS,
Rh/SiO2­TS, and Rh/SiO2­TDFDMS catalysts.
Figure 1. Product yields for the HDC reaction of CLAP over
Rh/SiO2 catalyst in 7:1 (v/v) water/ethanol using 1 MPa
hydrogen for 60 min at 373 K.
The catalyst (40 mg) in a 77-mL glass test tube was placed
into a magnetically stirred batch autoclave, as described
1604 | Chem. Lett. 2014, 43, 1604–1606 | doi:10.1246/cl.140539
© 2014 The Chemical Society of Japan

Table 1. Turnover frequency (TOF) for the HDC reaction of
CLAP over Rh/SiO catalysts
2
Organosilane
TOF
min
Metal particle sizea
/nm
Catalyst
numberb
¹
1
/
¹
2
/
group nm
Rh/SiO2
Rh/SiO2­TS
Rh/SiO2­BDMS
Rh/SiO2­DMOS
Rh/SiO2­DMODS
Rh/SiO2­TDFDMS
6.2
19.7
10.2
9.8
18.2
7.5
3.7
6.7
5.5
5.1
7.8
4.8
®
2.3
0.6
0.4
0.3
0.3
a
b
Calculated by carbon monoxide (CO) chemisorption. Calcu-
lated by elemental analysis.
Figure 2. Product yields for the HDC reaction of CLAP over
Rh/SiO2­TS in 7:1 (v/v) water/ethanol using 1 MPa hydrogen
for 60 min at 373 K.
1
05
Rh/SiO −TS
(
a) TG
2
Rh/SiO2
SiO2 (silica support)
(1-cyclohexenyl)-1-ethanol (CHEEL, <1% yield), and 4-chloro-
100
styrene (CS, <1% yield) (Table S1, Supporting Information).
Therefore, the reaction conversion and the overall dechlorinated
product yield amounted to 31% and 28%, where the reaction
conversion sums all product yields up, and the overall
dechlorinated product yield means total amounts of the other
products except for both CMBA and CS. Compared with the
bare and modified Pt/SiO2 catalysts,11 the Rh/SiO2 catalyst
provided trace amounts of CHEEL and CS as new compounds,
whereas the by-product ethyl α-methylbenzyl ether was barely
produced. This tendency shows a preferable catalytic activity for
the HDC, indicating that ethanol does not react with MBA at all
and promotes the HDC of CLAP on the Rh/SiO2 catalyst.
As a representative example involving an organosilane-
grafted catalyst, Figure 2 shows the conversion and product
yields as a function of reaction time for the HDC reaction of
CLAP over Rh/SiO2­TS. The results obtained for other organo-
silane-grafted catalysts are summarized in Table S1. A con-
version of 62% was achieved in 60 min using the Rh/SiO2­TS
catalyst. The detected products included MBA (27% yield),
CHEL (15% yield), ACH (8% yield), CMBA (5% yield), EB
4
.2%
9
9
5
0
400
600
800
1000
1200
Temparature /K
(
b) DTA
Rh/SiO2−TDFDMS
Rh/SiO2−DMODS
Rh/SiO2−DMOS
Rh/SiO2−BDMS
Rh/SiO2−TS
Rh/SiO2
SiO2 (silica support)
(4% yield), AP (3% yield), CHEEL (1% yield), as well as trace
amounts of ECH and CS, and their dechlorinated yields summed
up to 58%. As well as the bare catalyst, the by-product ethyl
α-methylbenzyl ether was barely detected, although an improve-
ment in the solid catalytic activity is generally liable to
accompany with the undesirable subreaction. These increases
in the reaction conversion and the overall dechlorinated product
yield observed for Rh/SiO2­TS are consistent with the results
obtained with the other modified Rh/SiO2. However, the overall
dechlorination yields, which are indicators of catalyst activity,
decreased in the order: Rh/SiO2­TS (58% yield) > Rh/SiO2­
DMODS (44% yield) > Rh/SiO2­TDFDMS (40% yield) µ Rh/
SiO2­BDMS (40% yield) > Rh/SiO2­DMOS (39% yield) >
4
00
600
800
1000
1200
Temparature /K
Figure 3. (a) TG and (b) DTA curves of bare and modified
Rh/SiO catalysts.
2
results for Pt/SiO2­DMODS.11 The number of organosilane
groups per catalyst surface area unit was calculated using the
methods by Helmy et al.
independent of the organosilane species but agreed well with the
TG weight losses in air. Figure 3a shows the weight loss of Rh/
SiO2­TS as a typical example. All the grafted catalysts showed
exothermic peaks between 500 and 800 K, while the silica
support and Rh/SiO2 barely showed the peaks (Figure 3b),
suggesting that the weight loss resulted from the pyrolysis of the
organosilanes.
1
3,14
15
and Li. This number was
Rh/SiO (28% yield).
2
The textural characteristics of the rhodium catalysts were
investigated (Tables 1 and S2) to further evaluate their differ-
ence. Because both bare Rh/SiO and Rh/SiO ­TS displayed a
2
2
metal loading of 3 wt %, similar metal loading was used for their
modified counterparts. First, the particle size of each rhodium
catalyst increased upon silylation, meaning that the metal
migrated and aggregated during the reaction, similar to previous
Turnover frequencies (TOFs) were calculated using the
metal surface areas derived by carbon monoxide (CO) chem-
isorption. Here, it is noted that the amount of adsorbed CO may
decrease due to the presence of organosilanes adjacent to Rh
Chem. Lett. 2014, 43, 1604–1606 | doi:10.1246/cl.140539
© 2014 The Chemical Society of Japan | 1605

particles, thereby showing slightly higher TOFs for the grafted
Rh/SiO2 catalysts. Based on the reaction conversion at 10 min,
all modified Rh/SiO2 catalysts showed higher TOF than the
and ECH (5%) especially increased in the product distribution.
On the other hand, for lower ethanol compositions (39:1 (v/v)
water/ethanol), reaction conversions and TOFs decreased to
¹1
¹1
bare Rh/SiO catalyst (6.2 min ). Rh/SiO ­TS exhibited the
35% and 11.6 min , respectively, which represents approx-
2
2
¹
1
highest TOF (19.7 min ), which was approximately 3.2 times
that of bare Rh/SiO2. The TOF values of Rh/SiO2­BDMS
10.2 min ), Rh/SiO ­DMOS (9.8 min ), and Rh/SiO2­
DMODS (18.2 min ) were inconsistent with their main carbon
chain length, which contained 4, 8, and 18 carbons, respectively.
Although Rh/SiO ­DMODS (0.3 groups nm ) displayed the
imately half of the values obtained using 7:1 (v/v) water/
ethanol. Also, the overall dechlorinated product yields decreased
to 20%, which correspond to one-third of those observed in 7:1
(v/v) water/ethanol. Therefore, ethanol in the solvent mixture
plays a significant role in the HDC reaction. It (1) lowers the
viscosity of the reaction mixture, (2) expands the hydrophobic
zone, and (3) assists the percolation of substrates through the
hydrophobic zone.
In conclusion, five different water-repellent rhodium cata-
lysts were synthesized by attaching their organosilyl substituents
to the silica support. All the modified catalysts successfully
enhanced the HDC reaction in the water/ethanol mixture using
1 MPa hydrogen at 373 K. This improvement may stem from
the hydrophobic zone created by the organosilyl substituents,
retaining many CLAP molecules near the catalyst active sites. In
¹
1
¹1
(
2
¹
1
¹
2
2
lowest number of organosilane groups compared with the other
two catalysts, its TOF was close to that of Rh/SiO2­TS,
implying that long organosilane carbon chains were somewhat
effective at enhancing the HDC activity. These results indicate
that hydrophobic organosilane on the catalyst support effectively
promote the HDC catalyst activity in the water/ethanol mixture.
Here, the organosilane groups may create a hydrophobic zone
that retains a large amount of CLAP close to the active sites,
enhancing the catalytic activity. This tendency is similar to cases
in which Pt/SiO catalysts were grafted on BDMS, DMOS, and
particular, the Rh/SiO ­TS catalyst showed high performance
2
2
11
DMODS. To verify this hypothesis, Rh/SiO2 was grafted
on TDFDMS, a fluorinated organosilyl reagent with a higher
hydrophobicity. Contrary to expectation, the dechlorination
activity and TOF of Rh/SiO2­TDFDMS amounted to 40% and
in the HDC reaction. This behavior may also result from the (1)
numerous 4-tolylsilyl substituents grafted on the support and (2)
aromatic π­π interactions between these 4-tolylsilyl substituents
and the aromatic ring of CLAP. All organosilane-grafted
catalysts prevented the by-product formation of ethyl α-
methylbenzyl ether from ethanol and MBA, meaning that
overcame the weakness for the modified Pt/SiO2 catalysts.
Detailed studies are underway to uncover the mechanism
involving this HDC catalyst.
¹
1
7
.5 min , respectively. Therefore, low activity for the Rh/
SiO2­TDFDMS catalyst may stem from the lower organosilane
¹
2
density on the support (0.3 group nm ) and the considerable
lipophobicity of the organofluorinated terminal configuration
because of the hindered percolation of CLAP and ethanol
near the metal active site in the organic zone. Therefore, an
appropriate organosilyl reagent is required to promote the HDC
reaction over organosilane-grafted Rh catalysts.
Supporting Information is available electronically on J-STAGE.
Rh/SiO ­TS showed the highest activity. This phenomenon
References
2
may originate from its high organosilane number (2.3
1
2
G. S. Pozan, I. Boz, J. Hazard. Mater. 2006, 136, 917.
E. Díaz, J. A. Casas, Á. F. Mohedano, L. Calvo, M. A.
Gilarranz, J. J. Rodrígues, Ind. Eng. Chem. Res. 2009, 48,
3351.
¹2
groups nm ), which facilitates the HDC reaction. To better
understand this activity, the reaction intermediates CMBA,
MBA, AP, and ACH were hydrotreated over Rh/SiO2­TS using
1
MPa hydrogen at 373 K for 60 min. Reaction conversions
3
4
5
6
C. Xia, Y. Liu, S. Zhou, C. Yang, S. Liu, J. Xu, J. Yu, J.
Chen, X. Liang, J. Hazard. Mater. 2009, 169, 1029.
D. Angeles-Wedler, K. Mackenzie, F.-D. Kopinke, Environ.
Sci. Technol. 2008, 42, 5734.
F.-D. Kopinke, D. Angeles-Wedler, D. Fritsch, K.
Mackenzie, Appl. Catal., B 2010, 96, 323.
and product distributions are listed in Table S3 (Supporting
Information). TOFs amounted to 18.9, 11.5, 36.7, and 1.1 min
¹
1
for CMBA, MBA, AP, and ACH, respectively, clearly showing
that the aromatic substrates (CMBA, MBA, and AP) smoothly
converted to their hydrogenation and/or hydrogenolysis deriv-
atives, whereas ACH barely reacted. Moreover, when cyclo-
hexanol was hydrotreated over Rh/SiO2­TS under the same
conditions, no conversion occurred. These results indicate that,
in addition to a hydrophobic zone on the catalyst surface, the
high activity of Rh/SiO2­TS depends on aromatic π­π inter-
actions between the 4-tolylsilyl substituent on the catalyst and
the aromatic ring of CLAP. Also, the steric hindrance of the
cyclohexyl ring might prevent ACH from approaching the
hydrophobic zone.
Finally, the effect of the ethanol ratio in the solvent mixture
on the HDC reaction was examined over Rh/SiO2­TS (Table S4,
Supporting Information). When the ethanol composition in-
creased to 3:1 (v/v) water/ethanol, all reaction conversions,
overall dechlorinated product yields, and TOFs increased up to
R. Navon, S. Eldad, K. Mackenzie, F.-D. Kopinke, Appl.
Catal., B 2012, 119­120, 241.
7
8
R. Nakao, H. Rhee, Y. Uozumi, Org. Lett. 2005, 7, 163.
Y. Uozumi, R. Nakao, H. Rhee, J. Organomet. Chem. 2007,
692, 420.
9
Y. Uozumi, Y. M. A. Yamada, Chem. Rec. 2009, 9, 51.
10 Y. Lan, L. Yang, M. Zhang, W. Zhang, S. Wang, ACS Appl.
Mater. Interfaces 2010, 2, 127.
11 T. Yoneda, T. Aoyama, T. Takido, K. Konuma, Appl. Catal.,
B 2013, 142­143, 344.
12 K. Konuma, N. Kameda, J. Mol. Catal. A: Chem. 2002, 178,
239.
13 R. Helmy, A. Y. Fadeev, Langmuir 2002, 18, 8924.
14 S. Marcinko, A. Y. Fadeev, Langmuir 2004, 20, 2270.
15 T. Li, J. Chromatogr. A 2004, 1035, 151.
¹
1
7
2%, 68%, and 27.3 min , respectively. Yields for EB (9%)
1606 | Chem. Lett. 2014, 43, 1604–1606 | doi:10.1246/cl.140539
© 2014 The Chemical Society of Japan