Kinetic Model Development for Dehydration of 2,3 ... - ACS Publications

Subscriber access provided by HKU Libraries

Article

Kinetic Model Development for Dehydration of 2,3Butanediol to 1,3-Butadiene and Methyl Ethyl Ketone over an Amorphous Calcium Phosphate Catalyst Daesung Song Ind. Eng. Chem. Res., Just Accepted Manuscript • DOI: 10.1021/acs.iecr.6b02930 • Publication Date (Web): 29 Sep 2016 Downloaded from http://pubs.acs.org on October 2, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Industrial & Engineering Chemistry Research is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 34

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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

Manuscript for Industrial & Engineering Chemistry Research

Kinetic Model Development for Dehydration of 2,3Butanediol to 1,3-Butadiene and Methyl Ethyl Ketone over an Amorphous Calcium Phosphate Catalyst

Daesung Songa,* a

Global Technology, SK innovation, 325 Exporo, Yuseong-gu, Daejeon 305-712, Republic of

Korea

To whom correspondence should be addressed: Daesung Song Tel: +82-42-609-8592. E-mail: [email protected]

1

ACS Paragon Plus Environment


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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Abstract A reaction network and kinetic model for the dehydration of 2,3-Butanediol(2,3-BDO) to 1,3Butadiene(1,3-BD) and Methyl Ethyl Ketone(MEK) on an amorphous calcium phosphate(a-CP) catalyst are herein proposed. The kinetic parameters of the model were estimated using experimental data obtained from a laboratory-scale fixed-bed reactor operated under isothermal conditions. Experiments were performed using a mixture of 2,3-BDO, 3-Buten-2-ol(3B2OL) and N2 as feed, at temperatures ranging from 304 to 334.5oC and gas hourly space velocity (GHSV) ranging from 1780 to 2222 hr-1. 2,3-BDO conversion varied from 6 to 100%, the selectivity of 1,3-BD ranged from 5 to 33 wt%, and the selectivity of MEK ranged from 31 to 34 wt%. Kinetic models based on the simple power law and Langmuir-Hinshelwood-Hougen-Watson(LHHW) were developed to describe the dehydration of 2,3-BDO to 1,3-BD and MEK. Statistical and physicochemical criteria are used to contrast the performance of the two kinetic approaches. The power law model showed the highest capacity to represent the tendency of experimental data obtained by changing temperature and GHSV. The kinetic parameters indicate: (i) the reaction orders(n1, n3, n4: 0.0187 and n2: 0.146) are very close to 0, meaning that reactor performance for 2,3-BDO dehydration is mainly determined by the temperature of the reactor, not by the concentrations of reactants; and (ii) the reaction routes that produce 1,3-BD demand higher activation energy than do others, explaining the quickly changing rate of 3B2OL and 1,3-BD selectivity and slowly changing rate of MEK and 2MPL selectivity with increasing temperature. 1. Introduction 1,3-BD and MEK are widely used in various industrial fields. 1,3-BD is used as hydrocarbon fuel, synthetic rubber, polymers, fibers, and plastics. On the other hand, MEK is an effective and common solvent and is also utilized in the synthesis of various fine chemicals and as a plastic

2


Page 2 of 34

Page 3 of 34

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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60



welding agent. The market sizes for 1,3-BD and MEK in 2015 are over 10 and 1.3 million metric tons, respectively[1]. However, these compounds are mainly prepared from petroleum, which is a finite resource and is recognized as a major cause of regional disparities and environmental pollution. Furthermore, as many gas crackers have been built, C4 oil fractions have decreased continuously and 1,3-BD is considered a high-value product. 2,3-BDO has been considered as a potential intermediate for the production of hydrocarbons including 1,3-BD and MEK, because 2,3-BDO can be produced through fermentation by various microorganisms, such as Klebsiella oxytoca, Enterobacter aerogenes, Bacillus licheniformis and Enterobacter cloacae[2]. Feedstock includes not only various biomasses but also synthetic gases (syngas) from coal gasification and industrial gas waste, e.g. from steel mills. Many companies are doing research on the production of liquid fuels using fermentation technologies, mainly ethanol; Lanzatech, for example, is a pioneer in gas fermentation to 2,3-BDO[3-4]. Research on the dehydration of 2,3-BDO to 1,3-BD and MEK using various catalysts has been reviewed by Makshina et al.[5] and research on the 1,3-BD production from biomass-derived C4 alcohols including the dehydration of butanediols to unsaturated alcohols were summarized by Hailing Duan et at.[6]. The dehydration of 2,3-BDO to 1,3-BD and MEK has been investigated since the 1940s. The catalysts used are bentonite clay[7], metal and earth oxides[8-12], zeolites[13-16], a perfluorinated resin with sulfonic acid groups[14], heteropolyacids[13,17], calcium phosphates[18-21], Cs/SiO2[22], and so on. Among them, the research group achieved the highest 1,3-BD Selectivity (>90%) using the SiO2-supported CsH2PO4[20]. Most methods focus on the dehydration catalysts or reaction conditions that create a high yield of target products including 1,3-BD, MEK, and 3B2OL using a special catalyst. An understanding of reaction kinetics is essential to commercialize the 2,3-BDO dehydration process, but, to our

3



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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


knowledge, research on reaction kinetics of the dehydration of 2,3-BDO to 1,3-BD and MEK has not been done. The purpose of this work is, therefore, to develop a suitable kinetic model for the dehydration of 2,3-BDO to 1,3-BD and MEK using an a-CP catalyst and to determine the kinetic parameters of a set of main reactions by fitting the experimental data obtained from a laboratory-scale fixedbed reactor operated under isothermal conditions. The kinetics described herein are novel as they characterize the a-CP catalyst in the reactions. Our experimental data, kinetic model, and analysis are expected to be very useful when the 2,3-BDO dehydration process is commercialized. 2. Kinetic modeling 2.1. Reaction pathway Figure 1. shows a probable reaction network using the a-CP catalyst based on experimental observations. The proposed reaction network is made based on a theoretic background [23]. It is possible to describe 2,3-BDO dehydration as a set of two parallel routes that produce the major products(1,3-BD, MEK, 3B2OL, 2MPN). The first route represents the traditional rearrangement mechanism, which occurs via a 1,3-hydride shift leading to MEK or a 1,2-methyl shift leading to 2MPN. The second route is a sequential 1,2-elimination of water leading to the formation of 3B2OL after the first dehydration and 1,3-BD after the second one[1]. In this study, we did not consider impurities in the reaction products, since the total amount of minor butene isomers and heavy compounds considered as impurities are less than 0.3wt% over all experiments, and disregarding impurities greatly simplifies the kinetic model.

4


Page 4 of 34

Page 5 of 34

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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60



-H2 O

Oligomerization Heavy compounds 1,3-BD

-H2 O

3B2OL

+H 2

-H2 O

-H2 O MEK

2,3-BDO -H2 O

? 2MPL

Heavy compounds

Figure 1. Major (solid arrows) and minor (dashed arrows) pathways of 2,3-BDO dehydration The major pathways of 2,3-BDO dehydration are described by the following reactions: r1 C4 H10O2   C4 H 8 O  H 2 O (2,3 BDO )

(1)

(3 B 2 OL )

r2 C4 H 8O   C4 H 6  H 2 O (3 B 2 OL )

(2)

(1,3 BD )

r3 C4 H10O2   C4 H 8 O  H 2 O (2,3 BDO )

(3)

( MEK )

r4 C4 H10O2   C4 H 8O  H 2O (2,3 BDO )

(4)

(2 MPL )

5



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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 34


2.2. Kinetic equations A power law and LHHW model are developed to characterize the reactions using the a-CP catalyst. 2.2.1. Power law expressions. The reaction kinetics of the dehydration of 2,3-BDO to 1,3-BD and MEK have not yet been studied. Therefore, the simple power law below was used to describe the reaction rates as a first step.

ri  ki Creact ,i ni

(5)

with

Cj 

Pj

(6)

RT

where the rate constants are modeled as the following Arrhenius equation with a reference temperature to reduce the degree of nonlinearity

ki  kTref ,i exp(

 Ei 1 1 (  )) R T Tref

(7)

2.2.2. LHHW expressions. The use of LHHW expressions was also considered, since they are based on a realistic description of a heterogeneous catalytic reaction mechanism, given the assumption of competitive adsorption of species on the same active sites on the catalyst. The rate equations become

6


Page 7 of 34

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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60



ri 

Ki Creact ,i ni

(8)

1   K jC j j

with j being the adsorbed species. The reaction rate constant, Ki, in the nominator is of the same form as in the power law(Eq.5), whereas the adsorption constant K is in the form

K j  K ref , j exp(

H j 1 1 (  )) R T Tref

(9)

with j being the adsorbed species. 2.2.3. Reactor mass balance. The laboratory-scale fixed-bed reactor was modeled for the

kinetic study under the following assumptions: 1) ideal plug flow to describe neither axial nor radial gradients, 2) ignoring external mass and heat transfer limitations, 3) ignoring internal mass and heat transfer limitations (by using around 100  m catalysts). Table 1. gives a summary of the results obtained after assessing the above assumptions with the information from reaction 2(r2) under the most severe conditions (a temperature of 334.5oC and a GHSV of 1780 hr-1), which lead to maximum 2,3-BDO conversion and 1,3-BD selectivity. Our results indicate that experiments were indeed performed under plug flow conditions and in the absence of external and internal mass and heat transfer resistances. Table 1. Summary of the results after assessment of our assumptions

1a 1b 2a

Criterion Negligible axial dispersion[24] Flat velocity profile[24] External mass transfer limitation[25]

Mathematical form L/dp

Reference 50

55.3 5.47E-05

kcC2,3 BDO

7



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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 34


2b

H rxn ,2 (r2 ) b R p E2

External heat transfer limitation [25]

0.15>

7.4E-06

2

hT R

Therefore, the steady state one-dimensional homogeneous model takes the form

0

d (us C j ) dz

rxn

 b  ji ri

(10)

i 1

The pressure drop in the reactor was calculated by the Eurgun equation[26]. The reactor model was integrated by employing DASOLV, a differential algebraic solver included in the General Process Modeling System (gPROMS) library. 2.3. Estimation of parameters

The reactor model in the previous section was used to fit the experimental data under various conditions. gEST, the parameter estimation tool in the gPROMS modeling package, was used to estimate the kinetic parameters (i.e. transformed Arrhenius kinetic constants, activation energies, reaction orders with respect to reactants, heats of adsorption, and transformed adsorption preexponential factors). The parameter optimization solver named MXLKHD uses the maximumlikelihood approach(Eq.11) to estimate all kinetic parameters simultaneously. This solver applies a sophisticated sequential quadratic programming(SQP) method to find the global optimum[27]. NE NVi NM ij (  yijk  yijk ) 2 N 1 2   ln 2  min{  [ln  ijk  ]}  ijk2 2 2  i 1 j 1 k 1

8


(11)

Page 9 of 34

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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60



The variance in the mole fraction of the components, with the exception of N2, was 0.001, whereas the variance in the mole fraction of N2 was 1000 to remove the effects of N2, used as a carrier gas, on the objective function in Eq. 11. 3. Experimental Procedures 3.1. Catalyst preparation

The a-CP catalyst(Ca/P =1.3) was synthesized. 8.62 g of H4P2O7 was dissolved in deionized water to prepare 200 mL of a diluted pyrophosphoric acid aqueous solution and it was added with 37.60 g of 28 wt% NH4OH aqueous solution by stirring for 30 minutes (Solution A). Separately, 23.74 g of Ca(OAc)2 was dissolved in deioninzed water to obtain 150 mL of a calcium precursor aqueous solution (Solution B). To Solution A, Solution B was slowly added at a rate of 3.5 mL/min at room temperature, giving a calcium phosphate slurry solution (Solution C). Solution C as the mixed slurry solution was maintained at a pH of about 10.0, and was sufficiently stirred for 24 hr and then purified by filtration with 2 L of deionized water to obtain a sold product in cake form. Then, the calcium phosphate cake was dried in the form of cakes at 80 o

C for 12 hours and thermally treated at 500 oC for 6 hours[19]. The real Ca/P ratio was 1.67 but

the Ca/P ratio was changed. Ca would be lost in the process of making the solid product in cake form before the calcination. The Ca/P ratio of the catalyst was checked by the result of XRF analysis and the catalyst was analyzed to be amorphous calcium phosphate by XRD in figure 2. The surface area and pore volume of the catalyst were 37.4 m2/g and 0.25 cc/g.

9




Intensity (a.u.)

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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 34

10

20

30

40 2 (degree)

50

60

70

Figure 2. XRD of the amorphous calcium phosphate(Ca/P =1.3)

3.2. Kinetic experiments 3.2.1. Experimental setup. The catalytic reactions were carried out in a conventional

continuous flow fixed-bed reactor(3/8'' quartz) that was operated isothermally under atmospheric pressure. 0.3g of catalyst subjected to a sieve to produce an average particle size of 100  m was loaded into the center of the reactor. 2,3-BDO(98%, Sigma-Aldrich) and the mixtures of 2,3BDO and 3B2OL(>97% Sigma-Aldrich ) were used to check the effect of the intermediate product (3B2OL) and were introduced via a syringe pump with an N2 flow(carrier gas) of 10cc/min(standard ambient temperature and pressure, SATP) that was controlled with a massflow controller. Before each experiment, fresh catalyst was pretreated in N2 flow(10cc/min) at 10


Page 11 of 34

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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60



400oC for 1hr. Catalytic activity was evaluated by averaging the conversion and selectivity data in the initial 10h. In order to estimate kinetic parameters, 18 experiments were performed. Table 2. shows the experimental conditions. Table 2. Experimental conditions

Run

Temperature [oC]

23-BDO [g/hr]

3B2OL [g/hr]

N2 [cc/min]

1

304

0.73

0

10

2

314

0.73

0

10

3

324.5

0.73

0

10

4

334.5

0.73

0

10

5

304

1.46

0

10

6

314

1.46

0

10

7

324.5

1.46

0

10

8

304

2.92

0

10

9

314

2.92

0

10

10

324.5

2.92

0

10

11

304

0.44

0.19

10

12

314

0.44

0.19

10

13

324.5

0.44

0.19

10

14

304

0.89

0.38

10

15

314

0.89

0.38

10

16

324.5

0.89

0.38

10

17

304

1.78

0.76

10

18

324.5

1.78

0.76

10

11



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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 34


3.2.2. Analysis of reaction products. Product compositions were analyzed in an online gas

chromatograph(GC, Agilent 7890A) with a DB-1 column(non-polar phase, 60m x 0250mm x 1  m ) and a FID(Flame Ionization Detector) for the analysis of hydrocarbon content. The temperature of the bottom of the reactor at the inlet of GC was maintained at 230oC to avoid condensation of the products. The composition of water was calculated by reaction stoichiometry of Eqs.1-4 based on the compositions of 3B2OL, 1,3-BD, MEK, and 2MPL in the products. The compositions of major components were normalized to remove the effects of the impurities. The conversion of 2,3-BDO and selectivity for each product was computed as follows:

X BDO 

Sn 

FBDO ,in  FBDO ,out FBDO ,in

Fn ,out Ftotal  FBDO ,out  FN2

100

(12)

 100

(13)

where F is the mass flow rate and n is a component of the product. 4. Results and discussion 4.1. Experiments

This section deals with the reactor performance obtained from run 1 to run 10 in Table 2., given changing temperature and GHSV. Figure 3. shows 2,3-BDO conversion as a function of temperature at different GHSVs based on the inlet conditions of the reactor. As the temperature was increased from 304 to 334.5oC, the conversion increased most sharply at the low GHSV of 1780 hr-1 because that was associated with the longest residence time in the reactor. On the other hand, the changing rate of the conversion lessens with increasing GHSV. At a temperature of

12


Page 13 of 34


334.5oC and a GHSV of 1780 hr-1, the conversion rate was 100%. Since it is difficult to know how the reactants produce the intermediate or final products under specific operation conditions with 100% conversion data, the operation temperature used for other experiments was lower than 334.5oC.

120

100

2,3-BDO Conversion %

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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


80

60

40

20

0 304

314

324.5

334.5

Temperature o C Figure 3. 2,3-BDO conversion as a function of temperature at different GHSV hr-1: 1780 (●) ,

1928(◆), 2222(■)

13



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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 34


Figure 4. displays the selectivity of the products as a function of temperature from 304 to 334.5oC. The selectivity of 1,3-BD and 3B2OL were highly dependent on temperature in the temperature range of 324.5 to 334.5oC. On the other hand, the selectivity of MEK and 2MPN shows a flat pattern that indicates less dependence on temperature. Even as the temperature was increased, selectivity decreased slightly because of the decreasing concentration of the reactant (2,3-BDO) with a sharp increase in the selectivity for1,3-BD and 3B2OL. The plots at higher GHSVs(1928 and 2222 hr-1) are not shown since the above mentioned results were very similar to those using a GHSV of 1780 hr-1 and the rate of change of product selectivity is lower than that at GHSVs of 1928 and 2222 hr-1.

14


Page 15 of 34


40 35 30

Mass Selectivity %

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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


25 20 15 10 5 0 304

314

324.5

334.5

Temperature oC Figure 4. Mass selectivity of the products as a function of temperature at GHSV 1780 hr-1: 1,3-

BD(◆) , MEK(✕) , 3B2OL(▲) , 2MPN(■), H2O(*)

Figure 5. represents 2,3-BDO conversion as a function of GHSV at different temperatures. As GHSV increases, the conversion decreases because of low residence time in the reactor. Interestingly, conversion decreased almost linearly with increasing GHSV. This means that the orders of the main reactions would be near 0, which indicates that the performance of a reactor for 2,3-BDO dehydration would mainly be determined by the temperature of the reactor, not by the concentrations of the reactants[28].

15




120

100

2,3-BDO Conversion %

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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 34

80

60

40

20

0 1780

1928

2222

GHSV hr-1 Figure 5. 2,3-BDO conversion as a function of GHSV at different temperatures: 304(◆),

314(●), and 324.5°C (■)

The GHSV dependency of the selectivity of the products is shown in Figure 6. The selectivity of 1,3-BD decreases sharply with increasing GHSV because the low residence time in the reactor leads to an increase in the selectivity of 3B2OL, which is an intermediate product in the formation of 1,3-BD. On the other hand, the selectivity of MEK and 2MPL showed a slight increase with the low 2,3-BDO conversion with increasing GHSV. Similar tests using silica-

16


Page 17 of 34


supported sodium phosphates at 400oC were performed and also major products are same[1]. The research group concluded that mass transport limitations are more apparent for 2,3-BDO -> 3B2OL -> 1,3-BD route than for 2,3-BDO -> MEK route. These results are well matched with the above test results.

40.00 35.00 30.00

Mass Selectivity %

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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


25.00 20.00 15.00 10.00 5.00 0.00 1780

1928

2222

GHSV hr-1 Figure 6. Mass selectivity of the products as a function of GHSV at a temperature of 324.5oC:

1,3-BD(◆) , MEK(✕) , 3B2OL(▲) , 2MPN(■), H2O(*)

17



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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 34


4.2. Assessment of kinetic parameters

The kinetic parameters were fitted to the experimental values with 18 steady-state experimental data points, including the experimental data explained in section 4.1, in the reactor under the range of conditions shown in Table 2. The estimated kinetic parameters of the power law model (detailed in section 2.2.1) are listed in Table 3. All kinetic parameters can be considered significant since each standard deviation is less than half of that of the parameter. The narrow confidence interval (95%) confirms the precise determination of the parameters and the reliability of the experimental data. Also, the correlation matrix showed no strong correlations among the parameters[29].

Table 3. Estimated kinetic parameters of the power law model Model Parameter

Value

Confidence Interval (95%)

Standard deviation

E1

2.33E+05

7.89E+03

3.98E+03

E2

2.82E+05

1.65E+04

8.32E+03

E3

1.93E+05

7.81E+03

3.94E+03

E4

1.66E+05

3.82E+04

1.93E+04

kTref,1

7.45E-04

1.83E-05

9.23E-06

kTref,2

4.41E-04

1.71E-05

8.61E-06

kTref,3

6.44E-04

1.47E-05

7.44E-06

kTref,4

1.27E-04

1.43E-05

7.21E-06

n1, n3, n4

0.0187

0.0106

0.00537

n2

0.146

0.0249

0.0126

18


Page 19 of 34

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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60



All reaction orders (between 0 and 0.15) and activation energies (between 166 and 282 kJ/mol) are within the normal range of values for heterogeneously catalyzed reactions. The reaction orders (0.0187) for the reactions r1, r3, and r4 are very close to 0, and the other reaction order (0.146) for the reaction r2 is also close to 0. The corresponding activation energies(E2, E1, E3, and E4) for the reactions (r2,r1, r3, and r4) were found to be 282, 233, 193, 166 kJ/mol. The activation energies (E1 and E2) of the reaction leading to 1,3-BD are higher than that of E3, for MEK, and E4, for 2MPN. This indicates that the selectivity for 1,3-BD increases sharply with increasing temperature, in contrast to MEK and 2MPN. The estimated values of the reaction order and activation energy explain quite well the tendencies of the experimental results obtained by changing temperature and GHSV as described in section 4.1. Figure 7. shows the parity plots of the calculated and experimental outlet molar fractions of all components. There is no tendency toward negative or positive deviations between the calculated and experimental data. The deviations are within 10% in the case of 2,3-BDO and MEK, apart from a few outliers. The deviations of the other components are larger than those of 2,3-BDO and MEK. However, for the most part, the deviations are within 20%.

19




0.30

(a)

0.25

0.20

X2,3-BDO, Mod

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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 34

+10% 0.15

-10% 0.10

0.05

0.00 0.00

0.05

0.10

0.15

0.20

X2,3-BDO, Exp

20


0.25

0.30

Page 21 of 34


0.05

(b)

0.04

0.03

+20% X 1,3-BD, Mod

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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


-20%

0.02

0.01

0.00 0.00

0.01

0.02

0.03

X 1,3-BD, Exp

21


0.04

0.05



0.04

(c)

0.04

0.03

0.03

+10% XMEK, Mod

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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 34

0.02

-10% 0.02

0.01

0.01

0.00 0.00

0.01

0.01

0.02

0.02

0.03

XMEK, Exp

22


0.03

0.04

0.04

Page 23 of 34


0.12

(d)

0.10

0.08

X3B2OL, Mod

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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


+20% 0.06

-20%

0.04

0.02

0.00 0.00

0.02

0.04

0.06

0.08

X3B2OL, Exp

23


0.10

0.12



0.008

(e)

0.007

0.006

0.005

X2MPA, Mod

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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 34

+20% 0.004

0.003

-20%

0.002

0.001

0.000 0.000

0.001

0.002

0.003

0.004

0.005

X 2MPA, Exp

24


0.006

0.007

0.008

Page 25 of 34


(f) 0.12

0.10

0.08

XWater, Mod

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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


+20% 0.06

-20%

0.04

0.02

0.00 0.00

0.02

0.04

0.06

0.08

0.10

0.12

XWater, Exp

Figure 7. Parity plots of the calculated and experimental outlet molar fractions of all components: 2,3-BDO(a), 1,3-BD(b) , MEK(c) , 3B2OL(d) , 2MPN(e), H2O(f)

2,3-BDO, 1,3-BD, and water were considered to be adsorption components in the LHHW model. The value of the objective function is smaller than the power law model and the fit for the experimental data with the LHHW model is as good as that of the power law model. However, all estimated parameters that are listed in Table 4. showed very high confidence intervals (95%) and the correlation matrix showed very strong correlations among the parameters. Given these results, we conclude that the effects of temperature on surface reaction terms and adsorption 25



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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 34


terms cannot be distinguished with the current data set. Therefore, we decided to use the power law model to express the kinetic equations. Table 4. Estimated kinetic parameters of the LHHW model Model Parameter

Value

Confidence Interval (95%)

Standard deviation

E1

8.19E+04

6.40E+06

3.20E+06

E2

2.18E+05

6.40E+06

3.20E+06

E3

1.15E+04

6.40E+06

3.20E+06

E4

1.00E+04

6.40E+06

3.20E+06

 H1,3-BD

6.66E+05

6.50E+06

3.30E+06

 H2,3-BDO

1.02E+02

6.50E+06

3.30E+06

 Hwater

5.61E+04

1.90E+07

9.80E+06

kTref,1

2.71E-02

7.20E-01

3.70E-01

kTref,2

1.27E-02

3.40E-01

1.70E-01

kTref,3

2.70E-02

7.20E-01

3.60E-01

kTref,4

6.32E-03

1.70E-01

8.50E-02

KTref, 1,3-BD

2.54E+02

6.90E+03

3.50E+03

KTref, 2,3-BDO

8.00E+00

2.20E+02

1.10E+02

KTref, Water

5.92E-01

3.70E+01

1.90E+01

n1, n2, n3, n4

0.212

0.0280

0.0140

5. Conclusion

Kinetic models based on the power law and LHHW have been built to explain the dehydration reaction of 2,3-BDO to 1,3-BD and MEK using a-CP as a catalyst. For the reaction network including four major reactions, the corresponding reaction rate equations were derived and

26


Page 27 of 34

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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60



kinetic parameters were fitted to the laboratory-scale experimental data obtained from the fixedbed reactor operated under isothermal conditions. The experimental data revealed effects of temperature and GHSV on 2,3-BDO conversion and selectivity of products. The conversion increased most sharply at the low GHSV of 1780 hr-1 because it was associated with the longest residence time in the reactor, and was 100% at a temperature of 334.5oC and a GHSV of 1780 hr-1. Conversion decreased almost linearly with increasing GHSV. It is known that the performance of a reactor for 2,3-BDO dehydration is mainly determined by the temperature of the reactor, not by the concentrations of reactants. The selectivity of 1,3-BD and 3B2OL highly depended on temperature, but the selectivity of MEK and 2MPN showed a flat pattern. From these results, we conclude that the reaction route for the production of 1,3-BD highly depends on temperature, and other routes are less dependent on temperature. The kinetic parameters based on the power law model show narrow confidence intervals and no strong correlations between the parameters. Also, the estimated values of the reaction order and activation energy were very well matched with the tendency of the experimental results obtained by changing temperature and GHSV. The resulting kinetic information can be used to better understand the physicochemical aspects of the reaction. On the other hand, the model based on LHHW showed low confidence levels and the correlation matrix of the model showed very strong correlations between the parameters, even though the value of the objective function is smaller than the power law model and the fit is as good as that of the power law model. The effects of temperature between surface reaction terms and adsorption terms could not be distinguished with the current data set.

27



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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Nomenclature

C

mole concentration, mol/m3

dp

diameter of a catalyst, m

dt

diameter of a reactor, m

E

activation energy, J/mol

h

heat transfer coefficient between catalyst and gas, W/m2 K

K

adsorption-desorption equilibrium constant , m3/mol

Kref

transformed adsorption pre-exponential factor, m3/mol

k

reaction rate constant, mol(1-n) m3(n-1) s-1

kc

mass transfer coefficient between catalyst and gas, m/s

kTref

transformed Arrhenius kinetic constant, variable units

L

reactor length, m

N

total number of measurements taken during all experiments

NE

number of experiments performed

NVi

number of variables measured in the ith experiment

NMij

number of measurements of the jth variable in the ith experiment

n

reaction order

28


Page 28 of 34

Page 29 of 34

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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60



R

ideal gas law constant, J/mol K

Rp

catalyst radius, m

S

mass selectivity, %

r

reaction rate, mol/kg-cat s

T

temperature, K

Tref

reference temperature, K

u

superficial fluid velocity, m/s

X

conversion, %

z

axial reactor coordinate, m

Eijk

kth model-predicted value of variable j in experiment i

Greek Letters Ёijk

kth measured value of variable j in experiment i

H

heat of adsorption, J/mol

 Hrxn heat of reaction, J/mol

b

bulk density of catalyst bed, kg/m3

 ji

stoichiometric coefficient of species j in reaction i



objective function

29



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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 30 of 34




set of parameters to be estimated

 ijk2

variance of the kth measurement of variable j in experiment i

Subscripts i

reaction i

j

species j

react

reactant

rxn

reaction

References

[1] Kim, W.; Shin, W.; Lee, K. J.; Song, H.; Kim, H. S. 2.3-Butanediol dehydration catalyzed by silica-supported sodium phosphates. Appl. Catal., A. 2016, 511, 156. [2] Zheng, Q.; Wales, M. D.; Heidlage, M. G.; Rezac, M.; Wang, H.; Bossmann, S. H.; Hohn, K. L. Conversion of 2,3-butanediol to butenes over bifunctional catalysts in a single reactor. J. Catal. 2015, 330, 222.

[3] Köpke, M.; Mihalcea, C.; Liew, F.; Tizard, J. H.; Ali, M. S.; Conolly J. J.; Al-Sinawi, B.; Simpson, S. D. 2,3-Butanediol production by acetogenic bacteria, an alternative route to chemical synthesis, using industrial waste gas. Appl. Environ. Microbiol. 2011, 77, 5467. [4] Daniell, J.; Köpke, M.; Simpson , S. D. Commercial Biomass Syngas Fermentation. Energies. 2012, 5(12), 5372.

[5] Makshina, E. V.; Dusselier M.; Janssens W.; Degrève J.; Jacobs P. A.; Sels B. F. Review of old chemistry and new catalytic advances in the on-purpose synthesis of butadiene. Chem. Soc.

30


Page 31 of 34

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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60



Rev. 2014, 43, 7917. [6] Duan, H.; Yamada, Y.; Sato, S. Future prospect of the production of 1,3-butadiene from butanediols. Chem. Lett. 2016, 45, 1036. [7] Bourns, A. N. THE CATALYTIC ACTION OF ALUMINIUM SILICATES: I. THE DEHYDRATION OF BUTANEDIOL-2, 3 AND BUTANONE-2 OVER ACTIVATED MORDEN BENTONITE. Can. J. Res. 1947, 25b, 80. [8] Winfield, M. E. The catalytic dehydration of 2,3-butanediol to butadiene. J. C. S. I. R. (Aust.) 1945, 18, 412.

[9] Winfield, M. E. The catalytic dehydration of 2, 3-butanediol to butadiene. II. Adsorption equilibria. Aust. J. Chem. 1950, 3(2), 290. [10] Kannan, S. V.; Pillai, C. N. DEHYDRATION OF MESO-AND DL-HYDROBENZOINS AND 2, 3-BUTANEDIOLS OVER ALUMINA. Indian J. Chem. 1969, 7, 1164. [11] Duan, H.; Sun, D.; Yamada, Y.; Sato, S. Dehydration of 2, 3-butanediol into 3-buten-2-ol catalyzed by ZrO2. Catal. Commun. 2014, 48, 1. [12] Duan, H.; Yamada, Y.; Sato, S. Efficient production of 1, 3-butadiene in the catalytic dehydration of 2, 3-butanediol.Appl. Catal., A. 2015, 491, 163. [13] Molnár, Á.; Bucsi, I.; Bartók. M. Pinacol Rearrangement on Zeolites. Stud. Surf. Sci. Catal. 1988, 41, 203.

[14] Bucsi, I.; Molnár, Á.; Bartók, M.; Olah, G.A. Transformation of 1, 2-diols over perfluorinated resinsulfonic acids (Nafion-H). Tetrahedron. 1994, 50, 8195. [15] Lee, J.; Grutzner, J. B.; Walters, W. E.; Delgass, W. N. The conversion of 2, 3-butanediol to methyl ethyl ketone over zeolites. Stud. Surf. Sci. Catal. 2000, 130C, 2603. [16] Zhang, W.; Yu, D.; Ji, X.; Huang, H. Efficient dehydration of bio-based 2, 3-butanediol to

31



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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


butanone over boric acid modified HZSM-5 zeolites. Green Chem. 2012, 14, 3441. [17] Török, B.; Bucsi, I.; Beregszászi, T.; Kapocsi, I. Transformation of diols in the presence of heteropoly acids under homogeneous and heterogeneous conditions. J. Mol. Catal. A: Chem. 1996, 107, 305.

[18] Han, Y.; Kim, H.; Hong, J.; Jho, H. Fabrication Method of 1,3-Butadiene and 2-Butanone from 2,3-Butanediol. K.R. Patent 10-1287167 B1, 2013. [19] Kim, H.; Lee, H.; Park, D. Amorphous calcium phosphate catalyst used for 1,3-butadiene and methyl ethyl ketone from 2,3-butanediol and preparation method thereof. K.R. Patent 102014-0167303, 2014. [20] Tsukamoto D.; Sakami S.; Ito M.; Yamada K.; Yonehara T. Production of Bio-based 1, 3Butadiene by Highly Selective Dehydration of 2, 3-Butanediol over SiO 2-supported Cesium Dihydrogen Phosphate Catalyst. Chem. Lett. 2016, 45, 831. [21] Nikitina M. A.; Ivanova I. I. Conversion of 2,3-Butanediol over Phosphate Catalysts. ChemCatChem. 2016, 8, 1346. [22] Kim T. Y.; Baek J.; Song C. K.; Yun Y. S.; Park D. S.; Kim W.; Han J. W.; Yi J. Gas-phase dehydration of vicinal diols to epoxides: Dehydrative epoxidation over a Cs/SiO 2 catalyst. J. Catal. 2015, 323, 85. [23] Nakamura, K.; Osamura, Y. Theoretical study of the reaction mechanism and migratory aptitude of the pinacol rearrangement. J. Am. Chem. Soc. 1993, 115, 9112 [24] Rase, H. F. Fixed-bed reactor design and diagnostics. Stoneham, MA Butterworths publishers, 1990. [25] Mears D. E. Tests for Transprot Limitations in Experimental Catalytic Reactors. Ind. Eng. Chem. Process Des. Dev. 1971, 10(4), 541.

32


Page 32 of 34

Page 33 of 34

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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60



[26] Ergun S. Fluid flow trough packed columns. Chem. Eng. Prog. 1952, 48(2), 89. [27] Process Systems Enterprise, gPROMS Advanced User Guide, 2004. [28] Levenspiel O. Chemcial Reaction Engineering. John Wiley & Sons, 1999. [29] Redlingshöfer, H. Reaktionstechnische Untersuchungenzur heterogen katalysierten Gashpasenoxidation von Propen. Ph.D. Thesis. University of Erlangen-Nuremberg, 2002.

33



0.30

(a)

0.25

0.20

X2,3-BDO, Mod

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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47

Page 34 of 34

+10% 0.15

-10% 0.10

An amorphous calcium phosphate h h t

0.05

0.00 0.00

0.05

0.10

0.15

X2,3-BDO, Exp


0.20

0.25

0.30

Kinetic Model Development for Dehydration of 2,3 ... - ACS Publications

Recommend Documents