Applied Catalysis, 31(1987) 113-118 113

Elsevier Science Publishers B.V., Amsterdam - Printed in The Netherlands

CATALYTIC REFORMING OF N-HEPTANE ON PLATINUM, TIN AND PLATINUM-TIN SUPPORTED ON

ALUMINA

J. BELTRAtlINIa and D.L. TRIMM

School of Chemical Engineering and Industrial Chemistry, University of New South
Wales, PO Box 1, Kensington, NSW 2033, Australia.
a0n leave from INCAPE (Instituto de Investigaciones en Catalisis y Petroquimica,
Santiago de1 Estero 2654, 3000 Santa Fe, Argentina.

(Received 26 August 1986, accepted 20 January 1987)

ABSTRACT
The catalytic reforming of n-heptane over Pt/A1203, Sn/A1203 and Pt-Sri/A1203
has been studied. Distinct activity and selectivity changes have been observed
during the operation of each catalyst. The use of the bi-metallic catalyst
results in the control of coke deposition, increased stability and improved
aromatics yield. These effects can be interpreted in terms of a change of pro-
perties of Pt that results from interaction with Sn, involving electronic modifi-
cation, ensemble effects or both.

INTRODUCTION
It is well recognised that the introduction of a second metal (Re, Ir, Sn, etc.)
can have a beneficial effect on the performance of a platinum on alumina reforming
catalyst [t], improving both selectivity and stability of the solid. In this group
of catalysts, the Pt-Sri/Al203 system is of particular interest, although the
role of Sn is far from clear. The addition of tin is known to enhance aromatisation
[2,3], inhibit hydrocracking and isomerisation [4] and decrease coke formation [5].
Tin has been suggested to inhibit acid catalysed reactions by depositing on the
most acidic sites of the support [6] and/or to form a solution with platinum to
produce ensembles that do not favour the production of carbonaceous deposits
[7,8]. Overall the effect is to reduce deactivation and to increase selectivity

[91.
As part of a study of bimetallic reforming catalysts, attention has been focused
on the performance of Pt/A1203, Sn/A1203 and Pt-Sn/A1203. The present paper
reports studies of the performance of these catalysts for the catalytic reforming
of n-heptane.

EXPERIMENTAL
Catalysts were prepared by impregnation of y-alumina (Norton SA 6176) with
solutions of H2PtC16 and/or SnC12 and with 0.6 M HCl. The catalysts were dried
(393 K: 12 h) and calcined in air at 773 K for 7 h. The solids were reduced by

0166-9834/87/$03.50 0 1987 Elsevier Science Publishers B.V.

114

TABLE 1
Properties of catalysts.

Composition/% Surface area

Catalyst Pt Sn Cl /m2 g-'

Pt/Al203 0.5 0.94 196

Sn/A1203 0.5 0.91 195

Pt-Sn/A1203 0.5 0.5 0.84 199

0.92 201
*'2'3

hydrogen (773 K: 12 h) before use. The essential properties of the catalysts are
reported in Table 1.
Catalyst testing was carried out using a flow system involving 2 g catalyst
maintained at pre-set temperature [TO]. Operating conditions included temperature:
773 K, pressure: 5 bar, WHSV: 2 h-' , molar ratio H2/C7H,6 = 3, length of run: 6 h.
Conversion was defined as moles of n-heptane reacted per mole introduced x 100%.
Selectivity was defined in terms of the products from a particular reaction per
mole of feed introduced.
The carbon contents of catalysts were determined by using the DuPont 952 Thermal
Analyser. Temperature programmed oxidation involved an increase from 343 to 773 K
-1 -1
at 20 K min in oxygen at 50 cc min ,

IOO-ollra-aPt-Sn/;Ka c=
0.68%

2 go o~o-04
--o-o 2.31%
z Pt /A$03
; 80-

I I I
0 100 200 300 )O
TIME on line (mid TEMPERATURE,"C

FIGURE 1 The conversion of n-heptane on mono- and bi-metallic catalysts. Standard

operating conditions given in text.

FIGURE 2 Temperature programmed oxidation of used catalysts.

 115

20
CH, C2H, &ho bH12

FIGURE 3 Distribution of C,-C4 products from reforming over fresh and used cata-
lysts. Standard operating conditions. Initial = zero time on line. Final = 360 min

on line.

RESULTS AND DISCUSSION

Conversion was measured as a function of time on line for the three catalysts
to determine the extents of deactivation (Figure 1). Initial deactivation of Pt/
Al203 was fast, but conversion tended to level out after ca. 4 h on line. Sn/A1203
showed some deactivation but overall conversion was low. Pt-Sri/Al203 showed high
and constant conversion throughout the run. Carbon contents of the used catalysts,
also reported in Figure 1, show that deactivation appears to be connected to the
formation of carbonaceous deposits on the solids. Similar results have been re-

ported by Sexton et al. [II].

Comparison of the moles of coke formed per total moles of n-heptane converted
(Pt/A$03 = 5.08 x 10 -', Sri/Al203 = 6.19 x IO-' , Pt-Sn/A1203 = 1.37 x IO-') show

that about the same selectivity to coke is found on the monometallic catalysts
but that much less coke is formed on the bimetallic catalyst. On the other hand,
the total amount of coke formed on Pt-Sri/Al203 was much higher than on Sn/A1203,
despite the fact that the deactivation was much less (Figure 1). This reinforces
the finding that it is not only the amount of coke but also the location of the
coke that has most effect on activity. Temperature programmed oxidation studies
of coke on the different catalysts showed that one peak maxima for combustion
of coke on tin containing catalysts occurred at approximately the same temperature
as that produced by combustion of coke on alumina (Figure 2). As a result, it
would appear that most of the coke is formed on the support with less coke, com-
pared with Pt/A1,03 being formed on the metal. Coke formed on Pt/A1203 is less
polymerised 1131, is more reactive (Figure 2) and tends to form more on the metal
site (Figure 2).
Attention was then focused on the selectivity of the catalysts for different
reactions (Figures 3 to 5). Hydrocleavage reactions produce C,-C4 products and
it is known that the relative efficiency of hydrogenolysis to hydrocracking can
116

100 ,
1
I Pt/A1203
I

70- - W/A1203
--- Sn/A1203

2 6O-
y: BENZENE
E
l

2 50-
2

. 9 co-
z
I w
>

O/f----- 30-

Pt -Sn/Al203 20-
X_-__X__-X---X__~__
_X__)(___X----X---X---

0 I I I
0 100 200 300 0 100 200 300
TIME (mid TIME (min)

FIGURE 4 The importance of different reactions during reforming on mono- and bi-
metallic catalysts. 0 Yields of hydrocracking products, x yields of isomerisation
products, A yields of aromatisation products, + yields of cyclisation products.

FIGURE 5 Aromatic products produced on different catalysts.

be assessed from the C,:C3 ratio [14]. In the fresh catalysts, the addition of
tin to PtIA1203 is found to decrease hydrogenolysis and to increase slightly the
hydrocracking activity (Figure 3). Deposition of coke reduced hydrogenolysis on
all catalysts and hydrocracking on Pt-Sn/A1203 but increased hydrocracking on
Pt/A1203 and A1203.
Isomerisation was found to be important on Pt/A1,03 and Sn/A1203 (Figure 4)
and to increase with time on line. The reaction was unimportant on Pt-Sn/A1203.
By contrast, dehydrocyclisation was very important on Pt-Sn/A1203, low on Sn/A1203
and passed through a maximum on Pt/A1203 (Figure 4). Toluene was the main product
over all catalysts (Figure 5).
It is seen that the main effect of addition of Sn to Pt is to reduce coke
formation and to increase selectivity for dehydrocyclisation. This effect is
not due merely to the presence of the second metal, since the activity/selectivity
 117

patterns of the monometallic catalysts are very different from those of the bi-
metallic catalyst, but must result from interactions between the two metals.
Most coke per mole of heptane converted is formed on Sri/A1203 (Figure 1) and
the coke is deposited mainly on the support (Figure 2). Hydrocracking and isomeri-
sation, favoured by acidic sites, remain high despite the formation of carbon
(Figure 4). As a result, there is little support for the suggestion that tin
poisons the most acidic sites on the catalyst [6].
The maximum in the yield of toluene on Pt/Al203 contrasts with the high and
constant yield over Pt-Sri/Al203 (Figure 5). Decreased yields observed on PtlA1203
as a function of time on line are not due to further reactions of toluene to
produce benzene and xylene since the amounts of the latter compounds produced are
too small (Figure 5). Toluene yield decrease must, then, be due to catalyst
deactivation. Coke is formed on Pt/A1203 on the metal and the acid function (Figure
Z), both of which are needed to produce aromatics [15].
The role of the second metal is thus seen to involve the production of a new
catalytically active site that possesses different selectivity directing properties
and remains coke-free for long periods. Part of the ability to decrease coke
formation may result from enhanced gasification of coke precursors by tin [16,17].
However, the changed selectivity pattern, and possibly the reduced coke formation,
must result from Pt-Sn interactions [18,19,20,21]. Biloen et al. [ZO] have
suggested that tin delineates ensembles on the surface that favour reforming but
do not favour coking, and Ponec et al. [21] note that hydrogenolysis and coking
require larger ensembles than dehydrocyclisation (Figure 4) at the expense of
hydrogenolysis (Figure 3) and coking (Figure 2). Thus it would seem that the
enhanced selectivity pattern, and possibly the reduced coking, is due primarily
to ensemble size control on the surface of the metal.

ACKNOWLEDGEMENTS
JB wishes to thank CONICET (Consejo National de Investigaciones Cientificas
y Tecnicas) of Argentina for financial support. The support of the Australian
Research Grant Scheme is also gratefully acknowledged.

REFERENCES
1 R.R. Cecil, W.S. Kmak, J.H. Sinfelt and L.W. Chambers, Oil & Gas J., 7 (1972)
50.
2 F.M. Dautzenberg, German Offenlegungsschrift, 2.121.763 (1971).
3 F.M. Dautzenberg and H.W. Kouwenkoven, German Offenlegungsschrift, 2.153.891
(1972)
4 B.H. Davis, G.A. Westfall, J. Watkins and J.J. Pezzanite, J. Catal., 42 (1976)
247.
5 J.K.A. Clarke, Chem. Oev., 75 (1975) 391.
6 R. Baccaud, P. Bussiere and F. Figueras, J. Catal., 69 (1981) 399.
7 F.M. Dautzenberg, J.N. Helle, P. Biloen and W.M.H. Sachtler, J. Catal., 63
(1980) 119.
8 F.G. Gault, 0. Zahraa, J.M. Dartigues, G. Maire, M. Peyrot, E. Weisang and
P. Engelhard, Proc. 7th Int. Congress on Catalysis, Tokyo, 1980, p. 199.
118

9 H.C. De Jongste and J.R. Anderson, J. Catal., 61 (1980) 140.

IO J.N. Beltramini, D.L. Trimm and M.S. Wainwright, Chemeca 86, Adelaide,
Australia, 1986, p.471.
11 B.A. Sexton, A.E. Hughes and K. Foger, J. Catal., 88 (1984) 466.
12 M. Masai, K. Honda, A. Kubota, S. Ohnaka, Y. Nishikawa, K. Nakahora, K. Kishi
and S. Ideda, J. Catal., 50 (1977) 419.
13 J. Barbier, G. Corro, Y. Zang, J.P. Bournouville and J.P. Franck, Applied
Catalysis, 13 (1985) 245.
14 J.L. Carter, J-A. Cusumano and J.H. Sinfelt, J. Catal., 20 (1971) 223.
15 E. Christoffel, F. Fetting and H. Vierrath, J. Catal., 40 (1975) 349.
16 D.A. Dowden, Chem. Sot, Spec. Pub. Catalysis, 2 (1978) 1.
17 M.R. Hatswell, W.R. Jackson, F.P. Larkins, M. Marshall, D. Rash and D.E. Rogers,
J. Catal., 59 (1980) 442.
18 J.H. Sinfelt, Accounts Chem. Res., IO (1977) 15.
19 H. Berndt, H. Mehmer, J. Volter and W. Meisel, F. Anorg. Allg. Chem., 429
(1971) 47.
20 P. Biloen, J.N. Helle, H. Verbeek, F.M. Dautzenberq and W.M.H. Sachtler,
J. Catal.; 63 (1980) i12.
21 V. Ponec and H.C. De Jongste, Stud. Surf. Sci. Catal., 7 (1980) 186.