CC BY
9УДК 544.47:544.344
ФИЗИКО-ХИМИЧЕСКИЕ И КАТАЛИТИЧЕСКИЕ СВОЙСТВА АЛЮМОХРОМОВОГО КАТАЛИЗАТОРА ДЕГИДРИРОВАНИЯ ПРОПАНА, ПРИГОТОВЛЕННОГО СПОСОБОМ ОДНОСТАДИЙНОГО СООСАЖДЕНИЯ
Л.П. Диденко, А.М. Колесникова, В.И. Савченко, И.А. Домашнев, Ю.М. Шульга, А.В. Куликов, М.С. Воронецкий, Л.А. Семенцова, Е.И. Кнерельман, Г.И. Давыдова
Институт проблем химической физики РАН, пр-т Акад. Семенова, 1, г. Черноголовка, Моск. обл., 142432 Россия e-mail: [email protected]
Алюмохромовый катализатор получен способом одностадийного соосаждения, при котором суспензию гидроксидов хрома и алюминия подвергали высокотемпературной обработке при 550° С. Образцы, содержащие 2,8, 5,5, 9,0 и 11,3 мас.% Cr, исследованы методами химического анализа, термопрограммированного восстановления водородом, ЭПР, рентгенофазового анализа, термогравиметрического анализа и низкотемпературной адсорбции азота. Кроме того, изучено влияние кальцинирования при 650° С и 900° С на свойства катализатора.
Представлены результаты изучения пористой структуры, фазового состояния и состава катализатора.
Исследована реакция дегидрирования пропана при Т = 550° С и времени контакта 2,7-9,0 с. Установлена высокая селективность образования пропилена и незначительное коксообразование.
Высказано предположение о преимущественной каталитической активности изолированных форм Cr3+ при содержании хрома менее монослойного покрытия.
PHYSICO-CHEMICAL AND CATALYTIC PROPERTIES OF THE ALUMINA-CHROMIUM CATALYST OF PROPANE DEHYDROGENATION PREPARED BY SINGLE-STAGE CO-PRECIPITATING TECHNIQUE
L.P. Didenko, A.M. Kolesnikova, V.I. Savchenko, I.A. Domashnev, Yu.M. Shulga,
A.V. Kulikov, M.S. Voronetskiy, L.A. Sementsova, E.I. Knerelman, G.I. Davydova
Institute of Problems of Chemical Physics, Russian Academy of Sciences, 1, Akad. Semenov av., Chernogolovka, Moscow region, 142432 Russia e-mail: [email protected]
Alumina-chromium catalyst was prepared by a single-stage co-precipitating technique, at which the suspension of chromium and aluminum hydroxides was treated at high temperature (550° C). The samples contained 2.8, 5.5, 9.0, and 11.3 wt.% of Cr have been investigated by such methods as: chemical analysis, temperature-programmed reduction with hydrogen (TPR), EPR, X-ray diffraction, thermogravimetry, and low-temperature nitrogen adsorption. In addition, effect of the calcination at 650° C and 900° C on the catalyst properties has been studied.
The results of studying the porous structure, phase state, and composition of the catalyst have been presented.
The reaction of propane dehydrogenation was investigated at T = 550° C and at the contact time of 2.7-9.0 sec. A high selectivity to propylene and a slight coke formation have been revealed.
The assumption about preliminary catalytic activity of the isolated Cr3+ forms at the chromium content less than monolayer covering has been made.
Introduction
Catalytic dehydrogenation of light hydrocarbon raw materials is one of the perspective methods of the production of light olefins.
This process is of great practical significance because it is an alternative to the petrochemical method of the production of various valuable chemical substances from
inexpensive and available gaseous and oil-gaseous raw materials. At present, resources of gaseous and oil-gaseous raw materials are just large, but their application is quite inefficient.
Light olefins (ethylene, propylene, butene) are in great demand in the world markets. Increasing facilities of the processes of pyrolysis and catalytic cracking generally used today can not satisfy it.
Commonly used industrial processes of dehydro-genation of light alkanes, the most known of which are the Catofin technology of ABB Lummus Global Company, the Oleflex technology of UOP Company, and the Yarsynthesis-Snamprogetty technology [1-4], are cost- and energy-consuming. First of all, it is caused by passing a side reaction of the coke formation leading to the fast deactivation of the catalyst. The necessity of frequent catalyst regeneration makes the process complicate, expensive, as well, requires from the catalyst to retain its initial activity even after the numerous oxidative regenerations.
The fact that alumina-chromium catalysts display activity in dehydrogenation of light alkanes has been revealed about 70 years ago [5, 6]. However, in spite of the use of these catalysts in the industrial processes of dehydrogenation, the investigations directed to improving their catalytic properties are still actively carrying out. The results of the studies of the structure and catalytic properties of the alumina-chromium catalysts are summarized in several reviews, for example, [7-9]. It has been found that the properties of the catalyst are strongly determined by conditions of the synthesis (the nature and ratio of the initial components, temperature regime, pH value). These parameters influence on the composition of the surface chromium forms, the strength of their bond with the support, and the surface acidity. Consequently, they effect on the catalytic activity and selectivity, resistance to the coke formation, and time-dependent stability.
There are three techniques commonly used to produce the alumina-chromium catalysts: 1) impregnation of the support with a chromium salt solution and its consequent thermal decomposition to chromium oxides; 2) co-sedimentation with the more or less simultaneous chromium oxides formation and their following drying and treating at high temperature (calcination); 3) atomic layer epitaxy - when a chromium salt (i.e., chromium acetylacetate Cr(acac)3) reacts with the support surface just from the vapor phase [10].
A wide variety of the catalysts described in the literature is determined by differences in the conditions of these techniques.
In the present work, the alumina-chromium catalyst has been obtained by the method of co-sedimentation from water solutions of chromium and aluminum nitrates at pH 9.5-10. Taking into account that the most active catalysts are formed from the thermodynamic instable deposits, which should be obtained in non-equilibrium conditions [11], we excluded the stages of filtration, rinsing and drying from the method. The suspension of the hydroxides was treated at 550° C - the temperature at which catalytic properties of the obtained samples had been studied. During the treatment, the processes of evaporation, decomposition, drying, and interaction of chromium and alumina oxides were carried out with a high rate. As a result of this simplified technique, we obtained the catalyst with a high specific surface, stable structure, high catalytic activity, and selectivity [12].
In order to study the effect of the synthetic conditions on the properties of the obtained catalyst, we carried out the physical-chemical analysis and investigated its catalytic behavior in the reaction of propane dehydrogenation.
Experiment
Catalyst preparation techniques A mixture of Al(NO3)3-9H20 and Cr(NO3)3-9H2O was dissolved in water and heated to the boiling point. Ammonia water solution was added to it until pH 9.510.0 had been achieved. The suspension obtained was held in a muffle at T = 550° C for 75 minutes. Then it was cooled to the room temperature, powdered, and the fraction of 0.2-0.4 mm was sifted out.
The ratios of Al(NO3)3-9H20 to Cr(NO3)3-9H2O were chosen so that the samples with the chromium content of 2.8, 5.5, 9.0, and 11.3 wt.% could be obtained. In the given article these samples were called "initial".
With the purpose to study the effect of the temperature treatment, the initial samples were calcinated in air at 650° C and 900° C for 1 hour.
Chemical analysis Chemical analysis was used for determining [Cr], [Cr3+], and [Cr6+] (the total chromium content and the concentrations of Cr3+ and Cr6+, respectively).
Determination of Cr6+ content A sample of the catalyst (0.1-0.3 g) was dissolved in 10 ml of diluted sulfuric acid (1:1) at heating. Then the solution was cooled, diluted with distilled water to 100 ml in a volumetric flask. An aliquot of the obtained solution was titrated with the 0.0125 N standard solution of the More salt (Fe(NH4)2(SO4)26H2O) in the presence of N-phenylanthranilic acid as the indicator.
Determination of the total chromium content A charge of the catalyst sample was placed in a quartz crucible, and 10 g of potassium pyrosulphate (K2S2O7) was added to it. The crucible was heated in a muffle to 700° C for 30 minutes till melting. Then the melt was cooled and dissolved in 20 ml of diluted sulfuric acid (1:1) at heating. The solution was placed into a volumetric flask and diluted with distilled water to 100 ml. An aliquot of the solution obtained was mixed with 10 ml of 0.25% AgNO3 solution (to oxidize chromium ions) and diluted with water to 300-400 ml. Then 3-5 g of (NH4)2S2O8 was added to it. After cooling, the solution was titrated with the 0.0125 N standard solution of the More salt (Fe(NH4)2(SO4)2"6H2O) in the presence of N-phenylanthranilic acid as the indicator.
Cr3+ content was calculated as a difference between the total chromium content and [Cr6+].
X-ray diffraction analysis X-ray diffraction analysis was used for determination of the phase state of aluminum and chromium oxides.
Roentgenograms were registered by using an ADP-2-01 diffractometer (Cu Ka-emission, Ni filter) equipped with a X-RAY program for the automatic data acquisition and processing.
Determination of specific surface and pore distribution in the samples Specific surface and pore distribution of the samples were determined by a method of low temperature nitrogen adsorption. Adsorption and desorption isotherms were obtained with the use of an Autosorb-1 analyzer (Quantachrome, USA). The samples were previously vacuumed at T = 280° C for 60 minutes. Their specific surface was determined from the adsorption curves by the Brunauer-Emmett-Teller (BET) method. The total pore volume was determined from the adsorption curves at the partial nitrogen pressure p/p0 = 0.99.
Thermogravimetric analysis The coke content in the catalyst after reaction was determined by a TG-DTG-DTA method by a Q-1500D thermogravimetre (MOM, Hungary). The sample weight was 500 mg. The samples were previously vacuumed at T = 300° C for 60 minutes. In the course of analysis, the samples were heated in the air atmosphere with the linear heating rate of 10° C/min.
Temperature-programmed reduction
Reduction of the catalyst with hydrogen was studied with the use of an Autosorb-1C analyzer (Quantachrome, USA). The samples of the catalyst were previously heated in the flow of He (30 ml/min) at T = 140° C for 60 minutes, then they were cooled in He flow to the room temperature. After this pre-treatment, the samples were reduced with the gaseous mixture of 5.5 vol.% of H2 and 94.5 vol.% of N2 in the temperature range of 40750° C (linear heating rate was 10° C/min, gas flow rate was 40 ml/min; sample weight = 1 g).
EPR investigations The presence of Cr3+ and Cr5+ ions in the samples was revealed by EPR.
EPR spectra of the catalyst samples, which have been placed into a thin-walled glass ampoule, were registered at the room temperature by a SE/X-2544 spectrometer (Radiopan, Poland).
Concentration of Cr5+ was estimated by comparison of the second integrals of the spectra of the investigated samples with those of a standard - CuSO4^5H2O.
The study of the catalytic activity of the samples To establish the catalytic activity of the samples in the reaction of propane dehydrogenation, the conversion and selectivity to the obtained products were determined at T = 550° C.
The reaction was carried out in a quartz flow reactor (inner diameter of 10 mm) with a fixed catalyst bed. Propane (high grade of purity) was supplied out of a gas cylinder with the rate adjusted by a gas flow regulator.
The reactor heated by an electric oven had two zones: a zone of the previous heating of the initial substances to 400° C, and a reaction zone where the required amount (0.6 g) of the catalyst was placed. The reaction temperature was maintained by a temperature gauge and controlled with a chromel-alumel thermocouple, which had been inserted into the catalyst bed and bounded to a digital detecting device. During the pre-treatment, the catalyst was vacuumed at the reaction temperature (550° C) for 1 hour to evacuate water and other volatile impurities. The gaseous mixture after reactor was analyzed online with the use of a Cristal-5000 gas chromatograph (Chromatec, Russia) equipped with a flame-ionization detector and a katharometer. The amount of hydrogen in the reaction products was determined at a column with the 13X molecular sieves (2 mm*2 m, 50° C, Ar as the carrier gas). The content of hydrocarbons in the reaction products was determined at the HP - Al/KCl column (0.5 mmx30 m, 80° C, He as the carrier gas). A method of absolute calibration was used to calculate concentrations of the reaction products.
The propane conversion (a, %) and the selectivity to the obtained products (S, %) were calculated according to the following formulas:
ym V — X V
a = C3H/ in ^C3H/ out _ 100 •
Xin V '
С,Ня in
S = ■
nXprodVout
•100.
3( Yin V — Y""' V )
V C3H8 in C3H8 out'
where Vin is the bulk inlet rate of propane (cm3/min); X" h is the bulk concentration of propane in the
gaseous flow at the input of the reactor (%); XC"^ is the
bulk concentration of propane in the reaction products at the output of the reactor (%); Vout is the bulk outlet rate of the reaction products (cm3/min); Xprod is the bulk concentration of the product in the reaction mixture at the output of the reactor (%); n is the amount of carbon atoms in the product.
The selectivity to the coke formation (Sc) was calculated as:
Sc (%) = 100 - X S(CH4+C2H4+C2H6+C3H6) .
The catalyst deactivation factor (D) was calculated according to the following formule:
D =
Conv. of C3H8 at 120 min (%) Conv. of C3H8 at 40 min (%)
The contact time (t, s) was determined as:
V • 60
t = ■
where V is the catalyst volume (cm ); Vin is the propane bulk inlet rate (cm3/min).
25
№
J ^
99
average pore radii (that is the radii corresponding to the maximums on the pore distribution curves) are about 2.0 nm. The form of the hysteresis of the adsorption-desorption curves is mainly typical for the slot pores.
The calcination reduces the catalyst specific surface (Ssp); this effect depends on both the calcination temperature (Tcalc), and the chromium content. After the calcination of the samples with different chromium content at 900° С their specific surface decreases about twice (Table 1). At the same time, the fraction of lager pores in the pore distribution increases (Fig. 1). The increase of the average pore size may arise as a result either of sintering of smaller mesopores or the pore enlargement because of the destruction of the walls between them. As can be seen from Table 1, the total mesopore volume after the calcination is not changed. So it can be supposed that the destruction of the pore walls is the main cause of the specific surface reduction.
Таблица 1
Результаты физико-химического исследования алюмохромовых катализаторов с различным
содержанием Cr
Table 1
The results of physico-chemical studies of the alumina-chromium catalysts with various Cr contents
Results
Physico-chemical studies of the catalyst It is known that the composition of the alumina-chromium catalyst is generally determined by the chromium content and the temperature of the air calcination [9, 13]. Therefore, we carried out physico-chemical studies of the catalyst by changing just these factors. The samples calcinated at 650° C and 900° C, with the total chromium content of 2.8, 5.5, 9.0, and 11.3 wt.%, were investigated by such methods as chemical analysis, X-ray diffraction analysis, EPR, thermo-gravimetry, temperature-programmed reduction, and low-temperature nitrogen adsorption. The results obtained are presented in Table 1.
One can see that the catalyst has quite a high specific surface (Ssp), which weakly depends on the chromium content, so it is determined by the structure of the support. According to the pore size distribution (see Fig. 1), the
Cr content т °г 1 calci ^ Chem. analysis data wt.% Surface measurement results EPR
wt.% g/m2-104 Ci3+ Cr6+ % Cr* SSp, m2/g V y pores signals
2.8 1.0 initial 1.0 1.8 35.7 275.4 0.35 Y
1.2 650 1.0 1.6 38.5 239.5 0.34 -
1.9 900 2.3 0.7 76.7 149.0 0.34 Y, 5
5.5 2.1 initial 2.0 3.5 36.4 255.8 0.34 Y, 5
- 650 - - - - - Y, 5
- 900 - - - - - Y, 5, ß
9.0 3.0 initial 4.6 4.4 51.1 296.0 0.26 Y, 5, ß
3.8 650 6.1 2.9 67.8 233.4 0.26 Y, 5, ß
7.1 900 8.5 0.5 94.4 126.9 0.25 5, ß
11.3 4.2 initial 6.9 4.4 61.1 270.2 0.26 5, ß
5.0 650 7.9 3.4 69.9 225.7 0.26 5, ß
9.0 900 10.1 1.2 89.4 125.9 0.27 5, ß
Рис. 1. Влияние температуры кальцинирования и содержания Сг на распределение пор по радиусам: —•— 11,3 мас.% Сг без кальцинирования; —А— 11,3 мас.%
- -•- - 9,3 мас.% Cr без кальцинирования; ■
Cr, Гкальц.= 650° С; - -■- - 9,3 мас.% Cr, "о" 2,8 мас.% Cr без кальцинирования; "V" 2,8 мас.% Cr,
' с;
-▼- - 9,3 мас.% 1 кальц.- 900° С;
Fig. 1. Effect of calcination temperature and chromium content
on the pore size distribution: -•- 11.3 wt.% Cr without calc; —A— 11.3 wt.% Cr, 7^.= 650° C; -■- 11.3 wt.% Cr, 7caic = 900° C; - -•- - 9.3 wt.% Cr without calc; --▼- - 9.3 wt.% Cr, 7calc = 650° C; - -■- - 9.3 wt.% Cr, 7calc = 900° C; ■"o- 2.8 wt.% Cr without calc; "V" 2.8 wt.% Cr, 7caic = 650° C; ■■□"■ 2.8 wt.% Cr, 7calc= 900° C
Cr, Ткальц = 650° С; -■- 11,3 мас.% Cr
/кальц.= 650° С; ■ "G- 2,8 мас.% Cr, Ткальц = 900° С
The calcination at the lower temperature (650° C) results in the pore radius increasing only for the samples with the low Cr content (2.8 wt.%). At the same time, the calcination at 650° C leads to decreasing the specific surface for all the samples regardless of the chromium content (see Table 1). It can be explained by sintering of the micropores. The presence of micropores in the sample can be revealed by the pore distribution curve -by the rise of the initial part of the curve above the axis of abscissas (see Fig. 1). After the treatment at 650 °C, all of these initial parts are declining regardless of the chromium content. In the case of the samples calcinated at 900° C, the initial parts of the distribution curves decreases down to the axis of abscissas what say about the entire micropore sintering.
So, the sintering of the micropores during the calcination of the samples with 9.0 and 11.3 wt.% of chromium at 650° C is the only reason of decreasing the Ssp values. At Tcaic = 900° C, both the sintering, and the destruction of the mesopores reduces Ssp value.
The fact that the mesopore structure of the samples with 9.0 and 11.3 wt.% of chromium remains undamaged after the calcination at 650° C can be explained by their stabilization with chromium. It is known that the monolayer cover is formed at 4 atoms of Cr per nm2 [9] what corresponds to about 9-10 wt.% of chromium in the investigated catalyst. Therefore, the stabilizing effect of chromium occurs at its mono- and a higher layer covering.
The results of X-ray diffraction analysis of the samples with 2.8-11.3 wt.% of chromium point to stability of the support structure: the presence of peaks at 46.1° and 67.8°, as well a broad peak between 36° and 38° point to the y-Al2O3 structure. The spectra of the samples after the calcination at 650° C and 900° C are close to the spectra of the initial catalysts.
In addition, X-ray diffraction data indicate that in the range of [Cr] = 2.8-11.3 wt.% and at ^ < 900° C chromium exists in amorphous state, since the peaks typical for the a-C^Os crystals (at 24.8, 32.8, and 54.6°) have not been revealed.
The curves of temperature-programmed hydrogen reduction (TPR) of the catalyst are shown in Fig. 2. The profile of H2 absorption is presented by one peak that points to the occurrence of the only reducible form of chromium, namely Cr6+ (in accordance with the chemical analysis data, which will be shown below). Taking into account EPR data (see later), which say about the presence of Cr3+ catalyst form and negligible amount of Cr5+, we can assume that hydrogen consumes to the reduction of Cr6+ to Cr3+ catalyst forms.
The reduction occurs in the temperature range of 276-355° C (Fig. 2). Temperatures of the reduction maximums (TMAX) for the samples with the different chromium content (curve 1 - 2.8 wt.% of Cr, curve 2 -11.3 wt.% of Cr) have close meanings equal to 382 and 384° C, respectively.
After the calcination at 900° C, TMAX values for the given samples slightly decrease - to 355° C and 361° C,
respectively (curves 3, 4), what may be resulted from some loosening of the bonds of the chromium surface forms with the support.
200-
150
>
100-
50-
0
ft -
i 2
\
ftl 4n№ /1 ^Ч—.
0 200 400 600 800 Temperature,' С
Рис. 2. TPR-профиль для алюмохромовых образцов: 1 - 2,8 мас. % Cr (исходный); 2 - 11,3 мас. % Cr (исходный);
3 - 11,3 мас. % Cr (Гкальц .= 900° С); 4 - 2,8 мас. % Cr (Ткальц .= 900° С). Условия TPR: скорость нагрева 10° С/мин, состав смеси 5.5 об.% Н2 в N2, скорость потока 40 мл/мин, масса образца 1 г Fig. 2. TPR-profile for the alumina-chromium samples: 1 - 2.8 wt.% of Cr (initial); 2 - 11.3 wt.% of Cr (initial); 3 - 11.3 wt.% of Cr (^calc. = 900° С); 4 - 2.8 wt.% of Cr (Т^с. = 900° С). TPR conditions: linear heating rate is 10° С/min; reducing mixture: 5.5 vol.% of Н2 in N2, flow rate is 40 ml/min; sample mass is 1.0 g
EPR spectroscopy is widely used for studying the alumina-chromium catalysts, although only a part of chromium ions gives EPR signals. The isolated Cr5+ ions stabilized on the Al2O3 surface display activity in the spectrum [9, 13]. They gives so-called y-signal with the center near g = 1.97 and the distance between the maximums of the first derivative (A) of 5.0-6.0 mT. A broad P-signal resulted from small Cr2O3 clusters (nanoclusters) is also often observed in the EPR spectra. Accordingly the literature data [14, 15], this signal has A = = 80-180 mT and g = 1.95-1.98. It has been shown [15] that enlargement of the Cr2O3 clusters brings to the narrowing of the P-signal, which gradually approaches to the values typical for the bulk а-С^Оз (A = 50 mT and g = 1.98).
The samples also display so-called 5-signal - the most complicated both in shape, and by origin. Its spectrum has two maximums, one of them is located in the region of g = 4.0-5.5 and another - at g = 2.6. Most of authors, e.g. [13, 14, 16], consider this signal to be arrived from the isolated surface Cr3+ ions.
Fig. 3 shows EPR spectra of the catalyst containing 2.8 wt.% of Cr. The spectrum of the initial sample is presented only by the y-signal, which is imposed upon a small signal of about 150 mT from the glass ampoule. It should be noted that the initial catalyst was obtained at T = 550° C. The estimation of the signal intensity shows that it corresponds to 3-4% of Cr5+ of the total chromium amount in the sample.
25
Ж.
101
*
2 \ *
Y Г
О LOO ' 300 ' 500 ' 700 Magnetic field mT
Рис. 3. Спектры ЭПР катализатора, содержащего 2,8 мас. % Cr; 1 - исходный образец (получен при 550° С); 2 - после кальцинирования при 900° С Fig. 3. EPR spectra of the catalyst with 2.8 wt.% of Cr; 1 - the initial sample; 2 - the sample after the calcination at 900° С
After the calcination at 900° С, the intensity of the Y-signal rises about triply, its halfwidth also increases, and a 5-signal arises (in Fig. 3 the peaks belonging to the 5-signal are marked with a star). It indicates appearance of the isolated Cr3+ forms.
At the rise of the calcination temperature, the intensity of the Y-signal from the sample containing 5.5 wt.% of Cr decreases (see Fig. 4). The concentration of the isolated Cr3+ ions increases with the rise of Tcalc. Moreover, a P-signal is arising at 900° C. Hence, the nanoclusters of Cr2O3 have already formed at this temperature.
л/\
0 100 300 500 700 Magnetic field, mT
Рис. 4. Спектры ЭПР катализатора, содержащего 5,5 мас. % Cr: 1 - исходный образец (получен при 550° С);
2 - после кальцинирования при 650° С;
3 - после кальцинирования при 900° С.
Fig. 4. EPR spectra of the catalyst with 5.5 wt.% of Cr: 1 - the initial sample;
2 - the sample after the calcination at 650° С;
3 - the sample after the calcination at 900° С
The following rise of the chromium concentration leads to the formation of the Cr2O3 nanoclusters (P-signal) even in the initial catalyst (T = 550° C), see Fig. 5. The calcination at 650° C decreases the summary intensity of
the EPR signal approximately by 40%, whereas the intensity of the y-signal reduces to quarter. At Tcac = 900° C, the intensities of the 5- and P-signals rise, while the y-signal decreases practically to zero.
3
t
0 100 300 500 700 Magnetic field, mT
Рис. 5. Спектры ЭПР катализатора, содержащего 9,0 мас. % Cr: 1 - исходный катализатор;
2 - после кальцинирования при 650° C;
3 - после кальцинирования при 900° С
Fig. 5. EPR spectra of the catalyst with 9.0 wt.% of Cr: 1 - the initial sample;
2 - the sample after the calcination at 650° С;
3 - the sample after the calcination at 900° С
The increase of the calcination temperature for the sample with the maximal chromium content (11.3 wt.%) accompanies with reducing the concentration of the isolated Cr5+ ions (Fig. 6). The calcination at 650° C leads to reducing the concentration of the amorphous Cr2O3 nanoclusters After the calcination at 900° C, their concentration rises again, and dimension of the nanoclusters also increase (the halfwidth of the P-signal reduces). At the same time, the P-signal with g = 1.98 and Д = 50 mT, what is typical for the crystal form of a-Cr2O3, has not been revealed.
So, the increase of the total chromium content in the initial catalyst leads to the appearance of the Cr3+ nanoclusters, which size grows with the rise of the Tcalc.
i—.—i—.—i—.—i—.—[—.—i—.—i—.—i 0 100 300 500 700
Magnetic field, mT
Рис. 6. Спектры ЭПР катализатора, содержащего 11,3 мас.% Cr: 1 - исходного катализатора; 2 - кальцинированного при 650° С; 3 - кальцинированного при 900° С Fig. 6. EPR spectra of the catalyst with 11.3 wt.% of Cr: 1 - the initial sample; 2 - the sample after the calcination at 650° С; 3 - the sample after the calcination at 900° С
102 International Scientific Journal for Alternative Energy and Ecology № 2 (70) 2009
© Scientific Technical Centre «TATA», 2GG9
The chemical analysis data presented in Table 1 show that the increase both of the total chromium content, and of Tcah leads to the rise of [Cr3+] in the catalyst.
The study of the reaction ofpropane dehydrogenation The reaction of propane dehydrogenation has been investigated in the presence of the sample contained 9.0 wt.% of Cr (mcat = 0.61 g; fraction of 0.2-0.4 mm) at T = 550° C. The propane flow rate was changed from 6 cm3/min to 20 cm3/min what corresponded to the contact time t = 2.7-9 sec. The results obtained are presented as the changes of the propane conversion (Fig. 7, a), the selectivity to propylene (Fig. 7, b), the coke formation, and the methane concentration (Fig. 8, a, b) with the reaction time.
100
80
э
I
О
4-
o >'
с о О
60
40
20
0
100-,
80-
—I—I-1-1—Г Ui-1-1-1-1-1—r
0 40 80 120 0 40 80 120
Time of reaction, min Time of reaction, min
a b
"¡o"
T
20-
0-
—i—
40
it by the high activity of the strong acidic surface centers in the reaction of the coke formation. After deactivation of them with the coke, the rate of this process noticeably drops.
100
^ 60-CJ
о •—
13 40 H
20-
10п
8-
£ 6-
ЗС " 4-1
2-
Рис. 7. Изменение во времени конверсии пропана (а) и селективности образования пропилена (b) при варьировании времени контакта.
Катализатор 9,0 мас. % Cr/Al2O3 (фракция 0,2-0,4 мм; m = 0,61 г). Т= 550° С, 100% С3Н8. ■ - 9 с; о - 6 с; * - 2,7 с
Fig. 7. The change of the propane conversion (а) and the selectivity to the propylene formation (b) with time at different
contact time values. The catalyst: 9.0 wt.% of Cr/Al2O3 (fraction of 0.2-0.4 mm; m = 0.61 g). Т = 550° С, 100% of С3Н8.
■ - 9.0 sec; о - 6.0 sec; * - 2.7 sec
It can be seen that during the first 10-20 minutes the most changes of the propane conversion and the selectivity to the formation of the products are observed. At the high т values (9 s, 6 s), the propane conversion value in the beginning of the reaction amounts to 58-65%, but the selectivity to С3Н6 is low (Fig. 7, a, b) because of the high extent of the coke formation (Fig. 8, a). When т decreases to 2.7 s, the selectivity to coke drops noticeably, the conversion and the selectivity to propane slightly rise with time and achieve constant values by 15-20 minutes.
These changes may be associated with the change of the catalytic surface under the action of alkane in the beginning of the reaction. The similar effect was described in the works [9, 15, 17]. The authors explained
1—1—1—1—1—г
0 40 80 120 0 40 80 120 Time of reaction, min Time of reaction, min а b
Рис. 8. Изменение во времени селективности образования кокса (а) и содержания СН4 (b) при варьировании времени контакта.
Катализатор 9,0 мас. % Cr/Al2O3 (фракция 0,2-0,4 мм; m = 0,61 г). Т = 550 °С, 100% С3Н8. ■ - 9 с; о - 6 с; * - 2,7 с Fig. 8. The change of the selectivity to the coke formation (а) and СН4 content (b) with time at different contact time values.
The catalyst: 9.0 wt.% of Cr/AhO3 (fraction of 0.2-0.4 mm; m = 0.61 g). Т = 550° С, 100% of С3Н8. ■ - 9.0 sec; о - 6.0 sec; * - 2.7 sec
The curves presented in Fig. 8, b show that in the first minutes of the reaction a greater concentration of СН4, which is decreasing with time, is observed. Methane is not a product of the side reaction of the propane cracking (1), since the concentration of С2-hydrocarbons in the products is negligible (0.10.3 vol.%):
С3Н + С2Н4
(1)
Obviously, the CH4 formation occurs as a result of the interaction of the coke, which has been accumulated in the first minutes of the reaction, with the hydrogen presented in the system:
С + 2Н2 = СН4
(2)
When the process reaches the steady stage, the contribution of this reaction becomes negligible, and methane together with ethylene and ethane is formed by the reactions (1), (3):
С2Н4 + Н2 = С2Н6
(3)
The increase of the selectivity of the formation of the cracking products with the rise of t is the cause of diminishing the propylene formation at the steady part of the reaction (Fig. 7, b).
Since the chromium content and the calcination temperature change the catalyst composition (see Table 1),
25
103
we have studied the effect of these changes on the propane conversion and the selectivity to propylene formation. The results obtained, as well the catalyst deactivation factors (D), are presented in Table 2. The catalyst deactivation factors were calculated as a ratio of the propane conversion by 120 minutes to the corresponding value by 40 minutes.
It can be seen that the selectivity to propylene formation amounts to about 90% and slightly depends on the chromium content and Tcaic. With the rise of the chromium content, D is changing from about 1 to 0.8 and slightly depends on Tcaic. This fact is generally
determined by the catalyst resistance to the coke formation, what is one of the main causes of the catalyst deactivation during the reaction, and indicates the high catalyst stability with time.
According to the results of thermogravimetric study of the sample with 9.0 wt.% of Cr, about 3 wt.% of the coke accumulates in the sample after 120 minutes of the reaction. The investigation of the pore structure of the samples after reaction shows that the coke accumulation does not change the mesopore distribution and occurs mainly on the micropore surface, what leads to a small (about 10%) reduce of the specific surface of the catalyst.
Таблица 2
Влияние содержания хрома и температуры кальцинирования на показатели реакции дегидрирования пропана. Т = 550° С; m кат = 0,61 г (фр. 0,2-0,4 мм); расход С 3 Н 8 = 13,6 см3/мин. Время реакции равно 40 мин
Table 2
Effect of chromium content and calcination temperature on the characteristics of propane dehydrogenation. Т = 550° С; mcat = 0.61 g (fraction of 0.2-0.4 mm); С3Н8 flow rate is 13.6 cm3/min;
reaction time is 40 min
Cr content, wt.% T °С 1 calc, ^ С3Н8 conversion, % D Selectivity, %
С3Н6 СН4 С2Н4 С2Н6 coke
2.8 initial 9.1 1.2 89.4 3.3 2.0 2.0 3.3
650 13.7 1.1 90.0 3.2 1.7 2.0 3.1
900 21.7 0.9 93.2 2.3 1.2 1.7 1.6
5.5 initial 20.6 0.9 88.9 4.7 1.3 3.0 2.1
650 25.2 0.9 86.8 4.6 1.1 4.0 3.5
900 17.6 0.9 93.0 3.0 1.0 2.0 1.0
9.0 initial 18.5 1.0 91.1 2.4 1.6 2.8 2.1
650 21.7 0.9 86.2 5.2 1.3 3.5 3.8
900 23.7 0.8 89.7 4.5 1.2 2.9 1.7
11.3 initial 20.8 0.8 91.2 4.2 1.4 2.3 0.9
650 27.4 0.8 88.1 4.8 1.3 3.8 2.0
900 24.3 0.8 91.6 2.8 1.2 2.6 1.8
As follows from DTA curves, the maximal exothermic effect corresponding to the coke burnup takes place at 400° C what points to the formation of easy-burnable coke forms.
When the reaction time increases to 360 minutes, some catalyst deactivation decreasing the propane conversion occurs. Thus, the propane conversion for the sample with 9.0 wt.% of Cr drops from 18.5% (in 40 minutes) to 16.0%. This sample was chosen for testing its oxidative regeneration. For this purpose, the sample after 360 minutes of reaction was treated with air at T = 550° C for 1 hour, then it was reduced with hydrogen at the same temperature for 30 minutes. Testing of the sample after regeneration showed that its activity was close to the initial one.
Since the catalytic activity is intrinsic to the surface chromium forms, it is interesting to investigate the dependence of the propane conversion on the surface content of Cr3+ and Cr6+ (Fig. 9). It can be seen that the presented curves contain linear parts, after which the propane conversion becomes practically independent on the surface chromium content. On conversion to the total chromium content, the linear parts correspond to 2.85.5 wt.% of Cr, that is less than the monolayer chromium covering. Linearity of the dependence indicates about the mononuclear nature of the catalytically active chromium forms in the range of [Cr] = 2.8-5.5 wt.%. As for the nature of the catalytically active chromium forms, we can suppose with care that they seem to be the isolated surface Cr3+ ions. The similar character of the
dependencies presented in Fig. 10 may point to it. These dependencies were obtained for the sample with 2.8 wt.% of Cr, in which, in accordance with the abovestated EPR data, all Cr3+ occurred as the isolated ions. Since the Cr3+ concentration and the propane conversion are changing with the temperature in similar manner, it can be concluded that the isolated Cr3+ forms display predominant catalytic activity.
Our assumption about the nature of the catalytically active Cr forms is only preliminary one and it requires following experimental confirmations.
Рис. 9. Зависимость конверсии пропана от поверхностного содержания хрома. Т = 550° С; m^r = 0,61 г (фр. 0,2-0,4 мм); расход С3Н8= 13,6 см3/мин. [Cr3+]s и [Cr6+]s рассчитаны по данным табл. 1 Fig. 9. The dependence of the propane conversion on the surface chromium content: Т = 550° С; mcat = 0.61 g (fraction of 0.2-0.4 mm); С3Н8 flow = 13.6 cm3/min. [Cr3+]s and [Cr6+]s were calculated on the base of the Table 1 data
Рис. 10. Зависимость конверсии пропана и поверхностного содержания Сг3+ от температуры кальцинирования
(образец 2,8 мас. % Cr). Ткальц = 550° С соответствует исходному катализатору Fig. 10. The dependence of the propane conversion and the surface Сг3+ content on the calcination temperature for the sample with 2.8 wt.% of chromium. Тсак = 550° С corresponds to the initial catalyst
Discussion
The results presented show that the investigated catalyst is similar in some features to those described in the literature, but it also has several differences which may be caused from the synthesis peculiarities.
In the course of the synthesis, a highly developed and quite stable porous structure is forming. Its specific surface is more 100 m2/g than the literature one, see e.g. [13, 18]. The structure stability is caused by the absence of y-Al2O3 to 0-Al2O3 and to a-Al2O3 phase transitions in the region of the investigated Tcah (650-900° C). It is known that the change of the support structure during the phase transition accompanies with the loosening of the bond of the catalyst surface forms with the support what leads to the catalyst sintering.
During the calcination, the mesopores increase what leads to the reduction of the sample specific surface. The monolayer chromium covering promotes stabilization of the porous structure and prevents it from the disordering under the calcination at 650° C.
Stabilization of the porous structure is discussed in some other works [13, 19]. However, the authors explain it by inhibition of the y-Al2O3 ^ a-Al2O3 phase transition with chromium.
The investigated catalyst is similar in the composition to the catalysts described in the literature. Like in the other alumina-chromium catalysts (e.g., [9, 13]), chromium during the synthesis is stabilized on the surface as two main forms, namely Cr6+ and Cr3+. The concentration of Cr5+ is negligible. The literature data concerned with the Cr6+ to Cr3+ ratio are contradictious which can be explained by different conditions of the synthesis. In some catalysts with the chromium content higher than the monolayer covering, approximately 98 wt.% of Cr3+ is being stabilized in the course of the synthesis [9], whereas in other catalysts [20] almost all chromium up to the monolayer occurs as Cr6+. It has been shown [13] that the ratio of the surface chromium forms is effected by the balance between the amount of mono- and polychromates, which has been produced during the synthesis of the catalyst. The balance depends on the chromium content, which increasing leads to the rise of the output of polychromates. Moreover, the temperature influence is also considerable, because polychromates are thermally instable and can be easily transformed to the surface Cr3+ ions. The peculiarities of the given catalyst preparation lead to the stabilization on the surface of a great amount of Cr3+ at the chromium content less than the monolayer covering. It rises with the growth of chromium content, and at [Cr] higher that the monolayer covering, the initial catalyst contains over 60% of Cr3+.
The surface chromium forms are strongly bonded with the support, and a slight loosening of the bonds is observed only after the calcination at 900° C. It makes possible the maintenance of a high concentration of the isolated surface Cr3+ forms, which we regard as catalytically active. Moreover, the strong bond with the
Ж.
J '<:<
105
support prevents to the formation of the big clusters (a-Cr2O3) what is one of the ways to decreasing the catalytic activity during the calcination [9, 13].
There is not an unified view on the nature of the active centers in the literature. The works [17, 21] give reasons for the predominant catalytic activity of the highly dispersed surface phase of Cr6+/5+ which being reduced with alkane in the reaction course generates the catalytic active Cr3+ forms. Some other authors make conclusion about a higher catalytic activity of Cr3+ stabilized on the surface during the synthesis. The authors [13, 22] assume the amorphous Cr3+ nanoclusters to be the most active, whereas the works [14, 15] point to the predominant catalytic activity of the isolated Cr3+ forms. The last of the mentioned data are in a good agreement with the results obtained in our work. That allows making an assumption about the predominant catalytic activity of the isolated Cr3+ forms stabilized on the surface during the synthesis of the catalyst.
The investigated catalyst is highly resistant to deactivation during the reaction. The main cause of it is the catalyst resistance to the coke formation. In the first place, the coke accumulates on the surface of the micropores and does not influence on the mesoporous structure of the catalyst. After 360 minutes of the reaction, a slight decrease of the propane conversion is observed. However, the catalyst can be easily regenerated under the air calcination.
Thus, the peculiarities of the synthesis lead to the formation of the catalyst with a high catalytic activity, stability to deactivation and a stable porous structure.
Conclusions
The studies of the alumina-chromium catalyst prepared by the technique of single-stage co-precipitating from water solutions of chromium and aluminum nitrates showed that:
1) the catalyst has a high specific surface (250300 m2/g) and an average pore radius of about 2.0 nm. During the sample calcination the sintering of micropores and the mesopore enlargement leading to the reduction of Ssp value occur. The mono- and a higher layer surface covering with chromium makes the mesoporous structure stable and prevents it from the destruction during the calcination at 650° C;
2) at the temperature range of 550-900° C the support structure corresponds to y-Al2O3, and the phase transition to a-Al2O3 has not been observed;
3) Cr3+ and Cr6+ ions appear to be the main catalytic forms. The rise both of the total chromium content and the Tcalc leads to increasing the Cr3+ concentration. At that, the concentration of the amorphous nanoclusters in the Cr3+ composition increases. The formation of a-Cr2O3 has not been revealed;
4) the catalyst displays a high time stability, what is generally determined by its resistance to the coke formation;
5) we assumed the predominant catalytic activity of the isolated Cr3+ in the reaction of propane dehydrogenation.
References
1. Sanfilippo D., Buonomo F., Fusco G., Lupieri M., Miracca I. // Chem. Eng. Sci. 47 (1992) 2313.
2. Sauma M. // Petrotech. 15 (1992) 66.
3. Badoni R.P., Kumar Y., Shanker U., Rao T.S.R.P. // Chem. Eng. World. 31 (1996) 105.
4. Brasin M.M., McCain J.H., Vora B.V., Imai T., Pujado P.R. // Appl. Catal. A. 221 (2001) 397.
5. Frey F.E., Huppke W.F. // Ind. Eng. Chem. 25 (1933) 54.
6. Frey F.E., Huppke W.F. US Pat. 2 098 959 (1937).
7. Poole C.P., Maciver D.S. // Adv. Catal. 17 (1967) 223.
8. Weckhuysen B.M., Wachs I.E., Schoonheydt R.A. // Chem. Rev. 96 (1996) 3327.
9. Weckhuysen B.M., Schoonheydt R.A. // Cat. Today. 51 (1999) 223-232.
10. Kytokovi A., Jacobs J.P., Hakuli A., Merilainen J., Brongersma H.H. // J.Catal. 162 (1996) 190.
11. Krylov O.V. Heterogeneous Catalysis. Moscow: Akademkniga, 2004. P. 207.
12. Aldoshin S.M., Balihin I.L., Berestenko V.I., Didenko L.P., Domashnev I.A., Kolesnikova A.M., Kurkin E.N., Savchenko V.I., Troitsky V.N. Patent RU 2318593 C1 (10.03.2008).
13. Puurunen R.L., Weckhuysen B.M. // J. Catal. 210 (2002) 418-430.
14. De Rossi S., Ferraris G., Fremiotti S., Garrone E., Ghiotti G., M.C. Campa, Indovina V.J. // Catal. 148 (1994) 36.
15. De Rossi S., Pia Casaletto M., Ferraris G., Cimino A., Minelli G. // Appl. Catal. A. 167 (1998) 257.
16. Khaddar-Zine S., Ghorbel A., Naccache C. // J. Mol. Catal. A150 (1999) 223.
17. Hakuli A., Kytokivi A., Krause A.O., Suntola T. // J. Catal. 161 (1996) 393.
18. Shivaji Thapliyal, Goutam Deo // Bull. Catal. Soc. India. 2 (2003) 29-33.
19. Simon S., van der Pol A., Reijerse E.J., Kentgens A.P.M., van Moorsel G.-J. M.P., de Boer E.J. // Chem. Soc. Faraday Trans. 91 (1995) 1519.
20. Vuurman M.A., Hardcastle F.D., Wachs I.E. // J. Mol. Catal. 84 (1992) 193.
21. Weckhuysen B.M., Bensalem A., Schoonheydt R.A. // J. Chem. Soc. Faraday Trans. 94 (1998) 2011.
22. Bruckner A., Radnik J., Hoang D.-L., Lieske H. // Catal. Lett. 60 (1999) 183.
