Table 2 Rate of 2,2,4-trimethylpentane cracking
This Ðnding suggests that zirconium may be deposited on the
external surface of the zeolite particles and/or located inside
the micropores of the material. The results obtained by mea-
suring the pore volume through n-hexane absorption clearly
indicate that for Zr-SAPO-11 the pore volume is reduced by
an important factor since the pore volume of Zr-SAPO-11 is
nearly 1/3 of that of SAPO-11. Since the XRD results indicate
that both solids are highly crystalline, the decrease of the pore
volume is probably due to partial pore blocking because of
the presence of zirconium oxide inside the micropores.
The catalytic results, in particular the initial catalytic
properties for n-butene isomerization, are in good agreement
with such a picture; it has been shown recently10 that for
medium pore zeolites or aluminophosphates, used for n-
butene isomerization, there is a good correlation between the
selectivity to isobutene and the pore diameter.
rate of 2,2,4-trimethylpentane
cracking/lmol h~1 (g solid)~1
sample
SAPO-11
Zr-SAPO-11
0.28
7 ] 10~3
T \ 723 K, P(trimethylpentane) \ 1.6 kPa, complement to atmo-
spheric pressure being nitrogen.
Therefore, the improved selectivity of Zr-SAPO-11 is attrib-
uted at least in part to the partial blocking of the pore by
zirconium oxide.
Catalytic testing of 2,2,4-trimethylpentane cracking has
shown that the number of acid sites located on the external
surface of the individual crystals is much larger for SAPO-11
than for Zr-SAPO-11. Since it is known that these external
sites are not selective for n-butene skeletal isomerization, the
improved isobutene selectivity of Zr-SAPO-11 sample can be
also attributed to this phenomenon.
It is observed that the stability with time on stream (TOS) is
increased by the addition of zirconium: after 6 h on stream,
SAPO-11 reacted with the same amount of reactant as the
zirconium-based sample after 12 h, but the conversion is much
less decreased on Zr-SAPO-11 catalyst.
Di†erent factors may be at the origin of this improvement
in stability with TOS. (1) A particle size e†ect. For a unidi-
mensional pore system like SAPO-11, the size of the individ-
ual SAPO grains are of importance because it is very likely
that the pore blockage, due to carbonaceous deposits, would
be more severe for large grains compared to small grains.
Examination of the SEM results indicates that the individual
Zr-SAPO-11 grains are smaller than the SAPO-11 crystals.
So, this could explain, at least partially, the better stability of
the Zr-SAPO-11 catalyst. (2) E†ect of external acid sites.
Another factor which could inÑuence the stability is the initial
selectivity compared to SAPO-11, this having been attributed
to higher pore constraints and to a lower number of external
acid sites. Thus, the higher the selectivity, the lower the forma-
tion of butene oligomers which could be at the origin of the
coke formation. Indeed, another important factor would be
the lower number of external acid sites for Zr-SAPO-11 com-
pared to SAPO-11: these external acid sites are probably at
the origin of the ageing of the catalyst: formation of coke on
the external surface of the grain would induce pore blockage
and thus would decrease the activity of the catalyst. (3) A pore
constraint e†ect. This e†ect will render the formation of car-
bonaceous deposits inside the pores more difficult, which will
contribute to the ageing of the catalyst.
Scheme 1
products are isobutane and a mixture, at the thermodynamic
equilibrium, of butenes (n and iso); this result is as expected
because of the following reaction scheme for 2,2,4-tri-
methylpentane cracking (Scheme 1) the reactivity of iC
being such in these solids that the following thermodynamic
4/
equilibrium is reached:
isobutene H` ] isobutene, cis/trans-but-2-ene and but-1-ene.
It appears from Table 2 that SAPO-11 has a larger number
of external acid sites than Zr-SAPO-11.
Since it is known that these external acid sites are not selec-
tive for n-butene isomerization into isobutene,14 this suggests
that part of the improvement in the initial isobutene selec-
tivity for Zr-SAPO-11 may be due to this phenomenon. It is
not known if this decrease in the number of acid sites for Zr-
SAPO-11 is due to a preferential location of silicon inside the
micropores or if the external acid sites are covered by ZrO
x
entities.
Discussion
As shown by the XRD spectra, the two solids under study,
SAPO-11 and Zr-SAPO-11, exhibit the AEL structure, and do
not show any contamination by other crystallized phases.
Scanning electron microscopy results do not indicate large dif-
ferences between the morphologies of the two solids but it
appears that the sizes of the Zr-SAPO grains are somewhat
smaller than those of the SAPO-11.
Characterization of the acidity, either by using IR spectros-
copy and/or by using ammonia TPD does not show large
modiÐcations due to the addition of zirconium: the number of
acid sites, as evidenced by the IR vibration at 3630 cm~1 and
by the second ammonia TPD peak, is smaller on Zr-SAPO
than on SAPO-11.
These three factors would explain the lower ageing rate of
Zr-SAPO-11 compared to SAPO-11.
Taking into account the results of elemental analysis and
those obtained by NMR (70% of the Si in Zr-SAPO-11 create
one proton), it can be calculated that the number of BrÔnsted
acid sites in Zr-SAPO-11 is 70% of the number of BrÔnsted
sites in SAPO-11, in good agreement with the IR and
ammonia TPD results, which gave values of 60 and 50%,
respectively.
Examination of the IR spectra and/or of the ammonia TPD
results indicates that the presence of zirconium has not modi-
Ðed the acid strength of the acid sites of SAPO-11. These
results, at this stage suggest that the zirconium is not incorp-
orated in the framework of SAPO-11, at least in a signiÐcant
manner, because we fail to detect any new type of acid center.
Conclusions
SAPO-11 and Zr-SAPO-11 have been synthesized. Character-
ization results indicate that zirconium is not incorporated into
the SAPO framework but is localized, at least partially, in the
pores of the solid, inducing a partial pore blockage. This
partial pore blockage is at the origin of the increase in iso-
butene selectivity and is one of the causes of the improvement
of catalyst stability with time on stream. In addition, a very
small number of external acid sites of Zr-SAPO-11 as com-
pared to SAPO-11 could also explain the improvement in iso-
butene selectivity and stability with time on stream.
J. Chem. Soc., Faraday T rans., 1997, V ol. 93
4205