Effect of Cofeeding Butane with Methanol on the Deactivation by Coke

ARTICLE pubs.acs.org/IECR

Effect of Cofeeding Butane with Methanol on the Deactivation by Coke of a HZSM-5 Zeolite Catalyst Andr
es T. Aguayo, Pedro Casta~no,* Diana Mier, Ana G. Gayubo, Martin Olazar, and Javier Bilbao Department of Chemical Engineering, University of the Basque Country, 48080-644 Bilbao, Spain. ABSTRACT: The deactivation by coke of a HZSM-5 zeolite catalyst has been studied in the transformation of methanol into hydrocarbons by cofeeding butane (n-butane). This reaction is of interest as an energy-neutral integrated process that enhances the activity in the cracking reaction and upgrades the paraffins formed as byproducts. The process was carried out in a fixed-bed reactor under the following conditions: temperature, 550 °C; pressure, 1 bar; space time, 2.4 and 4.8 (g of catalyst) h (mol of CH2)
1; time on stream, 5 h; methanol/butane molar ratio, up to 16/1. The coke was characterized using several analytical techniques (TG
TPO, FTIR, Raman, and NMR spectroscopies), and the effects of cofeeding butane on the coke composition and structure were determined. The results in terms of coke content and composition, are explained in terms of the different pathways of methanol and butane transformation.

1. INTRODUCTION The demand for light olefins is boosting selective processes such as the transformation of methanol or dimethyl ether, which are key intermediate compounds in the upgrading of alternative raw materials (natural gas, coal, and lignocellulosic biomass).1
4 The methanol-to-olefins (MTO) process, originally developed using HZSM-5 zeolite, is at a more advanced state of industrial implementation and involves a SAPO-34 catalyst in a fluidized bed using natural gas as the raw material.5 The Lurgi process, based on fixed-bed reactor technology and HZSM-5 zeolite catalyst for the transformation of dimethyl ether into olefins, also has good prospects for commercial application, as it gives a high yield of propylene.6 HZSM-5 zeolite has more suitable properties than other zeolites for obtaining a high yield of olefins with limited coke deposition:7
9 (i) a crystalline structure (channels of 0.53  056 nm and 0.51  0.55 nm) that hinders the formation of polyaromatic structures and the presence of many interconnections, which favors the flow of coke precursors toward the outside of the crystalline structure,10,11 and (ii) moderate acid strength and, consequently, reduced activity for hydrogen-transfer reactions.12 Numerous methods have been studied to improve these properties and increase the selectivity to propylene:13
16 (i) crystal size reduction, (ii) desilication in alkaline medium, (iii) ionic exchange, (iv) incorporation of metals or P, and (v) hydrothermal stabilization. Nevertheless, the selectivity to C2
C4 olefins in the transformation of methanol is lower using HZSM-5 zeolite catalysts than SAPO-34 catalysts, and gaseous paraffins are significantly formed as byproducts. These paraffins should be fed back into reactor, and consequently, the cofeeding of paraffins with methanol should be studied. The cracking of butane and the joint transformation of methanol and butane (model paraffin) on different acid catalysts has been studied in previous works.17,18 The results showed a suitable balance concerning activity, light-olefin selectivity, deactivation, and regenerability in the joint transformation of methanol and butane on a HZSM-5 zeolite catalyst with SiO2/ r 2011 American Chemical Society

Al2O3 = 30. The selectivity to C2
C4 olefins was 80% for a space time (referred to CH2 units in the reactant mixture) of 0.05 (g of zeolite) h (gCH2)
1.18 Nevertheless, it should be noted that the high values of acid strength and temperature (above 500 °C) required for cracking butane are not those required for minimizing deactivation by coke in the transformation of methanol, given that these severe conditions enhance secondary reactions from olefins (primary products).7
9,19,20 Furthermore, joint transformation in the same reactor integrates an exothermic process (methanol transformation) with an endothermic process (butane cracking), which allows for operation in an energy-neutral regime (for a methanol/butane molar ratio of 3 in the feed). The synergy attained by integrating the reaction schemes of the two processes has considerable advantages over the individual processes: (i) Butane cracking is enhanced (mainly by energy compensation on the acid sites). (ii) The autocatalytic mechanism of olefin formation is enhanced in the transformation of methanol (high olefin formation rate in the cracking of butane). (iii) Deactivation is attenuated in the transformation of methanol, which allows a higher yield of olefins to be obtained than when the two reactions are carried out separately.21 Furthermore, a kinetic model has been proposed for the integrated process22 that is based on models determined for butane cracking17 and for the MTO process at the temperatures required for butane cracking (above 500 °C).20 Deactivation by coke deposition on HZSM-5 zeolite catalysts in the transformation of methanol has been the subject of numerous studies, and the mechanisms of coke formation and evolution are reasonably well-established.7
9 Nevertheless, these studies correspond mainly to moderate temperatures (around 400 °C), with monoaromatics being the main constituents of the coke, whose formation occurs in parallel with the formation of Received: May 2, 2011 Accepted: July 15, 2011 Revised: July 8, 2011 Published: July 16, 2011 9980

dx.doi.org/10.1021/ie200946n | Ind. Eng. Chem. Res. 2011, 50, 9980–9988

Industrial & Engineering Chemistry Research olefins.23
26 Schulz27 emphasized the significance of the effect of reaction temperature on the formation of products and the relationship between temperature and reaction medium composition, on one hand, and coke formation and composition, on the other. Above 475 °C, the monoaromatics derived from the hydrocarbon pool form polyaromatics at the mouth of HZSM5 zeolite micropores, where shape-selectivity restrictions are less severe for coke growth. The viability of the joint upgrading of methanol and paraffins requires a better understanding of catalyst deactivation. To pursue this goal, this work addresses the deposition of coke, its effect on the properties of the catalyst, its structure, and its location in the catalyst.

ARTICLE

Table 1. Physical and Chemical Properties of the Fresh Catalyst property average acid strength, kJ 3 (mol of NH3) total acidity, (mol of NH3) 3 g
1 dp, Å

120

SBET, m2 g
1

220

Sm, m2 g
1

101

Vm, cm3 g
1

0.044

Vp [17 < dp (Å) < 3000], cm3 g
1

0.69

2.3  10
4 102

distribution of volume, vol %

2. EXPERIMENTAL SECTION 2.1. Catalyst. The catalyst was selected in a previous work.18 It

was prepared with a HZSM-5 zeolite, with SiO2/Al2O3 = 30, supplied in ammonium form by Zeolyst International, that was then calcined at 570 °C to obtain the acid form. The zeolite was agglomerated with a binder (bentonite, exaloid) (30 wt %) and alumina (Prolabo, calcined at 1000 °C to become inert) as an inert charge (45 wt %). The catalyst particles were obtained by wet extrusion, using a high-pressure hydraulic piston, through 0.8-mm-diameter holes. The extrudates were first dried at room temperature for 24 h and then sieved to a particle diameter of between 0.15 and 0.3 mm. The particles were dried in an oven at 110 °C for 24 h and then calcined at 575 °C for 3 h. This temperature was reached following a ramp of 5 °C min
1. The agglomeration of the active phase does not significantly reduce acidity, but provides the catalyst with a matrix with mesopores and macropores, which contributes to reducing deactivation by coke deposition16,28 and increasing the hydrothermal stability in the regeneration step by coke combustion.29 The physical properties of the catalyst were measured by N2 adsorption
desorption (Micromeritics ASAP 2010) and Hg porosimetry (Micromeritics Autopore 9220). The total acidity and acid strength of the catalyst were determined by the adsorption
desorption of NH3, combining the techniques of thermogravimetric analysis (TG) and differential scanning calorimetry (DSC), and by temperature programmed desorption (TPD), using a Setaram TG-DSC calorimeter (Setaram 111) connected online to a mass spectrometer from Balzers Instruments.12,30 The Br€onsted/Lewis (B/L) acid site ratio has been determined by analyzing the range of 1400
1700 cm
1 in the FTIR spectrum of adsorbed pyridine, which was obtained using a Specac catalytic chamber connected online to a Nicolet 6700 FTIR spectrophotometer. The Br€onsted/Lewis site ratio at 150 °C was determined from the intensity ratio of the pyridine adsorption bands at 1545 and 1450 cm
1, taking into account the molar extinction coefficients of the two adsorption bands (εB = 1.67 cm μmol
1 and εL = 2.22 cm μmol
1). Table 1 summarizes the physical properties and acidity values of the fresh catalyst. The micropore volume corresponds to the crystals of HZSM-5 zeolite, whereas the volume of meso- and macropores corresponds to the matrix of the catalyst (bentonite and alumina). The low Br€onsted/Lewis ratio is explained by the significant dehydroxylation of the HZSM-5 zeolite due to the high catalyst calcination temperature (575 °C). This treatment is required to keep the catalyst hydrothermally stable and maintain its kinetic performance throughout reaction
regeneration cycles.29,31

value
1

dp (Å) < 20

2.96

20 < dp (Å) 500

46.5 50.5

Br€onsted/Lewis site ratio at 150 °C

1.50

2.2. Reaction and Analysis Equipment. Experimental runs were carried out using equipment described in a previous article.21 The fixed-bed reactor, made of 316 stainless steel, has an internal diameter of 0.9 cm and a length of 10 cm. It is located inside a ceramic-covered stainless steel cylindrical chamber that is heated by an electric resistance and can operate up to 100 bar and 700 °C with a catalyst mass of up to 5 g. The bed consists of a mixture of catalyst and inert solid (carborundum with an average particle diameter of 0.16 mm) to ensure bed isothermality and attain sufficient height under low space time conditions. The temperature is controlled by a digital TTM-125 series controller and measured by a thermocouple (type K) situated in the catalyst bed. Two additional temperature controllers are used: one for the furnace chamber and one for the transfer line between the reactor and the micro-gas chromatograph. The operating variables are controlled by custom software (Process@ from PID Eng&Tech, Madrid, Spain). Experiments were carried out at atmospheric pressure under the following operating conditions: temperature, 550 °C; space time, 2.4 and 4.8 (gcatalyst) h (molCH2)
1; time on stream, 5 h. Runs were carried out with pure feeds of butane and methanol and with methanol/butane mixtures with molar ratios of up to 16/1. The specific conditions used in this work correspond to the state of an energy-neutral process (methanol/ butane molar ratio of 3) and states similar to it.18 A small fraction of the stream made up of unreacted reactants and products (diluted in a He stream of 17 cm3 min
1) was continuously analyzed online in a micro-gas chromatograph (Varian CP-4900). The remaining stream of reaction products passed through a Peltier cell at 0 °C. A level sensor controlled the amount of liquid condensate, and the noncondensable gas flow was vented. The micro-gas chromatograph (with Star Toolbar software) was provided with three analytical modules and the following columns: Porapak Q (PPQ) (10 m), where the lighter products were separated (CO2, methane, ethane, ethylene, propane, propylene, methanol, dimethyl ether, water, butanes, and butenes); a molecular sieve (MS-5) (10 m) where H2, CO, O2, and N2 were separated; and 5CB (CPSIL) (8 m), where the C5
C11 fraction was separated. The compounds were identified and quantified using calibration standards of known concentration. The balance of atoms (C, H, O) was closed in all runs above 99.5%. 2.3. Coke Characterization. The coke content deposited on the catalyst was determined by temperature-programmed 9981

dx.doi.org/10.1021/ie200946n |Ind. Eng. Chem. Res. 2011, 50, 9980–9988

Industrial & Engineering Chemistry Research

ARTICLE

oxidation (TPO) in a SDT 2960 thermobalance (TA Instruments). The procedure involved an initial sweeping with a He stream (Air Liquide, 99.9990% purity) at 550 °C for 30 min to guarantee the absence of O2 in the medium and stabilize the coke, to ensure reproducible results.32
34 The subsequent coke combustion step started by stabilizing the sample at 250 °C for 30 min in a He stream to ensure an inert atmosphere in the chamber until the oxidant mixture (25% O2 in He) was introduced. Combustion followed a temperature ramp of 3 °C min
1 to 550 °C. This temperature was maintained for 1 h to complete coke combustion. The values of temperature, weight, temperature difference between reference and sample, and weight loss derivative were monitored throughout the run. The H/C ratio of the coke was determined by following the signals for CO, CO2, and H2O in a Thermostar mass spectrometer (Balzers Instruments). Calibration factors were used, as determined from the DTG signal registered during the decomposition in an inert atmosphere of a given amount of monohydrate calcium oxalate for CO2 signal and by introducing reference samples of CO and CO2 in known concentrations. The FTIR analysis of the coke components was conducted in a Specac catalytic chamber connected in series to a Nicolet 6700 FTIR spectrophotometer. The procedure started with the preparation of a pellet of catalyst powder pressed onto a KBr support. The sample was heated to 150 °C under a vacuum to remove impurities, and subsequently, an IR spectrum was recorded prior to combustion. Air was used for combustion following a ramp of 3 °C min
1 to 550 °C, and this temperature was maintained for 60 min. IR spectra were recorded every 15 min throughout the entire process.35 Raman spectroscopy was performed using a laser (excitation source) with a wavelength of 514 nm, Renishaw confocal microscope. The analysis was undertaken over 3
5 mg of spent catalyst, performing at least 3 analyses at different positions and reducing the exposure to air for avoiding coke oxidation. The fluorescence caused by coke was subtracted using a baseline of 4 points. To analyze the coke soluble components, extraction from the catalyst was carried out by dissolving the zeolite structure in aqueous hydrofluoric acid and then applying a direct Soxhlet procedure with CH2Cl2 to recover the soluble coke compounds.36,37 The coke molecules soluble in CH2Cl2 were analyzed by two-dimensional gas chromatography/mass spectrometry (GC  GC/MS) and 1H NMR spectroscopy. The equipment used for the analysis was a GC  GC/MS system comprising an Agilent 7890A gas chromatograph (with DB5-MS and HP-INNOWAX columns) connected online to an Agilent 5975C inert XL MSD mass spectrometer. The temperature regime for the chromatographic analysis consisted of a stabilization step at 35 °C for 4 min, followed by a ramp of 3 °C min
1 to 260 °C and temperature stabilization at 260 °C for 5 min. The 1H NMR analysis of the coke was carried out in a Bruker Avance 500 instrument.

3. RESULTS 3.1. Effect of Cofeeding Butane on Deactivation. The reaction products at the reactor outlet were grouped into the following lumps: C2
C4 olefins (ethylene, propylene, and butenes), C2
C4 paraffins (ethane, propane, and isobutane), C5+ aliphatics (olefins and paraffins), aromatics (benzene, toluene, and xylenes), and methane. Another arrangement of the lumps according to the boiling points of the products is as follows: gas (methane, light olefins, and light paraffins) and

Figure 1. Evolution with time on stream of product yields in the transformation of (a) methanol, (b) butane, and (c) methanol/butane with a molar ratio of 3. Reaction conditions: 550 °C, 2.4 gcatalyst h (molCH2)
1.

gasoline (C5+ aliphatics and aromatics). Furthermore, a coke residue formed on the catalyst. 9982

dx.doi.org/10.1021/ie200946n |Ind. Eng. Chem. Res. 2011, 50, 9980–9988

Industrial & Engineering Chemistry Research

ARTICLE

Figure 2. Effect of feed composition on the yields of C2
C4 olefins and gasoline (C5
C11) at zero time on stream and on the content of the coke deposited on the catalyst after 5 h time on stream. Reaction conditions: 550 °C, 2.4 gcatalyst h (molCH2)
1.

Figure 1 shows the evolution of product yields with time on stream for the three different feeds: methanol (Figure 1a), butane (Figure 1b), and methanol/butane with a molar ratio of 3 (Figure 1c). The reaction conditions were a temperature of 550 °C and a space time of 2.4 gcatalyst h (molCH2)
1. It should be noted that, under the conditions used with the mixture (methanol/butane molar ratio of 3 and conditions for complete conversion of both reactants; Figure 1c), the process is energyneutral, that is, the exothermic steps of methanol transformation are balanced by the endothermic steps of butane cracking.21 The yield of each lump was calculated as the ratio between the molar flow rate of the lump in the product stream (Fi) and the molar flow rate in the feed (F0), with both terms being expressed in C-atom content. Yi ¼

Fi  100 F0

ð1Þ

In the transformation of methanol (Figure 1a), the yields of aromatics and paraffins decreased with time on stream as a result of catalyst deactivation, whereas the yield of olefins increased to 42 wt % for 5 h time on stream. In the cracking of butane (Figure 1b) and also when a feed of methanol/butane was used (Figure 1c), the yields of olefins were steady with time on stream, at around 10 and 20 wt %, respectively. The main products were light olefins in the transformation of methanol and methanol/ butane and paraffins in the cracking of butane. For the three feeds, the yield of paraffins (excluding butane when it is a reactant) was the one that was most influenced by catalyst deactivation. Furthermore, deactivation in the transformation of methanol (Figure 1a) increased the yield of methane (34% for 5 h time on stream), which is formed by the thermal decomposition of methanol (this step occurs with simultaneous dehydrogenation of other reactants, products, or coke). The formation rate of methane in the transformation of methanol/ butane (Figure 1c) is lower because of the attenuation of deactivation by cofeeding butane.21

Figure 2 shows the effects of the methanol/butane molar ratio in the feed on both the yields of product fractions (olefins and C5
C11 fractions, corresponding to zero time on stream) and the coke content deposited for 5 h time on stream (determined by TPO analysis). This allows the extent of coke deposition to be related to the reaction medium composition. The results show that the coke content increased as the methanol/butane molar ratio in the feed increased, in parallel with the increase in the yields of olefins and the C5+ aliphatic fraction and its concentration in the reaction medium. Furthermore, Mier et al.21,22 indicated that the process is energy-neutral for a methanol/butane molar ratio of 3. Accordingly, the reaction is endothermic for molar ratios below 3 and exothermic above this value, which can have an influence on the temperature in the proximity of the acid sites and, consequently, on the mechanisms of coke formation and growth on these sites. These mechanisms will presumably be enhanced under conditions in which the process is exothermic. The evolution of product fraction yields with time on stream was studied in a previous work for a wide range of space time values and in the 400
550 °C temperature range, and a expression was determined for the deactivation kinetics by coke (for a methanol/butane molar ratio of 3 in the feed)22


da ¼ ½kd1 ðyM þ yD Þ þ kd2 ðyO þ yG Þa dt

ð2Þ

The concentrations of the components in eq 2 (yi) are expressed in molar fraction. Activity (a) is defined as the ratio of the reaction rates at time t and zero time on stream a¼


rA ð
rA Þt ¼ 0

ð3Þ

Equation 2 shows that deactivation is dependent on the concentration in the reaction medium of the lumps of oxygenates (methanol and dimethyl ether, yM + yD), C2
C4 olefins (yO), and C5
C11 fraction or gasoline (yG). Because increasing the proportion of butane leads to a decrease in the the concentrations of oxygenates, olefins, and gasoline, the activity (eq 2) decreases less rapidly. This dependency between deactivation and the concentration of certain reactants or products has previously been reported for the transformations of methanol19,38,39 and ethanol40 into hydrocarbons on HZSM-5 zeolite. 3.2. Deterioration of Catalyst Properties. Table 2 reports the physical properties of the catalyst used in the runs with the three feeds in Figure 1. A comparison between these values and those for the fresh catalyst (Table 1) shows that the surface area decreased considerably, from 220 to 138
149 m2 g
1, in the three experiments. This decrease was caused by the reduction in the micropore area (from 101 to 44
49 m2 g
1). The micropore volume decreased by 50% (from 0.044 to 0.020
0.025 cm3 g
1) and, consequently, the average pore diameter increased (from 102 to 121
135 Å). This selective deterioration of micropore properties is evidence that coke deposition preferably took place on them. The coke content deposited on the catalyst (determined by TG
TPO analysis) is of the same order for the runs of methanol and methanol/butane transformation (with a molar ratio of 3), at 4.8 and 4.6 wt %, respectively, and is much lower for butane cracking (0.65 wt %). This difference is not consistent with the similar deterioration of surface area observed for the three feeds 9983

dx.doi.org/10.1021/ie200946n |Ind. Eng. Chem. Res. 2011, 50, 9980–9988

Industrial & Engineering Chemistry Research

ARTICLE

Table 2. Physical Properties of the Catalysts Deactivated after 5 h of Reaction and Coke Content SBET (m2 3 g
1)

Vm (cm3 3 g
1)

Vp (cm3 3 g
1)

Sm (m2 3 g
1)

dp (Å)

coke (wt %)

methanol

141

0.025

0.52

48

121

4.8

butane

138

0.020

0.51

46

135

0.65

methanol/butane = 3

149

0.022

0.56

49

132

4.6

feed

Figure 3. Effect of feed composition on the TPO curves corresponding to the combustion of the coke deposited on the catalyst. Reaction conditions: 550 °C, 2.4 gcatalyst h (molCH2)
1, 5 h time on stream.

and is explained by the fact that the coke deposited on the catalyst mesopores and macropores has a considerable surface area under the low-temperature conditions used for the measurement of this property by N2 adsorption
desorption. On the other hand, the yield of coke in the worst-case scenario was lower than 0.2 wt %, allowing us to exclude its values from those shown in Figure 1. 3.3. Coke Properties. TG
TPO Analysis. Figure 3 shows TG
TPO profiles corresponding to the combustion of the coke in the catalysts used with four different feeds (those corresponding to the experiments in Figure 1 and to a feed of methanol/ butane with a molar ratio of 1.5). Only one peak was observed for the four feeds at approximately 550 °C. This result of only one peak is characteristic of a uniform composition of coke, which is explained by the high reaction temperature (550 °C), which enhanced coke development toward a condensed structure with a low H/C ratio.7,10,27,41 Nevertheless, whereas the TPO curve for butane cracking was in a narrow temperature range, when the methanol content in the feed increased, the amplitude of the TPO curve also increased, which is evidence of a more heterogeneous coke. FTIR Spectroscopy. Figure 4 shows two regions of molecular vibration in the infrared spectrum of the coke. These results are shown as an example and correspond to a catalyst spent under the following operating conditions: temperature of 550 °C, space time of 2.4 gcatalyst h (molCH2)
1, feeding butane 5 h time on stream. Figure 4a corresponds to the 2800
3100 cm
1 region, in which the stretching vibrations of aliphatics (paraffins and olefins) are observed, and Figure 4b corresponds to the 1300
1700 cm
1 region associated with the unsaturated coke components (polyolefins and condensed aromatics) and bending vibrations

Figure 4. FTIR spectra of the coke deposited on the catalyst: (a) 2800
3100 and (b) 1300
1700 cm
1. Reaction conditions: 550 °C, 2.4 gcatalyst h (molCH2)
1, 5 h of time on stream.

of aliphatics. The IR vibrational spectrum was deconvoluted into several Lorentzian peaks, and each one was assigned to the vibration of molecular bonds in the coke as follows:42 1410 and 1470 cm
1, identified but not assigned, as a variety of bands overlapped in this region; 1580 cm
1, coke or polyaromatics; 1640 cm
1, olefins; 2860 cm
1,
CH2 groups; 2895 cm
1,
CH3; 2930 cm
1,
CH and
CH2; and 2960 cm
1,
CH3. Figure 5 shows the effect of feed composition on the intensity of each band characteristic of coke and, thereby, the effect on its structure. It is noteworthy that the band at 1580 cm
1 (the one of higher intensity even when the methanol content in the feed 9984

dx.doi.org/10.1021/ie200946n |Ind. Eng. Chem. Res. 2011, 50, 9980–9988

Industrial & Engineering Chemistry Research

ARTICLE

Figure 5. Effect of feed composition on the intensity of the FTIR spectral bands characteristic of coke. Reaction conditions: 550 °C, 2.4 gcatalyst h (molCH2)
1, 5 h time on stream.

was low) associated with the catalytic origin43 increased as the methanol/butane ratio in the feed increased. The coke formed in the cracking of butane is constituted mainly by long branched aliphatics, with a high proportion of simple and conjugated double bonds and a higher concentration of
CH2 and CH groups than
CH3 groups. As the methanol content in the reaction medium increased (and likewise the concentration of aromatics and olefins, Figure 2), the fraction of polyaromatics (coke band) increased, and the fractions of olefinic and paraffinic groups in the coke decreased. When there was methanol in the feed, the coke was mainly aromatic, with a small number of olefinic groups and an even smaller number of paraffinic groups. The bands of
CH groups (2930 cm
1) and olefins (1640 cm
1) exhibited a minimum at methanol/butane molar ratio of 3, which is an indication of a change in the reaction mechanism under these conditions. Raman Spectroscopy. Figure 6 shows the results of Raman spectroscopy for the catalysts deactivated in Figure 1 runs. The Raman spectrum intensity decreased as the methanol content in the feed increased, and accordingly, an enlarged ordinate scale was used as the content of methanol in the feed became higher. The spectra were deconvoluted into five Lorentzian peaks:44 At 1200
1220 cm
1, there is a band of C
H bond vibrations (νC
H). At 1340 cm
1, there is a D band caused by the “breathing” mode of disordered aromatic structures. At 1520
1550 and 1610 cm
1, there are two peaks caused also by disordered structures and called D3 and D2, respectively. At 1595 cm
1, there is a band characteristic of the in-plane stretching of sp2 carbons of aromatics and olefins (known as G from graphite), forming highly ordered coke. The intensity of νC
H vibrations increased when the concentration of butane in the feed increased, which is evidence of a higher concentration of aliphatic compounds in the coke. From the point of view of coke structure, the more interesting peaks are the G band (highly ordered coke) and D band (poorly ordered coke).42,45 As observed in Figure 6, the intensity of the G band (an index of coke ordering degree) increased as the methanol/ butane ratio in the feed increased. Furthermore, the D band also increased, so that the D/G ratio for the different feeds had the

Figure 6. Raman spectra of the coke deposited on the catalyst in the transformation of (a) butane, (b) methanol/butane with a molar ratio of 3, and (c) methanol. Reaction conditions: 550 °C, 2.4 gcatalyst h (molCH2)
1, 5 h time on stream.

same value of around 2, which corresponds to a coke particle size of 2 nm.46 H/C Molar Ratio. As an example of the results obtained for coke combustion, Figure 7 shows those for the weight change derivative (DTG) and heat generation (Figure 7a) and the evolution with time of the formation rate of combustion gases (CO2 and H2O) (Figure 7b). These results correspond to the catalyst deactivated under conditions of a temperature of 550 °C, a space time of 4.8 gcatalyst h (molCH2)
1, and a methanol/butane molar ratio of 3, and they allow for the determination of the H/C ratio of the coke deposited both for methanol/butane (Figure 7) and for pure methanol and butane in the feed. The ratio for methanol in the feed is H/C = 0.68, which corresponds to an evolved coke of aromatic nature. The ratio for butane in the feed is H/C = 1.68, which corresponds to coke of a paraffinic nature. When butane and methanol were cofed, the H/C ratio decreased to 0.90, corresponding to a coke with a significant aromatic content, albeit less evolved toward condensed structures than in the case of pure methanol. Soluble Coke Analysis. It should be noted that the coke was almost completely soluble in CH2Cl2, given that the fraction of insoluble coke under all conditions was below 2 wt % of the total coke. Figure 8 shows the contour plot of a chromatogram for the GC  GC analysis (using a flame ionization detector for quantification and a mass spectrometer for identification) of the coke deposited under given reaction conditions: temperature of 550 °C, space time of 4.8 gcatalyst h (molCH2)
1, and methanol/ butane molar ratio of 3. Different molecule families are identified: (i) aliphatic
naphthenic compounds of low polarity, which means that they are in the lower section of the figure; (ii) 9985

dx.doi.org/10.1021/ie200946n |Ind. Eng. Chem. Res. 2011, 50, 9980–9988

Industrial & Engineering Chemistry Research

ARTICLE

Figure 7. Results of coke combustion for a catalyst deactivated in the transformation of a methanol/butane mixture with a molar ratio of 3: (a) rate of weight loss and heat generation, (b) formation of combustion products. Reaction conditions: 550 °C, 4.8 gcatalyst h (molCH2)
1, 5 h time on stream.

single-ring aromatics; (iii) two-ring aromatics (mainly naphthalene and biphenyl); (iv) three-ring aromatics (anthracene and phenantrene) and even traces of four-ring molecules (