Isomerization of cyclohexane and hexane over silica-embedded triflate derivatives catalysts

Isomerization of cyclohexane and hexane over silica-embedded tert-butyldimethyl-silyltrifluoromethanesulfonate and lanthanum triflate is reported. Immobilization of these triflate derivatives was found to give materials exhibiting at least similar catalytic performances to the free molecules. Using these catalysts, the reactions occured at considerably lower temperatures than those prevailing in industrial processes. The catalysts could be recovered and reused. Traces of water had a favorable influence on the reaction, but led to more cracking to isobutane. Larger amounts of water caused catalyst deactivation


Introduction
The use of strong acid catalysts in organic synthesis was introduced by the work of Friedel and Crafts 1 in alkylation of aromatic hydrocarbons.Afterwards, Nenitzescu 2 was the first who reported the isomerization of paraffins over the same kind of catalysts.He showed that aluminum chloride was able to isomerize hexane in the presence of traces of water.
Since then, isomerization of n-and cyclo-paraffins has found an increased interest because of practical applications mainly in the refinery and petrochemical industry, but also for the synthesis of fine chemicals.Most interesting to industry are the processes, which lead to the formation of the more highly branched, higher octane and dimethylbutane isomers.The development of the process followed two directions, the one in heterogeneous catalysis and the other one in homogeneous catalysis.Because these reactions are favored at low temperatures, catalysts active at low temperatures are desirable.4][5] Pt supported on mesoporous aluminosilicate MCM-41 materials were also found to exhibit catalytic activity in this reaction. 6ore recently, other new solid acid materials have been reported to be active catalysts in this reaction even at temperatures lower than 520 K, which is typical for Pt/zeolites.Sulfated zirconias have been investigated mainly as catalysts for the acid catalyzed isomerization of light alkanes at relatively low temperatures. 7,8The most active catalyst in this family is sulfated zirconia doped with Fe and Mn oxides.This catalyst is able to operate n-butane isomerization above room temperature, 9,10 but it quickly deactivates.Addition of Pt to sulfated zirconia increases the time stability. 11][14][15] In homogeneous catalysis, the liquid superacids based on HF are very effective catalysts, able to carry out selectively this reaction at room temperature. 16,17 rifluoromethanesulfonic acid was also reported as an efficient homogeneous catalyst for this reaction.It was tested in butane, 18 isobutane, 19 and 3-methylpentane 20 isomerization.
Both kinds of processes have some inconveniences.Heterogeneous catalysts require rather high temperatures while most of the homogeneous catalysts are active even at low temperature but they are very corrosive and not easily recoverable.Therefore, a logical solution might reside in some hybrid catalysts resulting from the immobilization of these active compounds.But the insertion of solid superacids such as SbF 5 into graphite 21 or its deposition on SiO 2 22 led to catalysts which easily deactivate.The aim of this paper is to present results of the isomerization of hexane and cyclohexane on new hybrid catalysts resulting from the embedding of triflate derivatives in silica using a sol-gel procedure.

Results and Discussion
Figure 1 shows the conversion of cyclohexane over silica-embedded tertbutyldimethylsilyltrifluoromethanesulfonate catalysts (BDMSTFMS) (series A1-A4) after 2 h of reaction as a function of temperature.An increase of temperature leads to an increase of conversion.The selectivity was 100%, no other reaction products than methylcyclopentane being identified.The dependence of the yield on the reaction time is in concordance with the thermodynamics of the system.A similar dependence was determined for the silica-embedded lanthanum triflates (series B1-B4), except that these catalysts were less active than the similarly prepared BDMSTFMS catalysts (Figure 2).The conversion-time curves of Figure 3 show that, even after 4 h of reaction, the conversions stay quite low.In the absence of boundary layer and inter-and intraparticle diffusional influences, the conversion of cyclohexane, determined for levels smaller than 12%, indicates that the isomerization of cyclohexane over these catalysts followed the requirements of Bassett-Habgood 23 kinetic treatment for first order reaction rate kinetic processes in which the ratedetermining step is the surface reaction.The pores of these materials are large enough (about 2 nm) as to cause any mass transfer limitation.Table 1 gives the calculated apparent rate constants.Separate catalytic tests were carried out in homogeneous conditions over pure triflate derivatives, using the same amount of triflate as in embedded catalysts.They exhibit a different catalytic behavior, intermediate between the catalysts 1 and 2-4.For pure BDMSTFMS at 25 o C and after 2 h of reaction, the conversion was about 5% in the absence of water, and 8.7% in the presence of 90 mg water.It is worth to note that in the presence of water, cracking to isobutane amounted to 4.7%, in perfect agreement with previous data obtained over silica-supported aluminum chloride. 24At 75 o C, the conversion increased at 12.9% in the absence of water and at 22.4% in the presence of 90 mg H 2 O. Lanthanum triflate was less active.At 75 o C and after 4 h, the conversion was about 16% but in the presence of 90 mg water.At room temperature, no conversion of cyclohexane was detected on this catalyst.
The behavior of the immobilized catalysts was a heterogeneous one.The liquor separated after 4 h reaction remained untransformed for another 4 h reaction.As shown above, the addition of a very small amount of water (90 mg) had a positive effect on the reaction.The conversion of cyclohexane to methylcyclopentane was improved, irrespective of the reaction conditions used (Figures 2 and 3).For pure lanthanum triflate derivatives, the increase of the conversion caused by the addition of water was very small, cca.1%.It is worth to note that the embedded catalysts were not altered by the addition of a small amount of water.Reuse of the catalysts gave the same catalytic performances.NH 3 -FTIR spectra indicated the presence of Lewis acid sites as inferred from a band at 1611 cm -1 , but also of Brønsted acid sites (bands in the range 1460-1490 cm -1 ).The bands at 1784-1782 and 1750-1747 cm -1 are assigned to imine groups resulting from the same ammonia chemisorption (Figure 4).The band at 1460-1490 cm -1 may indicate that a limited part of the triflate-derivative is hydrolyzed, leading to chemisorbed triflate.The sensibility of triflates to moisture is well known. 25An increase of the water content to 10 mmoles caused the catalysts deactivation.The catalytic tests using catalysts pre-calcined at 300 o C to remove the surfactant indicated a slightly lower activity than the non-calcined catalysts.This temperature was chosen according to results obtained from thermogravimetric analysis.Figure 5 illustrates such a behavior for B catalysts.The addition of water also improved the catalytic activity.The structural investigation using FTIR, 13 C-NMR and XPS suggested that the triflate derivatives were not decomposed under these conditions (Table 2).The XPS F/S ratio was near 3 in all the investigated embeddedtriflates, which is in a perfect agreement with those obtained by chemical analysis.The binding energies of sulfur and fluorine slightly increased in each series indicating more oxidized species. 13C-CP/MAS NMR signals due to CF 3 were found at the same positions as for the pure triflates.
The decrease of the conversion must be associated with some textural modifications caused by the surfactant decomposition.The calcined samples had surface areas about 15% smaller compared to the values given in Table 2, which might suggest that some pores were blocked by the surfactant decomposition.As it was stressed above, an additional increase of the amount of water to 90 mg was not effective also for the calcined catalysts.Under such conditions a second aqueous phase started to be formed and the conversion was drastically decreased.Similar experiments were carried out using n-hexane.Isomerisation of hexane to 2-and 3methylpentanes ocurreded only over the A catalysts, and in the range of temperatures used, the conversions were smaller than 5%, irrespective of the reaction conditions.For A1, the selectivities were as follows: 10.5% to 2-methylpentane, 6%-to 3-methylpentane, 42% to methylcyclopentane, and 41% to cyclohexane.These data may account for a different reaction mechanism according to the molecule which is reacted.Schemes 1 and 2 describe the reaction pathways which may correspond to the reaction of cyclohexane and hexane on the investigated embbed triflates.For n-hexane, the ionic-radical mechanism 26 is more plausible (Scheme 2).Summarizing, the immobilization of triflate derivative catalysts on silica led to systems able to catalyze the skeletal isomerization of cyclohexane at temperatures much lower than reported for zeolites. 27The system is comparable with silica-supported aluminum chloride 24 but it is environmentally more friendly.The difference of the catalytic reactivity between cyclohexane and n-hexane is not a thermodynamical one because the free Gibbs energies of the reactions are similar, about -1 kcal/mol in both cases.The reason of the reactivity might be related, as suggested by the reaction mechanism, to the different ionization potential of the two molecules.On the investigated catalysts, only cyclohexane is able to surpass this barrier, leading to reasonable conversions.
The differences among the investigated catalysts result from the nature of the triflate and its accesibility.It is worth to note that the conditions in which the immobilization is carried out may lead to higher performances than free triflate derivatives (see A1 and B1 catalysts).Traces of water improve the performances, reediting Nenitzescu chemistry. 2Water was found to adsorb on these catalysts without altering the acid character.As proposed by Kobe at al. 28 for the isomerization of butane over sulfated zirconia catalysts, water may lower the energy barrier through which the catalytic cycle must proceed.Excessive catalyst hydration leads to the suppression of the catalytic activity very probably because of an extensive hydrogen bonding at the active sites.

Experimental Section
General Procedures.The analysis of the reaction products was made using a HP gas chromatograph, model 5890, series II equipped with a Chrompack capillary column.FTIR (Fourier Transform-Infrared Spectroscopy) spectra were recorded with a Bruker Equinox 55 instrument, equipped with KBr optics and a DTGS detector.Pure samples (15 mg) were pressed and placed inside a commercial temperature-controlled environmental chamber and evacuated at 10 -5 bar.Ammonia was introduced in this chamber at 50 mbar and the spectra were recorded at room temperature, 100, 250 and, 300 °C.The XPS (X-ray Photoelectron Spectroscopy) spectra were obtained with a SSI X probe FISONS spectrometer (SSX -100/ 206) with monochromated Al-Kα radiation.The spectrometer energy scale was calibrated using the Au 4f 7/2 peak (binding energy of 84.0 eV).For the calculation of the binding energies, the C 1s peak of the C-(C,H) component at 284.8 eV of adventitious carbon was used as an internal standard.The composite peaks were decomposed by a fitting routine included in the ESCA 8,3 D software.The surface composition of the investigated samples was determined using the same software.The F 1s , S 2s , Si 2p , La 3d , and O 1s peaks were investigated.Solid-state 13 C CP/MAS NMR spectra were recorded on a Bruker MSL-300 spectrometer, equipped with a double-bearing probe.The rotor (7 mm internal diameter) was charged with the samples and sealed by a Kel-F inset.The optimal contact time for 13 C cross-polarization was 2-3 ms.The spinning rate was between 2 and 4 kHz.The external standard for 13 C NMR was adamantane [ (CH 2 )= 38.40, relative to tetramethylsilane].Thermal curves were recorded using a SETARAM TGA 92.16.18 equipment.The samples in an amount of 40 mg were heated in a high purity helium stream (Air Liquide) from room temperature to 500 o C at a heating rate of 5 o C min -1 .

Figure 4 .
Figure 4. NH 3 -FTIR spectra of A catalysts at 100 o C.

Figure 5 .
Figure 5. Variation of the conversion of cyclohexane as a function of water content for calcined catalysts (30 mg catalyst, 10.1 g cyclohexane, 4 h, 75 o C, 800 rpm).

Scheme 1 .
Scheme 1.The reaction mechanism in isomerization of cyclohexane.

3 Scheme 2 .
Scheme 2. The reaction mechanism in isomerization of hexane.

Table 1 .
Apparent rate constant (k a ) at 300 K

Table 2 .
XPS, CP/MAS NMR and surface area characteristics of the investigated catalysts