Selective Water Removal by Sorption from Butanol ... - ACS Publications

Apr 17, 2017 - Divya Jayaprakash, Ravi Dhabhai, Catherine H. Niu,* and Ajay K. Dalai ... technique known as pressure swing adsorption (PSA) has been...
1 downloads 0 Views 1MB Size
Subscriber access provided by UB + Fachbibliothek Chemie | (FU-Bibliothekssystem)

Article

Selective water removal by sorption from butanol-water vapor mixtures: analyses of key operating parameters, and site energy distribution Divya Jayaprakash, Ravi Dhabhai, Catherine Hui Niu, and Ajay K Dalai Energy Fuels, Just Accepted Manuscript • DOI: 10.1021/acs.energyfuels.7b00310 • Publication Date (Web): 17 Apr 2017 Downloaded from http://pubs.acs.org on April 24, 2017

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

Energy & Fuels 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 44

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

Energy & Fuels

1

Selective water removal by sorption from butanol-water vapor mixtures:

2

analyses of key operating parameters, and site energy distribution

3

Divya Jayaprakash, Ravi Dhabhai, Catherine H. Niu*, and Ajay K. Dalai

4

Department of Chemical and Biological Engineering, University of Saskatchewan, 57 Campus

5

Drive, Saskatoon, Saskatchewan, S7N 5A9 Canada

6 7 8 9 10 11 12 13 14 15 16 17 18

*

19

[email protected].

Corresponding Author: Dr. Catherine H. Niu, Tel. +1-306-9662174; Fax: +1-306-9664777; E-mail:

1

ACS Paragon Plus Environment

Energy & Fuels

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

1 2

In the present paper, selective water removal from butanol-water vapor mixture was

3

carried out in a pressure swing adsorption (PSA) system using canola meal (CM) biosorbent.

4

Five operating parameters (temperature, pressure, feed butanol concentration, feed flow rate, and

5

CM particle size) were studied by the orthogonal array design method and range analysis to

6

obtain the favorable process conditions for butanol drying. The performance of butanol

7

dehydration was evaluated using five indices - water uptake; butanol uptake; water selectivity;

8

butanol recovery; and maximum butanol concentration in the effluent. The obtained favorable

9

dehydration conditions resulted in the maximum effluent butanol concentration of >99 v/v%,

10

water uptake of 0.48 g/g-ads, water separation factor of 5.4, and butanol recovery of 90%. The

11

Dubinin-Polanyi model for large pores fit the water adsorption isotherms reasonably well.

12

Furthermore, site energy distribution of water adsorption was also estimated. Average site energy

13

(3.33 kJ/mol) and standard deviation of the site energy distribution (2.36 kJ/mol) were

14

determined and applied to analyze the interaction between the biosorbent and adsorbate, and

15

adsorbent surface energy heterogeneity. Saturated CM was regenerated at 110°C under vacuum

16

and reused for more than 16 cycles.

17 18

Keywords: Butanol; Water adsorption; Pressure swing adsorption; Canola meal; Dubinin-

19

Polanyi theory; Site energy distribution.

20

2

ACS Paragon Plus Environment

Page 2 of 44

Page 3 of 44

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

1 2

Energy & Fuels

1. Introduction Diminishing supplies of crude oil, coupled with increasing greenhouse gas emissions and

3

carbon footprints, have led to a massive interest in renewable biofuels such as ethanol and

4

butanol.1 In recent years, biobutanol is preferred to bioethanol and other alcohols, mainly

5

because of its superior fuel properties that are very similar to gasoline. It is less corrosive and can

6

be easily transported through existing pipelines. Butanol also has higher combustion value, and

7

octane rating with less ignition problems.2 Most importantly, biobutanol can be used in place of

8

gasoline without vehicle modifications – that means it can be integrated seamlessly into the

9

existing petroleum infrastructure. Butanol-gasoline blends of up to 85% butanol can be used in

10 11

unmodified petrol engines.3 Biobutanol is often produced through acetone-butanol-ethanol (ABE) fermentation.

12

However, as butanol is toxic to microorganisms above 2 v/v% in the fermentation broth, 4 it is

13

difficult to obtain its high concentration. Thus, it is imperative to purify butanol from diluted

14

aqueous media. Unless concentrated to over 99 v/v%, biofuels can neither be mixed with

15

gasoline nor be used as a stand-alone fuel. Conventionally, for purification of butanol from ABE

16

process, distillation is carried out which produces azeotropic vapor of about 55 v/v% butanol and

17

45 v/v% water followed by decantation. However, it is a costly and energy intensive process due

18

to the need of multiple distillation and decantation steps.5

19

Commonly employed alternative techniques for biobutanol recovery and purification are

20

adsorption, gas stripping, and membrane pervaporation.6 Various researchers have selectively

21

adsorbed butanol from the fermentation broth using butanol adsorbing selective sorbents such as

22

activated carbon, 7 macroporous resins such as poly (styrene-codivinyl benzene) and cross-linked

23

polystyrene framework named KA-I ,8,9 and microporous mordenite framework inverted (MFI) 3

ACS Paragon Plus Environment

Energy & Fuels

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

1

type zeolite.10 Some have used integrated butanol production and recovery by gas stripping 11 or

2

two stage energy efficient gas stripping.12

3

Membrane based process pervaporation is also being widely employed for the integrated

4

butanol purification. Polydimethylsiloxane composite membranes, either in combination with

5

ceramic or polyvinylidene fluoride coating, have been used to efficiently separate butanol from

6

fermentation broth.13,14 Integrated or two stage processes such as combination of gas or vapor

7

stripping–pervaporation or two stage pervaporation processes have also been employed for

8

butanol recovery.15,16

9

Distillation, followed by adsorption could be a cost-effective method in terms of energy

10

requirement. A specific adsorption technique known as pressure swing adsorption (PSA) has

11

been put to practice by most of the bioethanol industries to dry ethanol due to its low energy

12

requirement to achieve anhydrous ethanol.17,18 Most commonly, zeolites are used as adsorbents

13

in the PSA process owing to their high adsorption capacity. However, the regeneration of water

14

saturated zeolite bed is energy intensive and their disposal can be a threat to environment. These

15

shortcomings could be overcome with a more effective approach of using biomaterials as

16

adsorbents as they are biodegradable, reusable,18,19 and safe,20 require relatively low temperature

17

for regeneration, and do not pose a threat to be disposed of into the environment. However, there

18

has been scarce literature for drying of butanol in a PSA process.

19

Agriculture byproduct like canola meal can be a potential biosorbent, as canola is

20

abundantly cultivated in Canada.21,22 This byproduct has not been thoroughly investigated as an

21

adsorbent for drying of butanol, although canola meal has been reported to have a relatively high

22

water adsorption capacity of about 303-390% of its initial dry weight.23 Research has

23

demonstrated that canola meal, before 24 and after 25 protein extraction, is capable of drying 4

ACS Paragon Plus Environment

Page 4 of 44

Page 5 of 44

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

Energy & Fuels

1

ethanol, thus, showing its potential for drying other alcohols. There is a great incentive in

2

examining the potential of CM biomaterial for drying of butanol.

3

In this work, a pressure swing adsorption process using canola meal based biosorbent was

4

investigated to selectively remove water from lower grade butanol-water vapors including the

5

simulated azeotropic butanol concentration (55 v/v %) from preliminary distillation in biobutanol

6

production industry, to optimize the crucial parameters affecting the performance of drying

7

butanol, and to explore the water selective adsorption mechanism. An important aspect of this

8

study was the approach of butanol drying by selective water adsorption from the butanol-water

9

vapor mixture on a biosorbent packed PSA column which produced fuel grade (>99 v/v%)

10

butanol. The process was aimed to be an alternative to the use of multiple decantation and

11

distillation units in the biobutanol production industry. In addition, the site energy distribution of

12

water adsorption was analyzed based on the Dubinin-Polanyi isotherm model which helped to

13

understand the adsorbate-adsorbent interactions and the water adsorption process.

14

2. Materials and Methods

15

2.1 Biosorbent preparation. Canola meal (CM) after protein extraction was obtained from

16

Bunge Global Innovation, White Plains, NY, USA. The material was oven dried at 105°C for 24

17

h followed by sieved using Canadian Standard Sieves Series (Combustion Engineering Canada

18

Inc.) to collect particles in the size range of 0.43-1.18 mm. This size range of CM were used in

19

the present work based on the available surface area and ease of operation. In addition, CM

20

pellets were made with the particles in the size range of 0.43-1.18 mm using a California Pellet

21

Mill (CPM-Laboratory Model CL-5, California Pellet Mill Co., Crawfordsville, IN, USA). The

22

pellet diameter was about 4.7 mm and length of 7-10 mm.

5

ACS Paragon Plus Environment

Energy & Fuels

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

1

2.2 Feed solution preparation. Butanol solutions of different concentrations were prepared by

2

mixing butanol (Fisher Scientific, ACS reagent grade; >99.4 v/v %) with deionized water.

3

2.3 Physico-chemical characterization of biosorbents. The composition of CM was

4

determined by Intertek Labs, Saskatoon, Canada as per AOAC International. The organic

5

elemental content was analyzed using an Elementar Vario III CHNS analyzer. 4-6 mg of sample

6

was weighed and packed along with a tin boat and placed in the designated chambers for

7

analysis. Sulfalinic acid (C6H7NO3S) was used as a standard for analysis. Fourier transform

8

infrared (FTIR) spectroscopy (Bruker Vertex 70 FTIR spectrometer, MA, USA) analysis was

9

carried out to identify the significant functional groups in biomass with respect to adsorption.

10

Each spectrum was the average of 16 co-addition of scans with a total scan time of 15 s in the IR

11

range of 400–4000 cm-1 at 4 cm-1 resolution. The devolatilization characteristics of the biomass

12

with temperature were investigated by thermogravimetric analysis (TGA) (Perkin Elmer Pyris

13

Diamond TG/DTA) in the range of 22°C to 400°C at the heating rate of 5°C/min. Particle size of

14

CM was measured using the Mastersizer MS-64 sample dispersion analyzer by means of dry

15

method. A 1000F lens was utilized and the particle size analysis was performed using 10,000

16

sweeps, and the obtained particle obscuration was comprehended between 10% and 30%.

17

2.4 Design of experiments by orthogonal array design (OAD). In order to evaluate the effects

18

of operating parameters on butanol drying, the orthogonal array design (OAD) was used.26 Five

19

operating parameters including temperature (A), pressure (B), butanol feed concentration (C),

20

feed flow rate (D), and size of adsorbent particles (E) were chosen for the present study as shown

21

in Table 1.

22

The butanol feed concentrations of 55 and 95 v/v% were chosen to mimic the azeotropic

23

distillate butanol concentration and the high end of concentration, and the corresponding boiling 6

ACS Paragon Plus Environment

Page 6 of 44

Page 7 of 44

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

Energy & Fuels

1

points (95 and 111°C) were chosen to study the effect of operation temperature. A pressure range

2

of 135-201 kPa was chosen to avoid high pressure operation demanding high energy

3

consumption.

4

The effect of these parameters on performance indices, including water uptake, butanol

5

uptake, water selectivity, butanol recovery, and maximum effluent butanol concentration were

6

determined using a statistical treatment called the “range analysis”.27 This analysis provided

7

relevant information to optimize the butanol dehydration performance by choosing appropriate

8

operation conditions. The details of range analysis are given in the supplementary information.

9

All the experiments were replicated and the average results with standard deviation were

10

reported.

11

2.5 Adsorption/regeneration experiments. The PSA system used in this study has been used

12

previously for ethanol drying and the schematic drawing is presented in Fig. 1. The PSA system

13

has been described in detail in previous reports.18,25 In order to simulate the butanol-water vapor

14

generated by preliminary distillation in biobutanol production industry, the butanol-water

15

solution stored in a jacketed sealed feed tank with stirring was pre-heated to 90°C. Then it was

16

pumped through a digital piston pump (Cole- Parmer, RK-74930-05) through a nebulizer to

17

stainless steel tubing equipped with heating tapes (Cole-Parmer; 50-60Hz, 120 V, 624W, 5.20A).

18

It was then mixed with preheated nitrogen (carrier gas) at a flow rate of 850 mL/min to

19

completely vaporize the feed butanol solution before it is reached to the adsorption column. The

20

316-stainless steel adsorption column with dimensions (length 500 mm, internal diameter 47.5

21

mm, wall thickness 1.65 mm) contained randomly packed canola meal adsorbents. The vapor

22

feed entered the column from the top through a three-way valve. The temperature and pressure of

23

the bed were measured both at the top and the bottom of the column. The pressure drop along the 7

ACS Paragon Plus Environment

Energy & Fuels

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

1

column during the adsorption process was 2.1-3.2 kPa, which is negligible compared with the

2

operation pressure range of 135-201 kPa. Water got selectively adsorbed onto the adsorbent bed

3

and the dried butanol vapor effluent exited from the bottom of the column. The effluent sample

4

was then condensed and collected for water content analysis by a Karl-Fischer (KF) titrator. The

5

sample was collected at an interval of 5 min and the adsorption process lasted for a total duration

6

of about 2.5 h till equilibrium (bed saturation) was achieved. After adsorption, desorption

7

process was carried out by purging heated nitrogen gas at 110°C (850 mL/min) from the bottom

8

of the column with a vacuum 33 kPa for 5.5 h. Desorbed water was condensed and collected.

9

Desorption process conditions were confirmed adequate not only to dry the wet column, but also

10 11

to preserve the chemical properties and stability of the adsorbent. Water or butanol uptake was determined by the mass balance, i.e. the water/butanol input

12

subtracted from the water/butanol output per g of dry adsorbent in the column. Recovery of

13

butanol was defined as the ratio of the amount of butanol in effluent to the amount of butanol in

14

influent. Water separation factor over butanol (α) was calculated according to eq. 1:

15

α = ( / ) / ( / )

16

where  is mole fraction of water in the adsorbed phase,  is the mole fraction of butanol in

17

the adsorbed phase,  is the mole fraction of water in the vapor phase and  is the mole

18

fraction of butanol in the vapor phase.

19

(1)

In the adsorption process, the feed butanol concentration varied from 55 to 95 v/v%, and

20

the rest is water. The lower end of the feed butanol concentration was chosen in order to simulate

21

the butanol concentration in the butanol-water azeotrope obtained from the initial distillation

22

process in industry. The feed of butanol-water liquid solution was 1.5-3 mL/min in order to

23

provide sufficient amount of water and butanol to the adsorption column. The pressure of 1358

ACS Paragon Plus Environment

Page 8 of 44

Page 9 of 44

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

Energy & Fuels

1

201 kPa and temperature of 95-111oC were chosen to investigate their effects, while ensuring

2

adsorption of butanol and water to be operated in vapor-solid phases without burning the

3

adsorbents. The amount of CM packed in the adsorption column depended on the particle size of

4

CM. In the case of CM particles (0.425-1.18 mm), the full packing required about 215 g CM,

5

while in the case of CM cylindrical pellets (4.7 mm in diameter and 5 mm in length) about 515 g

6

were packed. Each experiment was run at least in duplicate. The results were presented in

7

average and standard deviation.

8 9

2.6 Analytical methods. The water content in the feed and effluent was determined by an

10

automated KF coulometric titrator (Mettler Toledo DL 32) using methanol as diluent. Butanol

11

mass in the effluent samples was calculated by subtracting the mass of water from the total mass

12

of sample, these results were cross-validated with that measured by a gas chromatograph (GC)

13

(Agilent Technologies 7890A GC System; 7683B Series Injector) with a flame ionization

14

detector. The coefficient of variation (CV) for water content determination by the KF titrator

15

was 0.05%, while for GC analysis, CV was 1.5%, based on the analysis of a standard of 1%

16

water or butanol in 4 trials. Butanol analysis by GC was carried out at the following conditions -

17

flow rate 2.6 mL/min, average velocity 40 cm/s, hold up time 1.25 min, inlet temperature 150°C,

18

oven temperature 40°C and detector temperature 250°C. 25 µL sample was injected into the

19

column with a split ratio of 100:1.

20 21 22 23

3. Dubinin-Polanyi isotherm and site energy distribution The Dubinin-Polanyi model used in this work is based on the adsorption potential theory, which has been recognized as a useful model both gas and aqueous phase adsorption on 9

ACS Paragon Plus Environment

Energy & Fuels

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

1

energetically heterogeneous surfaces such as biomaterials. As per this model, for any molecule,

2

the magnitude of adsorption potential varies within the adsorption space depending on its

3

proximity to atoms on the adsorbent surface.28 Polanyi theory assumes that the adsorption

4

potential, Ƹ is independent of temperature and the adsorbed gas phase molecules have similar

5

properties as the corresponding bulk gas phase.29 Meanwhile, the adsorption potential Ƹ is given

6

by

7

Ƹ= RT ln

8

The Dubinin–Polanyi equation for microporous and large pore materials is described in eqs. 3

9

and 4, respectively, as shown below.

10



( ) 

ln q = ln -

β



(2)



)] 2

(3)



)]

(4)

[RT ln (



11

ln q = ln -

12

where q is the mass adsorbed per unit mass of adsorbent (mol/g adsorbent), q0 denotes the

13

limiting mass for adsorption (mol/g adsorbent),  and  are pore constants for micropore and

14

large pore materials, β is an affinity coefficient, Pi represents partial pressure of the adsorbate

15

(kPa), and Ps is saturated vapor pressure of the adsorbate (kPa). These relationships can be

16

helpful to estimate the adsorption capacity and affinity, and to evaluate the potential application

17

of CM as a sorbent.

β

[RT ln (



18

Site energy distribution can be determined from the isotherm representing the

19

experimental equilibrium data.30 The interaction forces between solutes and sorbents could be

20

indicated by the mean of the site energy distribution, while the surface energy heterogeneity of

21

sorbent could be interpreted by the width of the site energy distribution.31 Most biosorbents are 10

ACS Paragon Plus Environment

Page 10 of 44

Page 11 of 44

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

Energy & Fuels

1

composed of multiple components and thus are heterogeneous in their structure. The basic

2

integral equation defining adsorption distribution is:

3

() = ∫ (ℎ) (, )()



(5)

4 5

As per eq. (5), the total adsorption, , by a heterogeneous surface is the integral of

6

energetically homogeneous isotherm ( (ℎ)) multiplied by a site energy frequency distribution,

7

(). E is the difference between the solute and solvent adsorption energies for a given site. The

8

maximum (Emax) and minimum (Emin) limits of energy space are generally ∞ and 0 which are

9

directly related to the partial pressure of adsorbate in the isotherm for a gas phase adsorption. Eq.

10

(5) is difficult to solve and has no analytical solution, however, approximate solutions have been

11

developed. The most employed method in literature is the condensation approximation method,

12

originally proposed by Cerofolini.32 It estimates site energy distribution function from the

13

isotherm equation. The partial pressure of adsorbate can be correlated with the energy of

14

adsorption as:

15

 =  ! . exp &−

16

Defining E*=E-Es gives

17

 =  ! . exp &− *+,

18

where  and  ! are the partial and saturation pressure of adsorbate at a fixed temperature. The

19

 ! reference state for E represents the lowest physically realizable sorption energy (kJ/mol).

20

Incorporation of eq. (7) into eq. (5) will lead to an approximate site energy distribution,

21

( ∗ )(g.mol/g.J), which is the differentiation of the isotherm, . ( ∗ ), with respect to  ∗ :

22

( ∗ ) =

()(  *+

,

(6)

(∗

(7)

/01 (( ∗ ) /(( ∗ )

(8) 11

ACS Paragon Plus Environment

Energy & Fuels

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

1

As such, the area under the distribution curve is equal to the maximum adsorption capacity, 2 ,

2

and is given by:

3

∫ ( ∗ )( ∗ ) = 2

4

Eq. (8) can be used to estimate ( ∗ ) by differentiating the corresponding isotherm expression

5

with respect to  ∗ . The value of ( ∗ ) was determined by plugging eq. (7) into eq. (4) and then

6

incorporating into eq. (8):

7

( ∗ ) = & 3.

8 9



 4

(9)

.  ∗ , . exp &−

 4

. ∗ ,

(10)

The average site energy µ(E*) (kJ/mol) and the surface energy heterogeneity in the tested range of temperature and partial pressure were also estimated. µ(E*) can be calculated by

10

integrating the value of ( ∗ ) in the minimum and maximum range (0 and ∞, respectively):33

11

μ( ∗ ) = ∫  ∗ ( ∗ ) ∗ /∫ ( ∗ ) ∗

12





Incorporating eq. (10) into the above equation, and integrating, the average site energy

13

for the present case can be determined,

14

μ( ∗ ) = 2/( 4 )

15

(11)

7

(12)

With site energy distribution model, sorption site heterogeneity can also be determined

16

by calculating the standard deviation (σ*e) (kJ/mol) which can be determined by first calculating

17

μ( ∗ ) which is given by:

18

μ( ∗ ) = ∫  ∗ ( ∗ ) ∗ /∫ ( ∗ ) ∗

19





(13)

μ( ∗ ) = 6/(92/:)

(14)

20

Once μ( ∗ ) and μ( ∗ ) are determined, the standard deviation can be simply calculated as: 33

21

σ∗. = above 8 kJ/mol, the adsorption process is considered to be predominantly 17

ACS Paragon Plus Environment

Energy & Fuels

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

1

chemisorption while the process is predominantly physisorption for the values of > lower than 8

2

kJ/mol.45 The mean free energy obtained in this work for the large pore model was 0.04 kJ/mol,

3

indicating the water adsorption is physisorption. It is consistent with water adsorption studies on

4

corn meal 43 and canola meal 25 as shown in Table 6.

5 6

4.5 Site energy distribution. In the section 4.4, the isotherm modeling results demonstrated that

7

the D-P model for the larger pores fit the experimental data better, indicated by R2 of 0.96. For

8

this reason, site energy distribution analysis was carried out by using the D-P model for larger

9

pores only. The site energy,  ∗ as a function of the equilibrium water uptake at different

10

temperature, is shown in Fig. 7.  ∗ decreased as the water uptake increased, revealing that water

11

molecules first occupied the high-energy sorption sites on CM at low water uptake, then spread

12

to low-energy sorption sites. The site energy distribution for water adsorption is presented in Fig.

13

8. It is to be pointed out that the shape and intensity of site energy distribution curve calculated

14

using the condensation approximation method will vary depending on the isotherm model

15

applied.46 In the present case, site energy distribution curve was constructed using the D-P

16

isotherm model for larger pores. It is important to note that no similar site energy distribution

17

analysis for the adsorption of water molecules on biomaterials was found in literature. Thus, the

18

discussion of the results has been developed with reference to the available site energy

19

distribution studies.

20

Site energy distribution curve was found to be the typical bell shaped unimodal curve

21

obtained in other similar studies.30,33,46 Site energy curve had two characteristic regions: (1) a

22

sharp peak area corresponding to high energy binding sites and, (2) a decaying region relating to

23

higher energy sites. As the site energy (E*) increased, the frequency function f(E*) increased up 18

ACS Paragon Plus Environment

Page 18 of 44

Page 19 of 44

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

Energy & Fuels

1

to a point, and then asymptotically decreased to zero at about 14,000 J/mol. This implied that

2

CM had negligible sites with high energy (e.g., >15,000 J/mol) in the tested range of parameters.

3

Similar observations were obtained for site energy distribution of gases and other organic

4

compounds.30,31,46 Theoretically, the area under the curves in Fig. 8 reveals the number of the

5

available sorption sites in a specific energy range as depicted in eq. 9 which is equal to the

6

theoretical maximum adsorption capacity qm.

7

In order to deduce the adsorbent-adsorbate interaction, average site energy μ( ∗ ) was

8

calculated as described in eq. 12. The average site energy μ( ∗ ) can be used as a measure of the

9

affinity of the adsorbate for the adsorbent surface. The higher the value of the average site

10

energy, the higher the sorption affinity.33 In the present case, μ( ∗ ) was 3.33 kJ/mol, which

11

implies that the water adsorption can be predominant physical adsorption. Yan and Niu 33

12

obtained μ( ∗ ) in the range of 26.3-28.1 kJ/mol for the adsorption of levofloxacin on pretreated

13

barley straw in which chemisorption was considered to be predominant. In this work, the average

14

site energy was much lower, demonstrating low affinity and binding force between water and the

15

biosorbent surface. This is consistent to the result of mean free energy determined by the D-P

16

isotherm model in this work which indicates physisorption in the system.

17

The standard deviation (σ∗. ), which is a measure of the adsorbent site energy

18

heterogeneity, was calculated to be 2.36 kJ/mol. The higher the value of σ.∗ , the higher is the

19

heterogeneity. As described in section 3.1, CM is composed of cellulose, hemicellulose, lignin,

20

protein, and starch which consists of carbonyl, hydroxyl and other oxygen containing groups.

21

This may account for the adsorbent site energy heterogeneity. For adsorption of antibiotics on

22

pretreated barley straw, Yan and Niu 23 obtained value of σ∗. in the range of 1.4-4.2 kJ/mol which

23

again indicates the heterogeneous nature of biomaterials. 19

ACS Paragon Plus Environment

Energy & Fuels

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

1

4.6 Regeneration and reusability. It is important to regenerate the water saturated column for

2

reuse. For that end, it is essential to ensure that the bed has been sufficiently dried for the next

3

cycle of adsorption. Regeneration of the column was done by purging heated nitrogen gas at

4

110°C (850 mL/min) from the bottom of the column with a vacuum 33 kPa. The temperature

5

profile for both adsorption and desorption is shown in Fig. 6. During the regeneration process,

6

initially, the temperature decreased, indicating that water desorption is an endothermic process.

7

After a while, when the water content in the bed decreased, the temperature started increasing

8

until the initial bed temperature was attained. A similar trend was reported by Simo 17 for

9

regeneration of type 3A molecular sieves for ethanol dehydration.

10

The adsorbent was examined for 16 cycles and are still used without deteriorated quality.

11

Fig. 9 shows examples of the butanol production profiles of fresh, and regenerated CM for the

12

1st, 2nd, 3rd reuse, 15th, and 16th reuse. The breakthrough curves for the first three reuse are

13

overlapped showing that the fresh and the regenerated CM biosorbents have similar

14

performance, and are capable of producing fuel grade butanol of over 99 v/v%. Though the

15

breakthrough curves for the 15th, and 16th resue were slightly below the other three runs, above

16

96 v/v% butanol was produced. The slightly decrease of the performance may be because more

17

water was accumulated in the column, and longer regeneration time is required. This can form an

18

area of future research. Table 3 shows that the elements composition of fresh and regenerated

19

CM is very much similar, once again confirming CM is stable after regeneration.

20

The results demonstrated that CM was easily regenerated at 110°C, a temperature much

21

lower than that required for regenerating molecular sieves commonly used in dehydration in

22

ethanol production industry, around 220°C-240°C.17 In addition, the water selective adsorption

23

approach was unlike the butanol selective adsorption approach that normally has regeneration 20

ACS Paragon Plus Environment

Page 20 of 44

Page 21 of 44

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

Energy & Fuels

1

issues like sequential heating desorption method at high temperature leading to adsorbent

2

damages, incomplete butanol recovery 47 and use of an external agent like methanol for

3

regeneration.46 The high water uptake capacity, coupled with lower regeneration temperature,

4

and a relatively easy disposal makes CM to be a promising material for drying of butanol vapor.

5 6

5. Conclusion

7

It was demonstrated that CM has the capability to dry butanol from the azeotropic butanol

8

concentration 55 v/v% to high purity butanol of 99 v/v%. Pressure was found to be the most

9

significant factor at the tested conditions, affecting butanol uptake, water selectivity, butanol

10

recovery, and maximum effluent butanol concentration. The optimum conditions obtained from

11

the statistical design resulted in a water separation factor of 5.4, butanol recovery of 90%, water

12

uptake of 0.48 g/g-ads and fuel grade butanol of >99 v/v%. The Dubinin- Polanyi (D-P) model

13

based on the adsorption potential theory for large pore materials gave a better fit to the water

14

adsorption isotherms. The mean free energy indicated that water adsorption is predominantly

15

physisorption. The approximate site energy distribution based on the D-P isotherm elucidated the

16

uptake of water on the heterogeneous CM biosorbent. The high-energy sorption sites were first

17

occupied at low concentration, followed by the low-energy sorption sites. Site energy

18

distribution curve revealed that CM had negligible sorption sites with very high energy (e.g.,

19

>15,000 J/mol) in the tested range of parameters. The average site energy μ( ∗ ) was 3.33

20

kJ/mol, which again indicated physical nature of water adsorption, while the standard deviation

21

σ∗. of 2.36 kJ/mol indicated the heterogeneous nature of biosorbents. Saturated CM was

22

regenerated at 110°C under vacuum and reused for more than 16 cycles.

23 21

ACS Paragon Plus Environment

Energy & Fuels

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

1

Notes

2

The authors declare no competing financial interest.

3 4

Acknowledgments

5

The authors sincerely thank Richard Blondin and RLee Prokopishyn for their technical support.

6

Funding: This work was supported by Mitacs; Saskatchewan Canola Development Commission;

7

Saskatchewan Ministry of Agriculture; University of Saskatchewan; and Western Grains

8

Research Foundation.

9

Nomenclature

10

English Letters

11

Å

angstrom

12

C

water content at time t (wt%)

13

Co

initial water content (wt%)

14

Ce

equilibrium water concentration in vapor phase (g/L)

15

>

mean free energy of adsorption (kJ/mol)

16

Es

reference state for E, representing the lowest physically realizable sorption energy (kJ/mol)

17 18

E*

E-Es (kJ/mol)

19

f (E)

site energy frequency distribution (dimensionless)

20

f (E*) approximate site energy distribution (g.mol/g-J)

21

k1 , k2

22

average values of the performance index for each of the parameters at levels 1 and 2 in OAD table (units of respective performance index) 22

ACS Paragon Plus Environment

Page 22 of 44

Page 23 of 44

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

Energy & Fuels

1

K1, K2 pore constants for microspore and large pore materials (dimensionless)

2

P

pressure (atm)

3

Pi

partial pressure of the adsorbate (atm)

4

Ps

saturated vapor pressure of the adsorbate (atm)

5

q

mass adsorbed per unit mass of adsorbent (g/g-ads)

6

qe

equilibrium water uptake (g/g-ads and also expressed as mol/g)

7

qo

limiting mass for adsorption (g/g-ads and also expressed as mol/g)

8

R

universal gas constant (J/mol K)

9

r2

correlation coefficient (dimensionless)

10

T

absolute temperature (K)

11



mole fraction of water in the adsorbed phase (dimensionless)

12



mole fraction of butanol in the adsorbed phase (dimensionless)

13



mole fraction of water in the vapor phase (dimensionless)

14



mole fraction of butanol in the vapor phase (dimensionless)

15

Greek Letters

16

α

separation factor (dimensionless)

17

β

affinity coefficient (dimensionless)

18

ε

adsorption potential (J/mol)

19



range value in statistical range analysis (units of the performance index)

20

Abbreviation

21

ARE

Average Relative Error

22

Bu-OH

Butanol

23

CM

Canola meal 23

ACS Paragon Plus Environment

Energy & Fuels

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

1

CNPs

Carbon nanoparticles

2

FT-IR

Fourier Transform Infrared Spectroscopy

3

GC

Gas chromatography

4

OAD

Orthogonal array design

5

ppm

Part per million

6

PAH

Polycyclic aromatic hydrocarbons

7

PSA

Pressure Swing Adsorption

8

TGA

Thermogravimetric analysis

9 10

References

11

[1] Shah, R. Y.; Dhrubo, J. S. Int. J. Cur. Sci. Res. 2011, 2, 57-62.

12

[2] Visioli, J. L.; Enzweiler, H.; Kuhn, C. R.; Schwaab, M.; Mazutti, A. M. Sustainable Chem.

13

Processes. 2014, 2, 1-9.

14

[3] Nigam, P.; Singh, A. Progr. Energ. Combust. Sci. 2011, 37, 52-68.

15

[4] Moreira, A. R.; Ulmer, D. C.; Linden, J. C. Biotechnol. Bioengin, Symp. 1981, 11, 567-579.

16

[5] Gupta Kumar, V.; Schmoll, M.; Makki, M.; Tuohy, M.; Antonio, M. M. Applications of

17

Microbial Engineering; first ed., CRC Press: Austria, 2013; pp 282-296.

18

[6] Qureshi, N.; Blaschek, H.P. Renew. Energ. 2001, 22, 557–564.

19

[7] Xue, C.; Liu, F.; Xu, M.; Tang, I.-C.; Zhao, J.; Bai, F.; Yang, S.-T. Bioresour.Technol. 2016,

20

219, 158–168.

21

[8] Wiehn, M.; Staggs, K.; Wang, Y.; Nielsen, D. R. Biotechnol. Prog. 2014, 30(1), 68-78.

22

[9] Liu, D.; Chen, Y.; Ding, F.-Y.; Zhao, T.; Wu, J.-L.; Guo, T.; Ren, H.-F.; Li, B.-B.; Niu, H.-

23

Q.; Cao, Z.; Lin, X.-Q.; Xie, J.-J.; He, X.-J.; Ying H.-J. Biotechnol. Biofuels 2014, 7(5), 1-

24

13.

24

ACS Paragon Plus Environment

Page 24 of 44

Page 25 of 44

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

1 2

Energy & Fuels

[10] Faisal, A.; Zarebska A.; Saremi, P.; Korelskiy, D.; Ohlin, D.; Rova, U.; Hedlund, J.; Grahn, M. Adsorption 2014, 20, 465–470.

3

[11] Qureshi, N.; Cotta, M. A.; Saha, B. C. Food Bioprod. Process. 2014, 92, 298–308.

4

[12] Xue, C.; Zhao, J.; Liu, F.; Liu, L.; Xu, M.; Yang, S.-T.; Bai, F. Bioresour. Technol. 2013

5

135, 396–402.

6

[13] Xue, C.; Du, G.-Q., Chen, L.-J.; Ren, G.-J.; Bai, F. J. Biotechnol. 2014, 188, 158–165.

7

[14] Dong, Z.; Liu, G.; Liu, S.; Liu, Z.; Jin, W. J. Memb. Sci. 2014, 450, 38–47.

8

[15] Cai, D.; Hu, S.; Miao, Q.; Chen, C.; Chen, H.; Chen, H.; Zhang, C.; Li, P.; Qin, P.; Tan, T.

9 10 11 12 13

Bioresour. Technol. 2017, 224, 380–388. [16] Vane, L. M.; Alvarez, F. R.; Rosenblumb, L.; S. Govindaswamy, J. Chem. Technol. Biotechnol. 2013, 88, 1448–1458. [17] Simo, M. Pressure Swing Adsorption Process for Ethanol Dehydration; first ed., ProQuest: USA, 2008, pp 102-107.

14

[18] Tajallipour, M.; Niu, C.; Dalai, A. Energ. Fuel. 2013, 27, 6655-6664.

15

[19] Sun, J.; Wang, W.; Wan, P.; Hao, L.; Luo, X.; Gao, H. Adsorpt. Sci. Technol. 2013, 31,

16

829-844.

17

[20] Boonfung, C.; Rattanaphanee, P. World Acad. Sci. Eng.Technol. 2010, 4, 557-560.

18

[21] Canadian Canola: Agriculture and Agri-Food Canada; [updated 2015 Mar 18 ; accessed

19

2017 Jan 20]: http://www.agr.gc.ca/eng/industry-markets-and-trade/exporting-and-buying-

20

from-canada/buying-canadian-food-products/canadian-canola/?id=1426170869200.

21

[22] Newkirk, R. Canola Meal Feed Industry Guide; fourth ed., Canola Council of Canada:

22

[updated 2009 Dec 31; accessed 2014 May 11]: http://www.canolacouncil.org/.

23

[23] Aider, M.; Barbana, C. Trends in Food Sci. Technol. 2011, 22, 21-39.

24

[24] Baylak, T.; Kumar, P.; Niu, H. C.; Dalai, A.; Energ. Fuel. 2012, 26, 5226-5231.

25

ACS Paragon Plus Environment

Energy & Fuels

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

1 2 3 4 5 6

[25] Ranjbar, Z.; Tajallipour, M.; Niu, H. C.; Dalai, A. Ind. Eng. Chem. Res. 2013, 52, 1442914440. [26] Medina, A. R.; Gamero, P. M.; Almanza, J. M. R.; Cortes, D. A. H.; Vargas, G. G. J. Chil. Chem. Soc. 2009, 54, 244-251. [27] Sharma, P.; Verma, A.; Sidhu, R. K.; Pandey, O. P. J. Mater. Process. Technol. 2005, 168, 147-151.

7

[28] Polanyi, M. Z. Elektrochem. 1920, 26, 370-374.

8

[29] Chang, H.; Yuan, X.; Tian, H.; Zeng, A. Ind. Eng. Chem. Res. 2006, 45, 3916-3921

9

[30] Kumar, K. V.; De Castro, M. M.; Escandell, Sabio, M. M.; Reinoso, M. M. F. R. Phys.

10 11 12

Chem. Chem. Phys. 2011, 13, 5753–5759. [31] Shen, X.; Guo, X.; Zhang, M.; Tao, S.; Wang, X.; Environ. Sci. Technol. 2015, 49, 48944902.

13

[32] Cerofolini, G. F. Thin Solid Films, 1974, 23, 129–152.

14

[33] Yan, B.; Niu, C. H. Chem. Eng. J. 2017, 307, 631–642.

15

[34] Okewale, A. O.; Babayemi, K. A.; Olalekan, A. P. Int. J. Appl. Sci. Technol. 2013, 3, 35-42.

16

[35] Beery, K. E.; Ladisch, M. R. Enzyme Microb. Technol. 2001, 28, 573-581.

17

[36] Himmelsbach, S. D.; Khalili, S.; Akin, E. D. J. Sci. Food. Agr. 2002, 82, 685-696.

18

[37] Theivandran, G.; Ibrahim, M. S.; Murugan, M. J. Med. Plants. Stud. 2015, 4, 30-32.

19

[38] Xu, F.; Yu, J.; Tesso, T.; Dowell, F.; Wang, D. Appl. Energ. 2013, 104, 801-809.

20

[39] Raveendran, K.; Ganesh, A.; Khilar, K. C. Fuel, 1996, 75, 987-998.

21

[40] Biagini, E.; Barontini, F.; Tognotti, L. Ind. Eng. Chem. Res. 2006, 45, 4486-4493.

22

[41] Wu, P.; Gao, H.; Sun, J. S.; Ma, T.; Liu, Y.; Wang, F. Bioresour. Technol. 2012, 107, 437-

23

443. 26

ACS Paragon Plus Environment

Page 26 of 44

Page 27 of 44

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

Energy & Fuels

1

[42] Ou, H.; Tan, W.; Niu, C. H.; Feng, R. Ind. Eng. Chem. Res. 2015, 54, 6100-6107.

2

[43] Chang, H.; Yuan, X.; Tian, H.; Zeng, A. Chem. Eng. Technol. 2006, 29, 454-461.

3

[44] Lin, X.; Wu, J.; Fan, J.; Qian, W.; Zhou, X.; Qian, C.; Jin, X.; Wang, L.; Bai, J.; Ying, H. J.

4 5 6 7 8 9

Chem. Technol. Biotechnol. 2012, 87, 924-931. [45] Ruthven, D. M. Principles of Adsorption and Adsorption Processes; John Wiley & Sons: first ed., USA, 1984, pp 323-329. [46] Kumar, K. V.; De Castro, M. M.; Escandell, M. M.; Sabio, M. M.; Reinoso, F. R. Chem. Phys. Chem, 2010, 11, 2555-2560. [47] Barshad, I. Thermodynamics of water adsorption and desorption on montmorillonite; A.

10

Swineford (Ed.)., Clays and Clay Minerals: Proc. Eighth Nat. Conf., Pergamon, New York,

11

1960.

12

13

14

15

16

17

18

19 27

ACS Paragon Plus Environment

Energy & Fuels

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

1

2

Figure captions

3

Fig. 1 Schematic diagram of the experimental PSA setup

4

Fig. 2 FTIR spectra of canola meal

5

Fig. 3 TG/DTA analysis of canola meal

6

Fig. 4 (a) Butanol and (b) water breakthrough curves for canola meal. All runs were at a

7

temperature of 100°C, pressure of 135 kPa, butanol feed concentration of 79 v/v %, feed

8

flow rate of 3 mL/min and particle size of canola meal in the sieved range of 0.43−1.18

9

mm.

10 11

Fig. 5 Dubinin−Polanyi model for (a) micropores and (b) large pores: (●) experimental data; (−) model

12

Fig.6 Bed temperature profile during adsorption-desorption cycle in PSA

13

Fig. 7 Dependence of site energy on equilibrium water uptake at different temperature

14

Fig. 8 Site energy distribution of water adsorption on CM

15

Fig. 9 Butanol breakthrough curves to evaluate reusability of CM as adsorbent. All runs were at

16

a temperature of 111°C, pressure of 201 kPa, butanol feed concentration of 55 v/v %,

17

feed flow rate of 3 mL/min and particle size of canola meal in the sieved range of

18

0.43−1.18 mm

19 20 21 22 28

ACS Paragon Plus Environment

Page 28 of 44

Page 29 of 44

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

Energy & Fuels

1 2

Tables

3 4

Table 1. Factors and levels in orthogonal array design experiments

5

Levels L1 L2

A Temperature (°C) 95 111

B Pressure (kPa) 135 201

Factors C D E Bu-OH Feed Feed flow rate (mL Particle sizes (mm) Concentration (v/v %) /min) 55 1.5 0.425-1.18 95 3.0 4.7

6 7

29

ACS Paragon Plus Environment

Energy & Fuels

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

Table 2. Composition of canola meal

1

Content in CM Protein Starch Acid detergent fibre (Cellulose +Lignin) Hemicellulose (%) Ash Moisture 2

Page 30 of 44

*

Composition (wt. %) 27.1±0.53* < 1.5 26.4±0.42 6.3±0.56 3.6±0.17 4.2±0.9

All results were presented in average ± standard deviation.

3

30

ACS Paragon Plus Environment

Page 31 of 44

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

Energy & Fuels

1 2

Table 3. Organic elemental composition of fresh and used canola meal (in wt. %) Element

C H N S 3

*

Fresh (unused) CM

46.34±0.04 6.15±0.03 3.78±0.08 0.35±0.06

CM after adsorption and regeneration (after 16th reuse)* 46.85±0.13 6.60±0.04 3.85±0.07 0.33±0.01

All results were presented in average ± standard deviation.

4 5

31

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 Exp. 6 No. 7 8 9 10 11 12 13 1 14 15 2 16 17 3 18 19 4 20 5 21 22 6 23 24 7 25 26 8 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 32 of 44

Table 4. Experimental combinations using OAD and the corresponding results

1

Factors

Experimental results

A

B

C

D

E

Temp (°C)

Pressure

Inlet BuOH Conc.

Inlet flow rate (mL/min)

Particle size (mm)

(kPa)

(v/v %)

Water Uptake*

Butanol uptake*

Equilibrium water separation factor**

Recovery (%)

Max. Effluent BuOH conc. (v/v %)

95

135

55.99 ± 1.03

1.5

0.425-1.18

0.36±0.02

0.21±0.02

1.77±0.03

58.03±3.17

99.17 ± 0.09

95

135

55.39±0.14

3.0

4.7

0.16±0.00

0.06±0.02

3.22±0.92

83.56±5.57

96.65 ± 0.01

95

201

95.33±0.02

1.5

0.425-1.18

0.02±0.00

0.80±0.00

0.40±0.11

6.88±0.03

88.82 ± 1.89

95

201

94.02±0.76

3.0

4.7

0.05±0.01

0.59±0.00

1.05±0.02

15.40±0.17

99.19 ± 0.11

111

135

95.72±0.10

1.5

4.7

0.01±0.00

0.11±0.00

1.95±0.10

71.16±0.82

99.13 ± 0.03

111

135

94.96±0.06

3.0

0.425-1.18

0.03±0.00

0.25±0.01

2.05±0.03

85.63±0.91

99.17 ± 0.55

111

201

56.50±0.33

1.5

4.7

0.14±0.00

0.10±0.00

1.42±0.04

49.79±1.18

93.67 ± 2.44

111

201

55.50±0.91

3.0

0.425-1.18

0.64±0.01

0.32±0.11

2.13±0.67

67.48±10.26

98.86 ± 0.41

2

*

3

**

g adsorbed/ g dry net weight of adsorbent at equilibrium conditions. All results were presented in average ± standard deviation.

4 5

32

ACS Paragon Plus Environment

Page 33 of 44

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

1 2 3 4 5

Energy & Fuels

Table 5. Validation test results for the most favorable process conditions for drying of butanol [Experimental conditions: Temperature 111°C, pressure 135 kPa, butanol concentration 55 v/v%, feed flow rate 3 mL/min, and particle size 0.425-1.18 mm] Water uptake (g/gads)* 0.48±0.02

6

Butanol uptake (g/g-ads) 0.09±0.00

Water separation factor 5.43±0.08

Recovery (%) 90.11±0.26

*All results were presented in average ± standard deviation.

7 8

33

ACS Paragon Plus Environment

Max. Butanol concentration achieved (v/v %) 99.20 ±0.79

Energy & Fuels

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 34 of 44

Table 6. Modeling results of Dubinin-Polanyi equations

1

Model Type

K/β

Mean free energy of adsorption (kJ/mol)

Micropore in this work

1E-07

2.24

0.016

0.89

11

Large pore in this work

6E-04

0.04

0.028

0.96

7

Canola meal for drying ethanol15

4E-04

0.04

0.038

0.97

5

Corn meal for drying ethanol19

3E-04

0.04

0.009

0.96

-

q0 (mol/gads)

*ARE - Average Relative Error

34

ACS Paragon Plus Environment

r

2

ARE % *

Page 35 of 44

Energy & Fuels T

V2

V1 Nebulizer

Heating tape  wrapped on  feed tube V3

Flow meter

V5

Pump T1

P1

Feed tank

T P

Condenser

Adsorption column

Nitrogen gas cylinder

P1 & P2 Pressure indicators T1 & T2 Temperature indicators V1‐V5 three way valves V6 Check valave 

Waste tank

T2

P

P2

Pressure regulator

T

V4 V6

Waste  Condensate  Storage 

Vacuum pump

Condenser

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

Fig. 1 ACS Paragon Plus Environment

Sample tube

Energy & Fuels

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

Page 36 of 44

-C-H Alkanes -COOCarboxylic -C-N Protein -C=O Lignin, Hemicellulose

-CHO carbohydrate -CO carbohydrate

ACS Paragon Plus Environment

Fig. 2

-O-H Stretching

100

0.4

95

0.35

90

0.3

85

0.25

Weight (% )

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

Energy & Fuels

Weight (%)

80

0.2

75

0.15

70

0.1

65

0.05

60

0 0

100

200 300 Temperature (⁰C)

400

Figure 3

ACS Paragon Plus Environment

Deriv. weight change (% / ⁰C )

Page 37 of 44

Energy & Fuels

100

(a)

Butanol (v/v %)

95

90

85

80 10

30

50

70 90 Time (min)

110

130

150

Figure 4(a)

1

(b)

0.8 0.6

C/C0

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 38 of 44

0.4 0.2 0 10

30

50

70 90 Time (min)

110

130

Figure 4 (b) ACS Paragon Plus Environment

150

Page 39 of 44

-3

(a)

ln q

-4 -5

R² = 0.89

-6 -7 -8 0

10000000

20000000

30000000

Ƹ2

Fig. 5a -3.00

(b) -4.00

ln q

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

Energy & Fuels

R² = 0.96

-5.00

-6.00

-7.00 0

2000

Ƹ

4000

Fig. 5b

ACS Paragon Plus Environment

6000

Energy & Fuels

130 T1- temp at top of column

Temperature (°C)

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 40 of 44

120 T2-temp at bottom of column 110

100 ADSORPTION

90 0

100

REGENERATION

200

300

400

Time (min)

Figure 6

ACS Paragon Plus Environment

500

Page 41 of 44

14000

111⁰C

12000

100⁰C 95⁰C

E* (J/mol)

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

Energy & Fuels

10000

8000

6000

4000 0.000

0.005

0.010

0.015

0.020

Equilibrium water uptake (mol/g)

Fig 7

ACS Paragon Plus Environment

0.025

Energy & Fuels

0.012 0.01

f(E*) (g.mol/g.J)

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 42 of 44

0.008 f(E*) 0.006 f(E*) exp. 0.004 0.002 0 0

4000

8000

12000

16000

20000

E* (J/mol)

Fig 8

ACS Paragon Plus Environment

24000

Page 43 of 44

120

Butanol concentration (v/v %)

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

Energy & Fuels

100 80 Fresh Run

60

1st reuse 2nd reuse

40

3rd reuse 15th reuse

20

16th reuse 0 0

20

40

60

80

100

120

140

160

Time (min) Figure 9 new

Regenerability and Reusability of PECM

Experimental conditions: All runs- 111°C,201 kPa,55%Bu-OH,3 ml/min FR,PECM 0.424-1.18 mm

ACS Paragon Plus Environment

180

Energy & Fuels

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 ACS Paragon Plus Environment

Page 44 of 44