A. Alazman et al. / Journal of Catalysis 357 (2018) 80–89
83
mole of n-hexane converted and quoted in mole per cent. The
mean absolute percentage error in conversion and selectivity was
ꢀ5% and the carbon balance was maintained within 95%. Reaction
ꢂ1
rates (R) were determined as R = XF/W (in mol gꢂ1
h
), where X is
cat
the fractional conversion of n-hexane. In most cases, the reaction
was carried out at differential conditions (X ꢀ 0.1), where X is
directly proportional to the reaction rate. In some cases, the cata-
lysts were diluted with silica in order to achieve low conversion.
3. Results and discussion
3.1. Acid-catalysed isomerisation of n-hexane
Fig. 2. Plot of ln(TOF) (TOF in hꢂ1) versus
catalysed by HPA catalysts (0.20 g catalyst, 200 °C, 5.78 kPa n-hexane partial
pressure, 20 mL minꢂ1 H2 flow rate): 15%HPW/TiO2 (1), 15%HSiW/SiO2 (2), 15%
HPW/SiO2 (3), Cs2.25PW (4), CsPW (5), HSiW (6), HPW (7).
D
HNH3 for n-hexane isomerisation
Bulk and supported heteropoly acids HSiW and HPW were
found to have small activity in n-hexane isomerisation. Bulk acidic
Cs salts of HPW, i.e. CsPW and Cs2.25PW, showed better activities in
this reaction (Table 3) despite their weaker acidity compared to the
bulk HPW (Table 1). This can be attributed to the larger surface
area hence larger proton site density of the Cs salts (Table 1).
2MP and 3MP were the main reaction products, which formed with
65–67% and 29–31% selectivity, respectively, together with 3–5% of
cracking products (mainly C3–C5 hydrocarbons). Double-branched
23DMB isomer was formed in less than 1% selectivity. Strong cat-
alyst deactivation was observed, which can be assigned to coke
deposition. Initially white, CsPW catalyst turned black after reac-
tion, with a carbon content of 0.6 wt% as determined by combus-
tion chemical analysis for the reaction at 180 °C after 6 h time on
stream (Table S1 in the Supporting Information). As seen from
the time course of reaction with CsPW at 200 °C (Fig. 1), n-
hexane conversion is strongly affected by catalyst deactivation,
however without changing reaction selectivity. Practically the
same results were obtained when using N2 instead of H2 as the car-
rier gas.
The rates of acid-catalysed isomerisation of n-hexane (R) and
turnover frequencies (TOFH) per surface proton site were calcu-
lated using the values of n-hexane conversion for 1 h time on
stream (Table 3). The required densities of surface proton sites
are given in Table 1; these were estimated as described elsewhere
[14]. For supported HPA catalysts, which contained HPW or HSiW
at a sub-monolayer loading of 15%, all HPA protons were assumed
to be equally available for reaction. For bulk HPW, HSiW and Cs
salts of HPW, the number of surface proton sites was calculated
using a Keggin unit cross section of 144 Å2 [6,7] and the catalyst
surface areas from Table 1.
Table 3
Acid-catalysed isomerisation of n-hexane.a
The TOFH values obtained ranged from 0.1 hꢂ1 for 15% HPW/
TiO2 to 5.2 hꢂ1 for bulk HPW (Table 3) indicating a strong effect
of catalyst acid strength on the reaction turnover rate. Fig. 2 shows
the relationship between the activity of catalysts in n-hexane iso-
merisation, ln (TOFH), and their acid strength represented by the
d
Catalyst
Conversionb
%
103 Rc mol gꢂ1
TOFH
hꢂ1
hꢂ1
CsPW
2.7
2.8
0.38
0.41
5.1
3.7
Cs2.25H0.75PW12O40
(Cs2.25PW)
HPW
initial enthalpy of ammonia adsorption,
DHNH3 (Table 1). Although
0.7
0.10
5.2
HSiW
1.0
0.15
3.6
there is considerable scatter of points, which is probably caused by
catalyst deactivation, this relationship clearly indicates that
Brønsted acid sites play important role in n-hexane isomerisation
over heteropoly acid catalysts as can be anticipated from the reac-
tion mechanism shown in Scheme 1.
15%HPW/SiO2
15%HPW/TiO2
15%HSiW/SiO2
0.34
0.14
0.16
0.049
0.020
0.023
0.32
0.13
0.11
a
200 °C, 0.20 g catalyst, 5.78 kPa n-hexane partial pressure in H2 flow (20 mL
minꢂ1).
b
n-Hexane conversion at 1 h time on stream (mean of two parallel runs).
Reaction rate calculated from R = XF/W, where X is the fractional conversion of
c
3.2. Bifunctional metal-acid catalysed isomerisation of n-hexane
n-hexane, W is the catalyst weight (0.20 g) and F is the molar flow rate of n-hexane
(W/F = 69.2 g h molꢂ1).
d
As expected, Pt/CsPW bifunctional catalysts in the presence of
H2 were more efficient in n-hexane isomerisation than the acid cat-
alyst CsPW. The Pt/CsPW catalysts showed higher catalytic activity
and displayed much less catalyst deactivation compared to CsPW
(Fig. 3). The amount of coke in spent Pt/CsPW catalysts was below
the detection limit after reaction at 180 °C (6 h on stream)
(Table S1). When using N2 as the carrier gas in the absence of H2,
the activity of Pt/CsPW was much lower and strong catalyst deac-
tivation was observed (Fig. S1). The activity of Pt/CsPW increased
with increasing Pt loading (Table 4, entries 1–3, 5 and 6). Thus
5.78%Pt/CsPW gave a tenfold higher n-hexane conversion than
CsPW at 200 °C (entries 2 and 6). The reaction products included
2MP (64–69% selectivity), 3MP (30–32%) and 23DMB (0.8–1.2%)
together with 2–5% of cracking products (mainly C3-C5 hydrocar-
bons) at n-hexane conversion of 4–22%. Similar product selectivity
was observed with CsPW without Pt (Table 4, entries 1 and 2). This
indicates that Pt does not affect b-migration of methyl group in the
protonated cyclopropane intermediate (Schemes 1 and 2). The low
23DMB selectivity can be explained by the low n-hexane conver-
Turnover frequency per surface proton site; proton site densities are given in
Table 1.
Fig. 1. Isomerisation of n-hexane catalysed by CsPW (0.20 g) at 200 °C, 5.78 kPa n-
hexane partial pressure in H2 flow (20 mL minꢂ1).