J. Phys. Chem. A 2010, 114, 3811–3817
3811
Kinetics of Liquid Phase Catalytic Hydrogenation of Dicyclopentadiene over Pd/C Catalyst† Miaoli Hao, Bolun Yang,* Haiguo Wang, Gong Liu, and Suitao Qi Department of Chemical Engineering, State Key Laboratory of Multiphase Flow in Power Engineering, Xi’an Jiaotong UniVersity, Xi’an 710049, China
Jianming Yang, Chunying Li, and Jian Lv Xi’an Modern Chemistry Research Institute, Xi’an 710065, China ReceiVed: June 27, 2009; ReVised Manuscript ReceiVed: August 28, 2009
To investigate the kinetics behaviors of dicyclopentadiene hydrogenation, a series of experiments were performed at different temperatures (323-353 K) under varying hydrogen pressure (0.5-1.5 MPa) with a range of Pd/C catalyst loading (0.25-1.00 wt %) using ethanol as solvent in a batch reactor. The time dependent concentration variations for each component were traced under the conditions of removing both the internal and external diffusion effects. The Langmuir-Hinshelwood mechanism was proposed with the consideration of the noncompetitive adsorption between the organic species with hydrogen, and the surface reaction was the rate-determining step. The kinetic equations for the sequence reaction were derived on the basis of the analysis of mechanisms, and the model parameters were determined by fitting the experimental data in differential temperature using the method of Runge-Kutta. The reaction activation energies for the first and second steps are 3.19 and 31.69 kJ · mol-1, respectively, and the reliability of the model was verified by these experimental results to change hydrogen pressure, reactant concentration and catalyst loading. The simulation results agreed well with the experimental data. 1. Introduction With the development of aerospace engineering, the need for higher performance fuel for aircrafts has been raised; therefore, the study on the higher volumetric energy density of the fuels is attracting more and more attention in the world.1,2 Exotetrahydrodicyclopentadiene (exo-THDCPD) is a fuel with higher volumetric energy density used widely in aircrafts such as military jets and Tomahawk Cruise Missiles,3-5 which are the products of the isomerization reactions of endotetrahydrodicyclopentadiene (endo-THDCPD).6,7 Therefore, it is important to study the synthesis of endo-THDCPD. C5 resources are the byproduct of catalytic cracking process, among which cyclopentadiene and its derivative dicyclopentadiene (DCPD) can be used for the production of endo-THDCPD by catalytic hydrogenation.8,9 In addition, a certain quantity of DCPD can be obtained from the separation of coal tar. The available resources of dicyclopentadiene in the world are about 1 million ton in total annually.10,11 Therefore, the synthesis of endo-THDCPD by the hydrogenation of the DCPD is an attractive and technically feasible route. The catalytic hydrogenation of DCPD is a basic part in the manufacture of endo-THDCPD. Although plenty of work has been done on the catalytic hydrogenation of DCPD, major topics are mainly centralized on the process optimization of operation conditions. The kinetics and reaction mechanisms of the hydrogenation of DCPD to endo-THDCPD have seldom been reported. Liu and Mi8 indicated that the adsorption of organic species and dissolved hydrogen was more appropriate, a Eley-Rideal type model over the Pd/Al2O3 catalyst was proposed and the kinetic † Part of the special issue “Green Chemistry in Energy Production Symposium”. * Corresponding author. Tel.: +86 29 826683189. Fax: +86 29 82668789. E-mail:
[email protected].
parameters were regressed. Skala and Hanika9 established kinetics models based on the Langmuir-Hinshelwood (L-H) mechanism, and the model parameters were developed with the consideration of the nondissociation adsorption of hydrogen. On the other hand, some different work reported on catalytic hydrogenation indicates that the hydrogen is believed to be a kind of chemical and the dissociation adsorption happens on the active sites of the catalyst, and the active hydrogen atoms are combined with C-C carbon atoms of unsaturated hydrocarbons to afford products. Moreover, since the difference in molecular sizes between hydrogen and other reactants is sufficiently large, hydrogen can be adsorbed on the surface of the voids between the bulky organic molecules, and thus the absorption can be assumed to be noncompetitive.12,13 It can be known from these analyses that investigation of the kinetics of hydrogenation to fully understand the reaction essentials is still needed. The objectives of this work are to study the reaction mechanism, to achieve the kinetics of DCPD hydrogenation into endo-THDCPD in the presence of Pd/C catalyst using ethanol as solvent, and to provide a useful tool for large-scale production of fuel with higher performances. 2. Experiments 2.1. Materials. DCPD with a purity of 99% or higher was purchased from Puyang shenghuade Chemical Co., Ltd. of Henna, China. Analytic reagent ethanol was obtained from Tianjin Chemical Co., Ltd. Hydrogen with a purity of 99.99% or higher was supplied by chang’an Gas service Co. of Xi’an China. Pd/C catalyst with a diameter of e40 µm was provided by Baoji Rock Co., Ltd. The surface area was more than 800 m2/g and the initial amounts of Pd and C were defined as 5% and 95%, respectively. 2.2. Experimental Procedures. The hydrogenation experiments were performed in a 1 L stainless steel autoclave operated
10.1021/jp9060363 2010 American Chemical Society Published on Web 09/29/2009
3812
J. Phys. Chem. A, Vol. 114, No. 11, 2010
Figure 1. Catalytic hydrogenation apparatus of DCPD: (a) N2 steel cylinder; (b) H2 steel cylinder; (c) autoclave controller meter; (d) sampling port; (e) pressure gauge; (f) electromagnetic stirrer.
Figure 2. Effect of agitation speeds on the conversion of endoTHDCPD at T ) 333 K, PH2 ) 1.0 MPa, solvent/reactant ratio is 5:2.
in the batch mode under different hydrogen pressures. The reactor was equipped with a stirrer, an internal cooling coil, a dip tube for liquid-phase sampling, and a dual-action programmable temperature controller. A simplified schematic of the reactor setup is shown in Figure 1. DCPD, solvent, and catalyst were added into the autoclave, which was were sealed and purged by nitrogen to 2.0 MPa at room temperature. It was retained for 30 min to test the air tightness, and the pressure was released to displace the air. This process was repeated until the air in the autoclave was exhausted. The whole system was then flushed by hydrogen to replace nitrogen before initiation of the operation. When the system was heated to the desired reaction temperature, the start time of the reaction was noted by a point. The stirring speed and the hydrogen pressure were maintained constant during the reaction. A special sampling port was equipped to ensure that the pressure was kept as a constant for sampling. The samples were collected in definite time and analyzed immediately. At the end of the experiment, the reactor was allowed to equilibrate to room temperature and pressure. The reaction mixtures were taken out and the products and catalysts were separated by the centrifugal filtration. The solvent was separated from products by distillation in a water bath heating at atmospheric pressure. 2.3. Products Analysis. The products were analyzed by an Agilent 4890 GC system equipped with an HP-5 capillary column (length is 30 m; diameter is 0.32 mm; film thickness is 0.25; manufactured by J&W Scientific, Agilent Technologies)
Hao et al.
Figure 3. Effect of catalyst loading on the initial rate of DCPD at T ) 333 K, PH2 ) 1.0 MPa, solvent/reactant ratio is 5:2.
and a flame ionization detector. The carrier gas was nitrogen, and the oven, vaporization chamber, and detector temperatures were 393, 493, and 523 K, respectively. Integration of the GC results was performed by HP Chemstation software. 2.4. Experiment Conditions. Pd/C and Raney Ni catalysts are well-known as the classical catalysts in hydrogenation reactions. However, the use of Raney Ni catalysts was restricted because of their drawbacks such as short life, bad reusability and difficulty to regenerate. Therefore, the Pd/C catalyst was selected in all kinetics experiments. The effects of different solvents such as methanol, ethanol, and 2-propanol on the hydrogenation reaction were considered. The yield in the pre-experiment was very low when 2-propanol was used as solvent due to the formation of byproduct of small molecules. The effect of methanol and ethanol on the yield of the reaction was basically similar. However, in view of the low boiling point and toxicity of methanol, ethanol was selected as the reaction solvent in this study. 3. Results and Discussion 3.1. Verification of Kinetic Conditions. DCPD hydrogenation is a gas-liquid-solid multiphase catalytic reaction. The reaction rate may be influenced by external and internal masstransfer resistance. These mass-transfer effects must be excluded by choosing suitable agitation speeds and particle sizes of the catalyst. The conversion of endo-THDCPD with time under different agitation speeds is used to verify the external mass transfer limitations. Figure 2 presents the results obtained at 333 K when the solvent/reactant ratio is 5:2 and the hydrogen pressure is 1.0 MPa under different agitation speeds. As is shown in this figure, the initial reaction rate of DCPD increased with the increase in agitation speeds at the beginning, but remained almost constant when the mechanical rotation rate exceeded 850 rpm. It shows that the influence of the external diffusion resistance can be neglected when the agitation speed is over 850 rpm. Figure 3 shows the variation of initial reaction rates of DCPD is directly proportional to the catalyst loading. These results also suggest that the effect of external mass-transfer resistance has been excluded, and all the kinetics experiments thus were carried out with the agitation speed of 850 rpm. For the cases of hydrogenation, the influence of internal diffusion can be omitted when the average particle size of the catalyst is less than 50 µm.12 As is described above, the
Hydrogenation of Dicyclopentadiene over Pd/C Catalyst
J. Phys. Chem. A, Vol. 114, No. 11, 2010 3813
Figure 4. Variation of the system concentration with the reaction time at different temperatures with PH2 ) 1.0 MPa, catalyst loading is 1.00 wt %, solvent/reactant ratio is 5:2.
Pd/C catalyst particle size in this work was about 10-20 µm. It is considered that the influence of internal diffusion can be ignored. 3.2. Kinetics Experiment. The kinetics experiments study the changes of concentration with the reaction time for every component. The concentrations were measured at 323, 333, 343, and 353 K, respectively. Figure 4 shows the evolution of the concentration of all the components with reaction time at different temperatures. As can be seen, the concentration of DCPD decreased constantly with reaction time, and the endo-THDCPD increased steadily; however, the concentration of dihydrodicyclopentadiene (DHDCPD) increased initially and decreased after reached the maximum. It might be considered that this reaction system is a consecutive process in which the DCPD hydrogenated to intermediates of DHDCPD and the DHDCPD further hydrogenated to get the target product of endoTHDCPD. It should be noted that the reactant concentration sharply decreases in the initial stages of the reaction. However, the reaction rate became slower with the increase in time and the product growth rate slowed down simultaneously. The variation form of the reactant concentration appears to be an exponential function. These facts showed that this reaction process followed the first-order consecutive reaction. 3.3. Mechanisms and Kinetics Model of DCPD Hydrogenation. From the analysis of the experiments, the hydrogenation of DCPD over the Pd/C catalysts can be considered as follows. The main reactions are Since no C5H6, C5H8, C5H10, or C10H18 was detected during the reaction, the following side reactions can be neglected in this system.
Like most organic liquid hydrogenation reactions, kinetics models in this system can be established by the LangmuirHinshelwood theory with the dissociation adsorption of hydrogen. Because the molecular size of hydrogen is much smaller than those of organic species, it can pass through the gap among the DCPD molecules for adsorption freely, so the absorption of hydrogen and DCPD can be viewed to
3814
J. Phys. Chem. A, Vol. 114, No. 11, 2010
Hao et al.
be noncompetitive.12,13 Kinetics models thus can be formulated with the following steps:
As is stated above, the influences of internal and external diffusion resistances were eliminated in this work, and the adsorption of hydrogen is usually easier than the adsorption of organic species. The adsorption or the desorption of organic species on the surface of catalysts might be considered as the rate-limiting step, and the reaction rate thus is correlated only with the concentrations of organic species. But it is independent of the hydrogen pressure, as shown in eqs 13 and 14 from the theory of Langmuir-Hinshelwood:
r ) kaCjθV - kdθj
(13)
r ) kaθj - kdCjθV
(14)
or
However, since the experimental results showed that the reaction rate was enhanced by the increase in both the concentration of DCPD and the pressure of hydrogen, two surface reaction steps can be assumed as the rate-limiting step.
Figure 5. Comparison of experimental results and simulation results at different hydrogen pressures. Solvent/reactant ratio is 5:2, catalyst loading is 1.00 wt %, T ) 333 K. Simulation: solid lines. Measurement: symbols.
Since the adsorption equilibria for hydrogen and organic species are achieved, and the relationship between the liquid phase concentration and the gas phase frication pressure for hydrogen is obeyed with Henry’s Law, the following equations can be obtained:
θC10H12 ) KC10H12CC10H12θV
(17)
θC10H14 ) KC10H14CC10H14θV
(18)
θC10H16 ) KC10H16CC10H16θV
(19)
Hydrogenation of Dicyclopentadiene over Pd/C Catalyst
θH2 ) (KH′ 2CH2)0.5θδ
J. Phys. Chem. A, Vol. 114, No. 11, 2010 3815
dnC10H12
(20)
Wdt
The reaction rates thus can be given by
dnC10H14 r1 )
k1KC10H12KH2CC10H12PH2
√
W dt
k2KC10H14KH2CC10H14PH2
√
)
V dCC10H12 W dt
V dCC10H14 W dt
) -r1
(23)
) r1 - r2
(24)
(21)
(1 + KH2PH2)2(1 + KC10H12CC10H12 +
dnC10H16
KC10H14CC10H14 + KC10H16CC10H16)
r2 )
)
W dt
)
V dCC10H16 W dt
) r2
(25)
where W is the catalyst mass and V is the volume of the liquid phase. Combining eqs 21-25 with the following initial conditions:
(22)
(1 + KH2PH2)2(1 + KC10H12CC10H12 + KC10H14CC10H14 + KC10H16CC10H16)
where k1 and k2 are reaction rate constants, θV and θδ are the vacant adsorption sites of the species and the hydrogen, θjand θH2 are the adsorption sites of the species j and hydrogen, Kj is the adsorption equilibrium constant of the species j, KH2 is given by KH2 ) K′H2/H, K′H2 is the adsorption equilibrium constant of hydrogen, PH2 is the hydrogen pressure, and Cj is the concentration of the species, j ) C10H12, C10H14, or C10H16. Then, the differential equations describing the variations in concentrations of liquid-phase organic reactants in the reactor can be written as
t ) 0,
CC10H12 ) C0C10H12,
CC10H14 ) CC10H16 ) 0 (26)
yields the kinetic model of catalytic hydrogenation of DCPD over a Pd/C catalyst. 3.4. Calculation of Kinetics Model Equation. Kinetic parameters were evaluated by the curve fitting method using the experimental data of concentration-time obtained under different temperatures with calculated results. The fourth-order Runge-Kutta method was used to solve the differential equations (19)-(24) by the Matlab 7.0 software. Then, the rate constants at different temperatures were correlated using the Arrhenius equation to obtain activation energies and the preexponential factors, respectively.
Figure 6. Comparison of experimental results and simulation results at different catalyst loading. PH2 ) 1.0 MPa, T ) 333 K, solvent/reactant ratio is 5:2. Simulation: solid lines. Measurement: symbols.
3816
J. Phys. Chem. A, Vol. 114, No. 11, 2010
Hao et al.
TABLE 1: Regressed Kinetics Parameters parameters
k1
k2
KC10H12
KC10H14
KC10H16
KH
pre-exponential factor activation energy or adsorption heat
2.35
370.08
1.40
48.83
46.62
0.13
3.19
31.69
11.66
11.00
12.14
0.43
All kinetics parameters are summarized in Table 1. As is shown in this table, the activation energies were calculated to be 3.19 and 31.69 kJ · mol-1 for two hydrogenation steps, respectively. This result shows that the first step of hydrogenation is more rapid than the second step. The adsorption and dissolution heat of hydrogen (0.43 kJ/mol) is too small and
is close to zero. It is illustrated that the assumption of the dissociation adsorption of hydrogen is correct.14 3.5. Validation of Kinetics Model. The kinetics model was verified by the consistency of the calculation results with experimental observations under different operation conditions with the changes in the hydrogen pressure, the initial DCPD concentration, and the catalyst loading. 3.5.1. Effect of Hydrogen Pressure. The results of different hydrogen pressures (0.5, 1.0, and 1.5 MPa) are shown in Figure 5. The solid lines and the dots in these figures represent the predicated results and experiments. It can be seen that the model fits these experimental data well. As seen from Figure 5, the DCPD concentration decreased rapidly with the increase in hydrogen pressure. The reason is that the organic liquid hydrogenation reaction is a process with decreasing molar numbers; the increase in the hydrogen pressure can enhance the reaction. 3.5.2. Effect of Catalyst Loading. The simulation and experimental results are presented with the change of catalyst loading (0.25, 0.50, 0.75, and 1.00 wt %) in Figure 6. It shows that the simulation result matches well with the experimental data. The time of the endo-THDCPD concentration reaching the equilibrium value is shortened with the increase in the catalyst loading. However, no change in the maximum yield of DHDCPD with catalyst loading is in accordance with eq 24. 3.5.3. Effect of Initial DCPD Concentration. Figure 7 shows the comparison between the experimental data and simulation results with the change in the initial DCPD concentration. The time of the endo-THDCPD concentration reaching the equilibrium value was shortened with the increase in the initial DCPD concentration. It is because the productive capacity increased with the increase in the reaction feed. 4. Conclusions The kinetics of TCPD hydrogenation into endo-THDCPD over the Pd/C catalyst has been studied with the elimination of both the internal and external diffusion effects. The overall reaction consists of two consecutive elementary reactions with DHDCPD as an intermediate. The kinetic model was established based on the L-H mechanism with the noncompetitive adsorption of the dissociated hydrogen and the organic species on catalyst active sites. The responding model parameters were calculated, and the reaction activation energies for the first and second steps are 3.19 and 31.69 kJ · mol-1, respectively. The kinetic model was validated by different operation parameters such as hydrogen pressure, the initial DCPD concentration, and the catalyst loadings. Comparison of the results shows that the experimental data agree well with the simulation. Acknowledgment. Financial support from Xi’an Modern Chemistry Research Institute is gratefully acknowledged. References and Notes
Figure 7. Comparison of experimental results and simulation results at different initial DCPD concentrations. PH2 ) 1.0 MPa, T ) 333 K, catalyst loading is 1.00 wt %. Simulation: solid lines. Measurement: symbols.
(1) Norton, R. V.; Howe, S. C. U.S. Patent, 4,320,238, 1982. (2) Chung, H. S.; Chen, C. S. H.; Kremer, R. A.; Boulton, J. R. Energy Fuels. 1999, 13, 641–649. (3) Schneider, A.; Ware, R. E.; Janoski E J. U.S. Patent, 4,086,284, 1978. (4) Chenoweth, K.; van Duin, A.; Dasgupta, S.; Goddard, W. A. J. Phys. Chem. A 2009, 113, 1740–1746. (5) Osmont, A; Gokalp, I; Catoire, L. Propellants, Explos., Pyrotech. 2006, 31, 343–354. (6) Imanari, M.; Ikeda, M. JP Patent, 60,209,536, 1985.
Hydrogenation of Dicyclopentadiene over Pd/C Catalyst (7) Navratilova, M.; Sporka, K. Appl. Catal. A Gen. 2000, 203, 127– 132. (8) Liu, G. Z.; Mi, Z. T.; Wang, L.; Zhang, X. W. J. Ind. Eng. Chem. Res. 2005, 44, 3846–3851. (9) Skala, D.; Hanika, J. Petroleum Coal 2003, 45, 105–108. (10) Li, S. Q.; He, S. Z.; Bian, Y. Q. Techno-economics Petrochem. 2002, 18, 72–75. (11) Zhou, J. S.; Feng, J. C.; Zhang, Z. Y.; Gui, L.; Yu, H. C. Chem. Propellants Polym. Mater 2003, 2, 17–21.
J. Phys. Chem. A, Vol. 114, No. 11, 2010 3817 (12) Simakova, I. L.; Solkina, Y.; Deliy, I.; Waerna, J.; Murzin, D. Y. Appl. Catal. A Gen. 2009, 356, 216–224. (13) Jere, F. T.; Jackson, J. E.; Miller, D. J. Ind. Eng. Chem. Res. 2004, 43, 3297–3303. (14) Zhou, Z. M.; Cheng, Z. M.; Li, Z.; Yuan, W. K. J. East China UniV. Sci. Technol. 2004, 30, 1–5.
JP9060363
