Experimental and Mathematical Simulation of Noncompetitive and


Experimental and Mathematical Simulation of Noncompetitive and...

0 downloads 135 Views 1MB Size

Subscriber access provided by Kaohsiung Medical University

Separations

Experimental and Mathematical Simulation of Noncompetitive and Competitive Adsorption Dynamic of Formic AcidLevulinic Acid-5-Hydroxymethylfurfural from Single, Binary and Ternary Systems in a Fixed-bed Column of SY-01 Resin Jiayi Zheng, Baoying Pan, Jiangxiong Xiao, Xianda He, Zhe Chen, Qianlin Huang, and Xiaoqing Lin Ind. Eng. Chem. Res., Just Accepted Manuscript • DOI: 10.1021/acs.iecr.8b01283 • Publication Date (Web): 03 Jun 2018 Downloaded from http://pubs.acs.org on June 3, 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 45 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

Experimental and Mathematical Simulation of Noncompetitive and Competitive

Adsorption

Dynamic

of

Formic

Acid-Levulinic

Acid-5-Hydroxymethylfurfural from Single, Binary and Ternary Systems in a Fixed-bed Column of SY-01 Resin Jiayi Zheng,† Baoying Pan,† Jiangxiong Xiao,† Xianda He,† Zhe Chen,† Qianlin Huang,‡,§ and Xiaoqing Lin*,†,‡ †

School of Chemical Engineering and Light Industry, Guangdong University of

Technology, No. 100 Waihuan Xi Road, Panyu District, Guangzhou 510006, People’s Republic of China ‡

Guangzhou Institute of Energy Conversion, Chinese Academy of Sciences, No. 2

Nengyuan Road, Tianhe District, Guangzhou 510640, People’s Republic of China §

University of Chinese Academy of Sciences, No. 19A Yuquan Road, Beijing 100049,

People’s Republic of China

∗ Corresponding author: Xiaoqing Lin (X. Lin) Tel.: +86 20 39322172 E-mail: [email protected]

1

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research 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 Levulinic acid (LA) is a versatile platform chemical in the modern concept of the biorefinery and can be used to synthesize a broad range of desirable chemicals and fuel additives. Unfortunately, since LA released from biomass hydrolysate is accompanied by formic acid (FA) and 5-hydroxymethylfurfural (5-HMF), it is also important to investigate the binary and ternary adsorption equilibrium, as well as competitive dynamic fixed-bed column adsorption from the viewpoint of industrial application. Batch adsorption experiments showed that the affinity of SY-01 resin toward FA-LA-5-HMF were in the order of 5-HMF>LA>FA under noncompetitive and competitive system. The highest adsorption capacity were 7.54 mg/g wet resin for FA, 103.51 mg/g wet resin for LA, and 107.73 mg/g wet resin for 5-HMF. Interestingly, the presence of FA has a synergistic effect on the adsorption of LA and 5-HMF onto SY-01 resin in binary or ternary mixtures system, leading to a slight increase in adsorption uptakes. Furthermore, a mathematical model based on the general rate model coupled with the noncompetitive single component and competitive multi-component Langmuir isotherm was successfully developed to simulate the breakthrough curves of FA-LA-5-HMF from single, binary, as well as ternary components mixtures. The proposed methodology for fixed-bed column multi-component competitive adsorption model can be successfully implemented to completely design the separation unit of LA from aqueous solution or biomass hydrolysate. Furthermore, it also has the potential to expand the application to the actual biomass hydrolysate, saving a lot of manpower and material resources.

Keywords: Adsorption; Competitive breakthrough curve; Fixed-bed column; Mathematical model; Resin adsorbent

2

ACS Paragon Plus Environment

Page 2 of 45

Page 3 of 45 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

1 Introduction In order to deal with the shortage of petroleum resources and the environmental protection requirements of energy saving and emission reduction, lignocellulosic biomass, which does not compete with the food supply, is recognized as one of the most important ways to solve these problems, attracting extensive worldwide attention.1 The renewable biomass resource can be converted into glucose through enzymatic2, 3 and acid hydrolysis4 and can be utilized to produce high value-added functional chemical intermediates and renewable fuels.5-7 Among those chemicals, levulinic acid (LA) and 5-hydroxymethylfurfural (5-HMF) are two versatile platform chemicals in the modern concept of the biorefinery and can be used to synthesize a broad range of desirable chemicals and fuel additives.8-10 Numerous patents and articles so far have focused on the preparation of LA through acid-catalyzed dehydration and enzymatic hydrolysis of several different substrates in terms of monosaccharide, polysaccharides and renewable biomass resource.3, 6, 11-14 However, to the best of our knowledge, significant attention has been paid to selection of catalyst system as well as preparation and optimization of catalyst, while, on the other hand, surprisingly little attention has been devoted to separation and purification of LA from aqueous solutions or actual biomass hydrolysate.10, 15, 16 As a result, the efficient separation and removal of LA from aqueous solution and actual biomass hydrolysate have received worldwide attention in recent years. In the past two decades, various separation and purification techniques including vacuum distillation,6 solvent extraction17, 18, esterification19, membrane separation,20 electrodialysis,21 ionic liquids22 and adsorption15,

16, 23-26

have been developed and

applied to remove LA from aqueous solutions or actual biomass hydrolysates. Each method possesses its respective advantages, resulting in a significant in LA removal.

3

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research 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

However, there are some limitations, such as membrane fouling, clogging, operational simplicity, energy requirement, toxicity, capacity, scale up and cost investment, which have been depicted in detail previously.10 Among the above-mentioned methods, adsorption has been considering as an efficient treatment strategy and garnering much attention due to its wide application scope, easy operation, environmentally friendly, cost-effective implementation, high efficiency and low energy consumption.27, 28 In spite of these advantages, adsorption has certain limitations such as it could not achieve a good status at commercial levels. Ideal adsorbent for actual biomass hydrolysate separation should exhibit high selectivity and optimal adsorption capacity for LA and 5-HMF at relevant conditions. Accordingly, various adsorbent, including active carbon,29, 30 ion-exchange resin,24-26 and hyper-cross-linked adsorption resin,15, 16, 31

have been extensively studied for recovering LA for aqueous solutions or actual

biomass hydrolysate. Active carbon, possessing the important characteristics in terms of high specific surface area, the surface charge, as well as predominant proportion of micropores, exhibits an excellent performance in adsorption of LA and 5-HMF.29, 30, 32 Dornath and Fan reported selective adsorption of HMF, LA and fructose onto the four carbon material at concentrations ranging from 0.5 to 100 g/L at room temperature.30 Unfortunately, the concentration of adsorbate is much larger than that of the actual biomass hydrolysates and how to effectively regenerate and reuse the used activated carbon has become an obstacle for large scale production of LA and 5-HMF. Ion exchange resin is a common choice for removal of LA, but large amounts of acid and alkali required in the processes of pretreatment and regeneration, also generating a great deal of acid and alkali wastewater, limits its use at an industrial level.24-26 In recent years, hyper-cross-linked polymers, as excellent alternatives for activated carbon and zeolite, are attracting much interest for potential application in efficient

4

ACS Paragon Plus Environment

Page 4 of 45

Page 5 of 45 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

separation and purification of different bio-based chemicals by distinguishing the differences in molecular sizes, shapes, polarities, and so on.15, 16, 33-36 In our previous work, a hyper-cross-linked microporous adsorption resin, SY-01, 37

possessing amide groups and a well-developed pore structure in both the region of

micropores and mesopores as well as a high specific surface area exhibited a great potential for practical application in product recovery of LA from biomass hydrolysate due to its high capacity of LA, high separation selectivity, as well as easy desorption and regeneration.15 The adsorption equilibrium, thermodynamic, kinetic analysis was determined experimentally and simulated.15 Furthermore, the fixed-bed column breakthrough behavior of LA from aqueous solution onto SY-01 resin at various operating conditions, such as feed flow rate (Qf=1.0-5.0 mL/min), initial LA concentration (cf=1.0-10.0 g/L), fixed-bed column length (Lc=5.65-16.96 cm), and fixed-bed column diameter (Dc=1.6-5.5 cm), were systematically investigated.16 The general rate model (GRM) was successfully applied to simulate the breakthrough curves and to determine the related column kinetic constants.16 However, all the above-mentioned studies have been focused on the adsorption behaviors of pure LA, the detailed competitive adsorption isotherms and kinetic dynamics of FA-LA-5-HMF from single, binary and ternary systems in a fixed-bed column of SY-01 resin have not been reported to date. As a matter of fact, since LA released from biomass hydrolysate is accompanied by FA and 5-HMF, it is also important to investigate the binary and ternary adsorption equilibrium, as well as competitive dynamic fixed-bed column adsorption from the viewpoint of industrial application. Herein, as a continuation of our previous work, the main objective of this study is to use a hyper-cross-linked microporous adsorption resin, SY-01, for adsorption equilibrium of FA, LA, as well as 5-HMF from single,

5

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research 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

binary and ternary component mixtures. After that, as a preliminary step in fixed-bed column kinetic analysis, the general rate model (GRM) was developed to simulate the LA, FA and 5-HMF single noncompetitive breakthrough curves, and LA-FA, LA-5-HMF, and FA-5-HMF binary mixture competitive breakthrough curves, as well as the LA-FA-5-HMF ternary mixture competitive breakthrough curves. Furthermore, the model parameters in terms of the axial dispersion, external mass transfer, and pore diffusion coefficients were calculated using a series of empirical equations. 2. Mathematical model 2.1 General rate model (GRM) Various mathematical models used to describe the breakthrough curves during the chromatographic processes have received extensive attention since 1960.38 Among these models, the GRM is one of the most detailed chromatographic models for all kinds of chromatographic operations,39 and the detailed derivation of the GRM was described systematically by Gu et al.40 The GRM model is based on the mass balances of each component in bulk liquid and adsorbent particle phases, which considers axial dispersion (Dax), external film mass transfer (kfilm), internal mass transfer (Dpore), multi-component linear/nonlinear isotherms and sometimes even finite rate of adsorption reaction.39, 41-43 In this paper, the following assumptions are made in order to simplify the GRM complexity; (i) the chromatographic column operates under isothermal conditions; (ii) the resin particles are porous, spherical in shape and uniform in size; (iii) the column void fraction and the linear velocity of liquid phase remain constant along the length of the chromatographic column; (iv) the diffusion in the radial direction in the fixed-bed column is negligible; (v) the diffusion and mass transfer coefficients remain constant; (vi) the equilibrium exists for each species between the pore surface and the liquid phase inside the pores; (vii) the concentration

6

ACS Paragon Plus Environment

Page 6 of 45

Page 7 of 45 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

throughout the fixed-bed column is independent on the bulk liquid phase and intra-particle phase. 2.1 The dimensionless form of the GRM Based on the above assumptions, the following dimensionless partial differential equations (PDEs) can be formulated from the differential mass balances for each species in the bulk liquid and adsorbent phases. ∂Ci

∂τ

εp

∂Ci , p ∂τ

+

∂Ci

∂X

(

)

+ ϕi Ci − Ci , p |R =1 =

+ (1 − ε p )

∂C i*, p ∂τ

= ηi

1 ∂ 2 Ci Pe ∂X 2

1 ∂  2  ∂Ci , p   R   R 2 ∂R   ∂R  

(1)

(2)

With the dimensionless initial and boundary conditions:

τ = 0 , 0 < X < 1 ; Ci = Ci (0, X )

(3)

τ = 0 , 0 < R < 1 ; Ci , p = Ci , p (0, X , R)

(4)

 ci , f (t )  = Pei Ci −  ∂X ci , f  

(5)

X = 0, τ > 0;

∂Ci

X =1, τ > 0;

=0

(6)

=0

(7)

=Bii ( Ci − Ci , p |R =1 )

(8)

R = 0 , τ > 0;

R = 1 ,τ > 0 ;

∂Ci , p ∂R

∂Ci ∂X ∂Ci , p ∂R

where

c*i , p ci , p ci * Ci = , Ci , p = , Ci , p = , ci , f ci , f ci , f

R=

r υt x , τ= , X= rp Lc Lc

7

ACS Paragon Plus Environment

(9)

(10)

Industrial & Engineering Chemistry Research 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

Pei =

υ Lc Di ,ax

, Bii =

ki , film rp

ε p Di , pore

,

ηi =

Page 8 of 45

ε p Di , pore Lc 3Biiηi (1 − ε b ) , ϕi = 2 rp υ εb

(11)

The adsorption isotherm is one of the most important thermodynamic parameters in the chromatography44 and reflects the special relation between the adsorbate concentration and its degree of accumulation on adsorbent surface at constant temperature.45 The adsorption equilibrium is described by the multi-component competitive Langmuir isotherm model:46

c*i , p =

ai ci , p

(12)

1 + ∑ j bj ⋅ c j, p N

The multi-component competitive Langmuir isotherm model in dimensionless form is described as follows: Ci*, p =

Ai Ci , p

(13)

1 + ∑ j =1 B j Ci , p N

2.2 Determination of GRM parameters The relevant GRM parameters in terms of fluid dynamics parameters (Dax), and adsorption kinetics parameters (kfilm and Dpore) were determined through some empirical formulas.47-49 Suzuki and Smith correlation: Di ,ax =0.44Di ,m + 0.83Ud p 1.09 Di , m  v ⋅ d p Wilson and Geankoplis correlation: ki , film = ε ε b d p  b Di ,m Mackie-Meares correlation: Di , pore =

εp Di ,m (2 − ε p ) 2

Wilke-Chang correlation: Di ,m = 7.4 ×10

−8



M i,s ) T

(14)

  

0.33

(15)

(16)

0.5

A

µVi ,0.6 m

(17)

2.3 Numerical simulation of the model equations The detailed method and process of numerical simulation could be referred to 8

ACS Paragon Plus Environment

Page 9 of 45 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

our previous work.16 Meanwhile, the average percentage error (APE) between experimental data and predicted result was used to evaluate quantitatively the accuracy of GRM in the single, binary and ternary components systems.

APE =

1 N

N

ci / ci ,0 exp − ci / ci ,0 cal

i =1

ci / ci ,0 exp

∑|

| ×100%

(18)

3. Materials and methods 3.1 Materials Analytical grade formic acid (FA, CH2O2, molecular

weight 46.02,

purity≥98.0%), levulinic acid (LA, C5H8O3, molecular weight 115.11, purity≥98.0%) and

5-Hydroxymethylfurfural

(5-HMF,

C6H6O3,

molecular

weight

126.11,

purity≥98.0%) used in this study were purchased from Sinopharm Chemical Reagent Company (Shanghai, China). All solutions used in this study were prepared by measuring and dissolving the required amounts of solutes in deionized water. The hyper-cross-linked resin adsorbent, SY-01, was supplied by Guangzhou Institute of Energy Conversion, Chinese Academy of Science (Guangzhou, China). The detail characterization and analysis processes of SY-01 in terms of scanning electron microscope (SEM), surface area (SBET), pore volume (Vp), average pore diameter (dp), elemental analysis (EA), as well as Fourier transform infrared (FTIR) spectroscopy were made in our previous work37 and the typical properties of SY-01 are also listed in Table 1. Prior to packing the fixed-bed column, the SY-01 resin was pretreated with 70% (V/V) ethanol for 2 h and then repetitively washed to neutral pH with deionized water.

3.2 Methods 3.2.1 Static equilibrium adsorption experiments Batch equilibrium adsorption experiments were carried out to investigate the single/binary/ternary-component equilibrium adsorption isotherms of LA (or other 9

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research 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 45

solutes, FA and 5-HMF) from single/binary/ternary-component systems onto SY-01 resin at the temperature of 298 ± 1 K. In typical batch adsorption experiments, 1.0 g of wet SY-01 resin was added to 50 mL single/binary/ternary-components systems of different initial concentrations (Table S1) in 100 mL Erlenmeyer flasks and kept shaking at 150 rpm for 6 h at 298 ± 1 K in constant temperature incubator shaker (ZQZY-80BS, Shanghai Zhichu Instrument Co., Ltd., Shanghai, China). The adsorption time was kept constant to ensure the adsorption reach the equilibrium, which was confirmed by previous kinetic investigation.15 The experiments were carried out in triplicate, and samples were withdrawn from the supernatant fluid with a syringe and filtered through 0.45 µm membranes to determine the concentrations of FA, LA and 5-HMF using high performance liquid chromatography (HPLC). The experimental data in this section were performed in triplicate to ensure for consistency of the results. The uptake of individual solute i adsorbed by SY-01 resin from single/binary/ternary-component systems at equilibrium (qe, mg/g wet resin) was calculated by the following equation:

qi ,e =

(ci ,0 − ci ,e ) ⋅V m

(19)

3.2.2 Dynamic fixed-bed column adsorption experiments Dynamic fixed-bed column adsorption experiments were carried out to evaluate column performance for adsorption of single/binary/ternary components (FA, LA and 5-HMF) on SY-01 resin. Experiments were conducted in a glass column (16 mm in diameter, 200 mm in length, Shanghai HuXi Analysis Instrument Factory Co., Ltd., Shanghai, China), which allow for packing about 30 g of the SY-01 resin. The column was equipped with a water jacket connected to a super-heated water bath (BBL111C, Yamato Scientific ChongQing Co., Ltd., Chongqing, China) to keep the temperature

10

ACS Paragon Plus Environment

Page 11 of 45 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

constant. Fixed-bed operation was performed isothermally at 298 ± 1 K and the feed stream containing a desired initial amount of one or more components (FA, LA and 5-HMF) was pumped into the column with SY-01 resin in down-flow mode as shown in Fig. 1. The flow rate was controlled by a constant-flow peristaltic pump (BT100-2J, Hebei, China). Samples were collected by automatic sampling instrument (BS-100A, Shanghai HuXi Analysis Instrument Factory Co., Ltd., Shanghai, China) at the column outlet at pre-determined time intervals and the concentrations of raffinate were periodically analyzed by HPLC. All the experiments in this section were conducted in triplicate and the average results are reported. In fixed-bed studies, the adsorption performance of a column is typically evaluated by analyzing the form of cout,t versus time or volume throughput curves (i.e., breakthrough curves). The area above this curve, in the case of symmetric curves only, is equal to Qf×t1/2×cf. The contribution of non-selective fluid volumes was removed from the experimental breakthrough curves by subtracting the residence time in the surface layer volume of the unit (ts=Vs/Qf), the residence time in the void volume of the fixed-bed column (tR=εb×Vc/Qf) and the residence time in the pore volume of the fixed-bed column (tp=εp×(1-εb)×Vc/Qf).50 The saturated adsorbed quantity of FA, LA and 5-HMF per mass of SY-01 resin (qi) was determined using the following equation:

qi =

ci , f ⋅ [ti ,1/2 − (ti , s + ti , R + ti , p )] ⋅ Qi , f m

(20)

However, the saturated amount of FA, LA and 5-HMF adsorbed on SY-01 resin (qi) in cases in which the breakthrough curve is not symmetric was calculated by numerical integration of the area above the breakthrough curve using the following equation:

11

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research 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

qi =

[∫

ttotal

0

(ci , f − ci ,out ,t )dt − ci , f ⋅ (ti ,s + ti , R + ti , p )] ⋅ Qi , f

Page 12 of 45

(21)

m

3.2.3 Analytical method The concentrations of FA, LA and 5-HMF were quantified by HPLC (Waters 2685 systems, Waters Corp., USA) equipped with a refractive index detector (Waters 2414). An Aminex HPX-87H anion exchange column (300 mm × 7.8 mm, Bio-Rad Corp., USA) was used at 328 K with 5 mM sulfuric acid as the mobile phase (0.5 mL/min).51 The quantity of sample was 10 µL. External standards were established for calibration.

4. Results and discussion 4.1 Adsorption isotherm of FA, LA and 5-HMF in the single component system The single-component adsorption isotherms of FA, LA and 5-HMF onto SY-01 resin measured at 298 ± 1 K are displayed in Fig. 2 and fitted by both Langmuir isotherm (Eq. 22) and Freundlich isotherm model (Eq. 24). The corresponding isotherm models constants and correlation coefficients are listed in Table S2. Obviously, the regression coefficients indicates that the adsorption equilibrium data of each solute onto SY-01 resin are well-described by Langmuir isotherm model, confirming the monolayer coverage of FA, LA and 5-HMF onto SY-01 resin. The estimated maximal uptakes of FA, LA and 5-HMF derived from the Langmuir isotherm model were 7.54, 103.51, and 107.73 mg/g wet resin. Furthermore, the values of RL (calculated by Eq. 23) were between 0 and 1, suggesting that the adsorption was favorable for both of these samples.52 These deviations are mainly caused by adsorption affinity of each adsorbate on SY-01 resin and can be attributed to the hydrophobic interaction occurring between the aromatic building blocks and alkyl chain of LA as well as furan ring of 5-HMF. Generally, hydrophobic interaction

12

ACS Paragon Plus Environment

Page 13 of 45 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

was facilitated by an increased carbon content of adsorbate, and thus increased the adsorption loading.28 Furthermore, hydrogen bonding between the amide groups of SY-01 resin particle and carboxyl functional group of LA and aldehyde functional group 5-HMF also played an important role in the adsorption.15 Clearly, SY-01 resin shows the strongest affinity for 5-HMF, followed by LA, and then FA in the experimental investigation of concentration range. It is interesting that this trend follows the pKa values of the adsorbate (5-HMF:12.89 > LA:4.5 > FA:3.74) .51, 53 Consequently, SY-01 resin with high surface area, large micropores volume, as well as amide groups is suited for the applications in removal of LA and 5-HMF from biomass hydrolysates. Langmuir isotherm model:54 qi ,e =

ki , L ⋅ qi ,m ⋅ ci ,e 1 + ki , L ⋅ ci ,e

=

ai ⋅ ci ,e 1 + bi ⋅ ci ,e

(22)

The affinity between the adsorbate and the adsorbent was described using the separation factor, RL, which can be calculated using the following equation:55

1 1 + k L c0

(23)

qi ,e = ki , F ci ,e1/ n

(24)

RL = Freundlich isotherm model:56

4.2 Competitive adsorption isotherm of FA, LA and 5-HMF in the binary and ternary components system Recently, many researchers have reported the routes for production of LA from biomass by catalytic conversion technology.12, 57 However, the routes mostly reported via a conversion of polymeric carbohydrates to hexoses sugars (e.g. glucose and fructose). Of the hexoses, the glucose is first isomerized to fructose, and then dehydrated to 5-HMF intermediate by Bronsted acid catalysts, which is in turn 13

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research 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

rehydrated into LA and FA. Accordingly, it is also important to investigate the competitive adsorption isotherms of FA-LA-5-HMF binary and ternary mixtures, as well as competitive dynamic fixed-bed column adsorption from the viewpoint of industrial application. Fig. 3 displays the binary and ternary-components competitive adsorption isotherms of FA, LA and 5-HMF onto SY-01 resin measured at 298 ± 1 K and fitted by the competitive Langmuir isotherm model. The specific model parameters are listed in Table S3. It can be seen from Fig. 3 and Table S3 that the predicted values of the model are in agreement with the experimental data, suggesting that the competitive Langmuir adsorption isotherm model can be used to describe the competitive adsorption mechanism of FA, LA and 5-HMF in both binary and ternary components mixture system. For the sake of comparison, the single-component adsorption isotherms of FA, LA and 5-HMF are also presented in Fig. 3. As illustrated in Fig. 3A and 3B, the addition of FA (weak adsorbed component) did not result in the reduction of adsorption capacities of LA and 5-HMF (strong adsorbed components) onto SY-01 resin. This is because the pKas of LA and 5-HMF at 298 K are 4.5 and 12.89,53 respectively, and the structures of LA and 5-HMF are stable under acidic conditions. Contrarily, compared with single system, the uptakes of LA and 5-HMF in binary components system (FA-LA, FA-5-HMF) were slightly increased in the presence of FA. This interesting phenomenon could be attributed to the fact that the effect of hydrogen bonding on the adsorption of LA and 5-HMF in acidic condition was significant, resulting in increasing the adsorption capacity. Moreover, the amount of FA onto SY-01 in binary components system was greatly reduced compared with its single component system due to the competitive effect. Fig. 3C shows competitive LA-5-HMF adsorption onto SY-01 resin. As could be expected, the amounts of LA

14

ACS Paragon Plus Environment

Page 14 of 45

Page 15 of 45 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

and 5-HMF adsorbed from LA-5-HMF binary component system and FA-LA-5-HMF ternary system were dropped. Compared to noncompetitive adsorption, the adsorption capacities of LA and 5-HMF in LA-5-HMF binary system were dropped about 17.33% and 65.67%, respectively. The similar behavior was reported by Dornath and Fan30 when they investigated the adsorption isotherm of fructose, LA, and 5-HMF onto activated carbon, in which the capacity of LA adsorbed in the presence of 5-HMF decreased by 50% due to the competition of LA and 5-HMF. A similar trend has been reported by Chaline Detoni et al.36 who studied the competitive adsorption of LA and 5-HMF using nonpolar nanoporous hyper-cross-linked polymers (HCP) as adsorbent and also found a similar behavior trend with Dornath and Fan. It worth noting that the adsorption capacities of FA, LA and 5-HMF in FA-LA-5-HMF ternary mixtures system were dropped about 80.47%, 15.85% and 63.56%, respectively. It is proved again that the presence of FA has a synergistic effect on the adsorption of LA and 5-HMF onto SY-01 resin in binary or ternary mixtures system. 4.3 Simulation of the adsorption behaviors of FA, LA and 5-HMF single system Fig. 4 displays the noncompetitive breakthrough curves (effluent/initial feed concentration versus volume) of FA, LA, and 5-HMF on the fixed-bed column packed with SY-01 resin at 298 ± 1 K in the single component system, along with the simulation curves predicted by GRM. The breakthrough and saturated points were defined as the phenomenon when the effluent concentration at the outlet of fixed-bed column is approximately 3-5% and 95-100% of the initial feed concentration, respectively. It can be observed that the simulation results were found to be in accordance with the experimental data fairly well, suggesting the GRM was proved to be right and could be used to predict the noncompetitive adsorption behaviors on the SY-01 resin. The related model parameters, composition of feed mixtures, as well as

15

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research 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

fixed-bed column characteristic are listed in Table S4. Also can been seen from Fig. 4 that FA shows the weakest adsorption affinity to the SY-01 resin, the fixed-bed reaches breakthrough and saturated sate quickly and breakthrough profile for FA is sharp. The breakthrough and saturated volume of FA were 2.500 and 3.333 bed volume (BV), respectively. The breakthrough and saturated adsorption capacity of FA were 6.44 and 7.30 mg/g wet resin, respectively. Whereas the adsorption affinity of 5-HMF to the SY-01 resin is strongest, the largest volume required to reach the state of penetration and saturation. The breakthrough and saturated volume of 5-HMF were 20.833 and 45.833 BV, respectively. The breakthrough and saturated adsorption capacity of 5-HMF were 32.72 and 56.54 mg/g wet resin, respectively. Since the adsorption affinity of LA onto SY-01 resin is between FA and 5-HMF, the breakthrough and saturation volumes of LA are also between FA and 5-HMF. The breakthrough and saturated volume of LA were 8.333 and 16.667 BV, respectively. The breakthrough and saturated adsorption capacity of 5-HMF were 60.25 and 80.74 mg/g wet resin, respectively. Overall, the above experimental results were consistent of with the results obtained from the noncompetitive adsorption equilibrium data (see Fig. 2). 4.4 Simulation of the adsorption behaviors of FA and LA in binary components mixture systems One of the objectives of this work was to validate the competitive breakthrough curves simulated by the extension of the multicomponent competitive GRM based on the coefficients determined for single component system. Binary FA-LA competitive breakthrough curves were carried out for mixtures with FA and LA in a fixed-bed column. The experimental and predicted results are plotted in Fig. 5 and the detailed fixed-bed operating conditions as well as model parameters are described in Table S4.

16

ACS Paragon Plus Environment

Page 16 of 45

Page 17 of 45 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

There is a good agreement between the experimental competitive adsorption breakthrough curves of FA and LA in binary components mixtures and the simulated data predicted by extended multicomponent GRM model. Since FA is only slightly absorbed (weak adsorbed component), it breaks through the fixed-bed column very fast, and then turns to the LA. The breakthrough volumes of FA and LA were 1.667 and 8.333 BV, respectively. Compared with the noncompetitive breakthrough curves of FA and LA in single component system, the breakthrough volume of FA decreased from 2.500 to 1.667 BV, however, the breakthrough volume of LA was remained constant, indicating that the FA interaction with the SY-01 resin was slight enough that it did not interfere with the adsorption of LA. This result is consistent with the competitive isotherm of FA-LA in binary components system (see Fig. 3A) determined in the static method. The adsorption amount of FA and LA at the breakthrough state were 4.29 and 59.14 mg/g wet resin. In addition, it is remarkable that a significant bump exists in the transient breakthrough curve of FA between the saturated sate of FA and saturated sate of LA, which is a typical characteristic of the competitive adsorption.58-60 In our work, the transient concentrations of FA at the outlet of fixed-bed column were abnormally larger than initial FA concentration after FA reaches the adsorption equilibrium. Then, the transient concentrations of FA gradually decreased as the state of LA changed from breakthrough to saturation. We attribute this interesting performance to the existence of the most strongly retained adsorbate and displacement occurs at the active sites in SY-01 resin. Due to the occurrence of competitive adsorption, the saturated adsorption capacity of FA onto SY-01 resin was reduced to 2.30 mg/g wet resin. However, the saturated adsorption amount of LA onto SY-01 resin was kept constant, 79.25 mg/g wet resin. 4.5 Simulation of the adsorption behaviors of FA and 5-HMF in binary component

17

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research 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

systems The experimental and model-predicated breakthrough curves for FA and 5-HMF in binary component systems were obtained in fixed-bed column packed with SY-01 resin at an initial FA and 5-HMF concentration of 1.825 and 1.114 g/L, respectively. The detailed operating conditions in terms of bed height and feed flow rate are listed in Table S4. From the Fig. 6, it can be observed that the model simulated results complied fairly well with the experimental data, thereby reflecting the correctness and reliability of the extended multicomponent GRM model. As is evident from the Fig. 6, the breakthrough volume of FA in FA-5-HMF binary components mixture was found to be significantly lower (by 66.68%) compared to noncompetitive single component adsorption, whereas, the penetration volume of 5-HMF remained unchanged, indicating that the presence of FA has no negative effect on the adsorption of 5-HMF onto SY-01 resin. Furthermore, due to the fact that the affinity of 5-HMF onto SY-01 is stronger than FA, a feature of small roll-up of FA breakthrough curve was observed. It is noteworthy that the convexity degree of FA breakthrough curve in FA-5-HMF binary components mixture is lower than that in FA-LA binary components mixture. This interesting phenomenon is mainly caused by the following two reasons. One reason is the affinity of 5-HMF is greater than that of LA, leading to a greater reduction in the penetration volume of FA. In other words, the competitive adsorption force of 5-HMF with FA is stronger than of LA with FA. Another reason is that the sustained time of saddle-backing phenomenon of FA breakthrough curves in FA-5-HMF binary component mixture system is much longer than that in FA-LA binary component mixture system. Moreover, the adsorption amounts of FA and 5-HMF at the breakthrough state were 2.14 and 32.69 mg/g wet resin. Meanwhile, the saturated adsorption capacities of FA and 5-HMF were 4.24 and 55.84 mg/g wet resin.

18

ACS Paragon Plus Environment

Page 18 of 45

Page 19 of 45 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

This is consistent with competitive isotherm of FA-5-HMF in binary components mixture system (see Fig. 3B). 4.6 Simulation of the adsorption behaviors of LA and 5-HMF in binary component systems The experimental (data points) and predicated (solid lines) competitive LA-5-HMF adsorption breakthrough curves of LA and 5-HMF in fixed-bed column packed with SY-01 resin (see Run 6 in Table S4, cLA=5.133 g/L, c5-HMF=1.160 g/L) are plotted in Fig. 7. The experimental and simulated values of the LA and 5-HMF competitive breakthrough curves in Fig. 7 shows the prediction of extended multi-components GRM are quite well. The parameters of GRM in terms of Pei, Bii, ηi, and φi was calculated and listed in Table S4. The adsorption breakthrough curves of LA and 5-HMF from single component solution were also displayed in Fig. 7 as a comparison. Notably, the capacities of the LA and 5-HMF from the binary component solutions were less than that from the individual component solution, revealing an outstanding competition for available adsorption sites between LA and 5-HMF.36 The breakthrough volume of 5-HMF in LA-5-HMF binary components mixture system was significantly reduced from 20.833 BV to 12.50 BV. However, the penetration volume of LA was slightly decreased from 8.333 BV to 7.500 BV. In other words, when in the presence of LA (5.133 g/L), the amount of 5-HMF adsorbed onto SY-01 at penetration state was decreased from 32.72 to 20.43 mg/g wet resin. Accordingly, the capacity of LA adsorbed onto SY-01 at breakthrough state was decreased from 60.25 to 54.22 mg/g wet resin in the presence of 5-HMF (1.160 g/L). Furthermore, the competitive adsorption between LA and 5-HMF also resulted in a notable roll-up phenomenon at the transient breakthrough curve of LA (see Fig. 7). This roll-up phenomenon can be ascribed to the fact that the 5-HMF molecules competed with the

19

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research 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

adsorbed LA on the adsorptive sites, resulting in plenty of LA molecules were discharged and leading the higher effluent concentration of LA. This result indicated that the adsorption affinity of 5-HMF onto SY-01 was stronger than that of LA, which was in agreement with the experimental results of the LA-5-HMF competitive adsorption isotherm in binary components mixture (see Fig. 3C). Consequently, the amounts of LA and 5-HMF adsorbed onto SY-01 resin at saturate state were 63.90 and 40.85 mg/g wet resin, respectively. Compare with the single component system (Run 2 and 3 in Table S4), the amounts of LA and 5-HMF display 20.85% and 27.75% decrease, respectively. 4.7 Simulation of the adsorption behaviors of FA, LA and 5-HMF in ternary components systems The experimental (data points) and predicated (solid lines) results of the ternary FA-LA-5-HMF breakthrough curve (see Run 7 in Table S4, cFA=1.816 g/L, cLA=5.106 g/L, c5-HMF=1.173 g/L) are reported in Fig. 8. It is evident in Fig. 8 that the extended multicomponent competitive GRM can successfully simulated the experimental data. The adsorption breakthrough of 5-HMF in the presence of FA and LA follows a similar behavior presented by adsorption of 5-HMF in the presence of LA. Clearly, the first adsorbate to breakthrough the fixed-bed column is FA, followed by LA, and finally 5-HMF. During the early stage of the fixed-bed column operation (cumulative effluent volume < 0.833 BV), the concentrations of FA, LA, and 5-HMF at the effluent were very low, which is due to the fact that the SY-01 resin provided a lot of adsorption sites for the incoming FA, LA and 5-HMF and all the FA, LA and 5-HMF molecules were adsorbed onto the SY-01 resin. During the intermediate stage of the fixed-bed column operation (0.833 BV < cumulative effluent volume < 33.333 BV), the adsorbed FA molecules were replaced by the incoming LA and 5-HMF molecules

20

ACS Paragon Plus Environment

Page 20 of 45

Page 21 of 45 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

because of weaker affinity of FA to the SY-01. While the adsorbed LA molecules were replaced by the incoming 5-HMF molecules because of the affinity of LA to the SY-01 resin was weaker than 5-HMF. Consequently, the competitive adsorption resulted in two significant bumps exist in the transient breakthrough curve of FA and LA, and the effluent FA and LA concentrations higher than their initial feed concentration. The existence of such a high concentration bumps zone was also found by Chern and Chien.61 During the final stage of fixed-bed column operation (cumulative effluent volume > 33.333 BV), all the adsorption sites were occupied with the incoming FA, LA and 5-HMF molecules, thus the SY-01 resin column lost its adsorption capacity. The effluent FA, LA and 5-HMF concentrations therefore equaled to the initial feed concentrations. The final amounts of FA, LA, and 5-HMF in ternary components mixtures were calculated by Eq. 21, and the capacities of FA, LA, and 5-HMF onto SY-01 resin were 2.47, 68.24 and 40.90 mg/g wet resin, which correspond to the values obtained from the ternary components system competitive adsorption isotherm (see Fig. 3D). The presence of FA makes the saturated adsorption capacities of LA and 5-HMF slightly increase. The detailed analysis of this interesting phenomenon has been discussed in Section 4.2. For a better understanding of the competitive adsorption process, Fig. 9 schematically displays the development of the transient breakthrough curves of FA-LA-5-HMF at different fixed-bed column positions. Clearly, LA and 5-HMF were adsorbed onto SY-01 resin initially, and the weakest adsorbate, FA, propagated with the faster moving front whereas LA and 5-HMF are preferentially retained in the region of the fixed-bed inlet. As this region is saturated with LA (the second strong adsorbate), this adsorbate trickles along the column so that most of FA that is initially adsorbed is displaced and released to the solution, resulting in the outlet

21

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research 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

concentrations higher than the concentration of the initial feed solution. Moreover, the entire breakthrough curves shifted from left to right as the position of fixed-bed column increased, especially, the sizes of bump region of FA and LA increased with the increase of the length of the fixed-bed column.

5. Conclusions This work is mainly focused on systematic investigation of the adsorption behavior of LA as well as its intermediate product (5-HMF) and by-product (FA) onto a microporous hyper-cross-linked adsorption resin (SY-01) from single component and multi-component mixtures solution using both static adsorption and fixed-bed column adsorption approach. The results revealed that the noncompetitive adsorption equilibrium data of each solute onto SY-01 resin were well-described by Langmuir isotherm model. Meanwhile, the estimated maximal amounts of FA, LA and, 5-HMF derived from the Langmuir isotherm model were 7.54, 103.51, and 107.73 mg/g wet resin, indicating that the adsorption affinity of FA, LA, and 5-HMF to the SY-01 resin follows the order: 5-HMF>LA>FA. Contrarily, compared with single system, the uptakes of LA and 5-HMF in binary components system (FA-LA, FA-5-HMF) were slightly increased in the presence of FA. However, the amounts of LA and 5-HMF adsorbed from LA-5-HMF binary component system and FA-LA-5-HMF ternary system were dropped. Furthermore, the GRM incorporated with parameters (Pei, Bii, ηi, and φi) as well as the noncompetitive single component and competitive multi-component Langmuir isotherm was successfully developed to predict the breakthrough curves of FA-LA-5-HMF single, binary, as well as ternary components mixtures. Most importantly, the verified chromatographic model can be used to optimize and design the process of adsorption and separation, saving plenty of manpower and material resources. Due to the superior properties of SY-01 resin, it is recognized that the synthesis of entirely novel hyper-cross-linked polymers would benefit fundamental research and provide new opportunities in emerging technological areas such as industrial hydrolysate detoxification or separation and purification of bio-based platform chemicals.

Supporting Information This information is available free of charge via the Internet at http://pubs.acs.org. 22

ACS Paragon Plus Environment

Page 22 of 45

Page 23 of 45 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

Table S1: The different initial concentrations of FA, LA, and 5-HMF in single/binary/ternary-components systems in typical bath adsorption experiment Table S2: Langmuir and Freundlich isotherm model parameters and correlation coefficients for FA, LA and 5-HMF in single component system Table S3: Isotherm parameters for FA, LA and 5-HMF adsorption on SY-01 resin in binary and ternary components system Table S4: Experimental conditions for column runs and some parameters of the column estimated from noncompetitive and competitive breakthrough experiments.

AUTHOR INFORMATION Corresponding Authors *X. Lin. E-mail: [email protected]. Tel. /fax: +86- 20-39322172. ORCID Xiaoqing Lin: 0000-0002-1751-5348

Notes The authors declare no competing financial interest.

Acknowledgments This work was supported by the financial support of the Project of National Natural Science Foundation of China (51508547), Pearl River S&T Nova Program of Guangzhou (201710010096), the Science and Technology Planning Project of Guangdong Province, China (2017A010103043, 2016A010104009), the Science and Technology Project of Huaian (HAS2016023), “One-Hundred Young Talents” Program of Guangdong University of Technology (220413185) and the National Training Programs of Innovation and Entrepreneurship for Undergraduates (201811845015, xj201811845070, xj201811845083).

23

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research 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

Ac = cross-section area of the fixed-bed column, cm2 a = the multi-component competitive Langmuir isotherm constant, mL/g b = the multi-component competitive Langmuir isotherm constant, L/g c = concentration in the liquid phase, g/L cf = the inlet adsorbate concentration, g/L dp = particle size of the resin, cm Dax = axial dispersion coefficient, cm2/min Dm = the molecular diffusivity, cm2/min Dpore = pore diffusion coefficient, cm2/min kfilm = the external mass transfer coefficient, cm/min Lc = the length of the fixed-bed column, cm m = the mass of the resin, g Ms = the molecular weight of the solvent, g/mol Pe = the Peclet number q = loading concentration in the stationary phase, mg/g wet resin

q = the average adsorbed phase concentration, mg/g wet resin Qf = volumetric flow rate of the liquid phase, mL/min rp = particle radius, cm t = time, min ts = the residence time in the surface layer volume of the unit (tS=VS/Qf), min tp = the residence time in the pore volume of the fixed-bed column (tp=εp×(1-εb)×Vc /Qf), min T = the temperature, K tR = retention time (tR =εbVc/Qf), min 24

ACS Paragon Plus Environment

Page 24 of 45

Page 25 of 45 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

x = the axial coordinate, cm Vc = volume of the fixed-bed column, mL Vs = volume of the surface layer upper the fixed-bed column, mL Vm = the molar volume of the liquid solute at its normal boiling point, cm3/(g mol) Greek Letters αA = solute-solvent molecular interactions (for water αA is given as 2.6) ρp = apparent resin density, g/L ρs = Skeletal resin density, g/L εb = bed column porosity εp = particle porosity ν = interstitial velocity of the liquid phase (v=Qf/Acεb), cm/min µ = viscosity of the solvent, cP τ = the dimensionless time Subscripts FA = formic acid LA = levulinic acid 5-HMF = 5-Hydroxymethylfurfural i = the adsorbate species in = inlet out = the effluent solution s = solvent t= time

25

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research 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

References (1) Steen, E. J.; Kang, Y. S.; Bokinsky, G.; Hu, Z. H.; Schirmer, A.; McClure, A.; del Cardayre, S. B.; Keasling, J. D. Microbial Production of Fatty-Acid-Derived Fuels and Chemicals from Plant Biomass. Nature 2010, 463, 559. (2) Zhang, Y. H. P.; Lynd, L. R. Toward an Aggregated Understanding of Enzymatic Hydrolysis of Cellulose: Noncomplexed Cellulase Systems. Biotechnol. Bioeng. 2004, 88, 797. (3) Zhao, X. B.; Cheng, K. K.; Liu, D. H. Organosolv Pretreatment of Lignocellulosic Biomass for Enzymatic Hydrolysis. Appl. Microbiol. Biotechnol.

2009, 82, 815. (4) Rinaldi, R.; Schüth, F. Acid Hydrolysis of Cellulose as the Entry Point into Biorefinery Schemes. ChemSusChem 2009, 2, 1096. (5) Yan, K.; Jarvis, C.; Gu, J.; Yan, Y. Production and Catalytic Transformation of Levulinic Acid: A Platform for Speciality Chemicals and Fuels. Renew. Sust. Energ. Rev. 2015, 51, 986. (6) Mukherjee, A.; Dumont, M.-J.; Raghavan, V. Review: Sustainable Production of Hydroxymethylfurfural and Levulinic Acid: Challenges and Opportunities. Biomass Bioenerg. 2015, 72, 143. (7) Himmel, M. E.; Ding, S. Y.; Johnson, D. K.; Adney, W. S.; Nimlos, M. R.; Brady, J. W.; Foust, T. D. Biomass Recalcitrance: Engineering Plants and Enzymes for Biofuels Production. Science 2007, 315, 804. (8) Corma, A.; Iborra, S.; Velty, A. Chemical Routes for the Transformation of Biomass into Chemicals. Chem. Rev. 2007, 107, 2411. (9) van Putten, R.-J.; van der Waal, J. C.; de Jong, E.; Rasrendra, C. B.; Heeres, H. J.; de Vries, J. G. Hydroxymethylfurfural, A Versatile Platform Chemical Made from

26

ACS Paragon Plus Environment

Page 26 of 45

Page 27 of 45 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

Renewable Resources. Chem. Rev. 2013, 113, 1499. (10) Morone, A.; Apte, M.; Pandey, R. A. Levulinic Acid Production from Renewable Waste Resources: Bottlenecks, Potential Remedies, Advancements and Applications. Renew. Sust. Energ. Rev. 2015, 51, 548. (11) Ya’aini, N.; Amin, N. A. S.; Asmadi, M. Optimization of Levulinic Acid from Lignocellulosic Biomass Using A New Hybrid Catalyst. Bioresour. Technol. 2012, 116, 58. (12) Zuo, Y.; Zhang, Y.; Fu, Y. Catalytic Conversion of Cellulose into Levulinic Acid by a Sulfonated Chloromethyl Polystyrene Solid Acid Catalyst. ChemCatChem 2014, 6, 753. (13) Alonso, D. M.; Gallo, J. M. R.; Mellmer, M. A.; Wettstein, S. G.; Dumesic, J. A. Direct Conversion of Cellulose to Levulinic Acid and Gamma-Valerolactone Using Solid Acid Catalysts. Catal. Sci. Technol. 2013, 3, 927. (14) Weingarten, R.; Cho, J.; Xing, R.; Conner, W. C.; Huber, G. W. Kinetics and Reaction Engineering of Levulinic Acid Production from Aqueous Glucose Solutions. ChemSusChem 2012, 5, 1280. (15) Lin, X. Q.; Huang, Q. L.; Qi, G. X.; Xiong, L.; Huang, C.; Chen, X. F.; Li, H. L.; Chen, X. D. Adsorption Behavior of Levulinic Acid onto Microporous hyper-cross-linked Polymers in Aqueous Solution: Equilibrium, Thermodynamic, Kinetic Simulation and Fixed-Bed Column Studies. Chemosphere 2017, 171, 231. (16) Lin, X. Q.; Huang, Q. L.; Qi, G. X.; Shi, S. L.; Xiong, L.; Huang, C.; Chen, X. F.; Li, H. L.; Chen, X. D. Estimation of Fixed-Bed Column Parameters and Mathematical Modeling of Breakthrough Behaviors for Adsorption of Levulinic Acid from Aqueous Solution Using SY-01 Resin. Sep. Purif. Technol. 2017, 174, 222. (17) Uslu, H.; Kırbaşlar, Ş. Đ. Equilibrium Studies of Extraction of Levulinic Acid by

27

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research 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 28 of 45

(Trioctylamine (TOA) + Ester) Solvents. J. Chem. Eng. Data 2008, 53, 1557. (18) Ayoub, P. M. Process for The Reactive Extractive Extraction of Levulinic Acid. U.S. Patent 7, 378, 549, 2008. (19) Fernandes, D. R.; Rocha, A. S.; Mai, E. F.; Mota, C. J. A.; Teixeira da Silva, V. Levulinic Acid Esterification with Ethanol to Ethyl Levulinate Production over Solid Acid Catalysts. Appl. Catal., A 2012, 425–426, 199. (20) Den Boestert, J. L. W. C.; Haan, J. P.; Nijmeijer, A. Process for Permeation Enhanced Reactive Extraction of Levulinic Acid. U.S. Patent 7, 501, 062, 2009. (21) Habe, H.; Kondo, S.; Sato, Y.; Hori, T.; Kanno, M.; Kimura, N.; Koike, H.; Kirimura, K. Electrodialytic Separation of Levulinic Acid Catalytically Synthesized from Woody Biomass for Use in Microbial Conversion. Biotechnol. Prog. 2017, 33, 448. (22) Xiao, S.; Liu, B.; Wang, Y.; Fang, Z.; Zhang, Z. Efficient Conversion of Cellulose

into

Biofuel

Precursor

5-hydroxymethylfurfural

in

Dimethyl

Sulfoxide–Ionic Liquid Mixtures. Bioresour. Technol. 2014, 151, 361. (23) Datta, D.; Uslu, H. Adsorption of Levulinic Acid from Aqueous Solution by Amberlite XAD-4. J. Mol. Liq. 2017, 234, 330. (24) Liu, B. J.; Hu, Z. J.; Ren, Q. L. Single-Component and Competitive Adsorption of Levulinic/Formic Acids on Basic Polymeric Adsorbents. Colloids Surf., A 2009, 339, 185. (25) Liu, B. J.; Zeng, L.; Mao, J. W.; Ren, Q. L. Simulation of Levulinic Acid Adsorption in Packed Beds Using Parallel Pore/Surface Diffusion Model. Chem. Eng. Technol. 2010, 33, 1146. (26) Liu, B. J.; Liu, S. W. Adsorption of Levulinic Acid from Aqueous Solution onto Basic Polymeric Adsorbent: Experimental and Modeling Studies. Sep. Sci. Technol.

28

ACS Paragon Plus Environment

Page 29 of 45 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

2011, 46, 2391. (27) Qureshi, N.; Hughes, S.; Maddox, I. S.; Cotta, M. A. Energy-Efficient Recovery of Butanol from Model Solutions and Fermentation Broth by Adsorption. Bioprocess Biosyst. Eng. 2005, 27, 215. (28) Nielsen, L.; Bandosz, T. J. Analysis of the Competitive Adsorption of Pharmaceuticals on Waste Derived Materials. Chem. Eng. J. 2016, 287, 139. (29) Ren, Q.; Liu, B.; Su, B.; Yang, Y. A Method of Separating Levulinic Acid by Active Aarbon. Chinese Patent CN200710067549.2, 2007. (30) Dornath, P.; Fan, W. Dehydration of Fructose into Furans over Zeolite Catalyst using Carbon Black as Adsorbent. Microporous Mesoporous Mater. 2014, 191, 10. (31) Ijzer, A. C.; Vriezekolk, E.; Ðekic Zivkovic, T.; Nijmeijer, K. Adsorption Kinetics of DowexTM OptiporeTM L493 for the Removal of the Furan 5-hydroxymethylfurfural from Sugar. J. Chem. Technol. Biotechnol. 2016, 91, 96. (32) Rolly Gonzales, R.; Hong, Y.; Park, J.-H.; Kumar, G.; Kim, S.-H. Kinetics and Equilibria

of

5-hydroxymethylfurfural

(5-HMF)

Sequestration

from

Algal

Hydrolyzate using Granular Activated Carbon. J. Chem. Technol. Biotechnol. 2016, 91, 1157. (33) Long, C.; Liu, P.; Li, Y.; Li, A.; Zhang, Q. Characterization of Hydrophobic Hypercrosslinked Polymer as an Adsorbent for Removal of Chlorinated Volatile Organic Compounds. Environ. Sci. Technol. 2011, 45, 4506. (34)

Huang,

J.;

Wang,

X.;

Wang,

X.;

Huang,

K.

Hyper-Cross-Linked

Polystyrene-co-divinylbenzene Resin Modified with Acetanilide: Synthesis, Structure, and Adsorptive Removal of Salicylic Acid from Aqueous Solution. Ind. Eng. Chem. Res. 2011, 50, 2891. (35) Liao, P.-Q.; Huang, N.-Y.; Zhang, W.-X.; Zhang, J.-P.; Chen, X.-M. Controlling

29

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research 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

Guest Conformation for Efficient Purification of Butadiene. Science 2017, 356, 1193. (36) Detoni, C.; Gierlich, C. H.; Rose, M.; Palkovits, R. Selective Liquid Phase Adsorption of 5-Hydroxymethylfurfural on Nanoporous Hyper-Cross-Linked Polymers. ACS Sustainable Chem. Eng. 2014, 2, 2407. (37) Lin, X.; Xiong, L.; Qi, G.; Shi, S.; Huang, C.; Chen, X.; Chen, X. Using Butanol Fermentation Wastewater for Biobutanol Production after Removal of Inhibitory Compounds by Micro/Mesoporous Hyper-Cross-Linked Polymeric Adsorbent. ACS Sustainable Chem. Eng. 2015, 3, 702. (38) Golshan-Shirazi, S.; Guiochon, G. The Equilibrium-Dispersive Model of Chromatography. In: Dondi F., Guiochon G., Eds.; Theoretical Advancement in Chromatography and Related Separation Techniques. NATO ASI Series (Series C: Mathematical and Physical Sciences), vol 383. Springer: Dordrecht, 1992; pp 35-59. (39) Guiochon, G.; Felinger, A.; Shirazi, D. G., Fundamentals of preparative and nonlinear chromatography. Academic Press: 2006. (40) Gu, T.; Tsai, G.-J.; Tsao, G., Modeling of Nonlinear Multicomponent Chromatography. In: Tsao G.T. Eds.; Chromatography. Advances in Bichemical Engineering/Biotechnology, vol 40. Springer: Berlin, Heidelberg, 1993; pp 45-71. (41) Gu, T.; Tsai, G. J.; Tsao, G. T. New Approach to a General Nonlinear Multicomponent Chromatography Model. AIChE J. 1990, 36, 784. (42) Gu, T.; Tsao, G. T.; Tsai, G. J.; Ladisch, M. R. Displacement Effect in Multicomponent Chromatography. AIChE J. 1990, 36, 1156. (43) Liu, Z.; Roininen, J.; Pulkkinen, I.; Sainio, T.; Alopaeus, V. Moment Based Weighted Residual Method—New Numerical Tool for a Nonlinear Multicomponent Chromatographic General Rate Model. Comput. Chem. Eng. 2013, 53, 153. (44) Seidel-Morgenstern, A. Experimental Determination of Single Solute and

30

ACS Paragon Plus Environment

Page 30 of 45

Page 31 of 45 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

Competitive Adsorption Isotherms. J. Chromatogr. A 2004, 1037, 255. (45) Lin, X.; Wu, J.; Fan, J.; Qian, W.; Zhou, X.; Qian, C.; Jin, X.; Wang, L.; Bai, J.; Ying, H. Adsorption of Butanol from Aqueous Solution Onto a New Type of Macroporous Adsorption Resin: Studies of Adsorption Isotherms and Kinetics Simulation. J. Chem. Technol. Biotechnol. 2012, 87, 924. (46) Markham, E. C.; Benton, A. F. The Adsorption of Gas Mixtures by Silica. J. Am. Chem. Soc. 1931, 53, 497. (47) Wilson, E.; Geankoplis, C. Liquid Mass Transfer at Very Low Reynolds Numbers in Packed Beds. Ind. Eng. Chem. Fundam. 1966, 5, 9. (48) Mackie, J.; Meares, P. The Diffusion of Electrolytes in a Cation-Exchange Resin Membrane. I. Theoretical. Proc. R. Soc. London Ser. A. 1955, 232, 498. (49) Wilke, C.; Chang, P. Correlation of Diffusion Coefficients in Dilute Solutions. AIChE J. 1955, 1, 264. (50) Silva, M. S. P.; Mota, J. P. B.; Rodrigues, A. E. Fixed-Bed Adsorption of Aromatic C8 Isomers: Breakthrough Experiments, Modeling and Simulation. Sep. Purif. Technol. 2012, 90, 246. (51) Lin, X.; Xiong, L.; Huang, C.; Yang, X.; Guo, H.; Chen, X.; Chen, X. Sorption Behavior and Mechanism Investigation of Formic Acid Removal by Sorption Using an Anion-exchange Resin. Desalin. Water Treat. 2016, 57, 366. (52) Alinejad-Mir, A.; Amooey, A. A.; Ghasemi, S. Adsorption of Direct Yellow 12 from Aqueous Solutions by an Iron Oxide-Gelatin Nanoadsorbent; Kinetic, Isotherm and Mechanism Analysis. J. Clean. Prod. 2018, 170, 570. (53) Liu, F. Separation and Purification of Valuable Chemicals from Simulated Hydrothermal Conversion Product Solution. Masters Thesis, Waterloo, Ontario, Canada, 2012.

31

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research 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

(54) Langmuir, I. The Constitution and Fundamental Properties of Solids and Liquids. Part I. Solids. J. Am. Chem. Soc. 1916, 38, 2221. (55) Weber, T. W.; Chakravorti, R. K. Pore and Solid Diffusion Models for Fixed-bed Adsorbers. AIChE J. 1974, 20, 228. (56) Freundlich, H. Über die Adsorption in lösungen. Zeitschrift fur Physikalische Chemie 1906, 57, 385. (57) Chen, S. S.; Maneerung, T.; Tsang, D. C. W.; Ok, Y. S.; Wang, C. H. Valorization of Biomass to Hydroxymethylfurfural, Levulinic acid, and Fatty Acid Methyl Ester by Heterogeneous Catalysts. Chem. Eng. J. 2017, 328, 246. (58) Cui, X. L.; Chen, K. J.; Xing, H. B.; Yang, Q. W.; Krishna, R.; Bao, Z. B.; Wu, H.; Zhou, W.; Dong, X. L.; Han, Y.; Li, B.; Ren, Q. L.; Zaworotko, M. J.; Chen, B. L. Pore Chemistry and Size Control in Hybrid Porous Materials for Acetylene Capture from Ethylene. Science 2016, 353, 141. (59) Lin, X.; Wu, J.; Jin, X.; Fan, J.; Li, R.; Wen, Q.; Qian, W.; Liu, D.; Chen, X.; Chen, Y.; Xie, J.; Bai, J.; Ying, H. Selective Separation of Biobutanol from Acetone–Butanol–Ethanol Fermentation Broth by means of Sorption Methodology Based on a Novel Macroporous Resin. Biotechnol. Prog. 2012, 28, 962. (60) Liu, S.; Chen, J.; Peng, Y.; Hu, F.; Li, K.; Song, H.; Li, X.; Zhang, Y.; Li, J. Studies on Toluene Adsorption Performance and Hydrophobic Property in Phenyl Functionalized KIT-6. Chem. Eng. J. 2018, 334, 191. (61) Chern, J.-M.; Chien, Y.-W. Competitive Adsorption of Benzoic Acid and p-nitrophenol onto Activated Carbon: Isotherm and Breakthrough Curves. Water Res.

2003, 37, 2347.

32

ACS Paragon Plus Environment

Page 32 of 45

Page 33 of 45 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

Table 1 Physical properties of the SY-01 resin particle.37

Parameter

unit

values

Grain shape

Spherical beads

Appearance

Brown

Matrix structure

Polystyrene divinylbenzene

Functional groups

Amide groups

Average resin diameter (dp)

m

0.8×10-3

Apparent resin density (ρp)

kg/m3

1050

BET specific surface area (SBET)

m2/g

1334

Micropore area (Smicro)

m2/g

995

Mesopore area (Smeso)

m2/g

126

Macropore area (Smacro)

m2/g

213

Total pore volumea (Vp)

m3/ kg

1.26×10-3

Micropore volume (Vmicro)

m3/ kg

4.2×10-4

Mesopore volume (Vmeso)

m3/ kg

6.9×10-4

Macropore volume (Vmacro)

m3/ kg

1.5×10-4

Average pore diameterb (dp)

m

1.13×10-9

Moisture content

%

60.82

Carbon content

%

85.30

Hydrogen

%

6.89

Nitrogen

%

0.02

Oxygen contentc

%

7.79

a

At P/P0=0.99; b Calculated by density functional theory (DFT);

c

Calculated by difference O (%) = [100 − (C + H + N + S )](%)

33

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research 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

Figure legends Fig. 1 The scheme of the fixed-bed column adsorption system with all the components. Fig. 2 Adsorption isotherms of FA, LA and 5-HMF onto SY-01 resin in the single component system at 298 ± 1 K. Fig. 3 Competitive adsorption isotherms of FA, LA and 5-HMF onto SY-01 resin in the binary components and ternary components system at 298 ± 1 K, (A) FA-LA binary components system; (B) FA-5-HMF binary components system; (C) LA-5-HMF binary components system; and (D) FA-LA-5-HMF ternary components system. Fig. 4 Noncompetitive adsorption dynamics of FA, LA and 5-HMF onto SY-01 resin at the outlet of the fixed-bed column in single component system at the temperature of 298 ± 1 K. Fig. 5 Experimental and predicted column competitive breakthrough curves for FA-LA separations onto SY-01 resin at 298 ± 1 K. The breakthrough experiment was performed in a column (Ф1.6×20, 18 mL SY-01 resin) at a flow rate of 2.0 BV/h. The contents of FA and LA in binary components mixtures were 1.826 and 5.039 g/L. Fig. 6 Experimental and predicted column competitive breakthrough curves for FA-5-HMF separations onto SY-01 resin at 298 ± 1 K. The breakthrough experiment was performed in a column (Ф1.6×20, 18 mL SY-01 resin) at a flow rate of 2.0 BV/h. The contents of FA and 5-HMF in binary components mixtures were 1.825 and 1.114 g/L. Fig. 7 Experimental and predicted column competitive breakthrough curves for LA-5-HMF separations onto SY-01 resin at 298 ± 1 K. The breakthrough experiment

34

ACS Paragon Plus Environment

Page 34 of 45

Page 35 of 45 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

was performed in a column (Ф1.6×20, 18 mL SY-01 resin) at a flow rate of 2.0BV/h. The contents of LA and 5-HMF in binary components mixtures were 5.133 and 1.160 g/L. Fig. 8 Experimental and predicted column competitive breakthrough curves for FA-LA-5-HMF separations onto SY-01 resin at 298 ± 1 K. The breakthrough experiment was performed in a column (Ф1.6×20, 18 mL SY-01 resin) at a flow rate of 2.0BV/h. The contents of FA, LA and 5-HMF in ternary components mixtures were 1.816, 5.106 and 1.173 g/L. Fig. 9 Predicted competitive breakthrough curves of FA, LA and 5-HMF by GRM at different fixed-bed positions (1/3 Lc, 2/3 Lc, and Lc) in ternary component system at 298 ± 1 K.

35

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research 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

Fig. 1 The scheme of the fixed-bed column adsorption system with all the components.

36

ACS Paragon Plus Environment

Page 36 of 45

Page 37 of 45 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

Fig. 2 Adsorption isotherms of FA, LA and 5-HMF onto SY-01 resin in the single component system at 298 ± 1 K.

37

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research 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

Fig. 3 Competitive adsorption isotherms of FA, LA and 5-HMF onto SY-01 resin in the binary components and ternary components system at 298 ± 1 K, (A) FA-LA binary components system; (B) FA-5-HMF binary components system; (C) LA-5-HMF binary components system; and (D) FA-LA-5-HMF ternary components system.

38

ACS Paragon Plus Environment

Page 38 of 45

Page 39 of 45 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

Fig. 4 Noncompetitive adsorption dynamics of FA, LA and 5-HMF onto SY-01 resin at the outlet of the fixed-bed column in single component system at the temperature of 298 ± 1 K.

39

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research 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

Fig. 5 Experimental and predicted column competitive breakthrough curves for FA-LA separations onto SY-01 resin at 298 ± 1 K. The breakthrough experiment was performed in a column (Ф1.6×20, 18 mL SY-01 resin) at a flow rate of 2.0 BV/h. The contents of FA and LA in binary components mixtures were 1.826 and 5.039 g/L.

40

ACS Paragon Plus Environment

Page 40 of 45

Page 41 of 45 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

Fig. 6 Experimental and predicted column competitive breakthrough curves for FA-5-HMF separations onto SY-01 resin at 298 ± 1 K. The breakthrough experiment was performed in a column (Ф1.6×20, 18 mL SY-01 resin) at a flow rate of 2.0 BV/h. The contents of FA and 5-HMF in binary components mixtures were 1.825 and 1.114 g/L.

41

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research 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

Fig. 7 Experimental and predicted column competitive breakthrough curves for LA-5-HMF separations onto SY-01 resin at 298 ± 1 K. The breakthrough experiment was performed in a column (Ф1.6×20, 18 mL SY-01 resin) at a flow rate of 2.0BV/h. The contents of LA and 5-HMF in binary components mixtures were 5.133 and 1.160 g/L.

42

ACS Paragon Plus Environment

Page 42 of 45

Page 43 of 45 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

Fig. 8 Experimental and predicted column competitive breakthrough curves for FA-LA-5-HMF separations onto SY-01 resin at 298 ± 1 K. The breakthrough experiment was performed in a column (Ф1.6×20, 18 mL SY-01 resin) at a flow rate of 2.0BV/h. The contents of FA, LA and 5-HMF in ternary components mixtures were 1.816, 5.106 and 1.173 g/L.

43

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research 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

Fig. 9 Predicted competitive breakthrough curves of FA, LA and 5-HMF by GRM at different fixed-bed positions (1/3 Lc, 2/3 Lc, and Lc) in ternary component system at 298 ± 1 K.

44

ACS Paragon Plus Environment

Page 44 of 45

Page 45 of 45 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

For Table of Contents Only

45

ACS Paragon Plus Environment