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