ISSN 20700504, Catalysis in Industry, 2014, Vol. 6, No. 2, pp. 134–142. © Pleiades Publishing, Ltd., 2014. Original Russian Text © M.D. Smolikov, V.B. Goncharov, E.M. Sadovskaya, K.V. Kazantsev, E.V. Zatolokina, D.I. Kir’yanov, E.A. Paukshtis, B.S. Bal’zhinimaev, A.S. Belyi, 2013, published in Kataliz v Promyshlennosti.

DOMESTIC CATALYSTS

Studying the Role of the State of Platinum in Pt/SO4/ZrO2/Al2O3 Catalysts in the Isomerization of nHexane M. D. Smolikova, b, V. B. Goncharovb, E. M. Sadovskayab, K. V. Kazantseva, E. V. Zatolokinaa, D. I. Kir’yanova, E. A. Paukshtisb, B. S. Bal’zhinimaevb, and A. S. Belyia, c aInstitute

of Hydrocarbon Processing, Siberian Branch, Russian Academy of Sciences, Omsk, 644040 Russia Boreskov Institute of Catalysis, Siberian Branch, Russian Academy of Sciences, Novosibirsk, 630090 Russia cOmsk State Technical University, Omsk, 644050 Russia email: [email protected]; [email protected]; [email protected]

b

Received September 24, 2013

Abstract—Samples of SO4/ZrO2/Al2O3 and Pt/Al2O3 Pt/Al2O3 catalysts and their physical mixtures are pre pared, and the catalytic properties of the samples in nhexane isomerization are studied. The considerable effect of the state of platinum on the catalytic performance of the samples is revealed. IR spectroscopy (COads), oxygen chemisorption, and oxygen−hydrogen titration show that the reduced catalysts contain ionic forms of platinum capable of adsorbing up to three hydrogen atoms per each surface atom of platinum. By means of H/D isotopic exchange, it is found that specific properties of ionic platinum are apparent in the for mation of the hydride form of adsorbed hydrogen. It is speculated that the activity and stability of catalysts based on sulfated zirconia in nhexane isomerization can be attributed to the involvement of ionic and metal lic platinum in the activation of hydrogen. The results can be used to develop effective catalysts for the isomer ization of C5−C6 gasoline fractions in order to obtain the isomerizate as a highoctane additive for modern gasolines. Keywords: isomerization, platinum catalysts, zirconia, isotopic exchange DOI: 10.1134/S2070050414020111

INTRODUCTION The skeletal isomerization of alkanes is one basis for producing components of environmentally safe gasolines [1]. In recent times, these processes have been catalyzed by such solid acids as zeolites and metal 2– oxides modified with SO 4 anions, primarily sulfated zirconia (SO4/ZrO2), that exhibit superacid properties and actively isomerize alkanes at low temperatures [2, 3]. Despite the high selectivity of the process over SO4/ZrO2, however, this system rapidly deactivates with an irreversible loss of activity. The introduction of platinum notably enhances the activity and stability of Pt/SO4/ZrO2 catalyst, in particular, during the isomerization of C5–C6 paraffins in a hydrogen envi ronment [4–6], since platinum is capable of suppress ing coke formation processes. It is speculated that after activation on platinum, hydrogen diffuses to the sup port and contributes to the formation of new acid sites [6–9]. In addition, Pt could participate in hydride transfer during the isomerization of paraffins on acid sites [10, 11]. In this work, we examine the effect the electronic state of platinum has on the catalytic activity of

Pt/SO4/ZrO2/Al2O3 acid catalyst during nhexane isomerization in a hydrogen environment. EXPERIMENTAL Preparation of Catalysts Our studies were conducted using platinumfree acid catalysts γAl2O3 and SO4/ZrO2/Al2O3, and platinum supported catalysts Pt/Al2O3 and Pt/SO4/ZrO2/Al2O3. Their chemical compositions and denotations are listed in Table 1. A sample of a mechanical mixture of Pt/Al2O3 and SO4/ZrO2/Al2O3 catalysts in a weight ratio of 1 : 1 was also studied. The SZA catalyst was prepared via alkaline precip itation of zirconium hydroxide from an aqueous solu tion of ZrO(NO3)2 with a concentration of 90 g/L in terms of ZrO2. The precipitation of zirconium hydrox ide was conducted at a temperature of 60–65°C with an ammonia solution being added to the zirconium salt solution in order to adjust pH to a value of 9.5–10. The resulting precipitate was washed, dried at 120°C, and then exposed to a 12% sulfuric acid solution. After wetting with water, the material was dried, admixed with aluminum hydroxide with a pseudoboehmite structure as a binder, dried at 120°C, and calcined in a

134

STUDYING THE ROLE OF THE STATE OF PLATINUM

135

Table 1. Chemical composition of our catalysts Content, wt % Sample γAl2O3 SO4/ZrO2/γAl2O3 Pt/γAl2O3 Pt/SO4/ZrO2/γAl2O3

Name Al2O3 SZA Pt/Al2O3 Pt/SZA

Pt

SO4

ZrO2

Al2O3

– – 0.40 0.40

– 6 – 6

– 24 – 24

100 70 99.60 Remainder

stream of dehydrated air at 650°C. After sulfation and calcination, the zirconia in the acid catalyst was in the form of tetragonal phase tZrO2. To prepare the Pt/SZA sample, SZA acid catalyst was impregnated with an aqueous solution of H2PtCl6, used in an amount that ensured a platinum content of 0.40 wt % in the finished catalyst. The sample was then dried at 120°C and calcined in a stream of dehydrated air at 450°C. The Pt/Al2O3 catalyst was prepared using γAl2O3, which was precalcined at 580°C and subjected to a vacuum in order to prevent the cracking of thin pores. The evacuated support was impregnated with an aque ous solution of H2PtCl6, used in an amount of 0.4% Pt of the weight of the support with the addition of a hydrochloric acid solution (3% HCl of the weight of the support) to ensure the uniform impregnation and homogeneous distribution of Pt over the granules of the support. The impregnated support was washed with distilled water to remove nonadsorbed solution components from the pore space, dried in air at 120°C, and calcined in a stream of dehydrated air at 500°C. Prior to our catalytic tests, the Pt/Al2O3 sam ple was reduced with hydrogen at 500°C. The mechanical mixture was prepared in an agate mortar by mixing and grinding the metal (Pt/Al2O3), and acid components (SZA) at a weight ratio of 1 : 1. The resulting powder was then compressed with a pressure of 9 MPa and ground in a mortar to select a fraction of 0.25–0.75 mm for the catalytic tests. Technique The catalytic tests of the samples in nhexane isomerization were conducted in a flow system with an isothermal fixedbed tubular reactor and a thermo couple mounted firmly along the reactor axis. The reactor was charged with 2 cm3 of the catalyst with a fraction of 0.25–0.75 mm. Before it was loaded into the reactor, the Pt/Al2O3 alumina–platinum catalyst was activated in a separate setup in a stream of purified hydrogen with a stepwise rise in temperature up to 500°C and held at this tem perature for 1 h. Prior to preparing the mechanical mixture for cat alytic tests, the SZA acid catalyst was calcined in a CATALYSIS IN INDUSTRY

Vol. 6

No. 2

2014

stream of dehydrated air at a temperature of 650°C. The catalysts and the mixture we further activated inside the catalytic unit in a stream of purified hydro gen at 270°C for 2 h. Catalytic testing was conducted in a temperature range of 140–400°C at a pressure of 1.5 MPa, a liquid hourly space velocity of 2 h–1, and a H2/nC6 molar ratio of 3/1. The feedstock (reagentgrade nhexane dried over a NaX molecular sieve) was fed with a dosing pump into a Tshaped pipe for mixing with hydrogen and then into the reactor. The reaction products were ana lyzed in real time using a Tsvet800 chromatograph equipped with a Petrocol DH 50.2 capillary column. IR spectra were recorded in the region of 4000– 1000 cm–1 on a Shimadzu 8300 spectrophotometer equipped with a DRS800 diffuse reflectance attach ment. Prior to spectroscopic studies, the samples were activated in a vacuum and reduced in a special setup using a portable cell. The spectra were recorded using a quartz cell equipped with a window of CaF2. To study the state of platinum, the pretreated and evacu ated samples (fraction, 0.25–0.75 mm) were exposed to CO at 30 Torr and a temperature of 25°C. After the spectrum was recorded, the sample was evacuated at room temperature, and the spectrum was measured again. The dispersion of platinum and the amounts of adsorbed oxygen and hydrogen were estimated via adsorption studies conducted with oxygen−hydrogen (O2–H2) titration and oxygen chemisorption [12, 13]. H/D isotopic exchange was conducted in the above tubular flow reactor at 200°C. The reactor was charged with 0.2 g of the catalyst with a fraction of 0.2–0.5 mm. The chemical and isotopic compositions of the gases at outlet of the reactor were recorded using a VG Seb sorlab 200D quadrupole mass spectrometer. The catalysts loaded into the reactor were activated in a stream of purified and dried hydrogen at 270°C for 10 h. The isotopic experiments were conducted in the isothermal mode at a temperature of 200°C via SteadyState Isotopic Transient Kinetic Analysis (SSITKA). A mixture of 1.2 vol % H2 in nitrogen was initially fed into the reactor; once adsorption−desorp tion equilibrium was established, the stream was switched to feed a mixture of 1.2 vol % D2 in nitrogen into the inlet of the reactor. The concentration of H2,

136

SMOLIKOV et al.

Table 2. Parameters of nhexane isomerization over the studied catalysts Yield of hydrocarbons, wt % T, °C

Catalyst

Conversion, % Selectivity, wt %

Σ iC6

ΣDMB

(ΣDMB/ΣC6) × 100%

SZA

140* 140 160 180 200 220

89.6* 3.9 5.6 12.5 11.6 14.9

92* 100 100 98 95 91

82.7* 3.9 5.6 12.2 11.0 13.6

40.2* 0.6 0.8 1.9 1.7 2.0

43.2* 0.6 0.8 1.9 1.7 2.0

Pt/SZA

140 160 180 200 220

89.8 89.3 89.5 90.3 93.1

97 94 91 83 76

86.8 83.7 81.5 74.7 71.2

38.1 33.7 28.2 26.4 21.1

39.3 35.7 30.7 31.3 27.0

SZA + Pt/Al2O3

180 220 260

30.2 65.8 83.3

99 93 65

29.8 61.0 53.9

4.9 11.9 13.4

4.9 12.5 19.0

Pt/Al2O3

360 400 450 500

25.1 51.4 82.6 97.5

90 86 53 6

22.5 44.5 43.6 6.3

2.7 8.1 10.1 1.4

12.0 18.2 23.2 22.2

* Parameters were measured 10 min after feeding in nhexane; the other parameters, after a run of 30 min. ΣDMB is the sum of 2,2 and 2,3 dimethylbutane isomers.

HD, and D2 was measured throughout the period of exchange, and the flow rate of the mixture was main tained at a level of 4 L/h in all cases [24]. RESULTS AND DISCUSSION Isomerization of nHexane Results from our catalytic tests of SO4/ZrO2/Al2O3 and Pt/SO4 /ZrO2/Al2O3 superacid catalysts, the mix ture (SO4/ZrO2/Al2O3 + Pt/Al2O3), and the Pt/Al2O3 catalyst are shown in Table 2. The SZA sample initially exhibited high catalytic activity during nhexane isomerization; the yield of hexane isomers was 82.7% after a 10min run. As the duration of the test run was raised to 30 min, the activ ity diminished considerably. The yield of isomeriza tion products at 140–160°C was ~4–5% at selectivi ties of up to 100%. As the temperature rose to 220°C, the yield of isomerization products grew to 11–14%, while the selectivity for isohexanes decreased owing to an increase in the yield of cracking products. Introducing platinum into SZA notably improved the stability of the Pt/SZA catalyst system. The yield of isomerization products at 140–160°C reached 83.7–86.8% at a selectivity toward isohexanes of 94– 97%. The reaction products contained 2methylpen tane and 3methylpentane, deep isomerization prod

ucts (2,2 and 2,3dimethylbutanes), and C1–C5 cracking products (Table 2). Raising the temperature to 200–220°C substantially reduced the yield of iso mers (to 71.2–74.7%), and the selectivity fell sharply (to 76–83%) due to the development of cracking reac tions. Introducing platinum into the catalyst thus notably improved conversion and the yield of hexane isomers. Tests on nhexane isomerization over a mixture of catalysts were conducted to determine the effect the state of platinum has on catalytic performance. The lowactivity acid component of SZA was admixed with an equal amount of Pt/Al2O3 alumina–platinum catalyst. Alumina–platinum catalysts in general are active isomerization catalysts. For a platinum catalyst, however, temperatures above 400°C are optimal in terms of activity during the isomerization reaction (see Table 2), as is characteristic of hightemperature isomerization by a bifunctional mechanism [14]. In its classical form, a bifunctional mechanism relies on the independent action of metal and acid sites. In [15–17, 21, 22], it was found that in Pt/Al2O3 catalysts modi fied with chloride (as well as bromide and fluorine) ions, the Pt atoms are stabilized by chemical reactions with surface sites of alumina (i.e., surface Al3+ cations; L denotes Lewis sites) and halogen ions to form a [PtOxXy] surface complex in which X = Br, Cl, F. CATALYSIS IN INDUSTRY

Vol. 6

No. 2

2014

STUDYING THE ROLE OF THE STATE OF PLATINUM

137

2202

2142

2120

0.05 2076

F(R)

2099

0.10

0 2000

1900

2100 Wavenumber, cm–1

2200

Fig. 1. Diffuse reflectance IR spectra of COads on the reduced surface (270°C) of the Pt/SZA catalyst.

Let us return to the results for the mixed catalyst summarized in Table 2. According to the figures for conversion and the yield of isomers, the mixed catalyst with ionic platinum occupies an intermediate position between the individual Pt/Al2O3 and Pt/SZA catalysts. Adding an acid component to the alumina–platinum catalyst in the mechanical mixture shifts the reaction to a range of temperatures much lower than 360−420°C, which is characteristic of the classical bifunctional mechanism of isomerization over PtX/Al2O3 catalysts. The temperature range of optimum isomerization activity with the maximum yield of hexane isomers for a mixed catalyst composed of the SZA acid compo nent and [PtOxCly] complexes in Pt/Al2O3 is 200− 240°C. For mechanical mixtures of SZA and a Pt/SiO2 catalyst (Pt0 metal atoms), the temperature range of isomerization is 350–420°C, as was shown in [18]. Our results suggest that the ionic state of Pt plays an important role in the catalysis of hexane isomeriza tion, over both an SZA acid system and alumina–plat inum catalysts (bifunctional catalysis). In mechanical mixtures, catalysts with ionic platinum are active at much lower temperatures than those for mixtures of the SZA acid component and a Pt/SiO2 catalyst (Pt0 metal atoms) [18]. Adsorption Properties and the State of Platinum in the Catalysts Our results from studying the state of Pt in Pt/SZA superacid catalyst by diffuse reflectance IR spectros copy using CO as a probe are shown in Fig. 1. After the adsorption of CO, the spectrum exhibits absorption bands that peak at 2076, 2099, 2120, 2142, CATALYSIS IN INDUSTRY

Vol. 6

No. 2

2014

and 2202 cm–1. The last band is characteristic of CO complexes with Lewis acid sites (LASes) on a surface of SZA [19, 20]. The band at 2142 cm–1 can be attrib uted to linear CO complexes with Pt+ ions contained in the CO–(Pt–O–SO3) surface complexes [19, 20]. The frequencies of 2099 and 2123 cm–1 can be attrib uted to the vibrations of CO adsorbed in linear form on (Pt–H)δ+ particles. The absorption band at 2076 cm–1 is attributed to the adsorption of CO on Pt0 metal atoms. The shift in the absorption frequency of adsorbed CO (from 2070–2080 cm–1 typical for the metal to 2099 cm–1) indicates the presence of metallic platinum, which is heavily affected by the acid sites of SO4/ZrO2 [20]. Note that no bridge form of the adsorbed CO was detected. The state of platinum in PtX/Al2O3 (where X = Br, Cl, F) alumina–platinum catalysts was studied by IR spectroscopy in [15, 16, 22]. A distinctive feature of halogencontaining catalysts is that their IR spectra exhibit absorption bands at 2128, 2123, 2167, and 2172 cm–1, corresponding to linear forms of adsorbed CO on oxidized platinum sites. It was concluded that platinum supported on alumina was not completely reduced in the alumina–platinum catalysts used in this study, even after treatment in H2 at 500°C. The obtained spectral characteristics provide strong evidence that the studied catalyst systems con tain ionic platinum. For alumina–platinum systems, ionic platinum was identified as a [PtOxXy] complex on LASes [21]. In the Pt/SZA acid catalyst, the ionic form of Pt is assumed to be located in the vicinity of Lewis sites in the form of (SO3–O)–Pt structures [19]. The [PtOxCly] surface complexes with platinum ions

138

SMOLIKOV et al.

Table 3. Adsorption characteristics of Pt in our catalysts: dispersity, oxygen absorption during chemisorption (OC) and titration (OT), and the amount of adsorbed hydrogen [22] Sample

Dispersity of Pts/Ptt, %*

Oxygen absorption, O atom/Ptt atom OT

OC

H/Ptt

H/Pts

PtBr/Al2O3

91

1.37

0.51

1.72

1.89

PtCl/Al2O3

100

1.50

0.50

2.0

2.0

PtF/Al2O3

100

1.79

0.60

2.38

2.38

16

0.24

0.02

0.44

2.75

Pt/SO4/ZrO2/Al2O3

* Pts and Ptt are the quantities of surface platinum atoms and total platinum atoms, respectively.

exhibit specific properties; in particular, they adsorb hydrogen with a H/Pt stoichiometry close to 2 [21]. Table 3 shows the oxygen absorption values obtained from the (O2–H2) titration and O2 chemi sorption data for a set of PtX/Al2O3 (where X = Br, Cl, F) catalysts modified with various halogens [22]. The adsorption data were used to calculate the amount of adsorbed hydrogen per surface atom of platinum H/Pts in each of the catalysts. It is evident that in the set of PtX/Al2O3 catalysts, the ability of surface Pt atoms to adsorb hydrogen increases in the halogen series Br → Cl → F. It is known that the acidic properties of gamma alumina increase in the same sequence upon the addition of halogens [14]. The specific chemisorption values are close for Pt/Al2O3 systems modified with bromine and chlorine (halogens with weak and medium acidity): H/Pts = 1.89–2.0. After modification with fluorine (a strongly acidic halogen) the chemisorption capacity for hydrogen rises to H/Pts = 2.38. A further increase in the amount of adsorbed hydrogen (up to H/Pts = 2.75) is observed upon switching to the Pt/SZA super acid system. Ionic platinum’s high capacity for the adsorption of hydrogen could help to explain the high catalytic performance of the isomerization reaction in a hydro gen environment. The reaction in an inert gas (nitro gen or helium) atmosphere was characterized by rapid deactivation (within several minutes) accompanied by a loss in activity, as in the case of isomerization using the SZA acid component without platinum (see Table 2). To explain the role of the state of Pt and its effect on the pattern of hydrogen adsorption, the H/D isotopic exchange on Pt/SZA and Pt/Al2O3 catalysts was therefore studied. SSITKA H/D Isotopic Exchange The dynamic features of deuterium exchange in SSITKA experiments become apparent after switch ing to dimensionless isotope variables [24], i.e., atomic isotopic fractions (αi):

2D2 + HD 2H 2 + HD , αH = , 2 (D2 + HD + H 2 ) 2 (D2 + HD + H 2 ) αD + αH = 1 and the molecular fraction of HD (fHD) in gasphase hydrogen: αD =

HD , H 2 + HD + D2 where H2, D2, and HD are the concentrations of gas phase hydrogen, deuterium, and the mixed form, respectively. f HD =

Pt/SZA and SZA Samples Figure 2 shows the time dependences of atomic isotopic fraction αH and molecular fraction of mixed hydrogen fHD, obtained by switching of the mixture flow from (H2 + N2) to (D2 + N2) over Pt/SZA and SZA catalyst samples. Over the SZA sample, the atomic isotopic fraction of deuterium in the gas phase (αD) and the molecular fraction of HD (fHD) almost immediately (at the same time as the argon label) exhibit a level corresponding to their content in the feed mixture. This means that the exchange rates are negligible. Over the Pt/SZA catalyst, an abrupt change in αD and fHD is observed at the instant when the mixture is switched and slowly approaches the isotopic composi tion of the feed mixture. The isotopic fraction of deu terium at the outlet of the reactor remains lower than in the feed mixture for some time, indicating an inter phase exchange and a large number of exchangeable hydrogen atoms on the surface of the catalyst. The amount of surface hydrogen involved in the exchange, as calculated from the area between the αD(t) curves, is 84 × 1019 atoms per gram of catalyst. Molecular isotopic fraction fHD initially grows and reaches its maximum at a 50% degree of surface hydro gen isotopic substitution. Afterward, fHD declines and gradually approaches the fraction of HD in the feed mixture. CATALYSIS IN INDUSTRY

Vol. 6

No. 2

2014

STUDYING THE ROLE OF THE STATE OF PLATINUM

139

Ar 1.0

αD

0.8

SO4/ZrO2/Al2O3 Pt/SO4/ZrO2/Al2O3

0.6 0.4 0.2 0 0

100

200

400

300

600

500

700

800

900

1000

900

1000

fHD

0.6 2 0.4 Pt/SO4/ZrO2/Al2O3

1 0.2

SO4/ZrO2/Al2O3

0 0

100

200

300

400 500 Time, s

600

700

800

Fig. 2. SSITKA on Pt/SZA and SZA (solid and dashed lines denote experimental results and calculations, respectively).

Pt/Al2O3 and Al2O3 Samples

Pt/Al2O3+ SZA Mechanical Mixture

Figure 3 shows the dependences of αD and fHD obtained by switching the mixture from (H2 + N2) to (D2 + N2) on the Pt/Al2O3 sample and the γAl2O3 support sample.

Figure 4 shows the dependences of αD and fHD obtained for a sample of the mixture with a weight ratio between the metal (Pt/Al2O3) and acid compo nents (SZA) of 1 : 1. If the isotopic exchange in the mechanical mixture occurred independently on each of the components, the dynamics of response on a sample of the mechan ical mixture would coincide with that of Pt/Al2O3 (see Fig. 3), since the exchange rate on the acid component is close to zero (see Fig. 2). This case is obviously dif ferent: the areas between the αD curves and the αD(t)

On Al2O3, the isotopic composition of the mixture at the outlet varies synchronously with the change in the argon concentration; i.e., the H/D exchange rate is negligible (as on SZA). On the Pt/Al2O3 sample, the isotopic fraction of deuterium at the outlet of the reactor is close to zero in the initial period of time after switching the mixture; i.e., there is total deuterium transfer from the gas phase onto the catalyst, suggesting that the interphase exchange rate is very high. The concentration of sub stituted hydrogen atoms on the surface of Pt/Al2O3 is 130 × 1019 atoms per gram of catalyst. The apparent exchange rate on Pt/Al2O3 is signifi cantly higher than on the Pt/SZA sample, as is evident from the slope of variation in the αD(t) plots over time. On the two types of supports (SZA and γAl2O3), virtu ally no H/D isotopic exchange is observed when there is no Pt. We may assumed that the observed differences in the exchange rates can be attributed to the state of platinum in the superacid system. CATALYSIS IN INDUSTRY

Vol. 6

No. 2

2014

inlet

and α D (t) curves for the mechanical mixture are about 1.5 times larger than for Pt/Al2O3. The time interval in which the degree of exchange of deuterated hydrogen in the gas phase is close to 100% (αD and fHD are close to zero) also increases approximately by a factor of 1.5, suggesting that the exchange rate on the acid component in the mixture is very high and is com parable to the exchange rate on the metal component. Hence, the hydrogen atoms resulting from dissociation migrate from the metal component to the acid compo nent. The number of hydrogen atoms inserted into the acid component of the mixture can be determined from the area between the αD(t) curves for the mixture and the pure Pt/Al2O3 catalyst. For a sample with a compo

140

SMOLIKOV et al. Ar 1.0 Al2O3

αD

0.8

Pt/Al2O3

0.6 0.4 0.2 0 0

100

200

Pt/Al2O3

400

300

500

600

700

800

500

600

700

800

KIE = 0 KIE = 1.2

fHD

0.4 0.2

Al2O3

0 0

100

200

400 300 Time, s

Fig. 3. SSITKA on Pt/Al2O3 and Al2O3 (solid and dashed lines denote experimental results and calculations, respectively).

1.0 Mechanical mixture 0.8 αD

0.6 0.4 0.2 0 0

100

200

300

400

500

600

700

800

0

100

200

300 400 Time, s

500

600

700

800

fHD

0.4 0.2 0

Fig. 4. SSITKA on the sample of the SZA + Pt/Al2O3 mechanical mixture (solid and dashed lines denote experiment results and calculations, respectively).

nent ratio of 1 : 1, this number is 70 × 1019 atoms per gram of the acid component, which is close to the con centration of substituted hydrogen on the Pt/SZA sam ple (84 × 1019 atoms per gram). This suggests that the exchange on the sample of the mechanical mixture involved most of the OH groups of the acid component. Two mechanisms (types) of interphase exchange of hydrogen on platinum are discussed in the literature

[23]: singlestage exchange between a hydrogen mole cule and an adsorbed hydrogen atom through the for mation of a threeatom complex (type I) and two stage exchange with a preliminary stage of the dissoci ation of molecular hydrogen (type II). Depending on the type of exchange, the rate of for mation of HD is expressed by one of the following equations: CATALYSIS IN INDUSTRY

Vol. 6

No. 2

2014

STUDYING THE ROLE OF THE STATE OF PLATINUM

type I:

= −bRPt

⎛f f ⎞ CH2 ⎜ HD + 1 HD ⎟ ⎝ ∂t τ ∂ξ ⎠ (1) g g Pt Pt α D 1 − α D + α D 1 − α D − f HD ,

( (

)

(

)

)

⎛f f ⎞ CH2 ⎜ HD + 1 HD ⎟ t ∂ τ ∂ξ ⎠ ⎝ type II: (2) Pt Pt = − bRPt 2α D 1 − α D − f HD . According to the literature, the H/D exchange on metallic platinum occurs by the first type of mecha nism only at very low temperatures (150 K and below) [23]. At higher temperatures, the exchange occurs through hydrogen dissociation by the second type of mechanism. As was expected, numerical analysis of the fHD(t) dependence on samples of the mechanical mixture and Pt/Al2O3 showed that the exchange occured via the dissociative mechanism (type II). We can describe the experimental results by introducing a small correc tion for the kinetic isotope effect (KIE), in accordance with Eq. (2) (see Figs. 3 and 4). An unexpected result was obtained in modeling the dynamics of response fHD(t) on the Pt/SZA superacid sample. Figure 2 shows the fHD(t) dependences calcu lated according to Eqs. (2) and (1) (curves 1 and 2, respectively), compared to the experimental curve of the response. It is evident that the observed dynamics of response in this case corresponds to the first type of mechanism (curve 2). In terms of the classical interpretation of the exchange mechanism of the first type, which excludes the stage of hydrogen adsorption and dissociation, it is unclear how the exchange on the surface of Pt/SZA involves the OH groups of the support. It was shown in this work that their rate of exchange with molecular hydrogen in the absence of platinum is negligible. In addition, it is known that hydrogen dissociation on metallic Pt occurs even at 100 K. The stage of interac tion involving platinum therefore cannot be excluded from the exchange scheme. The results from the H/D exchange for the Pt/SZA system can be interpreted in terms of the type I isoto pic exchange mechanism, which ensures that two adsorbed hydrogen species are formed at the stage of molecular hydrogen dissociation, one of which is involved in the isotopic exchange. This interpretation can be used to explain the observed dynamics of the H/D exchange on the Pt/SZA catalyst. According to IR spectroscopy using CO probe molecules, the supported Pt in this system is described by a set of states that include oxidized plati num Pt+ and metallic atoms Pt0, particularly those with additional charge Ptδ+ resulting from interaction with the protons of the Brønsted acid sites (BASes). We may assume that the Pt atoms form a sequence of charge states with different abilities to activate and retain adsorbed hydrogen; this facilitates the migra

(

(

CATALYSIS IN INDUSTRY

)

Vol. 6

)

No. 2

2014

141

tion of hydrogen atoms and the occurrence of H/D exchange: the deuterium molecule is initially adsorbed and dissociates on the platinum site with different states of Pt to form two species of absorbed hydrogen on the surface of the catalyst: Pt+ (Pt0) + D2 → [Pt+D–] + [PtD+] δ+ . The absorption bands of COads observed in the IR spectra at 2199 and 2120 cm–1 (see Fig. 1) can be attributed to the Pt/SZA catalysts containing (PtH)δ+ complexes in which the metal Pt atoms interact with protons from the BASes [19, 20]. We may assume that it is this metal species, originally associated with the protons, that exhibits the highest activity during the H/D exchange with the OH groups of the support: [PtD+]δ+ + HOZr/Al → [PtH+]δ+ + DOZr/Al. At the final stage, hydrogen undergoes recombina tion on the platinum sites to release the added hydro gen: [ Pt+D–] + [PtH+]δ+ → Pt+ (Pt0) + HD. The result is a triatomic isotopic exchange, which is described by Eq. (1). The results from our numerical analysis of the dependence of the molecular fraction of HD on the exchange time suggest that hydrogen dissociation and exchange on Pt/Al2O3 and Pt/SZA occur by different mechanisms. The first is characterized by the forma tion of a surface hydrogen species that can be easily transferred to the surface of the support, and this is responsible for the high rate of the H/D exchange. In the second, at least two different species of surface hydrogen are formed, and only one of them, the prod uct of interaction with metal platinum [PtD+]δ+, rap idly exchanges with the OH groups and is transferred onto the support. Considering the distribution of Pt in different charge states over the surface of Pt/SZA, we may assume that the Pt0, Ptδ+, and Pt+ states contribute to the formation of different species of adsorbed hydro gen. The heterolytic dissociation of molecular hydro gen on oxidized platinum leads to the emergence of protons and hydride ions on the surface of the catalyst. The migration of protons to the support, facilitated by the formation of hydrogen complexes of the (PtH+)δ+ type, is a familiar stage in the formation of acid sites (BASs) on the surface of zeolite catalysts [25]. The IR spectra of COads in our Pt/SZA catalyst exhibited absorption bands at 2099 and 2120 cm–1 that can be attributed to the platinum particles surrounded by BASes of varying strength [19, 20]. The hydride ions formed during the heterolytic dissociation of molecu lar hydrogen can be adsorbed and retained by oxidized Pt atoms in the (Pt+–O–SO3) complexes for a certain time and then separated and attached to the isomeric carbocation on neighboring acid sites. The transfer of hydride ions with the intermediate involvement of oxi

142

SMOLIKOV et al.

dized platinum atoms then plays an important role in the catalysis of the isomerization reaction at the final stage of hydride transfer. At the same time, molecular hydrogen is adsorbed on the Pt0 metal atoms to form atomic hydrogen that migrates over the surface of the support and hydrogenates the compaction products, thereby preventing polycondensation reactions that can poison the acid sites and lower catalyst activity. CONCLUSIONS The specific properties of platinum in a zirconium sulfate system that result in the activation of hydrogen adsorption in numbers of H/Pts up to 2–3 atoms/atoms are responsible for the high stability and selectivity of sulfated zirconia in the lowtemperature range of nhexane isomerization. The dissociative adsorption of hydrogen on ionic platinum, which results in the formation of hydride ions and protons, contributes to the modification of acid sites of both types (BASes and LASes). This is responsible for the regeneration and formation of new sites and changes in the acidity of existing sites under the conditions of a hexane isomer ization reaction in a hydrogen environment. At the final stage of isomerization, the ionic form of platinum can act as a source of hydride ions. On the other hand, the dissociative adsorption of hydrogen on metallic platinum is a source of atomic hydrogen for hydroge nating coke precursors to ensure the stability of zirco nium sulfate catalysts. Ionic platinum contained in the surface sites of a superacid catalyst should be regarded as a reactant of hexane isomerization that exhibits specific properties involved in the activation of hydrogen for hydride transfer or modification of the surface acid sites of the catalyst. Surface atoms of platinum in the metal state are activators and suppliers of hydrogen for hydroge nating coke precursors of polycyclic aromatic hydro carbons. The formation of an ensemble of states of Pt with different functions ensures high levels of activity and stability for lowtemperature isomerization cata lysts based on sulfated zirconia. REFERENCES 1. Kimura, T., Catal. Today, 2003, vol. 81, p. 57. 2. Weyda, H. and Koehler, E., Catal. Today, 2003, vol. 81, p. 51. 3. Yamaguchi, T., Appl. Catal., A, 2001, vol. 222, p. 237. 4. Hino, M. and Arata, K., Catal. Lett., 1995, vol. 30, p. 25. 5. Iglesia, E., Soled, S.L., and Kramer, G.M., J. Catal., 1993, vol. 144, p. 238. 6. Ebitani, K., Konishi, J., and Hattori, H., J. Catal., 1991, vol. 130, p. 257.

7. Ebitani, K., Konno, H., Tanaka, T., and Hattori, H., J. Catal., 1992, vol. 135, p. 60. 8. Hattori, H., Stud. Surf. Sci. Catal., 2001, vol. 138, p. 3. 9. Triwahyono S., Yamada T., and Hattori, H., Appl. Catal., A, 2003, vol. 242, p. 101. 10. Duchet, J.C., Guillaume, D., Monnier, A., Dujardin, C., Gilson, J.P., van Gestel, J., Szabo, G., and Nasci mentof, P., J. Catal., 2001, vol. 198, p. 328. 11. van Gestel, J., Nghiem, V.T., Guillaume, D., Gilson, J.P., and Duchet, J.C., J. Catal., 2002, vol. 212, p. 173. 12. Duplyakin, V.K., Belyi, A.S., Ostrovskii, N.M., Smolikov, M.D., Chalganov, E.M., and Nizovskii, A.I., Dokl. Akad. Nauk SSSR, Ser. Khim., 1989, vol. 305, no. 3, p. 648. 13. Belyi, A.S., Kiryanov, D.I., Smolikov, M.D., Zatolokina, E.V., Udras, I.E., and Duplyakin, V.K., Reac. Kinet. Catal. Lett., 1994, vol. 53, no. 1, p. 183. 14. Bursian, N.R., Tekhnologiya izomerizatsii parafinovykh uglevodorodov (Paraffin Hydrocarbon Isomerization Technology), Leningrad: Khimiya, 1985, p. 73. 15. Smolikov, M.D., Zaitsev, A.V., Zatolokina, E.V., Belyi, A.S., Borovkov, V.Yu., Duplyakin, V.K., and Kazanskii, V.B., Kinet. Katal., 1992, vol. 33, no. 3, p. 625. 16. Smolikov, M.D., Belyi, A.S., Kiryanov, D.I., Borovkov, V.Y., Zaitsev, A.V., Duplyakin, V.K., and Kazanskii, V.B., React. Kinet. Catal. Lett., 1994, vol. 53, no. 1, p. 161. 17. Belyi, A.S., Smolikov, M.D., Kiryanov, D.I., and Udras, I.E., Russ. J. Gen. Chem., 2007, vol. 77, no. 12, p. 2243. 18. Smolikov, M.D., Dzhikiya, O.V., Zatolokina, E.V., Kiryanov, D.I., and Belyi, A.S., Pet. Chem., 2009, vol. 49, no. 6, p. 473. 19. Ivanov, A.V., Vasina, T.V., Masloboishikova, O.V., KhelkovskayaSergeeva, E.G., Kustov, L.M., and Houzvicka, J.I., Catal. Today, 2002, vol. 74, p. 95. 20. Ivanov, A.V., Stakheev, A.Y., and Kustov, L.M., Stud. Surf. Sci. Catal., 2000, vol. 130, p. 263. 21. Belyi, A.S., Smolikov, M.D., Kiryanov, D.I., and Udras, I.E., Russ. J. Gen. Chem., 2007, vol. 77, no. 12, p. 2243. 22. Smolikov, M.D., Kazantsev, K.V., Zatolokina, E.V., Kiryanov, D.I., Paukshtis, E.A., and Belyi, A.S., Kinet. Catal., 2010, vol. 51, no. 4, p. 584. 23. Ozaki, A., Isotopic Studies of Heterogeneous Catalysis, Tokyo: Kodansha; New York: Academic Press, 1977. 24. Sadovskaya, E.M., Ivanova, Y.A., Pinaeva, L.G., Grasso, G., Kuznetsova, T.G., van Veen, A., Sadykov, V.A., and Mirodatos, C., J. Phys. Chem. A, 2007, vol. 111, p. 4498. 25. Fujimoto, K., Stud. Surf. Sci. Catal., 1999, vol. 127, p. 37.

Translated by M. Timoshinina

CATALYSIS IN INDUSTRY

Vol. 6

No. 2

2014