Catalysis Letters 45 (1997) 233^235 233

Selective dehydrogenation of cyclohexene to benzene using

Pd-exchanged -zirconium phosphate
R.B. Borade, B. Zhang and A. Clearfield
Department of Chemistry, Texas A&M University, College Station, TX 77843, USA

Received 14 January 1997; accepted 5 March 1997

The dehydrogenation of cyclohexene to benzene with  85% cyclohexene conversion and > 80% benzene selectivity has been
achieved over Pd-exchanged -zirconium phosphate at 200 C.
Keywords: dehydrogenation, cyclohexene, zirconium phosphate, MCM-41, zeolite beta

Catalytic dehydrogenation is an important step in 100 h. This treatment produces a product of high crystal-
the reforming process. In the reforming process n-paraf- linity which shall be referred to as -ZrP. Pd exchange of
fin molecules are mainly dehydrogenated to produce ole- -ZrP was carried out as follows. A weighed amount of
fins, diolefins and triolefins which then cyclize and dried solid -ZrP was slurried with propylamine
finally produce the aromatics [1,2]. The oxidation of (3 mmol/g) at room temperature for two days using an
cyclohexene has been used as a model reaction by many ultrasonic bath. The solid product was recovered by fil-
investigators to evaluate the catalyst [3^7]. Cation- tration and washing with deionized water and dried at
exchanged zeolites have been known to exhibit high cat- room temperature. To this solid product was added
alytic activity for oxidative dehydrogenation of cyclo- 0.1 N Pd(NH3 )4 Cl2 solution in water. The mixture was
hexane and cyclohexene [8^11]. However, selectivity of again ultrasonicated at room temperature for two days.
the reaction was found to be dependent on the nature of The yellowish solid product was obtained after washing
the ion-exchanged metals. For example, the most active and drying. As expected, incorporation of Pd(NH3 )4 Cl2
catalyst was Cu(II)-Y for benzene formation (80% selec- complex in -ZrP resulted in an increase in the interlayer
tivity and 10% conversion) whereas Pd(II)-Y was best distance from 7.6 to 11.6 Ð. The XPS analysis indicated
for carbon dioxide formation [8]. that Pd is present as Pd(II) and there are 4 nitrogen and
Cheung and Clearfield [12] reported the oxidative 1.6 chlorine atoms for each atom of Pd suggesting that
dehydrogenation of cyclohexene to benzene over the the Pd(NH3 )4 Cl2 complex is intact.
Cu(II) form of zirconium phosphate, CuZr(PO4 )2 . It The cyclohexene dehydrogenation reaction was car-
was observed that, in the absence of oxygen, dehydration ried out in a conventional glass reactor with 10 mm
takes place with reduction of Cu(II) to Cu metal forming diameter. A 50 mg sample of Pd-exchanged -ZrP was
benzene in the process. A similar hydrogen reduction loaded in a reactor which was supported on glass wool.
and dehydrogenation study indicated that the copper Prior to reaction the catalyst was calcined in situ at
metal once formed was reoxidized to Cu(II) in air and 200 C for 6 h in a N2 atmosphere. After activation the
then diffused back into the zirconium phosphate reduc- N2 flow was diverted to a saturator filled with cyclohex-
ing O2 to H2 O in the process. In the case of an erionite ene and then passed through the reactor. The saturator
catalyst [11] using an oxygen atmosphere, the conversion was maintained at different temperatures to yield
of cyclohexane at 450 C was 11.4% with 54% selectivity desired partial pressures. Analysis of the products was
to benzene whereas in the case of Cu(II)-ZrP 65^90% carried out by on-line gas chromatography (Varian
conversion with 65% selectivity in the temperature range 3400) using a J&W Scientific 50 m Alumina PLOT
260^350 C was achieved [12]. However, in the absence capillary column and FID detector. The outlet lines
of O2 , Cu(II)-ZrP showed a significant decrease in cat- were heated to ensure no product condensation within
alytic activity from 65% to 10% within 60 min of reaction the lines. Mass balances were within 2% of closure,
time. In the present case we report  85% cyclohexene i.e. the moles of carbon in equals to moles of carbon
conversion with > 80% benzene selectivity at 200 C out.
using Pd-exchanged -zirconium phosphate in the It was observed that the pretreatment/activation of
absence of oxygen. The catalyst was tested for one week the catalyst is an important step in determining the cat-
and showed no sign of deactivation. alytic activity of Pd-ZrP. Activation of Pd-ZrP in a N2
A zirconium phosphate gel, prepared by the method atmosphere for less than 6 h resulted in less active mate-
of Clearfield et al. [13], was refluxed in 12 M H3 PO4 for rial whereas the higher activity remained more or less the

Ä J.C. Baltzer AG, Science Publishers

234 R.B. Borade et al. / Cyclohexene dehydrogenation over Pd-ZrP

to be 8.5 kcal/mol [12]. This lower Ea may explain why

Pd-exchanged -ZrP is more active than Cu-exchanged
-ZrP. Davis and Somorjai reported [4] an apparent
activation energy for the same reaction of at least 8 kcal/
mol at 77 Torr but zero at low pressures in the tempera-
ture range 25^150 C for Pt supported on SiO2 surface.
In table 1 we have compared catalytic data of Pd-
ZrP, Cu-ZrP and Pd-exchanged beta (microporous) and
Pd-exchanged MCM-41 (mesoporous) catalysts for the
cyclohexene dehydrogenation reaction. The protonated
forms of MCM-41 and beta were obtained using proce-
dures described elsewhere [15^17]. A calculated amount
of Pd(NH3 )4 Cl2 solution in water was used to obtain
0.4% Pd-loaded MCM-41 or beta zeolite. From table 1,
It is clearly seen that Pd-ZrP is the most active and selec-
tive catalyst for the production of benzene. It also pro-
duces cyclohexane by hydrogenation. Note that
Figure 1. Cyclohexene conversion over Pd-ZrP at various tempera- benzene-to-cyclohexane (dehydrogenation/hydrogena-
tures. tion ratio) is > 4:5 and per cent deactivation is 1% < in
4 h. All silica Pd-MCM-41 also showed very high activ-
same for the catalyst pretreated for 6^12 h. The Pd-ZrP ity with only 8% deactivation. However, this catalyst
at 200 C (W =F ˆ 0:0185 min) gave cyclohexene conver- gave a benzene/cyclohexane ratio of 2. Although, in the
sion near 86%. Total conversion of cyclohexene as a case of Cu-ZrP the benzene/cyclohexane ratio was com-
function of contact time in the temperature range 120^ parable with that of Pd-ZrP, the activity of this catalyst
200 C is shown in figure 1. Dehydrogenation of cyclo- was much lower and the deactivation rate was much
hexene to benzene as well as hydrogenation of cylohex- higher. In addition to benzene and cyclohexane it pro-
ene to cyclohexane occurred simultaneously. No other duced significant quantities of 1,3-cyclohexene. On Pd-
products such as 1,3- or 1,4-cyclohexadiene were beta and Pd-MCM-41(Si=Al ˆ 15) not only benzene
detected. The formation of benzene from cyclohexene and cyclohexane but also C6 isomeric products such as
without forming 1,3-cyclohexadiene over Pt complexes methylcyclopentenes are formed. The activity of these
and Pt supported catalysts is reported in the literature catalysts is much lower than Pd-ZrP. It is known that
[11,14]. From the slope of the kinetic curves of cyclohex- beta and MCM-41 contain Brnsted acid sites resulting
ene conversion against contact time in figure 1, initial in formation of methylcyclopentenes in addition to ben-
rates were obtained. The apparent activation energy (Ea ) zene and cyclohexane. On Pd-MCM-41(Si=Al ˆ 15),
for cyclohexene conversion was 1.9 kcal/mol. Because isomerization is a main reaction, followed by hydroge-
the apparent activation energy value was very low, sev- nation and then dehydrogenation reactions, whereas on
eral experiments were carried out in this temperature Pd-beta, hydrogenation and isomerization reactions are
range to check the reproducibility of the results. Again, taking place simultaneously. Again the rate of deactiva-
Ea values were found to be within the same range of tion is very high for both catalysts. The benzene/cyclo-
experimental errors. The apparent activation energy for hexane ratio is 0.5 and 0.04 for Pd-MCM-
the same reaction over Cu-exchanged -ZrP was found 41(Si=Al ˆ 15) and Pd-beta, respectively.

Table 1
Cyclohexene dehydrogenation activity of various Pd-exchanged catalystsa

Catalyst Reaction Initial Deacti- Product distribution (wt%)

temp. ( C) con. vation b

(wt%) (%)
isomers
(mol)

Pd-ZrP 200 86.3 1< 19.0 ^ ^ 80.0 4.5

Cu-ZrP c 260 12.0 99 18.0 ^ 13.0 69.0 4.1
Pd-MCM-41 (all silica) 200 99.8 8 34.7 ^ ^ 65.2 2.0
Pd-MCM-41 (Si=Al ˆ 15) 200 25.8 96 22.8 66.6 ^ 10.5 0.5
Pd-beta (Si=Al ˆ 13) 200 37.0 97 58.4 39.4 ^ 2.1 0.04
a
Catalyst ˆ 50 mg; activation temp: ˆ 200 C/6 h; PCHE ˆ 23:9 mm; N2 ‡ cyclohexene flow ˆ 27 ml/min.
b
Defined as (A0 ÿ At =A0 †100, where A0 is activity at time 2 min and At is activity at time 4 h.
c
Data taken from ref. [12].