Subscriber access provided by University of Winnipeg Library

Kinetics, Catalysis, and Reaction Engineering

Coupled fluidized bed reactor for pyridine synthesis
Shuaishuai Zhou, Shanhe Liu, Yongfei Wei, Xihong Li, Chunxi Lu, mengxi liu, and Xiaotao Bi
Ind. Eng. Chem. Res., Just Accepted Manuscript • DOI: 10.1021/acs.iecr.8b01510 • Publication Date (Web): 15 May 2018
Downloaded from http://pubs.acs.org on May 18, 2018

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 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 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.

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 24 Industrial & Engineering Chemistry Research

1
2
3
4 Coupled fluidized bed reactor for pyridine synthesis
5
6 Shuaishuai Zhou1, Shanhe Liu2, Yongfei Wei2, Xihong Li2, Chunxi Lu1*, Mengxi Liu1*, Xiaotao Bi3
7
1. State Key Laboratory of Heavy Oil Processing, China University of Petroleum, Beijing 102249, PR China.
8 2. Costar Biochemical, Anhui 243100, PR China.
9 3. Department of Chemical and Biological Engineering, University of British Columbia, Vancouver BCV6T1Z3, Canada
10 *Corresponding author. Tel./fax: +86 10 89733803. E-mail address: [email protected], [email protected]

11 Abstract：To eliminate the feeding device coking deposit in commercial pyridine synthesis reactor, a
12
13 coupled fluidized bed reactor is proposed. The coupled reactor is composed of a feeding zone (FZ), a riser
14 reaction zone (RRZ) and a fluidized bed reaction zone (FRZ). During 15 days of continuous operation,
15
16 there is no coke deposit in the feeding device. Experimental results show that the product yield reaches as
17 high as 75 %. The selectivity is around 2.5 and is higher than that in commercial reactors, which is around
18
19 2.2. Core-annulus model and Dispersion model are proposed to model different zones of this coupled
20 reactor, and the predicted average deviation from experimental data is 12 %. The prediction results show
21
22 that the main reactions take place at FZ and FRZ. RRZ contributes the least conversion, because of the
23 limit by the mass transfer between the core and annulus.
24
25 Key words: Coupled, Fluidized bed reactor, Pyridine synthesis, Experiment, Modeling
26
27
28
29 1 INTRODUCTION
30 Pyridine and 3-picoline have been widely used for synthesis of pharmaceuticals and agrochemicals,
31
32 benefiting from high chemical reactivity as well as biological activity 1. Traditionally, pyridine was derived
33 from coal tar and suffered from high sulfur content. In 1920s, high market demand of pyridine encouraged
34
35 the development of new technologies for pyridine synthesis. Nowadays, most pyridine is produced from
36 the Chichibabin condensation process 1.
37
38 Research on pyridine synthesis has been rare and mainly focused on: catalysts, reaction mechanisms,
39 reaction kinetics and reactors. The development of catalyst can be divided into two stages 1, which are
40
41 silica-alumina catalysts and zeolite catalysts. Until now the most widely used zeolite catalysts are HZSM-5
42 catalysts, benefiting from the design of an appropriate shape selective system. Because of multiple side
43
44 reactions occurring in pyridine synthesis, the main reaction mechanism of the process has not been well
45 understood. It has only been confirmed that during the reaction process, imide intermediates is first
46
2-6.
47 produced, which is the key component to form pyridine There have been few reports on the reaction
48 kinetics of pyridine synthesis reaction. Reddy 7 carried out experiments in a 20 mm ID Pyrex reactor using
49
50 HZSM-5 catalysts, and their results suggested that the reaction rate is only determined by reactor operating
51 conditions and acetaldehydes concentration.
52
53 Fluidized bed reactor is favored for pyridine synthesis. Compared with fixed bed reactor, fluidized bed
54 reactor has advantages of high heat and mass transfer rates, good gas and catalyst contact and capable of
55
56 continuous catalysts regeneration. Besides, previous studies reported higher product yield of fluidized bed
57 reactors than fixed bed reactors for pyridine synthesis 8-11.
Feitler et al. 11
produced pyridine in both fixed
58
59 bed and fluidized bed reactors. Over a large range of reaction temperatures from 723 K to 773 K, the
60

*Corresponding author. Tel./fax: +861089733803. E-mail address: [email protected].

ACS Paragon Plus 1 Environment
 Industrial & Engineering Chemistry Research Page 2 of 24

1
2
3 fluidized bed reactor offered higher product yield than the fixed bed reactor. Due to those advantages of
4
5 fluidized bed reactors, some commercial scale pyridine synthesis units have been incorporated with
6 fluidized catalyst beds.
7
8 The gas-solid flow patterns in fluidized bed reactors ranged from bubbling fluidization to fast fluidization.
9 Usually, the flow pattern in commercial reactors for pyridine synthesis is bubbling fluidization. A problem
10
11 that results in frequent shut down of commercial reactors is the coking deposit inside the feeding device.
12 Coke is formed due to the high temperature around the feeding device 12,13
, which is installed inside the
13
12
14 reactor. To solve the problem, Zhou proposed to use a fast fluidized bed reactor to replace the bubbling
15 fluidized bed reactor for pyridine synthesis. By changing the flow pattern from bubbling to fast fluidization,
16
17 the diameter of the reactor decreases, and the feeding device could be installed outside the reactor to avoid
18 high temperature environment. Zhou 12 synthesized pyridine in a fast fluidized bed reactor with a 14 mm
19
20 ID and 3500 mm tall reactor, and the highest product yield is 65 %. Compared to the product yield of
21 commercial reactors, which is 73 %, the product yield decrease probably results from the short gas-solid
22
23 contact time or slow reaction rate.
24 To improve the performance of fast fluidized bed reactor, a coupled reactor is proposed in this paper. The
25
26 coupled reactor is composed of a feeding zone (FZ) with novel feeding scheme, a riser reaction zone (RRZ)
27 to increase the selectivity of pyridine over 3-picoline and a fluidized bed reaction zone (FRZ) for enough
28
29 gas-solid contact time. The details of this coupled reactor can be found in the experimental apparatus
30
section.
31
32 For coupled reactor, previous studies show that the flow patterns range from bubbling to fast fluidization in
33 14,15.
different sections along the bed height For feeding zone (FZ), the gas velocity is lower than that in
34
35 riser reaction zone (RRZ) above it, and the solid concentration is higher than that in RRZ. Previous studies
36
showed that non-uniform flow distribution in the FZ, which is caused by unsymmetrical catalyst feed, can
37
38 be eased by setting an enlarged section at the bottom of the riser. Liu 16
investigated the FZ of a riser by
39
experiment, and they found that the non-uniform flow problem is reduced and gas-solid contact is
40
41 enhanced. The riser reactor with FZ gives better reaction performance has been demonstrated in Fluid
42
Catalytic Cracking (FCC) process 17-20. In this paper, the arrangement of FZ mainly based on the following
43
44 considerations, which are to reduce the non-uniform flow phenomenon in the feed section and to eliminate
45
the coking problem in the feeding device.
46
47 In the RRZ, gas back mixing decreases. Besides, the particle residence time also decreases, which brings
48 21-23.
high gas-catalysts contact efficiency The flow pattern in RRZ is fast fluidization where the solid
49
50 concentration axial distribution exhibits from exponential shape to C shape depending on the reactor exit
51
52
configuration 24-26. Usually, a dense bed is formed at the bottom, with dense bed height determined by the
53 properties of the gas-solid system and operating conditions 27-31.
Based on experiment, Li 32
proposed a
54
55
semi-empirical model to predict the solid concentration axial distribution in fast fluidized bed reactors 33.
56 Along the radial direction in the fast fluidized bed reactor, there exists a core-annulus structure 34,35. In the
57
58 core region, the solid concentration is low, and the gas concentration is high. While, in the annulus region,
59 the structure is opposite. Usually, the mass transfer between the core and annulus, which is similar as the
60

2
ACS Paragon Plus Environment
Page 3 of 24 Industrial & Engineering Chemistry Research

1
2
3 mass transfer between dense phase and bubble phase in bubbling beds, significantly affects the reactor
4
36-38
5 performance. To quantify the mass transfer between the core and annulus, some useful correlations
6 have been developed based on experimental results.
7
8 In the upper FRZ, the gas velocity decreases to around 0.5 m/s. For the gas-solid system investigated in
9 this paper, the flow pattern is turbulent fluidization 39. By adjusting the height of the dense bed in FRZ, this
10
11 section serves either as a disengagement zone or as a reactor. Details on this zone will be presented in the
12 experiment apparatus section.
13
14 Numerous works have been reported on the turbulent fluidization and fast fluidization individually. Few
15 reports are found on the coupling of these flow patterns in one reactor. Wang 15 investigated the solid phase
16
17 flow hydrodynamics in a coupled fluidized bed reactor. The experiment apparatus employed in Wang’s
18 work is made of plexiglass, and the experimental results are obtained under cold operating conditions.
19
20 They found that the axial solid concentration distribution in the middle RRZ section is typically of C shape,
21 and the influence of the dense bed in FRZ on RRZ is mainly reflected by the solid concentration in RRZ.
22
23 With the increase of the dense bed height in FRZ, the solid concentration in RRZ is increased, especially in
24 the region near the riser exit. Based on experimental results, Wang proposed some empirical correlations to
25
26 predict the solid concentration axial distribution in the RRZ. This coupled reactor has been successfully
27 applied in the FCC gasoline olefins reduction process. While, no report is found on the application of the
28
29 coupled reactor in the pyridine synthesis process. This paper aims at investigating the performance of
30
coupled reactors for pyridine synthesis.
31
32
33 2 EXPERIMENT
34
35 Figure 1 shows the schematic diagram of the experimental apparatus (Costar Biochemical, Ltd). As
36
37 illustrated in Figure 1, the coupled reactor composes of FZ (feeding zone), RRZ (riser reaction zone) and
38 FRZ (fluidized bed reaction zone). The height of the FZ is 1.0 m, and the ID is 140 mm. For RRZ, the total
39
40 height is 13.5 m, and the ID is 84 mm. The FRZ is 5.74 m in height and has an ID of 311 mm. This
41 coupled reactor is made of stainless steel.
42
43 Formaldehydes and acetaldehydes feedstock flow were preheated to 373 K, with the flow rate measured
44
45 and adjusted by a regulating system, composed of flow meter, a valve and a control device (Costar
46 Biochemical, Ltd). They were mixed with ammonia, which was regulated by another regulating system.
47
48 The mixture of feedstock was introduced into the FZ by a nozzle, mixing with hot regenerated catalysts
49
from an adjacent commercial regenerator and vaporizing. The vapor and catalysts flowed upwards in the
50
51 coupled reactor and were separated by a cyclone at the top. The gas product was introduced to the
52
commercial reactor by products pipeline. To analyze the products composition, a sampling pipeline is
53
54 installed on the products pipeline, with a side stream of the products being withdrawn. The withdrawn
55
products were condensed and then collected into a glass bottle, with the liquid products composition being
56
57 determined by a gas phase chromatograph (GC7890II, Shanghai Tianmei Keji Instrument Co., Ltd., China).
58
The spent catalysts were stripped by water steam and then returned to the industrial reactor. After
59
60 regeneration, the spent catalyst was conveyed to coupled fluidized bed reactor through catalysts circulation

3
ACS Paragon Plus Environment
 Industrial & Engineering Chemistry Research Page 4 of 24

1
2
3 pipe.
4
5 The initial acetaldehydes concentration cA0 was calculated by eq (1) to eq (4).
6 nA0
7 cA0 = (1)
V
8
9 where nA0 is the mole flow rate of acetaldehydes, which was obtained by multiplying the mass flow rate of
10
11
aldehydes flow and its corresponding composition. V is the volumetric flow rate of both aldehydes flow
12 and ammonia at reaction conditions, which was derived based on ideal gas law.
13
14 PV = nRT (2)
15 where n is the mole flow rates of both aldehydes and ammonia and can be calculated by the following
16
17 equation.
18
19
n = nA0 / xA0 + nm (3)
20 where xA0 is the mole fraction of acetaldehydes in aldehydes flow, nm is the mole flow rate of ammonia
21
22 and was calculated by
23
24
nm = (1/ af + 1)nA0 / am (4)
25
where ηaf is acetaldehydes to formaldehydes ratio by mole (ATFR), ηam is ammonia to aldehydes molar
26
27 ratio (ATAR), which were obtained by the composition of both aldehydes and ammonia before experiment.
28
Reddy 7 suggested that the volume change in a small reactor can be neglected, therefore the gas velocity
29
30 was assumed to be constant in the coupled fluidized bed reactor. The superficial gas velocity in the
31
entrance of the coupled fluidized bed reactor is calculated by eq (5).
32
33 4V
34 ug = (5)
35
 D2
36 where D is the internal diameter of the coupled fluidized bed reactor.
37
38 The reaction temperature and pressure were measured by thermocouples and pressure gauges, all
39 thermocouples and pressure gauges were linked to a computer via a data acquisition system. The positions
40
41 of the thermocouples and pressure gauges are depicted in Figure 2. As shown in Figure 2, there were two
42 temperature taps, with one located in the FZ (2-1, T1) and the other one in the FRZ (2-2, T2). By neglecting
43
44 the wall friction and acceleration/deceleration, the average solid concentration in RRZ is calculated from
45 measured pressure drop by eq (6).
46
47 P1−1 − P1−2
 fs = (6)
48 ( z1−2 − z1−1 ) g
49
50
51 The solid concentration of the dense bed in FRZ is calculated by eq (7).
52
53 P1−3 − P1−4
54
bubs = (7)
( z1−3 − z1−2 ) g
55
56
57 The height of the dense bed in FRZ is calculated by eq (8).
58
59
60

4
ACS Paragon Plus Environment
Page 5 of 24 Industrial & Engineering Chemistry Research

1
2
3 P1−3 − P1−5
4 hd = (8)
5 bubs g
6
7 It was difficult to measure the catalysts circulation rates directly. Alternatively, the heat balance method
8
9 was used to calculate the catalysts circulation rates.
10
11 Figure 3 shows the energy balance volume used for catalysts circulation rate calculation. Based on heat
12
balance of the control volume under steady state, eq (9) is obtained. By solving eq (9), the catalysts
13
14 circulation rate could be obtained.
15
16
 Dtur 2Gs
17
18
 mi H i + 4
L(Tca − To ) =  mo H o (9)

19
20 Since only a small fraction of products was collected, it is impossible to calculate the product yield by
21
22 dividing the mass of pyridine in product with the mass of aldehydes in the feedstock. In this paper, the
23 product yield is calculated based on unit feedstock, with the details given as following. Firstly, the
24
25 aldehydes weight in unit feedstock is calculated by eq (10). Then the pyridine weight in unit product is
26 calculated by eq (11). At last, the product yield of pyridine is calculated by eq (12).
27
28
mAfe = M A / M fe (10)
29
30
31
cDpr cEpr
mApr = (  2M A +  2M A ) / (1 − cam ) (11)
32 MD ME
33
34 y = mAfe / mApr (12)
35
36 where mAfe is the mass of acetaldehyde in unit feedstocks, Mfe is the mass of feedstock, mApr is
37
38 the mass of acetaldehyde in unit products, cDpr is the concentration of pyridine in product, cEpr
39
40
is the 3-picoline concentration in products, cam is the ammonia concentration in product, y is
41 the product yield of pyridine and 3-picolines.
42
43
44
45 3. MODELING
46
47 In this paper, a coupled reactor model is proposed to investigate performance of this coupled reactor. The
48
details of the model are as following.
49
50
3.1 Reaction Kinetics
51
52 Up to now, the research on reaction kinetics of pyridine synthesis has been rare. Reddy 7
proposed a
53
reaction mechanism based on parallel series reactions. As shown in Figure 4, the feedstock (mixture of
54
55 acetaldehyde, formaldehyde and ammonia) first produce pyridine and 3-picoline, then a part of pyridine
56
and 3-picoline form coke at different rate. The main reactions are given by eq (13) and eq (14). The
57
58 byproducts in pyridine synthesis process are 2-picoline, 4-picoline, methylamine, dimethylamine and
59
trimethylamine, which can be minimized by varying the composition of the feed-stock and the ratios of
60

5
ACS Paragon Plus Environment
 Industrial & Engineering Chemistry Research Page 6 of 24

1
2
3 mixture.
4
5 2CH3CHO+HCHO+NH3 ⎯⎯
→ C5H5 N+3H2O+H2 (13)
6
7 2CH3CHO+2HCHO+NH3 ⎯⎯
→ C6H7 N+4H2O (14)
8
9 Based on the above mechanism and experimental results, Reddy 7
proposed the following reaction rate
10
expressions, with the fitted reaction constants and activation energy listed in Table S1.
11
12
13
−rA = (k1 + k2 )CA 2a (15)
14
15 rD = (k1CA 2 − k3CD )a (16)
16
17
18
rE = (k2CA 2 − k4CE )a (17)
19
where a is the reactivity of catalysts, and the value is unity for fresh catalysts. In this paper, the reactions
20
21 and the catalysts are the same as used in Reddy’s experiments, and the above reaction kinetics is adopted.
22
23
3.2 Reactor Model
24 For this coupled reactor, the gas-solid flow pattern varies from turbulent fluidization to fast fluidized bed 39
25
26 along the reactor axial direction. As shows in Table 1, for FZ and RRZ, the flow pattern is fast fluidization.
27 In FRZ, the flow pattern is turbulent fluidization.
28
29 Table 1 The flow patterns in different sections of the coupled reactor
30 ug (m/s) flow pattern
31 FZ 1.7 ~ 2.6 fast fluidization
32
33 RRZ 4.8 ~ 7.1 fast fluidization
34 FRZ 0.52 ~ 0.72 turbulent fluidization
35 For different flow patterns, various reactor models are proposed. For fast fluidization, the reactor could be
36
37 modeled by the core-annulus model 34-35. For turbulent fluidization, various models are proposed 40-41, and
38 these models could be divided into two main categories, which are CSTR in series model and dispersion
39
40 model. In this paper, different reactor models are chosen for different sections of the coupled reactor, with
41 the details given as following.
42
43 3.2.1 Reactor Model for FRZ
44 The flow pattern in the FRZ is turbulent fluidization. For turbulent fluidized bed reactors, the widely used
45
46 models are CSTR in series model and dispersion model. As shows in Figure 5, CSTR in series model uses
47 ideal mixed reactors to simulate the gas residence time distribution (RTD) 42.
While for the dispersion
48
49 model, gas back mixing is represented by an axial dispersion diffusion coefficient, which can be estimated
50 by the correlation proposed by Foka 43.
51
52 Pe = 0.071Ar 0.32 (d p / Dtur ) −0.4 (18)
53
54 Levenspiel 44 pointed out that for flow patterns not far from plug flow, the dispersion model should be used.
55
The Pe number was found to be around 3 under our experimental operating conditions, so the dispersion
56
57 model is adopted in this study. Based on gas phase mass balance, eq (19) can be derived for gas phase axial
58
dispersion flow 45.
59
60

6
ACS Paragon Plus Environment
Page 7 of 24 Industrial & Engineering Chemistry Research

1
2
3
dc d 2c
4 ug − Dzg 2 + r = 0 (19)
5 dz dz
6
By combing eq (19) with reaction kinetics, the following equations can be obtained for species A, D, E and
7
8 F, which form the model for FRZ.
9
10 dcA d 2cA
ug − Dzg − (k1 + k2 )c A 2  s = 0
11 dz dz 2

12
dc d 2 cD
13 u gtur D − Dzg + (k1c A 2 − k3cD )  s = 0
14 dz dz 2
(20)
15
dc d 2 cE
16 u gtur E − Dzg 2
+ ( k 2 c A 2 − k 4 cE )  c = 0
17 dz dz
18 dc d 2 cF
19 u gtur F − Dzg + ( k3 c D + k 4 c E )  s = 0
20 dz dz 2
21 To solve the above model equations, the catalyst concentration distribution along the axial direction is
22
46
23 required. In this paper, the correlation proposed by Lu is modified and adopted. Lu used three different
24 gas-solid systems to investigate the catalyst concentration axial distribution for turbulent fluidized bed
25
26 reactor by experiment. Based on experimental results, Lu proposed three different correlations to calculate
27 the catalysts concentration, which are given in Table 2.
28
29 Table 2 Correlations for the particle concentration calculation
30 ug (m/s) Particle Properties Correlation
31
32 0.535~1.328 ρpb=612 kg/m3，dp=63.3 μm ρs=-77.62z-141.56ugtur+503.39
33
34 0.536~1.315 ρpb=657 kg/m3，dp=59.5 μm ρs=-89.95z-126.49ugtur+502.64
35 0.826~1.519 ρpb=897 kg/m3，dp=62.4 μm ρs=101.62ugturz-387.15-252.94ugtur+925.3
36
37 Table 2 shows that the correlations proposed by Lu did not take the influence of gas-solid properties into
38 consideration, which in turn generated three correlations for three different gas-solid systems. For
39
40 simplification, in this study, the following correlation is proposed.
41
 s = Ar k (k2ugtur + k3 z + k4 )
1
(21)
42
43 By fitting Lu’s experiment data, the values of the parameters k1, k2, k3 and k4 are obtained, which are
44
45 k1=1.013, k2=-12.626, k3=-10.015, k4=48.936. Figure 6 is the comparison between the prediction result by
46 modified correlation and experimental data, with an average prediction deviation of 7 %.
47
48 3.2.2 Reactor Model for RRZ
49 The flow pattern in FZ and RRZ is fast fluidization. Core-annulus model is chosen to model this section.
50
51 Figure 7 shows the basic concept of core-annulus model, which divides the cross-sectional area into a core
52
and an annulus region. In the core, the solid concentration is low, and the gas phase could be treated as
53
54 plug flow. While in the annulus, the solid flow downward and the back-mixing is severe, and the gas phase
55
is treated as mixed flow. Based on mass balance in the core region and annulus region, the following
56
57 equations are obtained.
58
Core region:
59
60

7
ACS Paragon Plus Environment
 Industrial & Engineering Chemistry Research Page 8 of 24

1
2
3 dcc 4kac
4 u gc + (cc − ca ) + r = 0 (22)
5 dz Dr  g1/2
6
7 Annulus region:
8
dca 4ka  c1/2
9 u ga + (ca − cc ) + r = 0 (23)
10 dz Dr (1 −  c )
11 47 48
12 Marmo and Kagawa assumed the gas velocity in the annulus as zero. In this paper, the same
13 assumption is used. By combing equation (22) and equation (23) with the reaction kinetics, the following
14
15 equations could be obtained.
16
17
18 −m + m 2 + 4mnccA 4kac  c1/2
19 caA = (m = , n = (k1 + k2 )  sa )
20 2n D f (1 −  c )
21 dccA 4kac
22 = (− (ccA − caA ) − (k1 + k2 )ccA 2  sc ) / u gc
23 dz D f c1/2

24
25
mccD − k1caA 2  sa
caD =
26 m − k3  sa
27
dccD 4kac
28 = (− (ccD − caD ) + (k1ccA 2 − k3ccD )  sc ) / u gc (24)
29 dz D f c1/2

30
31 mccE − k2 caA 2  sa
caE =
32 m − k4  sa
33
34 dccE 4kac
= (− (ccE − caE ) + (k2 ccA 2 − k4 ccE )  sc ) / u gc
35 dz D f  c1/2
36
37 dccF
= (k3ccD + k4 ccE )  sc / u gc
38 dz
39
40 where Фc is ratio of the core area to cross sectional area and is calculated by the correlation proposed by
41 Patience 49.
42
43 1
44
c = (25)
1 + 1.1Fr (u ps / u gf ) 0.083 Fr
45
46 The solid concentration radial distribution could be treated as follows. The solid fraction in the core region
47 50-52.
48 is assumed as a constant, which is equal to 0.005 The solid concentration in the annulus could be
49 calculated by mass balance in eq (26).
50
51  c c + (1 −  c ) an =  av (26)
52
53 where the axial solid concentration εav is calculated by Li-Kwauk model 53,54, which is shows as following.
54  av −  bt 1
55 ln( ) = − ( z − zi ) (27)
56  '−  av Z0
57
The mass transfer coefficient is calculated by the correlation proposed by Patience 49,
58
59
60

8
ACS Paragon Plus Environment
Page 9 of 24 Industrial & Engineering Chemistry Research

1
2
3
kac D f  g 0.5  g ugf D f 0.75
4  g1/2 = 0.25( ) ( ) [Gs / (  pug )]0.25 (28)
5 Dg  g Dg  g  g (c )1/2

6
7 where Dg is the gas diffusion coefficient and is calculated by the correlation proposed by Fuller 55.
8
9 1 1
0.001T 1.75 +
10 MA MB
11 DAB = (29)
12 P[( vA )1/3 + ( vB )1/3 ]2
13
where DAB is the binary diffusion coefficient of gases A and B, ∑vA and ∑vB are the molecular diffusion
14
15 volume of A and B respectively.
16
17 Equations (24) to (29) form the reactor model for the RRZ.
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

9
ACS Paragon Plus Environment
 Industrial & Engineering Chemistry Research Page 10 of 24

1
2
3
4
5 4 EXPERIMENT RESULTS
6 4.1 Results of this pilot scale reactor
7
8 The experimental operation conditions are listed in Table 3. The physical properties of gases are obtained
9 from ASPEN Plus. The experimental apparatus successfully run for 15 days and then stopped for safety
10
11 check. There was no coke formation on the feed nozzle, which indicates the coking problem in the feed
12 introduction device has been partly eliminated.
13
14 Table 3. Variation range of operating parameters
15 Parameters Variation range
16
17 cA0 (mol/L) 0.003~0.009
18 T1-1 (K) 711~757
19 T1-2 (K) 700~734
20
P2-1 (kPa) 180~190
21
22 Gs (kg/m2/s) 10~80
23 hd (m) 0~2.2
24 dp (mm) 70
25
26 ρp (kg/m3) 2400
27 ρg (kg/m3) 0.71
28 μg (Pa.s) 2.56x10-5
29
30 Gs: catalysts circulation rate based on the RRZ cross sectional area
31
During the experiments, the values of ATFR and ATAR were in unity, ratios widely used in pyridine bases
32
33 industries. Table 4 and Table S2 showed the experimental production yield.
34
35 Table 4. Selectivity and Product yield under different operating conditions
36 cA0 T1-1 T1-2 Gs hd Selectivity yield (%)
N
37 (mol/L) (K) (K) (kg/m2/s) (m)
38
1 0.003 729 708 25 1.3 10.2 25
39
40 5 0.007 733 701 47 1.4 4.2 52
41 7 0.008 741 729 56 1.2 3.0 54
42
10 0.008 742 708 57 0 2.6 41
43
44 14 0.007 725 714 36 1.4 2.9 59
45 15 0.007 730 714 36 2.2 3.3 58
46 18 0.007 738 727 72 1.1 2.2 72
47
48 20 0.007 711 701 58 2.0 3.3 58
49
50
51 From Table 4 and Table S2, it is seen that the selectivity of pyridine over 3-picoline varies with production
52 yield, and the selectivity value is greater than 2.2. The selectivity under the highest production yield
53
54 operating conditions is 2.6, and is greater than that from commercial reactors, which is around 2. The
55 increase in selectivity mainly results from the increased gas velocities in this coupled reactor, which
56
57 shortens the residence time of the catalyst and increases its reactivity 12. The highest product yield is 75 %,
58 which is similar as that of commercial reactor, which is 73 %. This indicated that this coupled reactor has a
59
60 better performance than single fast fluidized bed reactor in terms of production yield.

10
ACS Paragon Plus Environment
Page 11 of 24 Industrial & Engineering Chemistry Research

1
2
3 From experimental data 1 and 5, it is seen that the increase in initial acetaldehyde concentration raises the
4
5 product yield, which agrees with the model prediction result of Zhou 12. However, for commercial reactors,
6 the feedstocks are fixed, it is impossible to change the initial acetaldehyde concentration. The influence of
7
8 initial aldehyde concentration on product yield is thus only of interest for lab scale experiment. Compare
9 experimental data 7 and 10, the product yield is seen to vary from 54 % to 41 % when the reaction
10
11 condition changes from Hb=1.2 to Hb=0. For Hb=0, the dense bed in FRZ disappears, the reactions take
12 place mostly in FZ and RRZ. For Hb=1.2, an additional reaction region-bubbling fluidization section forms,
13
14 and this region contributes 13 percent conversion. This result indicates that the product yield decrease
15 observed by Zhou may result from the limitation of gas-solids contact time under certain reaction
16
17 conditions. From experimental data 14 to 15, the dense bed height in the FRZ increases by 0.8 m, while the
18 production yield keeps almost the same. This indicates that the increase of residence time does not always
19
20 increases the product yield. The reason is that, with the increase of residence time, the product yield first
21 increases and then decreases due to over-reaction 12. Therefore, there must exist a suitable residence time,
22
23 which could be determined from experiment. Experimental data 18 and 20 show the influence of reaction
24 temperature on the product yield. It shows that increase of reaction temperature will increase the product
25
26 yield.
27 4.2 Comparison between the pilot scale reactor and commercial reactor
28
29 Table 5 shows the comparison between the pilot scale reactor and commercial reactor. As shows in Table 5,
30 compared with the commercial reactor, this pilot scale reactor has higher product selectivity and lower
31
32 spent catalysts coke contention with similar product yield.
33 Table 5. Comparison between the pilot scale reactor and commercial reactor performance
34 Pilot Scale Reactor Commercial Reactor
35
36 Product yield (%) 75 73
37 Product selectivity Above 2.5 2.2
38 Coke content in spent catalysts (w%) 0.92 1.46
39
40 The selectivity is determined by reaction kinetics and can be modified through adjusting the flow
41 hydrodynamics. Figure 8 shows the selectivity variation with reactants residence time under typical
42 commercial reaction condition. Figure 8 indicates that the selectivity increases with the increase of
43
reactants residence time. However, to optimize the product yield, the reaction should be terminated before
44
45 reaching the over-reaction zone 1. Under this reaction condition, the best reactants residence time is 3.1 s.
46 In that case, the reactants residence time should be kept at this value. However, due to the back-mixing, the
47 reactants residence time of real reactors is a distribution rather than a unique value. To narrow down the
48
49 reactants residence time distribution, the back-mixing should be minimized. For this coupled reactor, the
50 back-mixing in different reaction zones are varying owing to different gas-solid flow patterns. Since the
51 back-mixing of turbulent fluidization and fast fluidization is less than that in bubbling fluidization, the
52
53 coupled reactor has less back-mixing than the commercial reactor, which results in a higher selectivity of
54 the products.
55 Table 5 shows that the spent catalysts coke contention is higher in the commercial reactor. The spent
56
catalysts coke contention is closely related to the catalysts residence time 12, with the same reaction
57
58 conditions, the spent catalysts coke contention increases with the catalysts residence time. For the fast
59 fluidization, the catalysts residence time is around 10 s 12. While, for the bubbling fluidization, the catalysts
60 residence time is around 1 h. For this pilot scale reactor, the flow patterns are fast fluidization and turbulent

11
ACS Paragon Plus Environment
 Industrial & Engineering Chemistry Research Page 12 of 24

1
2
3 fluidization, and the catalysts residence time is less than that of the bubbling fluidization commercial
4
5 reactors, which decreases the spent catalysts coke contention.
6
7
8 5 MODELING RESULTS
9
10 Eqs. (18) to (29) formed the reactor model for this coupled reactor. Comparison between the model
11
prediction and experimental data of this coupled reactor is illustrated in Figure 9. The average prediction
12
13 deviation is 12 %.
14
Figure 10 shows the gas concentration variation along the axial direction of this coupled reactor, which
15
16 indicates that the main conversion takes place in FZ and FRZ. The conversion in RRZ only accounts for 10
17
percent of the total product yield. The superficial gas velocity in RRZ is 6.1 m/s, and the length of the
18
19 reaction zone is 13.5 m, then the residence time is 2.2 s. For FZ, the superficial gas velocity is 1.5 m/s, and
20
the length of the reaction zone is 1 m. In that case, the gas residence time in the FZ is 0.67 s. From the
21
22 residence time point of view, the conversion in the RRZ should be higher than that in the FZ. However, the
23
model prediction result gives the opposite conclusion. This indicates the superficial reaction rate in RRZ is
24
25 low. Damokler (Da) number is used to quantify the ratio of reaction rate from reaction kinetics to mass
26
27
transfer rate. The Da number is the ratio of the reaction rate from reaction kinetics to mass transfer rate at
28 initial reaction conditions. If the Da number less than 0.11, the reaction is the limit step. Figure 11 shows
29
30 axial distribution of the Da number in RRZ, which indicates that the mass transfer rate between the core
31 and annulus region is far less than the reaction rate from reaction kinetics, and the mass transfer between
32
33 the core and annulus region becomes the limit step for the pyridine synthesis reaction. For such a small
34 mass transfer rate, the corresponding conversion is low in RRZ.
35
36 The riser diameter is 84 mm, which has more severe wall effect when compared with commercial risers.
37 The severe wall effect leads to large annulus region along the riser radial direction, and in turn limits the
38
39 mass transfer between the core region and annulus region, which at last results in low conversion in the
40 RRZ. As for selectivity, as discussed in the Experimental Result section. The back-mixing in the riser is far
41
42 less than that in the commercial reactors, which increase the selectivity of the products.
43
44
45
46 6 CONCLUSIONS
47
48 A coupled fluidized bed reactor is proposed for gas phase pyridine synthesis. A pilot scale experimental
49 apparatus was commissioned, and the product yield under different operating conditions was measured. A
50
51 model is proposed to simulate the reaction process in the coupled reactor. During 15 days of continuous
52 operating, there is no coke deposit on the feeding device, which indicates this novel feeding scheme is
53
54 effective. The product yield in the coupled reactor reaches as high as that in the commercial reactors,
55 which means enough gas-solid contact time is ensured by this coupled reactor. The selectivity of pyridine
56
57 over 3-picoline is around 2.5 and is higher than that in commercial reactors, which is around 2.2.
58 The model accurately predicts the product yield with an average deviation from experimental data of 12%.
59
60 Based on model prediction, the conversion is mainly contributed by FZ and FRZ of the coupled reactor.

12
ACS Paragon Plus Environment
Page 13 of 24 Industrial & Engineering Chemistry Research

1
2
3 Limited by the mass transfer and solid concentration in the riser, the RRZ contributes less than 10 percent
4
5 to the total conversion.
6
7
8
9 SUPPORTING INFORMATION
10 Tables with the production yield under different operating conditions and the value of reaction kinetic
11
12 constants.
13
14 ACKNOWLEDGMENTS
15
16 This work was supported by NKBRDP (the National Key Basic Research Development Program) of China
17 (973, 2012CB215000).
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

13
ACS Paragon Plus Environment
 Industrial & Engineering Chemistry Research Page 14 of 24

1
2
3 NOMENCLATURE
4
5 Ar Archimedes number
6 a activity of catalysts, for fresh catalysts the value is unity
7 cA concentration of acetaldehydes in feedstock (mol/L)
8
9 cD pyridine concentration (mol/L)
10 cE 3-picoline concentration (mol/L)
11 cF coke concentration (mol/L)
12
D diameter (m)
13
14 Dg diffusion coefficient (cm2/s)
15 Fr Froud number
16 Gs catalysts circulation rate based on reactor cross sectional area (kg/m2/s)
17
18 g gravity constant, 9.8 N/kg
19 h height (m)
20 Hi the enthalpy of the feed stocks in the inlet of the computation node, (kJ/kg)
21
22 Ho the enthalpy of the feed stocks in the outlet of the computation node, (kJ/kg)
23 L the hear latent of the catalysts, (kJ/kg/K)
24 k1 reaction rate constant for reaction A to D (L/(mol.s))
25
k2 reaction rate constant for reaction A to E (L/(mol.s))
26
27 k3 reaction rate constant for reaction D to F (s-1)
28 k4 reaction rate constant for reaction E to F (s-1)
29 kac mass transfer coefficient from core to annulus region (m/s)
30
31 kbd mass transfer coefficient from bubble phase to dense phase (m/s)
32 Lb length of bubbling bed reaction zone (m)
33 m mass flow rate (kg/s)
34
35 M mole mass (kg/kmol)
36 N number of CSTRs
37 Nho number of holes per cm2 in gas distributor
38
P pressure (Pa)
39
40 Pe Peclet number
41 R ideal gas constant 8.314 (J/(mol.K))
42 r reaction rate (mol/(L.s)
43
44 rA reaction rate for acetaldehydes (mol/(L.s))
45 rD reaction rate for pyridine (mol/(L.s))
46 rE reaction rate for 3-picoline (mol/(L.s))
47
48 s product selectivity
49 T reaction temperature (K)
50 t residence time (s)
51
ug superficial gas velocity (m/s)
52
53 ups particle velocity, Gs/ρp (m/s)
54 V volumetric flow rate (m3/h)
55 v diffusion volume (cm3)
56
57 y pyridine and 3-picoline production yield
58 z height of the reaction zone (m)
59
60

14
ACS Paragon Plus Environment
Page 15 of 24 Industrial & Engineering Chemistry Research

1
2
3
4
5
6
7 Greek symbols
8
9 ηaf acetaldehydes to formaldehydes ratio
10 ηam ammonia to aldehydes ratio
11 △ difference
12
σ standard deviation
13
14 Фc the ratio of core area to cross area
15 ρs solid phase concentration (kg/m3)
16 ε solid fraction
17
18 μg gas phase diffusivity (Pa.s)
19 Subscripts
20 a annulus
21
22 ax axial
23 c core
24 ca catalyst
25
d dense phase
26
27 f fast fluidized bed
28 fe feedstock
29 g gas phase
30
31 i in
32 o out
33 0 initial
34
35 pb packed bed
36 pr product
37 s solid phase
38
tur turbulent fluidization
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60

15
ACS Paragon Plus Environment
 Industrial & Engineering Chemistry Research Page 16 of 24

1
2
3
4
REFERENCES
5 (1) Reddy, K. S. K.; Srinivasakannan, C.; Raghavan, K. V. Catalytic vapor phase pyridine synthesis: a
6 process review. Catal. Surv. Asia. 2015, 16, 28-35.
7 (2) Jin, F.; Cui, Y.; Rui, Z.; Li, Y. Effect of sequential desilication and dealumination on catalytic
8 performance on ZSM-5 catalyst for pyridine and 3-picoline synthesis. J. Mater. Res.. 2010, 25, 272-282.
9 (3) Calvin, J. R.; Davis, R. D.; McAteer C. H. Mechanistic investigation of the catalyzed vapor-phase
10 formation pyridine and quinoline bases using 13CH2O, 13CH3OH, and deuterium-labled aldehydes. Appl.
11 Catal., A. 2005, 285, 1-23.
12 (4) Baldev, S.; Sisir, K. R.; Krishnadeo, P. S.; Tarun, K. G. Role of acidity of pillared ineter-layered clay
13 (PILC) for the synthesis of pyridine bases. J. Chem. Technol. Biotechnol. 1998, 71, 246-252.
14 (5) Kalevaru, V. N.; Madaan, N.; Martin, A. Synthesis, characterization and catalytic performance of titania
15 supported VPO catalysts for the ammoxidation of 3-picoline. Appl. Catal., A. 2011, 391, 52-62.
16 (6) Sisir, K. R.; Banikar, G.; Shyam, K. R. Studies on the synthesis of 2 & 4-picoline –correlation of acidity
17 with the catalytic activity. Stud. Surf. Sci. Catal. 1998, 113, 713-719.
18 (7) Reddy, K. S. K.; Screedhar, I.; Raghavan, K. V. Kinetic studies on vapour phase pyridine synthesis and
19 catalyst regeneration studies. Can. J. Chem. Eng. 2011, 89, 854-863.
20 (8) Goe, G. I.; Davis, R. D. Pyridine base synthesis process and catalyst for same. U. S. Patent 5,218,122,
21 1993.
22 (9) Rao, R. R.; Kulkarni, S. J.; Subramanyam, M.; Rama Rao, A. V. Synthesis of pyridine and picolines
23 over modified silica-alumina and ZSM-5 catalysts. Reac. Kinet. Catal. Lett. 1995, 56, 301-309.
24 (10) Francis, E. Cislak; William, R. Wheeler. Synthesis of pyridine and 3-picoline. U. S. Patent, 2,807,618,
25 1957.
26 (11) Feitler, D.; Wolfgang, S.; Henry, W. Process for the production of pyridine or alkyl substituted
27 pyridines. U. S. Patent 4,675,410, 1987.
28 (12) Zhou, S.; Liu, Z.; Qin, D.; Yan, X.; Liu, M.; Lu, C.; Jin. G. Investigation of pyridine synthesis in a fast
29 fluidized bed reactor. Ind. Eng. Chem. Res. 2018, 57, 1179-1187.
30 (13) Luo, C.-W.; Chao, Z.-S. Unsaturated aldehydes: a novel route for the synthesis of pyridine and
31 3-picoline. RSC Adv. 2015, 5, 54090-54101.
32 (14) Wang, D.; Lu, C.; Yan. C. Effect of static bed height in the upper fluidized bed on flow behavior in the
33 lower riser section of a coupled reactor. Paricuology. 2009, 18, 19-25.
34 (15) Wang, D. Gas-solid flow behavior in a riser-fluidized bed coupled reactor. Ph. D. Dissertation, China
35 University of Petroleum (Beijing), Beijing, CHN, 2009.
36 (16) Liu, C.; Feng, Wei.; Zhang, Y.; Zhang, J.; Liu, X. Development of new pre-lifting construction for
37 FCC riser reactor. Petro. Refinery Eng. (in Chinese). 2007, 39, 24-27.
38 (17) Li, Y. Applied efficiency of pre-lift section in heavy oil riser pipe of catalytic cracking unit. J. Chem.
39 Ind. Eng. (in Chinese). 2011, 32, 46-49.
40 (18) Zhu, X.; Geng, Q.; Wang, G.; Li, C.; Yang, C. Hydrodynamics and catalytic reaction inside a novel
41 multi-regime riser. Chem. Eng. J. 2014, 246, 150-159.
42 (19) Zhu, X.; Li, C.; Yang, C.; Wang, G.; Geng, Q.; Li, T. Gas-solids flow structure and prediction of solids
43 concentration distribution inside a novel multi-regime riser. Chem. Eng. J. 2013, 232, 290-301.
44 (20) Zhu, X.; Yang, C.; Li, C.; Liu, Y.; Wang, L.; Li, T.; Geng, Q. Comparative study of gas-solids flow
45 patterns inside novel multi-regime riser and conventional riser. Chem. Eng. J. 2013, 215-216, 188-201.
46 (21) Harris, A. T.; Davidson, J. F.; Thorpe, R. B. Particle residence time distributions in circulating
47 fluidized beds. Chem. Eng. Sci. 2003, 58, 2181-2202.
48 (22) Wei, F.; Zhu, J.-X. Effect of flow direction on axial solid dispersion in gas-solid cocurrent upflow and
49 downflow system. Chem. Eng. J. 1996, 64, 345-352.
50 (23) Wei, F.; Jin, Y.; Yu, Z.; Chen, W.; Mori, S. Lateral and axial mixing of the dispersed particles in CFB,
51 J. Chem. Eng. Jpn. 1995, 28, 506-510.
52 (24) Zheng, Q.; Zhang, H. Experimental study of the effect of exit (end effect) geometric configuration on
53 enternal recycling of bed material in CFB combustor. In Circulating Fluidized Bed Technology IV; Avidan,
54 A. A., Eds.; AIChE Publishers, Inc.: New York, 1994: pp 145-151.
55 (25) Bai, D.-R.; Jin, Y.; Yu, Z.-Q.; Zhu, J.-X. The axial distribution of the cross-sectionally averaged
56 voidage in fast fluidized beds. Powder. Technol. 1992, 71, 51-58.
57 (26) Brereton, C. M. H.; Grace, J. R. End effects in circulating fluidized bed hydrodynamics. In
58 Circulating Fluidized Bed Technology IV; Avidan, A. A., Eds.; AIChE Publishers, Inc.: New York, 1994: pp
59 137-144.
60 (27) Chan, C.W.; Seville, J.; Yang, Z.; Baeyens, J. Particle motion in the CFB riser with special emphasis

16
ACS Paragon Plus Environment
Page 17 of 24 Industrial & Engineering Chemistry Research

1
2
3 on PEPT-imaging of the bottom section. Powder. Technol. 2009-a, 196, 318-325
4 (28) Van de Velden, M.; Baeyens, J.; Seville, J.P.K.; Fan, X. The solids flow in the riser of a CFB viewed
5 by Positron Emission Particle Tracking (PEPT). Powder. Technol. 2008, 183, 290–296.
6 (29) Parker, D. J.; McNeil, P. A. Positron emission tomography for process applications. Meas. Sci.
7 Technol. 1996, 7, 287-296.
8 (30) Bi, H.T.; Grace, J.R. Flow regime diagrams for gas–solid fluidization and upward transport. Int. J.
9 Multiphase Flow. 1995, 21, 1229–1236.
10 (31) Smolders, K.; Baeyens, J. Gas fluidized beds operating at high velocities: a critical review of
11 occurring regimes. Powder Technol. 2001, 119, 269-291.
12 (32) Jin, Y.; Zhu, J.; Wang, Z.; Yu, Z. Fluidization Engineering Principles (in Chinese). Tsinghua Publisher:
13 Beijing, 2008, pp 151-152.
14 (33) Luo, X. Catalyst density distribution in commercial FCC regenerator with fast fluidization. Journal of
15 the University of Petroleum (in Chinese). 1992, 16, 90-94.
16 (34) Bi, H. Some issues on core-annulus and cluster models of circulating fluidized bed reactors. Can. J.
17 Chem. Eng. 2010, 80, 809-817.
18 (35) Bai, D.; Zhu J.-X.; Jin, Y.; Yu, Z. Novel Designs and Simulations of FCC Riser regeneration. Ind. Eng.
19 Chem. Res. 1997, 36, 4543-4548.
20 (36) Patience, G. S.; Chaouki, J. Gas phase hydrodynamics in the riser of a circulating fluidized bed. Chem.
21 Eng. Sci. 1993, 48, 3195-3205.
22 (37) Li, J.; Zhang, X.; Zhu, J.; Li, J. Effects of cluster behavior on gas-solid mass transfer in circulating
23 fluidized beds. In Fluidization IX; Fan, L. S.; Knowlton, T. M., Eds.; Engineering Foundation Publishers,
24 Inc.: New York, 1998; pp 405-412.
25 (38) Kruse, M.; Schoenfelder, H.; Werther, J. A two-dimensional model for gas mixing in the upper dilute
26 zone of a circulating fluidized bed. Can. J. Chem. Eng. 1995, 73, 620-634.
27 (39) Bi, H. T.; Grace, J. R. Effect of measurement method on the velocities used to demarcate the onset of
28 turbulent fluidization. Chem. Eng. J. & Biochem. Eng. J. 1995, 57, 261-271.
29
(40) Jin, Y.; Zhu, J.; Wang, Z.; Yu, Z. Fluidization Engineering Principles (in Chinese). Tsinghua Publisher:
30
Beijing, 2008, pp 335-337.
31
(41) Bi, H. T.; Ellis, N.; Abba, I. A.; Grace, J. R. A state-of-the-art review of gas-solid turbulent
32
fluidization. Chem. Eng. Sci. 2000, 55, 4789-4825.
33
(42) van der Laan, E. T. Notes on the diffusion-type model for longitudinal mixing in flow. Chem. Eng. Sci.
34
1958, 7, 187–191.
35
(43) Foka, M.; Chaouki, J.; Guy, C.; Klvana, D. Gas phase hydrodynamics of a gas-solid turbulent
36
fluidized bed reactor. Chem. Eng. Sci. 1996, 55, 713-723.
37
(44) Octave Levenspiel. Chemical Reaction Engineering (3rd Edition). John Wiley & Sons Press: New
38
York, 1999, pp 94-96.
39
40 (45) Edwards, M.; Avidan, A. Conversion model aids scale-up of Mobil's fluid bed MTG process. Chem.
41 Eng. Sci. 1986, 41, 829-835.
42 (46) Lu, C.; Wang, Z. Study on Dense-Density in turbulent fluidized bed with FCC catalyst (In Chinese).
43 Petro. Chem. Eng. 1988, 17, 499-503.
44 (47) Marmo, L.; Rovero, G.; Manna, L. Modelling on circulating fluidized bed reactors. 44th Can. Chem.
45 Eng. Conf., Calgary, Alberta, Oct. 2-5.
46 (48) Kagawa, H.; Mineo, H.; Yamazaki, R.; Yoshida, K. A gas-solid contacting model for fast fluidized bed.
47 In Circulating Fluidized Bed Technology III; Basu, P.; Horio, M.; Hasatani, M., Eds.; Pergamon Press
48 Publishers, Inc.: Oxford, 1990; pp 551-556.
49 (49) Patience, G. S.; Chaouki, J. Gas phase hydrodynamics in the riser of a circulating fluidized bed. Chem.
50 Eng. Sci. 1993, 48, 3195-3205.
51 (50) Pugsley, T. S.; Berruti, F. A Predictive Hydrodynamic Model for Circulating Fluidized Bed Risers.
52 Powder Technol. 1996, 89, 57–69.
53 (51) Ouyang, S.; Li, X. G.; Potter, O. E. Circulating Fluidized Bed as a Catalytic Reactor: Experimental
54 Study. AIChE J. 1995, 41, 1534–1542.
55 (52) Buchyr, D. M. J.; Mehrotra, A. K.; Behie, L. A.; Kalogerakis, N. Modelling a Circulating Fluidized
56 Bed Riser Reactor with Gas–Solids Downflow at the Wall. Can. J. Chem. Eng. 1997, 75, 317–326.
57 (53) Li, Y.; Kwauk, M. The dynamics of fast fluidization. In: J.R.G. Fluidization, Matsen, J.M., Eds.;
58 Plenum Publishers, Inc.: New York, 1980, pp 537–544.
59 (54) Li, Y.; Chen, B.; Wang, F.; Wang, Y.; Kwauk, M. The correlations for the fast fluidization model’s
60 parameters (in Chinese). Chin. J. Proc. Eng. 1980, 4, 20-30.

17
ACS Paragon Plus Environment
 Industrial & Engineering Chemistry Research Page 18 of 24

1
2
3 (55) Fuller, E. N.; Schettler, P. D.; Giddings, J. C.. A new method for prediction of binary gas-phase
4 diffusion coefficients. Ind. Eng. Chem. 1966, 58, 18-27.
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

18
ACS Paragon Plus Environment
Page 19 of 24 Industrial & Engineering Chemistry Research

1
2
3 Figures
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 1: Self adjusting flow measuring system
30
31 Figure 1. Process of experiment apparatus
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

19
ACS Paragon Plus Environment
 Industrial & Engineering Chemistry Research Page 20 of 24

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 Figure 2. Axial positions of thermocouples and pressure gauges
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54 Figure 3. Computation node for catalysts circulation rate
55
56
57
58
59
60

20
ACS Paragon Plus Environment
Page 21 of 24 Industrial & Engineering Chemistry Research

1
2
3
4
5
6
7
8
9
10 A: Acetaldehyde; B: Formaldehyde; C: Ammonia; D: Pyridine; E: 3-picoline; F: Coke
11
Figure 4. Model reaction scheme for pyridine synthesis
12
13
14
15
16
17
18
19
20
21
22
23 Figure 5. CSTR in series model
24
25
26
27
28
29
30 Experiment data
31 600
32
33
Calculated ρb (kg/m3)

34
35 +15%
36 400
37
38
39
40
41 200 -15%
42
43
44
45
46 0
47 0 200 400 600
48
49 Experimental ρb (kg/m3)
50
51
52 Figure 6. Comparison between the modified correlation prediction result and experimental data
53
54
55
56
57
58
59
60

21
ACS Paragon Plus Environment
 Industrial & Engineering Chemistry Research Page 22 of 24

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19 Figure 7. Core-annulus model
20
21
22
23
24 0.008
CA (mol/L)
25 CD (mol/L) s: selectivity
26
CE (mol/L)
27 0.006
28 CF (mol/L) Over reaction
29 s/1000
30 0.004
31
32
33 0.002
34
35
36 0.000
37 T=728 K, CA0=0.008 mol/L, ρs=200 kg/m3/s
38
-1 0 1 2 3 4 5 6 7 8
39
40 t (s)
41
42
43 Figure 8. The selectivity variation with reactants residence time
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60

22
ACS Paragon Plus Environment
Page 23 of 24 Industrial & Engineering Chemistry Research

1
2
3
4
5
6
7 90
8
9 Experiment data
10
11
12
13 60
+15%
14
Calculated y

15
16
17
18
19 30
20 -15%
21
22
23
24
25 0
26 0 30 60 90
27
28 Experimental y
29
30
31
32
33
Figure 9. Comparison between the model prediction and experiment data
34
Fluidized bed
35
Reaction zone
36
37 acetaldehyde
38 pyridine
0.006
39 3-picoline
40 coke
41
Riser Reaction Zone
42 0.004
43
c (mol/L)

44
45 0.002
46
47
48
49 0.000
50
51 Feeding
52 Zone 0 3 6 9 12 15
53 z (m)
54
55
56 Figure 10. Concentration variation along axial direction of the coupled reactor by model
57
58 prediction
59
60

23
ACS Paragon Plus Environment
 Industrial & Engineering Chemistry Research Page 24 of 24

1
2
3
7
4 x 10
5 7
6
7
8 6
9
10
11 5
12
13
14 4
15
Da

16
17 3
18
19
20 2
21
22
23 1
24
25
26 0
1 1.5 2 2.5 3 3.5 4 4.5 5
27
28 z/0.1 (m)
29
30 Figure 11. The Da number distribution along axial direction in the RRZ
31
32
33
34
35
36 For Table of Contents Only
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60

24
ACS Paragon Plus Environment