Вы находитесь на странице: 1из 5

Catalysis Today 93–95 (2004) 723–727

A new approach of coke characterization on metal and support for


Pt/Al2 O3 by combination of Al2 O3 and Pt/SiO2
Sunee Srihiranpullop∗ , Piyasan Praserthdam
Department of Chemical Engineering, Center of Excellence on Catalysis and Catalytic Reaction Engineering,
Chulalongkorn University, Bangkok 10330, Thailand

Abstract

The main goal was to study the distribution and probability of chain growth of coke by the combination of Pt/SiO2 representing the
metal sites with 100–120 mesh mixed with Al2 O3 60–80 mesh representing the acid sites. TPO, ESR, IR and Soxhelt extraction analyzed
with GC-14B were employed for coke characterization. From the effects of time, temperature, H2 /HC ratio and modification of catalysts,
it was found that probability of chain growth on both sites were the same. However, the greater accumulation was occurred on the support.
This offered transformation on the gas phase of coke intermediates from the metal, which was primarily responsible for producing the coke
precursors. The roles H2 and promoter was more sensitive to reduce coke and probability of chain coke, especially the effect of K promoter.
© 2004 Elsevier B.V. All rights reserved.

Keywords: Coke; A physically mixed catalyst; Distribution and probability of chain growth of coke

1. Introduction is the criteria. To achieve this objective, the coke deposition


on catalysts must be recovered by a Soxhelt extraction and
Catalytic reforming has become one of the most important then analyzed through GC/MS in order to better understand
industrial applications of catalysis. Desired reactions are fa- the deactivation process of catalyst.
vored by high temperature and low pressure conditions that
also favor coke formation [1,2]. Analysis by temperature
programmed oxidation (TPO) generally indicated two types 2. Experimental
of coke. The first one burns at low temperatures (poly-
meric carbon), which is associated with the metallic phase. The experiments were carried out using reagents as fol-
The second type burns at high temperatures (amorphous low: H2 PtCl6 ·6H2 O (Wako), SnCl2 ·2H2 O (Fluka), KNO3
phases/graphitized carbon), corresponded to the support (Merck), Al2 O3 type NKH-3 (Sumitomo) and SiO2 (Merck).
[3,4]. However, the coke deposition completely separated Alumina and silica support were grounded to the require
on both the metal and the support site has not been inves- mesh sizes of 60/80 and 100/120, respectively. The Pt cata-
tigated sufficiently so far. Thus, coke formation is studied lyst, the Pt–Sn catalyst and the Pt–Sn–K catalyst were pre-
through a physically mixed system (Pt/SiO2 represented the pared by impregnation based on SiO2 and were calcined
metallic site and Al2 O3 represented the support site). The in air at 500 ◦ C for 3 h. The coked catalyst was performed
main goal was to determine the probability of chain growth the dehydrogenation with 10% n-hexane in N2 . The ratio
(usually called alpha value) of coke formation on both sites of physically mixed catalysts of the Pt-based catalyst and
using Schulz–Flory distribution. Theory of Schulz–Flory Al2 O3 was 1:1.
distribution presents the relationship between carbon atoms The beginning, 0.2 g of the catalyst was packed in quartz
and weight fraction or hydrocarbon distribution [5]. That the tube (6 mm in diameter). Next, the catalyst was in situ re-
rate-determining step of coke formation is polymerization duced in hydrogen with the flow rate of 100 cm3 /min at
500 ◦ C for 1 h. Then, the normal reaction condition was ad-
justed to 475 ◦ C and then the reactant was flowed to the
∗Corresponding author. Fax: +662 639 1830. system with the flow rate of 60 cm3 /min and the H2 /HC ra-
E-mail address: suneell@hotmail.com (S. Srihiranpullop). tio of 0 for 120 min. On the other hand, the effects of time

0920-5861/$ – see front matter © 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.cattod.2004.06.067
724 S. Srihiranpullop, P. Praserthdam / Catalysis Today 93–95 (2004) 723–727

Table 1
The amount, carbon density and probability of chain growth of coke
Effect % C of Coke Carbon density Probability of chain growth

Metal Support Metal Support Metal Support

Time (min)
10 0.28 1.47 2.27 × 105 1.45 × 106 0.29 0.32
30 0.31 1.64 3.73 × 105 2.28 × 106 0.38 0.38
60 0.31 2.10 4.70 × 105 2.33 × 106 0.38 0.47
120 0.36 2.50 8.55 × 105 3.02 × 106 0.67 0.65
Temperature (◦ C)
300 0.28 1.61 8.00 × 104 6.47 × 105 0.20 0.27
400 0.30 1.64 3.27 × 105 1.82 × 106 0.24 0.29
475 0.36 2.50 8.55 × 105 3.02 × 106 0.67 0.65
H2 /HC ratio
0 0.36 2.50 8.55 × 105 3.02 × 106 0.67 0.65
3 0.29 1.34 4.50 × 105 3.04 × 106 0.25 0.31
5 0.27 1.31 6.17 × 105 4.00 × 106 0.24 0.25
10 0.26 1.09 5.43 × 105 3.47 × 106 0.26 0.26
Type
Pt 0.36 2.50 8.55 × 105 3.02 × 106 0.67 0.65
Pt–Sn 0.32 1.73 5.18 × 105 2.07 × 106 0.36 0.29
Pt–Sn–K 0.21 1.26 3.17 × 104 1.83 × 106 0.29 0.29

(10–120 min), reaction temperature (300–400 ◦ C), H2 /HC have like Pt/Al2 O3 . This is confirmed by our previous work
ratio (0–10) and modified catalyst (Pt–Sn and Pt–Sn–K cat- [6]. Considering this research, coke was built up on both
alysts) were also studied. Finally, the coke on the mixed the metal site and the support site as listed in Table 1. Coke
catalyst was separated by sieve to the Pt-based catalyst and accumulation was extremely found to take place on the sup-
Al2 O3 with different sizes for TPO, IR, ESR and Soxhlet port. On behalf of TPO results as illustrated in Fig. 1, the
extraction analysis. two peaks around 300 and 425 ◦ C were corresponding to
coke oxidized at the metal phase as reported elsewhere [7].
However, coke accumulated over support sites shows only
3. Results and discussions a peak around 525 ◦ C. Moreover, the amount of coke radi-
cals on both sites as presented in Table 1 is parallel to the
3.1. Characterizations of carbon deposition, Cx Hy species TPO results. That is most of carbon radicals generated on
and the extracted composition of coked catalysts the support sites.
The IR spectra of coke samples are illustrated in Fig. 2.
It was found that the physical mixtures of Pt/SiO2 with One intense band was detected at approximately 1610 cm−1
100–120 mesh size and Al2 O3 with 60–80 mesh size be- which is noticed on both sites. Normally, this band in

Fig. 1. Example of TPO of carbonaceous deposits produced on: (A) the metal site and (B) the support site at 475 ◦ C and H2 /HC = 0 with various times
on stream; (䊏) 120 min, (䊊) 60 min, (䉱) 30 min, and (䉫) 10 min.
S. Srihiranpullop, P. Praserthdam / Catalysis Today 93–95 (2004) 723–727 725

Fig. 2. Example of IR spectra obtained after coking at various times over (A) the metal site and (B) the support site.

this region is specific to stretching mode of aromatic sys- that the soluble coke was mainly composed of C8 –C12 . How-
tem [8]. Whereas, there was a significant presence of the ever, the extracted coke on the metal sites was more consti-
characteristic group of C=C stretching in aromatic ring at tuted of a lower C content (C8 –C10 ) while the C12 fragments
1540 cm−1 on the support site. No IR adsorption bands of was dominantly presented on the support sites. Additionally,
aliphatic hydrocarbons at 2960 and 2800 cm−1 for CH3 the total amount of soluble coke was more extracted on the
and 2930 cm−1 for CH2 were observed for both sites of the support sites similar to TPO, ESR and IR results.
coked catalysts. Thus, only aromatic ring structure can be
identified. 3.2. The probability of chain growth
The chemical compositions of coke on the support and
metal sites were extracted by toluene for 24 h and then were From previous results, coke formation is taking place both
analyzed by capillary column (DB1) for the study of coke on the metal and on the support sites. Although many re-
distribution. As described in Fig. 3 and Table 2, it is noticed searchers studied so far, how the metal and support sites were
deactivated that is not clear. Concerning the Schulz–Flory
theory, the probability of chain growth of coke (α value)
as shown in Table 1 is calculated from the plot of carbon
compositions of the extracted coke. Interestingly, the close
probability of chain growth on both sites was found. It was
attribute to support the transportation of intermediate coke.
The soluble coke molecules were therefore intermediates in
the formation of insoluble coke, which was constituted of the
higher polynuclear aromatics [9,10]. Although coke forma-
tion cannot be avoided, the addition of H2 and promoter is
more sensitive to reduce coke accumulation and probability
of chain coke as illustrated in Table 1, especially the effect
of K promoter. It seems due to the ensemble and electronic
effects [11].

3.3. The relationship of coke formation between the metal


and support site

As mentioned above, the existence of aromatic structure


predominate in coke was confirmed by TPO, ESR and IR re-
sults. It was also found that the similar C8 –C12 structure of
extracted coke as coke precursor was presented on both sites.
Fig. 3. Example of chromatograph of the extracted coke. The carbon structures of extracted coke were corresponded
726 S. Srihiranpullop, P. Praserthdam / Catalysis Today 93–95 (2004) 723–727

Table 2
Carbon composition of extracted coke
Effect Metal Support

C8 C9 C10 C11 C12 C8 C9 C10 C11 C12

Time (min)
10 57.14 19.74 13.58 1.05 0.00 271.58 213.28 79.28 8.80 0.00
30 136.93 37.76 12.88 4.30 0.00 67.76 36.38 9.05 4.28 0.00
60 110.01 21.41 14.30 4.89 0.00 63.26 308.14 205.28 5.49 0.00
120 284.77 335.36 107.36 12.18 199.08 47.26 431.80 235.78 1.01 115.16
Temperature (◦ C)
300 221.01 121.18 41.88 2.52 0.00 68.73 18.63 9.79 1.09 0.00
400 256.85 32.69 9.41 1.95 0.00 180.6 28.23 12.63 3.83 0.00
475 284.77 335.36 107.36 12.18 199.08 47.26 431.80 235.78 1.01 115.16
H2 /HC ratio
0 284.77 335.36 107.36 12.18 199.08 47.26 431.80 235.78 1.01 115.16
3 415.82 212.30 62.51 4.27 0.00 214.59 143.31 56.21 4.45 0.00
5 159.92 69.96 25.59 1.36 0.00 282.61 138.64 42.32 2.97 0.00
10 138.07 98.55 235.78 1.60 0.00 408.72 388.60 173.57 4.72 0.00
Type
Pt 284.77 335.36 107.36 12.18 199.08 47.26 431.80 235.78 1.01 115.16
Pt–Sn 250.49 123.88 32.61 7.65 6.03 338.17 361.74 156.05 6.95 0.00
Pt–Sn–K 183.76 70.12 21.53 6.41 1.33 378.77 452.21 198.01 5.54 6.64

Fig. 4. Coke formation model.

to the aromatic coke as reviewed above. It is attributed to mixed catalyst, the transportation of coke precursors by gas
many researchers who said that C5 and more than carbon phase reaction may reasonable leading to coke formation
atom is the main structure of coke precursors [12–14]. The with the same probability of chain growth of coke on both
coke precursor is intermediate and is able to be transported sites. However, coke accumulation is more concentrated on
by the gas phase from the metal site to the support site. More- the support site because the heaviest components, C12 , were
over, the same probability of chain growth of coke indicates more deposited on the support site leading to more complex
that coke is made by the same kind of reactions (especially structure and higher amount of coke.
polymerization) accounted for the aromatic coke formation.
Therefore, a possible reaction for the coke formation is given
in Fig. 4. 4. Conclusion
Initially the coke precursor generated on the metal site
and then migrated by either gas phase or surface diffusion to Using the physically mixed catalysts and theory of
acid site. Consider the cooperative effect of the physically Schulz–Flory distribution, we can get the better knowledge
S. Srihiranpullop, P. Praserthdam / Catalysis Today 93–95 (2004) 723–727 727

of coke formation on the metal and support sites. We realize References


that at the polymerization is a domain reaction in the soluble
coke, C8 –C12 , at the beginning of the TOS. Surprisingly, the [1] D.L. Trimm, Appl. Catal. 5 (1983) 263.
probability of chain growth, α, of soluble coke is the same [2] G.F. Froment, Catal. Deact. (1997) 53.
[3] J. Biswas, P.G. Gray, D.D. Do, Appl. Catal. 32 (1987) 249.
on both the support site and the metal site. The aromatic [4] J. Barbier, Catal. Deact. (1987) 1.
coke was found on both sites. Whereas some researcher [5] P.J Flory, Principle of Polymer Chemistry, Cornell University Press,
said that coke structure on the metal site is preferred n fila- New York, 1953, pp. 317.
ment form. On the other hand, we found that the aromatic [6] S. Sunee, P. Piyasan, M. Tharathon, React. Kinet. Catal. Lett. 72
structures of coke on the metal site and the support site are (2001) 125.
[7] T.C. Basso, Z. Zhang, W.M.H. Sachtler, Appl. Catal. A: Gen. 79
different. The C=C aromatic structure is identically pre- (1991) 227.
sented on the support site leading to most of coke deposi- [8] C.L. Pieck, E.L. Jablonski, J.M. Parera, R. Frety, F. Lefebvre, Ind.
tion. Eng. Chem. Res. 31 (1992) 1017.
[9] G.A.D. Nassionou, P. Magnoux, M. Guisnet, Micro. Meso. Mat. 21
(1998) 389.
[10] X. Longya, L. Jinxiang, X. Yide, Y. Hong, W. Qingxia, L. Liwu,
Appl. Catal. A: Gen. 193 (2000) 95.
Acknowledgements [11] D.L. Trimm, Catal. Today 49 (1999) 3.
[12] R.J. Verderone, C.L. Pieck, M.R. Sad, J.M. Parera, Appl. Catal. 21
(1986) 239.
The authors would like to thank Thailand Research Fund [13] J. Biswas, P.G. Gray, D.D. Do, Appl. Catal. 32 (1987) 248.
(TRF) for financial support of this project. [14] J.N. Beltramini, T.J. Wessel, R. Datta, Catal. Deact. (1991) 119.

Вам также может понравиться