Professional Documents
Culture Documents
discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/244400962
CITATIONS READS
27 128
6 authors, including:
All content following this page was uploaded by Hua Wang on 20 December 2016.
Abstract
The etherification of methanol (MeOH) with 2-methyl-1-butene (2M1B) and 2-methyl-2-butene (2M2B) catalyzed by the macroporous cation
ion-exchange resin (D005II) was studied in a batch reactor. The thermodynamic properties and kinetic data were obtained under different initial
molar ratio of methanol and isoamylenes (0.52) in the temperature range between 323 and 353 K with pressure (1 MPa) constant. Because
of the nonideal property of the reaction, activities were adopted to calculate equilibrium constants and the activity coefficients were estimated
by Wilson method. There were three mechanisms adopted for simulating reaction rate equations, containing homogeneous reaction mechanism,
LangmuirHinshelwood (LH) type mechanism and EleyRideal (ER) type mechanism and the equations based on the LH type mechanism was found
to get better fit with the experimental data. The results showed that the experimental data for thermodynamics agreed with the theoretical predications
well, and the activation energy was 88.1 and 102.1 kJ mol1 for the etherification of 2M1B and 2M2B, respectively, by kinetic calculation.
2007 Elsevier B.V. All rights reserved.
0255-2701/$ see front matter 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.cep.2006.12.014
762 W. Mao et al. / Chemical Engineering and Processing 47 (2008) 761769
2.1. Apparatus
Fig. 2. Experimental unit: (1) gas cylinder; (2) (9) (12) valves; (3) thermocouples; (4) batch reactor; (5) bracket; (6) thermistor; (7) lab jack; (8) motor; (10) sampling
point; (11) cooler; (13) exhaust; (14) operating board.
W. Mao et al. / Chemical Engineering and Processing 47 (2008) 761769 763
2.3. Catalyst
Exchange capacity (mmol/g [H+ ]) 5.2. Fig. 3. Effect of agitation speed on yield of TAME (RM/IA = 1:1, T = 353 K).
Average porosity size (nm): 3845.
Particle size (mm): 0.351.25. 3.1.2. Effect of resin particle size on reaction rates
Specific area (m2 ): 4050. The catalyst was sieved as dry form into four size groups:
(I) 0.350.42 mm, (II) 0.420.50 mm, (III) 0.500.85 mm,
Before the reaction, the resin was treated several times by and (IV) 0.851.25 mm. Experiments were carried out at
methanol, then dried in a vacuum dryer at 80 C for 24 h and 400 rpm and other conditions were the same as agitation
cooled in a desiccator for 1 h. speed studying. The result of yield of TAME with different
particle size fractions were presented in Fig. 4 which indi-
cates that the particle size influences the reaction rate most
2.4. Materials remarkably when the particle size is over 0.85 mm, but no
significant effect is observed with the other three particle
The reagents utilized in this work were methanol (MeOH, sizes.
Fisher Scientific, p.a. Grade, 99.9 wt%) and isoamylenes To justify this result, the diffusion in D005II catalyst was
(8.22 wt% of 2M1B, 90.78 wt% of 2M2B, and 1.00 wt% of calculated following the methods published by Paakkonen and
other inert C5 compounds, Shangdong Hengyuan Petrochem- Krause [18] and Silva and Rodrigues [19]. As the result shown,
ical Corporation, China). For calibration solutions, methanol the effectiveness factor () decreases with the increasing of
(MeOH, 99.9 wt%), 2M1B (99 wt%, JK chemicals), TAME the particle size. When the particle size is between 0.85 and
(96 wt%, by distillation) were used. 1.25 mm, the effectiveness factor only equals to 0.81. That indi-
cates there is internal mass transfer effect in the catalyst to
3. Results and discussion
Table 1
Experimental reaction conditions (initial mole of methanol/isoamylene, temperature), mole fractions of methanol (1), 2-methyl-1-butene (2), 2-methyl-2-butene (3),
TAME (4), calculated activity coefficients and equilibrium constants
RM/IA T (K) Equilibrium molar fractions (%) Wilson activity coefficients Ka1 Ka2 Ka3
X1 X2 X3 X4 1 2 3 4
0.5 323 0.0229 0.0383 0.4787 0.4601 6.0257 1.0916 1.0304 0.9796 78.13 6.62 11.80
0.5 333 0.0310 0.0425 0.4822 0.4443 5.5927 1.0989 1.3069 0.9730 53.39 3.96 13.49
1.0 333 0.2140 0.0161 0.1915 0.5784 2.4636 1.2733 1.2555 1.0141 54.27 4.63 11.73
2.0 333 0.5652 0.0063 0.0770 0.3514 1.3536 2.1785 2.1364 1.4163 47.40 3.95 11.99
0.5 343 0.0406 0.0467 0.4909 0.4217 5.2368 1.1085 1.0445 0.9648 36.97 3.73 9.90
1.0 343 0.2425 0.0199 0.2090 0.5286 2.3106 1.3196 1.2944 1.0149 36.46 3.54 10.30
2.0 343 0.5916 0.0072 0.0769 0.3242 1.3094 2.2895 2.2378 1.4472 36.74 3.52 10.44
0.5 353 0.0798 0.0515 0.4826 0.3861 4.3643 1.1464 1.0798 0.9412 17.67 2.00 8.83
1.0 353 0.3101 0.0217 0.2155 0.4527 2.0206 1.4394 1.4027 1.0366 23.98 2.48 9.68
2.0 353 0.5652 0.0101 0.1007 0.3240 1.3516 2.1628 2.1078 1.3594 26.39 2.72 9.72
W. Mao et al. / Chemical Engineering and Processing 47 (2008) 761769 765
Table 3
Comparison with experimental and literature parameters in equilibrium constant
equations
ln Kai = Ai +Bi /T
A1 B1 A2 B2
2.916 102
Fig. 6. Vant Hoff plotting for the ln Kai vs. 1/T. lnKa2 = 24.9421 ln[T (K)]
T (K)
+ 7.356 102 T (K) 4.192 105 [T (K)]2
3527.151
lnKa2 = 9.091 + (3)
T + 1.255 102 (7)
1217.060
lnKa3 = 1.201 + (4) Table 2 shows there are some discrepancies between the
T
experimental data and the theoretical data, where the experimen-
As Ka3 could be calculated by Ka1 and Ka2 , only Ka1 and Ka2 tal data for Ka1 is a little smaller and the experimental data for Ka2
were compared to the theoretical equilibrium constants (Table 2) is a little greater than the counterparts. On average, Eqs. (2) and
in this work. The theoretical equilibrium constant for reaction i (3) provide values with errors of 10.4 and 13.7%, respectively,
could be written as a function of temperature by the form within the studied temperature range. Thus, a fair agreement is
obtained between the experimental data and the theoretical data.
H0 a 1 1 Also, these values are comparable to the ones published in the
lnKai (T ) = + lnT + b T + c T 2 + I
RT R 2R 6R literature which are listed in Table 3.
(5) On the other hand, besides the definition of Eq. (1), GoiT
could also be described by Eq. (8).
where the referential factors had been published by Syed et al.
[20]. The values were submitted into Eq. (5) to obtain the finial GoiT = RT lnKai = T Sio Hio (8)
form (Eqs. (6) and (7)) for the resulting equilibrium constants From Eqs. (1) and (8), the relationship between the reaction
for the TAME reaction network [7]. enthalpies and the equilibrium constants had been established
(Eq. (9)). From the same equations, the value of the standard
1.190 103 reaction entropies were also obtained, which were evaluated by
lnKa1 = 25.801 ln[T (K)]
T (K) the intercept multiplying the ideal gas constant R.
+ 7.599 102 T (K) 4.377 105 [T (K)]2 d(lnKai ) Hio
= (9)
d(1/T ) R
+ 1.299 102 (6)
As Table 4 shown, an excellent agreement is obtained
between experimental and literature data published lately except
Table 2 that So in R3 has deviations from values of literatures. The dif-
Comparison with experimental and theoretical equilibrium constants
ferences were likely due to the different thermodynamic data and
T (K) Experimental data Theoretical data method used in the evaluation.
Ka1 323 81.16 10.44 89.93
333 52.20 6.72 58.82 3.3. Kinetic modeling
343 34.45 4.43 39.29
353 23.27 2.99 26.75 In the published kinetic models for the synthesis of TAME,
Ka2 323 6.23 0.90 5.38 concentration was verified unfit for describing the strongly
333 4.49 0.65 3.88 nonideal kinetic behavior of the reagents, thus rate calcula-
343 3.29 0.47 2.84
tion should be set in terms of activities. In this paper, three
353 2.46 0.35 2.11
correlations based on different mechanisms were evaluated
766 W. Mao et al. / Chemical Engineering and Processing 47 (2008) 761769
Table 4
Liquid phase standard enthalpy and entropy changes of reaction obtained from experimental and literature data
Reaction Experimental data Literature data
1
Ho (kJ mol1 ) So (J mol1 K1 ) GoT (kJ mol ) Ho (kJ mol1 ) So (J mol1 K1 ) GoT (kJ mol1 )
k1 KM K1B (aM a1B (aT /Ka1 )) k1 KM (a1B aM (aT /Ka1 )) a2B
r1B = r1B = k3 a1B (20)
(1 + KM aM + K1B a1B + K2B a2B + KT aT )2 KM aM + KT aT Ka3
k3 K1B (a1B (a2B /Ka3 )) k2 KM (a2B aM (aT /Ka2 )) a2B
(14) r2B = + k3 a1B (21)
1 + KM aM + K1B a1B + K2B a2B + KT aT K M aM + K T aT Ka3
where
k2 KM K2B (aM a2B (aT /Ka2 ))
r2B = k1 KM k2 KM k3 K1B
(1 + KM aM + K1B a1B + K2B a2B + KT aT )2 Ka1 = , Ka2 = , Ka3 = .
k 1 KT k 2 KT k 3 K2B
k3 K1B (a1B (a2B /Ka3 )) Values of the kinetic constants ki were calculated by the
+ (15)
1 + KM aM + K1B a1B + K2B a2B + KT aT simplified model from the experimental data on the ini-
W. Mao et al. / Chemical Engineering and Processing 47 (2008) 761769 767
Table 5
Modeling result with D005II as catalyst
Mechanism Parameter Value S.D. S.D.% WSRS
tial rates of reactions at different temperature (323353 K) for the etherification of 2M1B and 2M2B, respectively, with
and different mole ratio of materials (RM/IA = 0.5, 1.0, 2.0). D005 catalyst produced in the same corporation as D005II cat-
The particle size of catalyst used in kinetic study was alyst. Since there are only one paper published with D005 type
0.350.50 mm to gain the more favorable data. Finally, the catalyst on kinetic studies, so the reaction energies were also
activation energies for the reactions were obtained from the lin- compared with those obtained by Amberlyst type catalyst and
ear regression fitted to Arrhenius type temperature dependence other catalysts. Rihko et al. [15] reported activation 72.6 and
(Eq. (18)). 94.1 kJ mol1 for the etherification of 2M1B and 2M2B. Piccoli
and Lovisi [4] published values of 85.6 and 92.5 kJ mol1 for
Eai reactions 1 and 2. The values of 77.0, 85.7 and 89.8 kJ mol1 ,
lnki = lnAi (22)
RT respectively, were obtained by Paakkonen and Krause [11] for
The parameters for the regression analysis based on the three the reaction system. Also some other kind resins have been used
kinetic models at 343 K are summarized in Table 5, which are for the kinetic studies of TAME synthesis, and the activation
the values of the parameters, their standard deviations (S.D.). energies published are comparative to the counterparts in this
To optimize the fitting, the modeling parameters were acquired work. In the study of Hoffmann [13], the etherification energies
by minimizing the weighted sum of residual squares (WSAS) were 89.5 kJ mol1 by lumping the isoamylenes together using
between the experimental and calculated rates of reaction and Lewatit SPC 118 catalyst. Paakkonen and Krause [11] also pub-
weight factor (wi ) equaled to 1 for every calculation. lished activation energies for 92.7 and 93.0 kJ mol1 for 2M1B
and 2M2B etherification using SMOPEX-101 catalyst. And the
WSRS = wi (ri,exp ri,calc )2 (23) values of 96.78 and 102.4 kJ mol1 , respectively, were obtained
with NKC-9 catalyst by Jin et al. [23].
According to the statistics, the LH model has the smallest Moreover, in order to validate the application of the cho-
WSAS (0.0204) and the homogeneous model largest (0.0455). sen kinetic mechanism, we used the kinetic equations based
These results are identical to those obtained by others [7,17], on three type mechanisms in order to see whether the derived
which indicated that the mechanism base on the two active sites model describes the experimental results within the experi-
are more appropriate. As the table shown, the value of kinetic mental conditions. In Fig. 7 the experimental and simulated
constant of R1, k1 , is dramatically greater than the value of k2 , molar fractions of the components are presented as a function
and the activation energy of R1 is less than that of R2. Thus, it is of reaction time. In the comparison experiments, the temper-
obvious that TAME synthesis from 2M1B is faster than TAME ature varies from 323 to 353 K and RM/IA contains 1:1, 1:2
synthesis from 2M2B which is in agreement with the conclusion and 2:1, to gain the reliable results. As the results shown in
in the thermodynamic study above. the figure, the homogeneous reaction mechanism is the least
Finally, the activation energies of the reactions, 88.1 kJ mol1 seemly model for its largest deviation. But the experimental
for the etherification of 2M1B, 102.1 kJ mol1 for the ether- results can be simulated by LH model accurately. Therefore,
ification of 2M2B and 99.1 kJ mol1 for the isomerization, we could conclude that the kinetic equations of LH and param-
respectively, are in good agreement with literature values. In eters in Table 5 can satisfactorily describe the synthesis of
1998, Yao et al. [22] reported values of 81.36 and 108.7 kJ mol1 TAME.
768 W. Mao et al. / Chemical Engineering and Processing 47 (2008) 761769
Fig. 7. Comparison of experimental and simulated molar fractions of TAME at different conditions by (a) homogeneous model, (b) LH model and (c) ER model.
Gi standard Gibbs free energy change of reaction i [4] R.L. Piccoli, H.R. Lovisi, Kinetic and thermodynamic study of the liquid-
(J mol1 ) phase etherification of isoamylenes with methanol, Ind. Eng. Chem. Res.
GiT standard Gibbs free energy change of reaction i at 34 (1995) 510515.
[5] S.P. Verevkin, Thermochemistry of branched ethers: experimental study
298.15 K (J mol1 ) of chemical equilibrium in the reacting system of tert-amyl alkyl ether
Hi standard enthalpy change of reaction i (J mol1 ) synthesis, J. Chem. Eng. Data 49 (2004) 576581.
ki kinetic constant for reaction i in the forward direction [6] N. Boz, T. Dogu, K. Murtezaoglu, G. Dogu, Mechanism of TAME and
(mmol h1 g1 cat (dry))
TAEE synthesis from diffuse-reflectance FTIR analysis, Catal. Today 100
Kai thermodynamic equilibrium constant of reaction i (2005) 419424.
[7] M.V. Ferreira, J.M. Loureiro, Number of actives sites in TAME synthe-
Kj adsorption equilibrium constant of component j sis: mechanism and kinetic modeling, Ind. Eng. Chem. Res. 43 (2004)
Kxi equilibrium constant based on molar fractions of reac- 51565165.
tion i [8] W.B. Su, J.R. Chang, Modeling and simulation of tubular reactor in the
Ki equilibrium constant based on activity coefficients of tert-amyl methyl ether synthesis process, Ind. Eng. Chem. Res. 39 (2000)
reaction i 41404147.
[9] L.K. Rihko, J.A. Linnekoski, A.O.I. Krause, Reaction equilibria in the
N number of components synthesis of 2-methoxy-2-methyl butane and 2-ethoxy-2-methylbutane in
ri rate of reaction i (mmol h1 g1
cat (dry)) the liquid phase, J. Chem. Eng. Data 39 (1994) 700704.
R ideal gas constant (8.3145 J mol1 K1 ) [10] L.K. Rihko-Struckmann, P.V. Latostenmaa, A.O.I. Krause, Interaction
RM/IA initial molar ratio of methanol versus isoamylenes between the reaction medium and an ion-exchange resin catalyst in the
Si standard entropy change of reaction i (J mol1 ) etherification of isoamylenes, J. Mol. Catal. A-Chem. 177 (2001) 4147.
[11] P.K. Paakkonen, A.O.I. Krause, Comparative study of TAME synthesis on
T temperature (K) ion-exchange resin beads and a fibrous ion-exchange catalyst, React. Funct.
xj molar fraction of component j Polym. 55 (2003) 139150.
xje equilibrium molar fraction of component j [12] L.K. Rihko-Struckmann, J.A. Linnekoski, A.O.I. Krause, Vapor-liquid and
chemical reaction equilibria in the synthesis of 2-methoxy-2-methylbutane
(TAME), J. Chem. Eng. Data 45 (2000) 10301035.
Greek letters
[13] C. Oost, U. Hoffmann, The synthesis of tertiary amyl methyl ether (TAME):
j liquid-phase activity coefficient of component j microkinetics of the reactions, Chem. Eng. Sci. 51 (3) (1996) 329340.
je equilibrium liquid-phase activity coefficient of compo- [14] J.A. Linnekoski, A.O.I. Krause, L.K. Rihko, Kinetics of the heteroge-
nent j neously catalyzed formation of tert-amyl ethyl ether, Ind. Eng. Chem. Res.
effectiveness factor 36 (2) (1997) 310316.
[15] L.K. Rihko, P. Kiviranta-Paakkonen, A.O.I. Krause, Kinetic model for
ji stoichiometric coefficient of component j in reaction i
the etherification of isoamylenes with methanol, Ind. Eng. Chem. Res.
36 (1997) 614621.
Subscripts [16] H. de Lasa, P. Fournier, A. Prakash, T. El Solh, MTBE synthesis in a novel
a in term of activity riser simulator, Can. J. Chem. Eng. 77 (1999) 413419.
e at equilibrium [17] P. Kiviranta-Paakkonen, A.O.I. Krause, Simultaneous isomerization and
etherification of isoamylenes with methanol, Chem. Eng. Technol. 26 (4)
i of reaction i
(2003) 479489.
j of species j [18] P.K. Paakkonen, A.O.I. Krause, Diffusion and chemical reaction in isoamy-
M methanol lene etherification within a cation-exchange resin, Appl. Catal., A 245
T tert-amyl methyl ether (2003) 289301.
x in term of molar fraction [19] V.M.T.M. Silva, A.E. Rodrigues, Kinetic studies in a batch reactor using ion
exchange resin catalysts for oxygenates production: role of mass transfer
1B 2-methyl-1-butene (2M1B)
mechanisms, Chem. Eng. Sci. 61 (2006) 316331.
2B 2-methyl-2-butene (2M2B) [20] F.H. Syed, C. Egleston, R. Datta, tert-Amyl methyl ether (TAME). Ther-
modynamic analysis of reaction equilibria in the liquid phase, J. Chem.
References Eng. Data 45 (2000) 319323.
[21] J.A. Serda, J.F. Izquierdo, J. Tejero, F. Cunill, M. Iborra, Equilibrium and
[1] D. Varisli, T. Dogu, Simultaneous production of tert-amyl ethyl ether and thermodynamics for 2-methyl-2-methoxy butane liquid-phase decomposi-
tert-amyl alcohol from isoamyleneethanolwater mixtures in a batch- tion, Thermochim. Acta 259 (1995) 111120.
reactive distillation column, Ind. Eng. Chem. Res. 44 (2005) 5227 [22] W. Yao, G. Lu, G. Zhang, B. Fang, L. Fu, Z. Yang, Study on kinetics
5232. of the liquid-phase etherification of iso-amylenes with methanol, Shiyou
[2] I. Muja, A. Toma, D.C. Popescu, I. Ivanescu, V. Stanisteanu, Thermody- Huagong 27 (6) (1998) 391395.
namic study of the methanol addition to isoamylene, Chem. Eng. Process. [23] H.-B. Jin, F.-R. Xiao, C.-Y. Yang, Z.-M. Tong, Micro-kinetics of the liquid-
44 (2005) 645651. phase synthesis of tertiary amyl methyl ether by ion-exchange resin with
[3] N. Boz, T. Dogu, Reflux-recycle-reactor for high yield and selectivity in isoamylenes and methanol, Shiyou Huagong Gaodeng Xuexiao Xuebao 15
TAME and TAEE production, AIChE J. 51 (2005) 631640. (1) (2002) 2632.