Professional Documents
Culture Documents
a r t i c l e
i n f o
Article history:
Received 21 January 2014
Received in revised form 16 April 2014
Accepted 22 April 2014
Available online 30 April 2014
Keywords:
Ethene
Propene intensication
MFI zeolite
CHA chabazite
Coke deactivation
a b s t r a c t
The deactivation of HZMS-5 and SAPO-34 catalysts has been studied in the transformation of ethylene
and 1-butene under propylene intensication conditions. The deterioration of spent catalysts physical
properties have been quantied and coke has been characterized by TPO and by several spectroscopic
techniques (Raman, 13C NMR, FTIR, FTIR-TPO), in order to determine the effect reaction medium composition and the severity of catalyst shape selectivity have on the nature and location of the coke in the porous structure. The results reveal that the mechanism for coke deactivation consists of two steps: one for
the formation of alkylated aromatics by oligomerization and another for the coke growth-condensation.
The rst step is analogous for both catalysts and it principally depends on the catalyst acid strength and
acid site density. The second step is different for both catalysts: the microporous structure of SAPO-34,
with cavities in the intersections, inhibits the diffusion of alkylated aromatics towards the outside of
the structure, thus blocking active acid sites; whereas, HZSM-5 structure, with a high connectivity and
without cavities, favors the diffusion of the aromatics that evolve for a longer time outside of the micropores. At process conditions, the results demonstrate that the coke formation is faster from ethylene than
from 1-butene, due to the lower reactivity of ethylene for oligomerization-cracking mechanisms as well
as its higher capability for coke formation.
2014 Elsevier Inc. All rights reserved.
1. Introduction
The increasing light olen demand is boosting the intensication of conventional processes for its production from crude oil,
by steam cracking and uid catalytic cracking (FCC) [1,2], and from
other fossil sources (natural gas, coal) by the MTO (Methanol to
Olens) process [3]. New catalytic routes from fossil fuels are currently at different developing stages, such as: (i) the transformation of DME (DTO process) obtained from natural gas or coal [4];
(ii) methane transformation (natural gas) via chloromethane [5];
(iii) coupled methanol-hydrocarbon cracking (CMHC) [6]. In addition, light olen production is a priority goal in biomass valorization routes, by ethanol or bio-oil catalytic transformation [79]
and in the valorization of polyolenic plastic wastes [10,11].
Although the aforementioned processes are selective for light
olen production, they do not satisfy the increasing propylene
demand (with an annual rate of 5.7%), due to the increase in
the production of petrochemicals, such as polypropylene and
Corresponding author. Tel.: +34 94601 8435; fax: +34 94 601 3500.
E-mail address: pedro.castano@ehu.es (P. Castao).
http://dx.doi.org/10.1016/j.micromeso.2014.04.040
1387-1811/ 2014 Elsevier Inc. All rights reserved.
285
Nomenclature
Roman symbols
B/L
Brnsted/Lewis ratio
Cc
coke content (wt.%)
D, G
bands in the Raman spectrum
dp
pore diameter ()
fc1
fraction of coke type I (%)
Fi
molar owrate of i lump in the product stream ((mol of
C) h1)
F0 and F butene molar owrate in the feed and in the outlet
stream ((mol of C) h1)
Gp
position of G band (cm1)
Gw
width of G band (cm1)
La
SBET
Sm
Si
Vm
Vp
X
Greek symbols
eL , eB
molar extinction coefcients (cm lmol1)
2. Experimental
2.1. Catalysts
HZSM-5 zeolite was supplied in ammonium form by Zeolyst
International (SiO2/Al2O3 = 80), and was calcined at 570 C to
obtain the acid form. SAPO-34 (of composition (SiO2)0.08
(Al2O3)2.8 (P2O5)0.98 (HCl)0.02 (C8H19N) 41H2O) was prepared following the method of Lok et al. [49], with tetraethyl ammonium
hydroxide (TEAOH) at 20 wt.% as template. The removal of the
template was carried out by calcination in a mufe furnace at
575 C for 6 h with a heating rate of 5 C min1.
The catalysts were obtained by agglomerating the active phase
(25 wt.%) by wet extrusion with bentonite (Exaloid, 30 wt.%) as a
binder and alumina (Martinswek, 45 wt.%) as inert charge. This
agglomeration is required in order to: (i) confer a suitable mechanical and hydrothermal resistance; (ii) obtain higher accessibility of
the reactants to the active phase, and; (iii) attenuate deactivation
by coke, since the coke deposition on the mesoporous matrix minimizes the blockage at the mouth of micropores [50]. The extrudates were rst dried (110 C, 24 h) and then were sieved to a
particle diameter between 0.15 and 0.3 mm. Finally, the catalysts
were calcined at 570 C for 2 h. This treatment is required in order
to achieve a balance of the acid sites, which allows keeping the catalyst hydrothermally stable and maintaining its kinetic performance throughout reaction-regeneration cycles [46].
The surface area and porous structure were measured by N2
adsorption-desorption (Micromeritics ASAP 2010). The total acidity and acid strength of the catalysts were determined by monitoring the adsorption-desorption of NH3, by combining the techniques
of thermo-gravimetric analysis and differential scanning calorimetry using a Setaram TGDSC calorimeter connected on-line with a
Thermostar mass spectrometer from Balzers Instruments [51]. The
Brnsted/Lewis (B/L) acid site ratio has been determined by analyzing the region of 14001700 cm1 in the FTIR spectrum of
adsorbed pyridine, which has been obtained using a Specac catalytic chamber connected on-line with a Nicolet 6700 FTIR spectrometer. The results have been determined from the ratio
between the intensity of pyridine adsorption bands at 1545 and
1450 cm1 and taking into account the molar extinction coefcients of both adsorption bands (eB = 1.67 cm lmol1 and
eL = 2.22 cm lmol1) [52].
Table 1 summarizes the physical and acid properties of the
fresh active phases and catalysts (after agglomeration). SAPO-34
shows a higher BET surface and micropore volume in comparison
to HZSM-5 zeolite. The catalyst micropore volume corresponds to
the active phase (zeolite or SAPO), whereas the volume of mesoand macropores corresponds to the matrix of the catalyst
286
Table 1
Properties of the active phases and agglomerated catalysts.
HZSM-5
SAPO-34
Active phase
SBET (m2 g1)
Sm (m2 g1)
Vp (m3 g1)
Vm (m3 g1)
Total acidity (mmolNH3 g1)
Acid strength (kJ mol1
NH3)
B/L ratio at 150 C
556
181
0.11
0.10
0.42
144
0.32
611
288
0.06
0.24
0.64
153
Agglomerated catalyst
SBET (m2 g1)
Sm (m2 g1)
Vp (m3 g1)
Vm (m3 g1)
dp ()
209
68
0.44
0.04
173
215
103
0.20
0.05
37
The coke deposited on the catalysts was characterized by different techniques developed in previous works [4547]. The coke
content was determined by temperature-programmed oxidation
(TPO) in a TGA Q5000 thermobalance (TA Instruments), following
a ramp of 8 C min1 from 250 C to 600 C. FTIR and combined
FTIR-TPO analysis were performed in a Nicolet 6700 (Thermo)
using a transmission cell (60 scans, and resolution of 4 cm1).
These analyses were carried out following a temperature ramp of
5 C min1 from 100 C to 550 C. Solid state CP-MAS (cross-polarization magic-angle spinning) 13C NMR spectra were obtained in a
Broker DXR 300. Raman spectroscopy was performed in a Reinshaw confocal microscope using an excitation source of 514 nm,
and subtracting the uorescence caused by coke.
3. Results
F0 F
100
F0
287
Si
Fi
100
F0 F
Table 2
Physical properties of the catalysts deactivated after TOS = 5 h.
Catalyst
Ethylene
HZSM-5
SAPO-34
91
89
1-Butene
HZSM-5
SAPO-34
97
76
Sm (m2 g1)
8.3
0
10
0
Vp (m3 g1)
Vm (m3 g1)
dp ()
0.35
0.14
0.005
0
197
62
0.37
0.12
0.006
0
185
63
Fig. 3. TPO proles for the coke deposited on the catalysts used in the transformation of ethylene and 1-butene.
288
that are activated by the acid sites mainly located within the
crystals.
TPO proles in Fig. 3 have been deconvoluted into two gaussian
peaks (as shown in the lower signal in Fig. 3), to quantify the total
coke content (Cc) and the amount of coke I fraction (fI) that are
summarized in Table 3. The coke content is similar for both catalysts in the transformation of ethylene (around 4.6 wt.%), even
though the effect of coke on deactivation is different (Fig. 1).
SAPO-34 catalyst activity is null at 5 h time on stream, whereas
for HZSM-5 catalyst, ethylene conversion is 78%. This higher stability of HZSM-5 zeolite (also observed by comparing the results in
Fig. 1 for 1-butene transformation), for the same coke content, is
due to its porous structure without cages in the crystalline channel
intersections, which avoids the formation of polyaromatic coke
structures within the crystalline structure [57,59]. The presence
of these cages in the SAPO-34 favors the retention of the alkylated
aromatics generated by oligomerization and hinders coke combustion; therefore the maximum coke combustion temperature is
20 C higher for the coke deposited on SAPO-34 (Fig. 3).
Cc (wt.%)
f C1 (%)
Ethylene
HZSM-5
SAPO-34
4.7
4.6
47.1
10.7
1-Butene
HZSM-5
SAPO-34
2.7
4.1
37.8
12.2
(a)
(b)
Fig. 4. FTIR spectra of the catalysts spent in the transformation of ethylene and 1-butene in the 1300-1800 (a) and 2800-3100 (b) cm-1 ranges.
289
between the maximums observed in TG-TPO (Fig. 3) and FTIRTPO (Fig. 6) analysis is due to the coke sweeping in the former,
which promotes coke aging reactions [62].
The combined results of TG-TPO and FTIR-TPO imply that coke I
burns easily due to its higher H/C ratio and its location in the mesopores of the catalyst. Coke II is either more condensed (as seen in
Fig. 6) or it is presumably located within the micropores of the zeolite. It burns more slowly and requires higher combustion temperatures; thus, the coke II deposited on SAPO-34 needs 20 C more to
burn in comparison to the one deposited on HZSM-5 zeolite,
because of the limited O2 access as well as O2-coke contact. An initial step for the formation of a coke of lower molecular weight
(alkylated aromatics) should be considered that is favored by the
high density of strong acid sites and because of the cages in the
intersection between micropores [63]. Moreover, the aliphatic
compounds of coke I probably evolve towards more condensed
structures, as a high mesopore volume is able to contain coke
molecules.
Fig. 5. Relative intensities of the FTIR bands corresponding to the catalysts spent in
the transformation of ethylene and 1-butene.
Fig. 6. Evolution of differential intensities of absorbance bands with time during the temperature programmed combustion of the coke deposited on H-ZSM5 catalyst in the
transformation of ethylene (a) and 1-butene (b).
290
Fig. 8. 13C NMR spectrum of the catalysts spent in the transformation of ethylene
and 1-butene.
Fig. 7. Raman spectra of the catalysts spent in the transformation of ethylene and
1-butene.
Table 4
Representative parameters of the Raman bands, corresponding to the coke deposited
on the catalysts.
a
b
Catalyst
Gp (cm1)
Gw (cm1)
La (nm)a
D/G
La (nm)b
Ethylene
HZSM-5
SAPO-34
1599.3
1595.9
43.9
69.4
1.12
1.04
2.03
1.18
1.92
1.46
1-Butene
HZSM-5
SAPO-34
1601.3
1594.3
44.9
53.4
1.12
1.09
1.97
1.50
1.89
1.65
Table 5
Fraction of the bands deconvoluted from the
coke deposited on the catalysts.
13
Band
Assignation
HZSM-5
SAPO-34
Ethylene
1014 ppm
1822 ppm
2832 ppm
129 ppm
141 ppm
CH3
CH2
CH
CAH
CPAH
0.180
0.219
0.240
0.351
0.011
0.010
0.090
0.014
0.886
0.000
1-Butene
1014 ppm
1822 ppm
2832 ppm
129 ppm
141 ppm
CH3
CH2
CH
CAH
CPAH
0.079
0.176
0.145
0.545
0.054
0.015
0.146
0.119
0.719
0.000
4. Discussion
The results in Section 3 allow analyzing the effect of the reaction medium composition and catalyst shape selectivity on the
magnitude of coke deposition and its structure. The higher coke
formation velocity in the transformation of ethylene is the result
of two circumstances: (i) the higher capability of ethylene for condensation in order to form coke compounds, with respect to other
olens; (ii) its lower reactivity for oligomerization-cracking mechanisms; thus, ethylene concentration in the reaction medium is
higher than the corresponding to 1-butene. This selective effect
of light olen lump concentration on deactivation provides
interesting information to be considered in a kinetic model for
the deactivation of olen interconversion reactions.
291
Fig. 9. Scheme of the internal coke precursor retention and of the composition of the external coke deposited on HZSM-5 zeolite and SAPO-34.
292
and olenic. Due to the low capability for sweeping alkylated aromatic compounds towards the exterior, the fraction of external
coke is small and the internal coke fraction is formed, based on
the deactivation prole (Fig. 1), at low values of time on stream.
Furthermore, the deactivation of HZSM-5 zeolite catalyst is less
severe (Fig. 1), owing to the expulsion of the alkylated aromatic
compounds formed in the rst step, which are deposited on the
mesopores and grow by aromatic and aliphatic condensation. The
nal composition of the coke is more condensed-structured, with
long parafnic chains (Fig. 8) and a lower olenic content. These
results of composition and growth are based on the external location of the coke outside of the micropores and justify the bimodal
distribution of coke I and II. The good qualities of HZSM-5 zeolite
for facilitating the circulation of the potential coke precursors
and preventing the micropore blockage have been highlighted in
the literature for parafn [41] and polyolen cracking [45]. In addition, using a high gas ow rate and agglomerating HZSM-5 zeolite
in a mesoporous matrix that houses the external coke are suitable
strategies for avoiding the blockage of the pore mouths [43].
Regarding the results obtained, catalyst selection is important
in order to attenuate the deactivation by coke in these reactions.
HZSM-5 zeolite deactivates slower and moderating acid strength
attenuates this deactivation. This strategy is feasible in the transformation of 1-butene at the studied temperature, on a zeolite with
high SiO2/Al2O3 ratio that can be steamed or doped in order to control its acid strength [9,20,21,23,26]. However, the transformation
of ethylene at the studied temperature requires a considerable acid
strength to ensure catalyst activity, thus, the attenuation of acid
strength should be accomplished by a raise in temperature for
maintaining a high conversion level. SAPO-34 deactivates more
rapidly and has more hurdles for modifying its porous structure
and acidity for attenuating the deactivation by coke [71].
The ability of HZSM-5 zeolite for avoiding micropore blockage
under the studied reaction conditions explains the lower deactivation rate of the catalyst than that expected for the coke content
obtained or the deterioration of surface area. HZSM-5 has a highly
connected micropore network that enhances the diffusion, without
cages in their intersections. In addition, the conned molecules
have a signicant mobility across the zeolite and their growth is
limited due to the relatively small diameter of the pores, thus hindering the condensation towards bigger molecules [41]. Furthermore, the mobility of coke precursors enhances their circulation
towards the outside of HZSM-5 zeolite, and therefore a signicant
fraction of the coke corresponds to that deposited on the outside of
the crystals. In the case of SAPO-34, the intersections between the
channels are cages that easily inhibit the circulation of the reactants. Thus, the outward circulation of intermediate compounds
responsible for coke formation is also inhibited and so the coke
deposited on this catalyst is mainly internal coke. Thus, the key
point is that the location of coke in HZSM-5 catalyst would not
be necessarily close to the acid sites; coke is denitively formed
on the acid sites, however it migrates to the exterior of the zeolite
and grows, whereas in the case of SAPO-34 it stays trapped within
the micropores.
In addition to the catalyst properties, the dilution of the feed
with inerts as N2 can contribute to attenuating deposition by coke,
thereby decreasing the concentration of coke precursors and the
velocity for their condensation [41]. Dilution with H2 would presumably be more effective for decreasing coke deposition since it
will also hinder the condensation of coke precursors that takes
place by dehydrogenation [57]. This hydrogenation can be periodically applied for catalyst rejuvenation between reaction steps,
with an expected partial recovery of activity, whose total recovery
will require coke combustion with air.
5. Conclusions
HZSM-5 zeolite and SAPO-34 catalysts are interesting alternatives for the selective light olen production and provide good
prospects for intensifying propylene production from ethylene
and 1-butene. However, their use is conditioned by the fast deactivation by coke. The results reveal that 1-butene shows better prospects than ethylene for obtaining propylene, owing to its higher
reactivity and lower capability for forming coke. Nevertheless, considering coke composition, no signicant differences are observed
that could indicate a different deactivation mechanism in the
transformation of ethylene and 1-butene.
Coke formation on both catalysts takes place following two
steps: the formation of alkylated aromatics and their growth to
form coke. The oligomers are presumably the intermediate compounds of alkylated aromatic formation and oligomerizationcracking reactions, which are crucial for interconverting olens.
The formation of alkylated aromatics as well as the reactivity of
intermediates is conditioned by the density and acid strength of
the sites, and these properties, among others, justify the amount
and impact of the coke.
The shape selectivity of SAPO-34 is a key factor for the high propylene selectivity giving way to high olen (propylene) selectivity
at zero time on stream. However, this catalyst shows a fast deactivation caused by the pore blockage due to the alkylated aromatic
retained in the internal cages of microporous intersections, giving
way to a coke with a high amount of non-condensed rings and of
olenic nature.
HZSM-5 zeolite deactivates more slowly as these alkylated aromatics are swept to the outside of the micropores, and therefore,
they are deposited and grow on the mesopores of the matrix. The
higher micropore diameter and the absence of cages in the intersections enable this external deposition. Consequently, the coke
deposited on HZSM-5 zeolite catalyst has a higher external coke
fraction and is mainly composed of condensed aromatics and long
aliphatic chains. The ability for the internal circulation of coke precursors is favored by agglomerating the HZSM-5 zeolite in a mesoporous matrix, which allows the deposition and evolution of the
coke on the outside of the zeolite thus avoiding the blockage of catalytic acid sites.
Acknowledgements
The nancial support of this work was undertaken by the Ministry of Economy and Competitiveness (MINECO) of the Spanish
Government (CTQ2010-19188 and CTQ2010-19623 projects), by
the Basque Government (Project IT748-13) and by the University
of the Basque Country (UFI 11/39). E. Epelde (BFI08.122) and M.
Ibaez (BFI-2012-203) are grateful for their Ph.D. Grants from the
Department of Education, University and Research of the Basque
Country. The technical and human support provided by SgIker
(UPV/EHU, MICINN, GV/EJ, ESF) is gratefully acknowledged.
References
[1]
[2]
[3]
[4]
293
[39] K. Qian, D.C. Tomczak, E.F. Rakiewicz, R.H. Harding, G. Yaluris, W.C. Cheng, X.
Zhao, A.W. Peters, Energy Fuels 11 (1997) 596600.
[40] H.S. Cerqueira, C. Sievers, G. Joly, P. Magnoux, J.A. Lercher, Ind. Eng. Chem. Res.
44 (2005) 20692077.
[41] M. Guisnet, L. Costa, F. Rama, J. Mol. Catal. A: Chem. 305 (2009) 6983.
[42] H. Hu, F. Cao, W. Ying, Q. Sun, D. Fang, Chem. Eng. J. 160 (2010) 770778.
[43] H. Schulz, Catal. Today 154 (2010) 183194.
[44] A.T. Aguayo, P. Castao, D. Mier, A.G. Gayubo, M. Olazar, J. Bilbao, Ind. Eng.
Chem. Res. 50 (2011) 99809988.
[45] P. Castao, G. Elordi, M. Olazar, A.T. Aguayo, B. Pawelec, J. Bilbao, Appl. Catal. B:
Environ. 104 (2011) 91100.
[46] M. Ibaez, B. Valle, J. Bilbao, A.G. Gayubo, P. Castao, Catal. Today 195 (2012)
106113.
[47] B. Valle, P. Castao, M. Olazar, J. Bilbao, A.G. Gayubo, J. Catal. 285 (2012) 304
314.
[48] M. Ibez, M. Artetxe, G. Lopez, G. Elordi, J. Bilbao, M. Olazar, P. Castao, Appl.
Catal. B: Environ. 148149 (2014) 436445.
[49] B.M. Lok, C.A. Messina, R.L. Patton, R.T. Gajek, T.R. Cannan, E.M. Flanigen, J. Am.
Chem. Soc. 106 (1984) 60926093.
[50] A.G. Gayubo, A. Alonso, B. Valle, A.T. Aguayo, M. Olazar, J. Bilbao, Fuel 89 (2010)
33653372.
[51] A.T. Aguayo, A.G. Gayubo, R. Vivanco, M. Olazar, J. Bilbao, Appl. Catal. A: Gen.
283 (2005) 197207.
[52] C.A. Emeis, J. Catal. 141 (1993) 347354.
[53] R.J. Quann, L.A. Green, S.A. Tabak, F.J. Krambeck, Ind. Eng. Chem. Res. 27 (1988)
565570.
[54] H. Oikawa, Y. Shibata, K. Inazu, Y. Iwase, K. Murai, S. Hyodo, G. Kobayashi, T.
Baba, Appl. Catal. A: Gen. 312 (2006) 181185.
[55] X. Zhu, S. Liu, Y. Song, L. Xu, Appl. Catal. A: Gen. 288 (2005) 134142.
[56] X. Tang, H. Zhou, W. Qian, D. Wang, Y. Jin, F. Wei, Catal. Lett. 125 (2008) 380
385.
[57] M. Guisnet, P. Magnoux, Appl. Catal. A: Gen. 212 (2001) 8396.
[58] F. Bauer, H.G. Karge, Mol. Sieves 5 (2007) 249364.
[59] B. Valle, A. Alonso, A. Atutxa, A.G. Gayubo, J. Bilbao, Catal. Today 106 (2005)
118122.
[60] H.G. Karge, W. Nieen, H. Bludau, Appl. Catal. A: Gen. 146 (1996) 339349.
[61] J. Robertson, Mater. Sci. Eng. R 37 (2002) 129281.
[62] A. Marcilla, A. Gomez-Siurana, F.J. Valdes, Appl. Catal. A: Gen. 334 (2008) 20
25.
[63] B. Wang, G. Manos, Ind. Eng. Chem. Res. 47 (2008) 29482955.
[64] J. Schwan, S. Ulrich, V. Batori, H. Ehrhardt, S.R.P. Silva, J. Appl. Phys. 80 (1996)
440447.
[65] A.C. Ferrari, J. Robertson, Phys. Rev. B 61 (2000) 1409514107.
[66] A.C. Ferrari, A. Libassi, B.K. Tanner, V. Stolojan, J. Yuan, L.M. Brown, S.E. Rodil, B.
Kleinsorge, J. Robertson, Phys. Rev. B 62 (2000) 11089.
[67] J.L. Bonardet, M.C. Barrage, J. Fraissard, J. Mol. Catal. A: Chem. 96 (1995) 123
143.
[68] Y. Jiang, J. Huang, V.R. Reddy, Microporous Mesoporous Mater. 105 (2007)
132139.
[69] G. Seo, J.H. Kim, H.G. Jang, Catal. Surv. Asia 17 (2013) 103118.
[70] M. Zokaie, D.S. Wragg, A. Gronvold, T. Fuglerud, J.H. Cavka, K.P. Lillerud, O.
Swang, Microporous Mesoporous Mater. 165 (2013) 15.
[71] D. Chen, K. Moljord, A. Holmen, Microporous Mesoporous Mater. 164 (2012)
239250.