(pyridine adsorption method). It can be seen, by referring, Fig 2 that

advertisement
Catalytic Cracking of a Mixture of Dodecane and 1,3,5 Triisopropyl-benzene over USY and ZSM-5 Zeolites Based Catalysts
N. Al-Baghli and S. Al-Khattaf*
Department of Chemical Engineering, King Fahd University of Petroleum & Minerals, P.O.
989, Dhahran 31261, Saudi Arabia., E-mail address: skhattaf@kfupm.edu.sa
The reaction of dodecane (C12) and 1,3,5.tri-isopropyl-Benzene (TIPB) was experimentally
investigated over catalyst SB (based on USY zeolite) and GKF-3 (based on ZSM5 zeolites).
Series of experiments with pure and 50-50 weight percent mixtures of C12 and TIPB were
conducted. The reaction conditions were adjusted to simulate a commercial FCC unit at 450
ºC, cat/oil ratio of 5, pressure of 1-1.5 bar, and reaction times of 3-15 sec. GKF-3 was
observed to give higher C12 conversion compared to SB. On the contrary, the kinetic of
TIPB showed a complicated behavior due to the mutual effects of zeolite acidity and pore
size. The conversion of pure C12 on SB was observed to be lower than the conversion of
TIPB while the opposite was true on GKF-3. This behaviour could be attributed to the higher
activation energy of paraffin catalytic cracking compared to alkyl-aromatic dealkylation.
While, the conversion of C12 was not affected by the presence of TIPB in both catalysts a
huge drop in the conversion of TIPB was observed in the presence of C12 over both catalysts.
This indicates that the diffusion of TIPB molecules is the rate determining step.
1. INTRODUCTION
The process of Fluid Catalytic Cracking (FCC) is one of the most important processes for
gasoline production. The catalyst in the FCC process is based on Y-zeolite as the main
component and ZSM-5 as an additive. These zeolites are bounded individually in a spraydried matrix such as silica-alumina. While diffusion of hydrocarbons in the catalyst matrix
belongs to the well-known Knudsen regime, diffusion in zeolites falls into the configurational
regime [1-5]. However, it essential not to view hydrocarbon molecules as rigid bodies [6].
Even if molecule critical diameters are greater than the zeolite super cage, these hydrocarbon
molecules are able to diffuse inside the zeolite structure. However, there is a cut off size for
each zeolite. Molecules with a size bigger than this cut off size can not transport inside this
particular zeolite [2,6]. For example, molecules with a critical diameter up to 10 Å can enter
Y-zeolite [2]. Thus it is possible for TIPB which has a critical diameter of 9.5 Å to enter into a
Y-zeolite cage [3,5] but not ZSM-5.
Regarding the catalytic cracking of isopropyl-benzene, it is established that the cleavage of
the propyl group from the benzene ring is the main reaction pathway with the benzene ring
remaining unaltered [7]. Recently Al-Khattaf and de Lasa [3,4] cracked cumene, 1,3-diisopropyl-benzene, and 1,3,5-TIPB using small and large Y-zeolite crystal size. They reported
that cumene has no diffusion limitation inside Y-zeolite while 1,3,5-TIPB showed a clear
diffusion obstacle at low temperature. N-hexadecane and 1,3,5-TIPB were used to confirm the
diffusivity effect on zeolite performance [8]. It was found that accessibility to internal acid
sites influences both catalyst activity and selectivity.
ZSM-5 is a typical FCC additive due to its acidity and shape selectivity. The performance
of ZSM-5 as FCC additive and the factors that can influence its activity and selectivity have
been addressed by many researchers [9-17]. The efficiency of ZSM-5 in converting normal
hydrocarbons decreases with increasing the number of carbon atoms due to the difficulty of
long chain to diffuse inside the tiny zeolite pores [1]. Furthermore, ZSM-5 has the advantage
2
of low deactivation tendency compared to the other zeolites [22]. Recently, Corma et al [23]
have observed that the addition of ZSM-5 to the FCC commercial catalyst has improved
significantly the octane number and the C3-C4 olefins of the FCC gasoline.
The objective of this paper was to study the catalytic performance of ZSM-5 and USYzeolites in cracking of TIPB and n-dodecane using the Riser Simulator. The study also
investigated the role of diffusion on conversion of these molecules. The catalytic cracking of
these compounds is of great interest because of its potential industrial applicability.
2. EXPERIMENTATION
2.1 Apparatus
All experimental runs were carried out in the riser simulator. This reactor is novel bench scale
equipment with internal recycle unit invented by de Lasa [23]. A complete description of this
unit can be found in work by Kraemer [27]. The products were analyzed in an Agilent 6890N
gas chromatograph with a flame ionization detector and a capillary column INNOWAX, 60-m
cross-linked methyl silicone with an internal diameter of 0.32 mm.
2.2 Materials
Both ZSM-5 and Y-zeolite were provided by Tosoh Co. The as-synthesized Na zeolite was
ion exchanged with NH4 NO3 to replace the Na cation with NH4+. Following this, NH3 was
removed and the H form of the zeolite was spray-dried using kaolin as the filler and a silica
sol as the binder. The resulting 60 μm catalyst particles had the following composition: 30 wt
% zeolite, 50 wt % kaolin, and 20 wt % silica. The process of Na removal was repeated for
the pelletized catalyst. Following this, the catalyst was calcined at 600oC for 2 h. Finally, the
fluidizable SB catalyst particles (60 μm average size) were treated with 100% steam at 810 oC
for 6 h. GKF-3 catalysts were not treated with steam.
2.3 Catalyst Characterization
The BET surface area was measured according to the standard procedure ASTM D-3663
using Sorptomatic 1800 Carlo Erba Strumentazione unit, Italy. The acid property of the
catalyst was characterized by NH3 temperature-programmed desorption (NH3-TPD). In all the
experiments, 50 mg of sample was outgassed at 400 oC for 30 min. in flowing He and then
cooled down to 100 oC. At that temperature, NH3 was adsorbed on the sample by injecting
pulses of 2 µl/pulse. The injection was repeated until the amount of NH3 detected was the
same for the last two injections. After the adsorption of NH3 was saturated, the sample was
flushed at 100 oC for 1 h. with He to remove excess NH3, and then the temperature was
programmed at 10 oC /min up to 850 oC in flowing helium at 30 ml/min. Flame ionization
detector was used to monitor the desorbed NH3.
3. RESULTS AND DISCUSSION
3.1 Catalyst Characterization
The physico-chemical properties of the catalysts used in this study are presented in Table 1.
Figure 1 shows the NH3-TPD for the two catalysts. This figure reveals the total amount of
acidity of each catalyst. It can be shown that the non-steamed GKF-3 catalyst has almost 8
times more acidity than SB catalyst. The ratio of Lewis acid sites and Bronsted acid sites was
measured by FTIR (pyridine adsorption method). It can be seen, by referring, Fig 2 that the
steamed SB catalyst has more Lewis acidity than Bronsted acidity. However, comparable
amount of both acid types was measured for the non-steam GKF-3 catalyst.
3
Figure 1. TPD of ammonia profiles of the catalysts.
Figure 2. FTIR spectra of pyridine adsorption of the catalysts.
4
Table 1. Characterization of Used Catalysts
Acidity
Lewis
Bron
Surface Area
Catalyst
(mmol/g)
sites %
sites %
(m2/g)
Type
GKF-3
0.233
44
56
70
ZSM-5
Negligible
SB
0.03
65
35
150
Y-Zeolite
Negligible
Catalyst
Na2O wt %
3.2 Conversion of TIPB and C12
The experimental results are reported as plots of conversion of C12 and TIPB versus
reaction time. All the experiments were carried out at 450 ºC, cat/oil ratio of 5, and pressure
of 1 bar. The reactor effluent was analysed at reaction times of 3, 5, 7, 10, and 15 sec.
A comparison of the conversion of pure C12 over GKF-3 and SB is shown in Figure 3. It
was observed that for both zeolites, the C12 conversion increases, as expected, with increase
in reaction time (5-15 s). At low reaction time both zeolites seem to have the same
conversion. However, as the reaction time continues to increase the difference in C12
conversion also increases. It is clear from Figure 3 that the conversion of C12 is higher in
GKF-3 than in SB. At reaction time of 15 sec, the C12 conversion over GKF-3 is almost
double that over SB. The higher activity of GKF-3 is attributed to its high acidity compared to
SB bearing in mind the small size of C12 molecules. On the contrary, the kinetic behaviour
of pure TIPB is more complicated than that of C12 as shown in Figure 4. It appears that both
the acidity and the pore size of both zeolites have a significant role to play. It is depicted from
Figure 4. that the TIPB conversion over both zeolites overlaps. It is well known that TIPB can
not penetrate inside the GKF-3 structure, thus it has to react on the external acid site (only 3
% of the total acid sites). On the other hand, SB has larger pores (steaming at 810 ºC for 6 hr
can create large pores). It is quite clear that each zeolite used in the present study has an
advantage over the other when TIPB was used as feedstock. This explains why both catalysts
have similar conversion.
30
% C o n v ersio n
% C o n v ersio n
30
20
20
10
10


SB
G K F -3


SB
G K F -3
0
0
0
5
10
15
20
0
5
10
15
20
T im e / sec
T im e / sec
F ig u re 3. C o n v ersion of p u re C 1 2 o n G K F -3 an d S B
F ig u re 4. C o n version o f p u re T IP B o n G K F -3 an d S B
5
The conversion of pure C12 is higher than the conversion of TIPB over GKF-3 (Figure 5)
while the opposite is true over SB (Figure 6). This behaviour can be interpreted based on the
higher activation energy of cracking reaction of paraffines in comparison to the corresponding
dealkylation reaction of alkyl aromatics in addition to the effect of catalyst acidity and pore
size.
The interaction of C12 and TIPB over both catalysts has been studied by injecting the
reactor with a feed that contains equal amounts of both components. The results are compared
to the pure feed runs. As shown in Figure 7 and Figure 8, the conversion of C12 is not
affected by the presence of TIPB over both zeolites. In contrast, the conversion of TIPB has
substantially decreased in the presence of C12 in both catalysts (Figure 9 and Figure 10). It
appears that the C12 molecules which have substantially higher diffusivity than the TIPB
molecules occupy the active sites faster leaving fewer active sites available for the TIPB
molecules. This is an indication that the diffusion is the rate determining step.
30
% C o n v ersio n
% C o n v ersio n
30
20
20
10
10


C 12
T IP B


0
0
5
10
15
C 12
T IP B
0
20
0
5
T im e / sec
10
15
20
T im e / sec
F ig u re 5. C o n version o f p u re C 1 2 an d p u re T IP B o n G K F -3
F ig u re 6 . C o n version o f p u re C 1 2 an d p u re T IP B o n S B
25
30
% C o n versio n
% C o n versio n
20
20
15
10
10


F eed: 100 % C 12
F eed: 50 % C 12 + 50 % T IP B
0
5


F eed : 10 0 % C 1 2
F eed: 50 % C 12 + 50 % T IP B
0
0
4
8
12
T im e / sec
16
F ig u re 7. C o n version o f C 1 2 o n G K F -3
20
0
4
8
12
T im e / sec
16
F ig u re 8. C o n version o f C 1 2 o n S B
20
6
3.3- Product Distribution.
30
30
20
20
% C o n v ersio n
% C o n v ersio n
Cracking of C12 was observed to produce mainly gases on both catalysts. No aromatic
compounds were detected by cracking pure C12. However, The final main products obtained
by cracking 1,3,5 TIPB are gases (mainly propylene) and benzene. The intermediate cracking
products were cumene and di-iso-propylbenzene (DIPB) isomers. Recently Mohgoub and AlKhattaf [28] discussed cracking of pure C12 and 1,3,5 TIPB on FCC catalyst based on ZSM-5
and USY zeolites. It was found that the catalyst based on ZSM-5 (with high acidity) crack
C12 more efficiently than catalyst based on USY-zeolite (with low acidity).
It is well known that 1,3,5 TIPB catalytic cracking undergoes a series of three steps
namely; 1) dealkylation of 1,3,5 TIPB into DIPB isomers and propylene, 2) dealkylation of
DIPB into cumene and propylene, and 3) dealkylation of cumene into benzene and propylene
[4]. These compounds were detected in cracking both pure 1,3,5 TIPB and mixture of C12
and 1,3,5 TIPB. Besides these three dominant products, other products have been detected
like ethylbenzene, xylenes, and toluene. These compounds are the product of the cracking of
the propyl group attached to the benzene ring. Regarding gases, it is difficult to determine
their sources because they can be produced from the cracking of both C12 1,3,5 TIPB.
It was observed that the benzene selectivity is higher for the mixture feed than that of the
pure 1,3,5 TIPB even though pure C12 cracking does not produce any aromatics at the present
reaction conditions. For both catalysts mixing C12 with 1,3,5 TIPB increased the benzene
selectivity. This result can be interpreted based on the low aromatic concentration when
mixture feed is used, which enhances all the 1,3,5 TIPB cracking steps. As a result, the yield
of the product gases and benzene increased. Furthermore, Table 2 shows clearly the high yield
of TEX (toluene, ethylbenzen and xylenes) for the mixture feedstock. Other products are
essentially cumene, DIPB (products of intermediate step catalytic cracking), 1,3,5 TIPB
isomers and some heavy compounds. Referring to Tables 2 to 5, it seems mixing C12 to 1,3,5
TIPB did not have any significant effect on the yield of these compounds.
10
10


F eed : 10 0 % T IP B
F eed : 50 % C 12 + 5 0 % T IP B
0


F eed : 10 0 % T IP B
F eed : 50 % C 12 + 5 0 % T IP B
0
0
5
10
15
T im e / sec
F ig u re 9. C o n version o f T IP B o n G K F -3
20
0
5
10
15
T im e / sec
F ig u re 10 . C o nversio n o f T IP B o n S B
20
7
Table 2. Product Distribution for the Cracking of TIPB on GKF-3 Catalyst
(Feed: 50 % TIPB + 50 % C12)
Time/s
3
5
7
10
15
% Conversion
of TIPB
7.02
9.62
12.50
16.14
22.01
% Yield of
Benzene
1.70
2.60
3.40
3.72
5.00
% Yield of
Cumene
0.16
0.24
0.34
0.42
0.56
% Yield of
DIPB
0.30
0.50
1.02
2.00
3.12
% Yield of
TEX
1.22
2.00
2.84
3.60
5.08
Table 3. Product Distribution for the Cracking of TIPB on GKF-3 Catalyst
(Feed: Pure TIPB)
Time/s
3
5
7
10
15
% Conversion
of TIPB
7.30
14.00
18.30
20.10
24.50
% Yield of
Benzene
1.10
2.50
3.30
3.90
4.60
% Yield of
Cumene
0.20
0.78
0.82
1.20
1.30
% Yield of
DIPB
0.10
0.91
1.20
2.20
3.10
% Yield of
TEX
0.20
0.45
0.60
0.75
1.00
Table 4. Product Distribution for the Cracking of TIPB on SB Catalyst
(Feed: 50 % TIPB + 50 % C12)
Time/s
3
5
7
10
15
% Conversion
of TIPB
5.50
8.90
12.50
17.50
24.89
% Yield of
Benzene
1.72
3.00
4.00
5.42
7.40
% Yield of
Cumene
0.42
0.86
1.50
2.40
4.00
% Yield of
DIPB
0.00
0.00
0.12
0.28
0.44
% Yield of
TEX
0.40
0.50
0.63
0.86
1.00
Table 5. Product Distribution for the Cracking of TIPB on SB Catalyst
(Feed: Pure TIPB)
Time/s
3
5
7
7’’
10
% Conversion
of TIPB
7.60
11.55
14.60
18.80
21.62
% Yield of
Benzene
1.10
2.00
3.00
3.30
4.00
% Yield of
Cumene
0.30
0.96
1.40
1.90
3.00
% Yield of
DIPB
0.00
0.15
0.13
0.30
0.51
% Yield of
TEX
0.10
0.30
0.40
0.40
0.53
8
4. CONCLUSIONS
The cracking behaviour of C12 and TIPB was experimentally investigated over GKF-3 and
SB catalysts. The conversion of pure C12 is higher than the conversion of TIPB over GKF-3
while the opposite is true over SB. The presence of C12 substantially affected the conversion
of TIPB over both catalysts while the conversion of C12 remained unchanged. The diffusion
of TIPB molecules is the rate determining step in its reaction over zeolites.
ACKNOWLEDGEMENT
The authors gratefully acknowledge King Fahd University of Petroleum & Minerals for the
financial support provided for this work under project # 255. We also wish to thank Mr.
Mariano Gica and Mr A. Iliyas for their useful collaboration on the experimental work under
the same project. Special thanks are due to Dr Shakeel Ahmed for his invaluable assistance
and suggestions on TPD measurement.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
Rossini, Stefano, Catalysis Today, 77, 2003, 467-484.
Karger, J., & Ruthven, D.M. John Wiley & Sons, Inc. 1992.
Al-Khattaf, S. Ph.D Dissertation, University of Western Ontario, London, Ontario, Canada,
2001.
Al-Khattaf, S. & de Lasa, H. Applied Catalysis A General, 2002, 226, 139-153.
Al-Khattaf, S.; Atias, J.A; Jarosch, K.; de Lasa, H. Chemical Engineering Science, 2002, 57,
4909.
Xiao, J.; Wei, J. Chem Eng Sci. 1992, 47, 5, 1123-1141.
Corma, A.; Wojciechowski, B.W. Cat. Rev. Sci. Eng. 1982, 24, 1, 1-65.
Roos, K.; Liepoid, A.; Koch, H., and Reschetilowski, W. Chem. Eng . Technol. 1997, 20, 326332.
Rajagopalan, K.; Young, G.W. Prepr.-Am. Soc., Div. Pet. Chem. 1987, 32, 3, 627.
Biswas, J.; Maxwell, I. E. Appl. Catal. 1990, 58, 1.
Madon,R. J. J. Catal. 1991, 129, 275.
Elia, M. F. ; Iglesias, E.; Martinez, A.; Pascual, M.A. Appl. Catal. 1991, 73, 195.
Miller,S. J.; Hsieh, C.R. Prepr.-Am. Chem. Soc. Div. Pet. Chem. 1990, 685.
Nalbandian, L.; Vasalos, I.A.; Kasapaki, A.; Vassilakis, K. Prepr.Am. Chem. Soc., Div. Pet.
Chem. 1993, 584.
Guerzoni, F.N.; Abbot, J. Appl. Catal. A 1994, 120, 55.
Nalbandian, I.; Lemonidou, A.A.; Vasalos, I.A. Appl. Catal. A. 1993, 105, 107.
Donnelly, S.P.; Mizrahi, S.; Sparrell, P.T.; Huss, A. Jr; Schipper,P.H.; Herbst, J.A. Prepr.Am.
Chem. Soc., Div. Pet. Chem. 1987, 32 (3), 621.
Dwyer, J.; Rawlence, D.J. Catal . Today, 1997, 18, 487.
Buchanan J.S. Catalysis Today, 2000, 55, 207-212.
Biswas, J.; Maxwell, I.E. Applied Catalysis A, 1990, 63, pp 197-258.
Aitani, A.; Yoshikawa, T.; Ino, T. Catalysis Today, 2000, 60, 111-117.
Herrman, C.; Haas, J.; Fetting, F. Applied Catalysis, 1987, 35, 299-310.
Corma, A.; Melo, F.; Sauvanaud, L.; Ortega, F. Applied Catalysis A: General 265, 2004, 195206.
de Lasa, H.I. U.S. Pat. 5 102 628, 1992.
Iliyas, A.; Al-Khattaf, S. Ind. Eng. Chem. Res., 43, 2004, 1349.
Iliyas, A.; Al-Khattaf, S. Appl. Catal. A: Gen 269, 2004, 225-236.
Kraemer, D.W. Ph.D. Dissertation, University of Western Ont., London, Canada, 1991.
Mohgoub, K.; Al-Khattaf, S. Energy & Fuel (Accepted).
Download