Professional Documents
Culture Documents
Dynamic Modelling of Glucose Oxidation W PDF
Dynamic Modelling of Glucose Oxidation W PDF
a r t i c l e i n f o a b s t r a c t
Article history: Heterogeneously catalyzed oxidation of glucose in liquid phase was studied as a model reaction under
Received 2 December 2008 conditions of reversible deactivation of a palladium catalyst. The proposed kinetics model fitted the exper-
Received in revised form 17 February 2009 imental data very well. The model was used to explore benefits of a CSTR operation under forced periodic
Accepted 18 February 2009
modulation of the gas feed stream. The periodic modulation makes it possible to optimize the cata-
lyst reaction–reactivation processes in CSTR, so that the highest possible reactor productivity could be
Keywords:
achieved for given input parameters. Under such operation of the multifunctional CSTR, the catalyst activ-
Glucose oxidation
ity is long-term steady. Based on simulations, conditions of the highest possible CSTR productivity were
Palladium catalyst
Reactivation
identified for given operational parameters.
CSTR © 2009 Elsevier B.V. All rights reserved.
Periodic operation
1. Introduction during the reaction course. The deactivation was proved reversible
[3].
In many practical applications gas–liquid reactions are catalyzed There are two major reasons behind the catalyst deactivation.
by a heterogeneous catalyst. Such reactions usually take place in a First, in diffusion regime the rate limiting step of the process is
three-phase slurry reactor. A good example of such reaction systems oxygen transport towards the catalyst surface. The deactivation is
are oxidations with oxygen, catalyzed by transient metals like wet caused by a strong adsorption of the substrate degradation prod-
air oxidation of pollutants in waste waters, production of chemical ucts that block the catalyst active centres [2,4]. On the other hand,
specialities, etc. in kinetic regime where the reaction runs in excess of oxygen, the
Glucose (Glc) oxidation with oxygen to gluconic acid (Glcac) catalyst deactivation results from formation of metal-oxide phases
is the model reaction in this study. It runs in an alkaline aque- [15–17]. They are less active compared to chemisorbed oxygen (oxy-
ous solution and it is catalyzed by palladium on activated carbon gen adsorbed above the first metallic layer), and so will be referred
commercial catalyst. This oxidation itself is interesting from the to as those responsible for change in the catalyst activity. In the
viewpoint of the reaction product utilization in food, pharmaceu- case of palladium catalysts this deactivation proceeds very quickly.
tical, textile and other industrial fields. However, to overcome the deactivation effect is challenging and
The research interest is mainly focused on several methods of interesting research task.
Glcac production (or structures similar to Glcac) – catalytic oxida- A classical approach to overcome the catalyst activity problem
tion [1–8], fermentation techniques [9–12] or enzymatic oxidation leads through the catalyst optimization. Earlier studies reported
[13,14]. The biotechnological routes of Glcac production currently on beneficial effect of doping the Pt-group metals with promoters,
prevail over the catalytic one on an industrial scale. Glc oxidation by such as Bi or Pb, that improve selectivity and the catalyst resistance
oxygen using noble metal catalysts runs selectively at ambient con- to overoxidation [4]. Lifetime of the catalyst remained a prob-
ditions. This fact makes this catalytic process attractive for research. lem, though, because of the promoters leaching under the reaction
Its problem, on the other hand, is fast deactivation of the catalyst conditions. Recently, very good activity, selectivity and long-term
stability were reported for supported gold catalysts [6–8].
Another approach to solve the problems with the catalyst activ-
ity deals with the process or reactor optimization. Beneficial effects
Abbreviations: CSTR, continuous stirred tank reactor; GLR, gas-lift reac- of composition modulation on a CSTR performance are presented
tor; LHHW, Langmuir–Hinshelwood–Hougen–Watson (kinetics); SSTR, semi-
continuous stirred tank reactor.
here. In the CSTR, the catalytic process is accompanied by fast
∗ Corresponding author. Tel.: +420 2 20390349; fax: +420 2 20920661. reversible catalyst deactivation. We analyzed possibilities of the
E-mail address: [email protected] (J. Hanika). multifunctional CSTR performance enhancement by optimization
1385-8947/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.cej.2009.02.020
224 Z. Gogová, J. Hanika / Chemical Engineering Journal 150 (2009) 223–230
Table 1
Elementary steps sequence of the proposed mechanism for the model reaction of Glc oxidation with the catalyst deactivation and reactivation.
A rate equation is separable if it can be expressed as a product of Value of kL a was for the experimental reactor calculated from
two terms – the reaction rate on the fresh catalyst and the catalyst correlation (14) [26], derived for oxygen (or air)–water system (coa-
activity in the following form: lescing system).
˙ w = ˙ w
0
(c̄)a(˛) [T ] = const. (5) kL a = 6.64 × 10−3 p0.675
V UG0.494 (14)
The total power input per unit volume of liquid, pV , was adjusted
Active fraction ˛ is defined as the ratio of the number of active
for geometry of the experimental reactor by using power number,
sites per unit mass of the catalyst and the number of all sites, i.e.:
NP , obtained from correlation (15) [27]:
˛=
Cch,tot − Cso
=
1 − so
(6)
x −0.2 D 0.065
R
Cch,tot 1 NP = 2.5 (15)
DM 1000
The rate equation (1) is separable. Combining Eqs. (1), (5) and Oxygen saturation concentration in the liquid phase, cO∗ , was
(6) gives the following relation between the catalyst activity and calculated according to data on oxygen solubility in Glc aqueous
the active fraction: solutions [28] by using the following regression equation:
cO∗ = p(O2 )(1.1624 × 10−5 − 2.3799 × 10−8 (cGlc + cGlcac )0.6907 ) (16)
2
a = (1 − so ) = ˛2 (7)
Thus the activity depends only on the amount of inactive oxygen Simulations of reactors provided in this paper were run using
species. The activity change in Eqs. (1) and (2) comes from the commercial software Matlab.
following relation:
3. Experimental
d˛ dso
− = (8)
dt dt 3.1. The model reaction kinetics study
2.2. Mathematical model of the batch reactor The model reaction kinetics was studied in the lab-scale semi-
continuous stirred tank reactor, SSTR (cylindrical vessel of volume
All the experiments were carried out in semi-continuous stirred 1 L, diameter DR = 0.1 m; liquid level HL = DR ; 4 baffles; mixer: 6-
tank reactor, SSTR (see the next chapter). Inlet and outlet concen- bladed Rushton turbine of diameter DM = 0.04 m and thickness
trations of oxygen in the SSTR gas streams were assumed identical. x = 1 mm). The gas phase was introduced and discharged contin-
Therefore the SSTR mathematical model consists of liquid phase uously. From the liquid and solid phases viewpoint the reactor was
material balances only Eqs. (9)–(13). operated batchwise. The reactor was equipped with thermocou-
√ 2 ple, pH-control, the stirring frequency control, pressure sensor and
dcGlc kw cGlc cO (1 − so ) polarographic oxygen probe for monitoring the oxygen concentra-
= −c √ (9)
dt (1 + KGlc cGlc + KGlcac cGlcac )(1 + KO cO )2 tion in the liquid phase. Ranges of reaction conditions are listed in
Table 2. Kinetics of the model reaction was studied in the kinetics
dcGlcac dc
= − Glc (10) regime – influence of internal and external mass transfer was neg-
dt dt ligible (see [29] for evidence). Properties of the Pd/C catalyst used
in this study are summarized in Table 3.
√ 2
dcO 1 kw cGlc cO (1 − so )
= kL a(cO∗ − cO ) − c
dt 2 (1 + KGlc cGlc + KGlcac cGlcac )(1 + KO √cO )2 Table 2
Experimental conditions.
(11)
Condition Range
Table 3
Characteristics of the Pd/C catalyst.
Property Value
Fig. 2. Initial reaction rate as a function of (a) Glc initial concentration, and
(b) oxygen concentration in bulk liquid. Dots, measured; lines, solution of the
Fig. 3. Simulations of four reaction cycles in CSTR operating in periodic O2 /N2 mode
0
kinetic model. Experimental conditions: SSTR; cGlcac = 0.5 mol m−3 ; c = 1 kg m−3 ;
for various oxygen molar fractions in the gas feed stream. Time course of (a) imme-
0
T = 303 K; pH = 8.1; (b) cGlc = 20 mol m−3 .
diate rate of Glcac formation, (b) the catalyst fractional coverage by inactive oxygen
species. Conditions: tR = 3600 s; tA = 1800 s; c = 1 kg m−3 ; VR = 860 cm3 ; cGlc = const.;
f f
XGlc = 2%; cGlc = 100 mol m−3 ; cGlcac = 0.
Fig. 5. Time course of the catalyst activity, expressed through ratio of the immediate
to the initial rate of Glcac formation in CSTR as a function of (a) oxygen molar fraction
in the gas feed stream, (b) Glc concentration in the reaction mixture. The case “(a)”
Fig. 4. Simulations of four reaction cycles in CSTR operating in periodic O2 /N2 mode corresponds to the conditions of the second reaction cycle of Fig. 3; the case “(b)” to
for various Glc concentrations in the reaction mixture. Time course of (a) immedi- similar conditions of Fig. 4.
ate rate of Glcac formation, (b) the catalyst fractional coverage by inactive oxygen
species. Conditions: tR = 3600 s; tA = 1800 s; c = 1 kg m−3 ; VR = 860 cm3 ; cGlc = const.;
f f
XGlc = 2%; YO = 0.21; cGlcac = 0. phase. With its raise, both the so steady-state value as well as the
transient one increase (Fig. 3b), and the catalyst’s transient and
steady-state activity decreases (Fig. 5a). On the contrary, unimpor-
examined further. For deeper insight into the model system under tant is the effect of Glc concentration in the reaction mixture on the
periodic mode of operation, the SSTR model (Eqs. (9)–(13)) was catalyst deactivation extent (Figs. 4b and/or 5b).
modified for conditions of CSTR operating at constant Glc and Glcac
concentrations in time and varying volumetric flow rate of the liq- 4.4. Optimization of the CSTR productivity
uid feed stream according to the actual reaction rate. Results of this
approach are presented in Figs. 3–5. This section deals again with CSTR running at constant Glc and
In the kinetics model (Eqs. (1) and (2)), the change in the catalyst Glcac concentrations in time and varying the volumetric flow rate
activity is expressed through the change in the fractional coverage of the liquid feed stream according to the actual reaction rate. CSTR
by inactive oxygen species, so . Relation between so and the activity of this section differs in the way of the gas feed flow modulation.
is given in Eq. (7). The reactor is now operated under on–off periodic mode, i.e. with
The CSTR application of the kinetics model enables, in contrast alternating cycles of switching on and off air feed stream. The cata-
with the SSTR one, to separate the effect of the reagents concentra- lyst reactivation in this case runs without assistance of nitrogen.
tions from the effect of the change in the catalyst activity itself on Fig. 6a shows course of the model reaction cycle-time-averaged
deviation of the reaction rate in time. Thus, it makes the study of reaction rate as a function of the period length and split value.
the deactivating reaction system more transparent. The CSTR model The period length is defined as a sum of reaction and reactivation
makes it possible to express the catalyst activity directly through times (P = tR + tA ). The split of period is the time the reaction takes in
the ratio of the immediate to the initial reaction rate (i.e. by using relation to the entire period length (S = tR /(tR + tA )). As can be seen
Eq. (4)), and reveals directly the progress in the extent of the catalyst in Fig. 6a, for any period length always a split value exists, where
deactivation. This approach is for the cases of the second reaction maximum reaction rate is achieved. The reaction rate is cycle-time-
cycle of Figs. 3 and 4 depicted in Fig. 5. averaged and represents the reactor productivity.
Comparison of Figs. 3b and 4b with 5a and b, respectively, leads The optimization task was to find conditions that guarantee the
through Eq. (7). The rate and the extent of the catalyst deactiva- highest reactor productivity at any period given. In other words,
tion are definitely influenced by oxygen concentration in the liquid the goal was to share optimally the reaction and reactivation times
Z. Gogová, J. Hanika / Chemical Engineering Journal 150 (2009) 223–230 229
reaction mixture has negligible effect on the catalyst deactivation [16] G.W. Simmons, Y. Wang, J. Marcos, K. Klier, Oxygen adsorption on Pd(1 0 0)
extent. surface: phase transformations and surface reconstruction, J. Phys. Chem. 95
(1991) 4522–4528.
[17] E. Lundgren, G. Kresse, C. Klein, M. Borg, J.N. Andersen, M. De Santis, Y. Gauthier,
Acknowledgements C. Konvicka, M. Schmid, P. Varga, Two-dimensional oxide on Pd(1 1 1), Phys. Rev.
Lett. 88 (24) (2002).
[18] P.L. Silveston, J. Hanika, Challenges for the periodic operation of trickle-bed
We acknowledge financial support of Grant Agency of the Czech catalytic reactors, Chem. Eng. Sci. 57 (2002) 3373–3385.
Republic (grant no.: 203/08/H032). Our thanks go to Prof. Jozef [19] P.L. Silveston, J. Hanika, Periodic operation of three-phase catalytic reactors,
Markoš from Slovak University of Technology in Bratislava for read- Can. J. Chem. Eng. 82 (2004) 1105–1142.
[20] J.G. Boelhouwer, H.W. Piepers, A.A.H. Drinkenburg, Advantages of forced non-
ing the manuscript and providing valuable feedback. steady operated trickle-bed reactors, Chem. Eng. Technol. 25 (2002) 647–
650.
References [21] V. Tukač, J. Hanika, V. Chyba, Periodic state of wet oxidation in trickle-bed
reactor, Catal. Today 79–80 (2003) 427–431.
[22] G. Liu, X. Zhang, L. Wang, S. Zhang, Z. Mi, Unsteady-state operation of trickle-bed
[1] I. Nikov, K. Paev, Palladium on alumina catalyst for glucose oxidation: reaction reactor for dicyclopentadiene hydrogenation, Chem. Eng. Sci. 63 (2008) 4991–
kinetics and catalyst deactivation, Catal. Today 24 (1995) 41–47. 5002.
[2] A.P. Markusse, B.F.M. Kuster, J.C. Schouten, Platinum catalysed aqueous methyl- [23] P.L. Silveston, Composition modulation of catalytic reactors, Gordon and Breach
a-D-glucopyranoside oxidation in a multiphase redox-cycle reactor, Catal. Science Publishers, 1998.
Today 66 (2001) 191–197. [24] S. Szépe, O. Levenspiel, Catalyst deactivation, in: Proceedings of the 4th Euro-
[3] J.H. Vleeming, B.F.M. Kuster, G.B. Marin, Selective oxidation of methyl ␣-D- pean Symposium on Chemical Reaction Engineering, Oxford, 1970.
glucopyranoside with oxygen over supported platinum: kinetic modeling in [25] J.B. Butt, E.E. Petersen, Activation, Deactivation and Poisoning of Catalysts, Aca-
the presence of deactivation by overoxidation of the catalyst, Ind. Eng. Chem. demic Press, Inc., 1988.
Res. 36 (1997) 3541–3553. [26] V. Linek, T. Moucha, J. Sinkule, Gas-liquid mass transfer in vessels stirred with
[4] T. Mallat, A. Baiker, Oxidation of alcohols with molecular oxygen on platinum multiple impellers. I. Gas–liquid mass transfer characteristics in individual
metal catalysts in aqueous solutions, Catal. Today 19 (1994) 247–284. stages, Chem. Eng. Sci. 51 (12) (1996) 3203–3212.
[5] M. Kunz, C. Recker, A new continuous oxidation process for carbohydrates, [27] W. Bujalski, A.W. Nienow, S. Chatwin, M. Cooke, The dependency on scale of
Carbohydr. Eur. 13 (1995) 11–15. power numbers of Rushton disc turbines, Chem. Eng. Sci. 42 (2) (1987) 317–
[6] A. Mirescu, U. Prusse, A new environmental friendly method for the preparation 326.
of sugar acids via catalytic oxidation on gold catalyst, Appl. Catal. B: Environ. [28] H. Eya, K. Mishima, M. Nagatani, Y. Iwai, Y. Arai, Measurements and correlation
70 (2007) 644–652. of solubilities of oxygen in aqueous solutions containing glucose, sucrose and
[7] N. Thielecke, K.D. Vorlop, U. Prusse, Long-term stability of an Au/Al2 O3 cata- maltose, Fluid Phase Equilib. 94 (1994) 201–209.
lyst prepared by incipient wetness in continuous-flow glucose oxidation, Catal. [29] Z. Gogová, J. Hanika, Experimental study of palladium catalyst reactivation
Today 122 (2007) 266–269. during glucose oxidation by molecular oxygen, Chemical Pap., 63 (2009), in
[8] S. Biella, L. Prati, M. Rossi, Selective oxidation of D-glucose on gold catalyst, J. press.
Catal. 206 (2002) 242–247. [30] P. Zámostný, Z. Bělohlav, A software for regression analysis of kinetic data,
[9] J. Klein, M. Rosenberg, J. Markoš, O. Dolgoš, M. Krošlák, L’. Krištofíková, Bio- Comput. Chem. 23 (1999) 479.
transformation of glucose to gluconic acid by Aspergillus niger—study of mass [31] V.R. Gangwal, J. van der Schaaf, B.F.M. Kuster, J.C. Schouten, Noble-metal-
transfer in an airlift bioreactor, Biochem. Eng. J. 10 (3) (2002) 195–205. catalysed aqueous alcohol oxidation: reaction start-up and catalyst deactivation
[10] M. Juraščík, M. Hucík, I. Sikula, J. Annus, J. Markoš, Influence of biomass on and reactivation, J. Catal. 232 (2005) 432–443.
hydrodynamics of an internal loop airlift reactor, Chem. Pap. 60 (6) (2006) [32] Z. Gogová, J. Hanika, Purpose tailored design of gas-lift reactor, in: Proceedings
441–445. of the 35th International Conference of SSCHE, Tatranské Matliare, Slovakia,
[11] O.V. Singh, R. Kumar, Biotechnological production of gluconic acid: future impli- 2008.
cations, Appl. Microbiol. Biotechnol. 75 (2007) 713–722. [33] Z. Gogová, J. Hanika, Model aided design of gas-lift reactor for oxidation reaction
[12] I. Sikula, M. Juraščík, J. Markoš, Modeling of fermentation in an internal loop with fast reversible catalyst deactivation, in: Proceedings of the 10th interna-
airlift bioreactor, Chem. Eng. Sci. 62 (2007) 5216–5221. tional Chemical and Biochemical Conference Chempor 2008, Braga, Portugal,
[13] J. Liu, Z. Cui, Optimization of operating conditions for glucose oxidation in an 2008.
enzymatic membrane bioreactor, J. Membr. Sci. 302 (2007) 180–187. [34] R.J. Berger, F. Kapteijn, J.A. Moulijn, G.B. Marin, J. De Wilde, M. Olea, D. Chen,
[14] I. Sikula, J. Markoš, Modeling of enzymatic reaction in an airlift reactor using an A. Holmen, L. Lietti, E. Tronconi, Y. Schuurman, Dynamic methods for catalytic
axial dispersion model, Chem. Pap. 62 (1) (2008) 10–17. kinetics, Appl. Catal. A: Gen. 342 (1–2) (2008) 3–28.
[15] G. Ketteler, D.F. Ogletree, H. Bluhm, H. Liu, E.L.D. Hebenstreit, M. Salmeron, In [35] M. Pekař, J. Koubek, Rate-limiting step. Does it exist in the non-steady state?
situ spectroscopic study of the oxidation and reduction of Pd(1 1 1), J. Am. Chem. Chem. Eng. Sci. 52 (14) (1997) 2291–2297.
Soc. 127 (2005) 18269–18273.