The Role of Mesopore Structure of Hierarchical Porous Carbons on

Subscriber access provided by UNIV AUTONOMA DE COAHUILA UADEC

Article

The Role of Mesopore Structure of Hierarchical Porous Carbons on the Electrosorption Performance of Capacitive Deionization Electrodes Turki Baroud, and Emmanuel P. Giannelis ACS Sustainable Chem. Eng., Just Accepted Manuscript • DOI: 10.1021/ acssuschemeng.8b05782 • Publication Date (Web): 21 Mar 2019 Downloaded from http://pubs.acs.org on March 21, 2019

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

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

Page 1 of 50 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 Sustainable Chemistry & Engineering

The Role of Mesopore Structure of Hierarchical Porous Carbons on the Electrosorption Performance of Capacitive Deionization Electrodes Turki N. Barouda,b, Emmanuel P. Giannelisa* a.

Department of Materials Science and Engineering Cornell University, 326 Bard Hall, 126 Hollister Drive, Ithaca NY 14853, USA

b.

Department of Mechanical Engineering, King Fahd University of Petroleum & Minerals, Dhahran 31261, Saudi Arabia * E-mail:: [email protected] (E.P. Giannelis)

Keywords: Porosity; Surface area; Pore size; Pore volume; Hard template; Ice templation; Electric double layer; Water treatment; Desalination.

Abstract Capacitive deionization, (CDI) is a promising alternative approach for water desalination and treatment. Hierarchical Porous Carbons, HPCs, have been viewed as a promising porous structure material for electrosorption purposes. However, limitations associated with the synthesis and porosity control of HPCs limit their utilization as model systems in correlating the textural characteristics and the CDI performance. Here we report for the first time a systematic investigation using a wide range of tightly control primary mesopore size, mesopore surface area, mesopore volume and high mesopore fraction synthesized by the ice templation approach and correlate to their CDI performance. Larger mesopores are preferable for faster ion removal as they can provide easier pathways for the ions to diffuse and establish the electric double layer. However, smaller mesopores are more preferable in order to achieve higher salt capacity. While for mesomacro HPCs the salt capacity scales up with the mesopore surface area, HPCs that contain all levels of porosity (i.e. micro-meso-macro) do not show such correlation. Besides the excellent CDI

ACS Paragon Plus Environment

1

ACS Sustainable Chemistry & Engineering 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 2 of 50

performance reported, the model systems allow to delineate the role of several materials design parameters and correlate with their electrosorption behavior.

Introduction The World Economic Forum has consistently identified water shortage among the top global risks over the last several years1. Thus, pushing the limits of existing and developing new techniques for desalination technologies have become a critical imperative towards resolving the shortage and inaccessibility to pure water. Capacitive deionization, CDI has emerged as a promising alternative to other more conventional desalination techniques2. The concept of capacitive deionization relies on the basis of electric double layer formation at the interface between the electrode surfaces and the ionic solution, which results in temporary immobilization of ions (Capacitive) and consequently reduces the ions concentration in the solution (deionization)3. Therefore, increasing the amount of the accessible surfaces to the ions is critical toward maximizing the capacitive deionization performance. Typically, carbon is used as the active material for the electrodes and is considered to be at the heart of this application. Up to now, various types of carbons have been synthesized and evaluated as active material for capacitive deionization electrodes. They include carbon aerogels4–6, carbon fibers7,8 , graphene 9– 11

, activated carbon12–14, mesoporous carbon15–17, hierarchical porous carbons 18–21, and composite

carbon22,23. From the materials side, high surface areas can be obtained by synthesizing porous carbons with abundant micropores, however, this level of porosity has been associated with two main issues; 1) kinetic resistance for the ion transport in/out of the carbon structure24, 2) electric double layer overlap, which might eliminate the contribution of many of the available micropores25. However, in 2006 Chimola et al. working against the common


2

Page 3 of 50 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


notion about the effectiveness of small pores they showed that an anomalous increase in the charge capacitance can be attained by pores that are smaller than the size of solvated ions26. Similarly, by using carbide-derived carbon Porada et al. showed the significance of sub-nanometer pores for ion adsorption27. While the impact of the electric double layer overlap in small pores during electrosorption is still debatable, many (including those that showed the relevance of sub-nanometer pores)28 believe that an electrode material with structure that posseses only micropores is not preferable for the electrosorption because such design leads to non-accessible surfaces by ions and high mass transfer resistance. Despite anticipated limitations that are associated typically with micropores, which might hinder ionic diffusion, it is still believed that micropores are an important part of the porous structure for higher ion adsorption due to the resulting higher surface area and higher electric fields. Thus, increasing the accessibility of the micropores is a key design parameter in order to improve the desalination performance of capacitive deionization electrodes. This goal can be achieved through several approaches including integrating the microporosity with other level of porosities (i.e meso and/or macro)21 as described in this work, and through careful engineering of the carbon particles. Interestingly, precise tailoring of the carbon particles by reducing the carbon particles size, enhancing the pore connectivity and generating accessible micropores can substantially enhance the electrosorption behaviour29,30. Accordingly, optimizing the design parameters of carbon particles will provide shorter diffusion length for easy ion transport in and out the carbon particles, and provide better utilization of the available micropores. On the other hand, mesoporous materials represent another class of porous carbons, which have been widely regarded by several researches as preferable for an optimum performance. Besides the role mesopores play toward the salt capacity, they also they play a crucial role during the


3


Page 4 of 50

electrosorption experiments by providing easy pathways for ion transport in/out of the carbon matrix. Hence, mesopores can increase the accessibility of the carbon surfaces and potentially enhance the salt removal rate. Up to now, various types of mesoporous carbons have been successfully synthesized and evaluated as capacitive deionization electrodes

15,16,31,32

. Some of

these studies have reported better electrosorption performance compared with activated, micropore bearing only carbons. However, mesoporous carbons typically exhibit low surface areas compared to those based on micropores. Recently, there have been several efforts to design hierarchical porous structures that integrate both micro- and mesoporosity as a means to develop efficient electrosorption materials

19,33,34

.

Hence, amongst different types of carbon structures, hierarchical porous carbons have been proposed as promising materials for capacitive deionization electrodes due to their excellent electrosorption performance. The textural characteristics of these hierarchical carbons suggest the significance of the coexistence of micropores and mesopores toward high salt capacity and fast salt removal. Despite the good deionization performance of the reported HPCs, there are number of challenges that are associated with the adapted approaches to synthesize HPCs. Most of the reported synthesis of HPCs typically involve complicated steps, costly procedures, and might result in secondary environmental pollution 35. Another issue with most of the adapted approaches is related to the limitation of synthesizing various carbons with wide range of pore characteristics. In spite of the importance of mesopores as part of the HPCs, still many HPCs lack the tight mesopore size control (instead they exhibit broad BJH peaks) as a result of template aggregations or other factors, along with a low mesopore fraction 21,36. Consequently, these challenges limit the HPCs wide utilization in correlating between the textural characteristics and the CDI performance in order to optimize and maximize the benefits of the HPCs as an electrode material.


4

Page 5 of 50 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


Several attempts have been reported to correlate the electrosorption performance with the carbon pore structure characteristics (i.e. pore volume, pore size, surface area). Wang et al. investigated the performance of six activated carbons with narrow range of pore volume (0.5-1.1 cm3g-1) and wide range of surface area (917-2030 m2g-1) and they showed that the salt capacity and mesopore volume positively correlate with the specific surface area. They also found that the carbon with the highest mesopore fraction (40%) exhibit better electrosorption performance

37

. Other

researchers reaches similar conclusions where they confirmed the superiority of structures that combine micropores and high fraction of mesopores 38,39. Porade et al. evaluated four microporous carbons (two carbide-derived carbons and two commercial carbons) with surface area ranging between from 1100 to 1700 m2 g-1, total pore volume (0.43-0.7 cm3 g-1) and pore size less than 2nm as capacitive deionization electrode27. Interestingly, they found that while both of the BET surface area and pore volume show negative correlation with salt capacity, the salt capacity positively correlates with pore volume of sub-nanometer pores. Similar conclusion was reached by Amir et al., where they compared the electrosorption performance of three activated biochar (dominated by micropores, mesopores or a combination of micropores-mesopores) with the mesopore fraction up to 72% and surface ranging from 971 to 1674 m2 g-1 mesopore volume (0.09 - 0.94 cm3 g-1) average pore size (1.9 - 2.97 nm)

40

. They found that carbons with similar sub-

nanometer pore volume exhibited similar amount of monovalent removal regardless of BET surface area and pore volume. Yeh et al. studied the impact of tuning the mesopore ratio of activated carbons with wide range of surface areas (903 - 2481 m2 g-1)12. Despite that their work was limited by narrow range of total pore volume (0.99 - 1.5 cm3 g-1) and only had mesopore fraction less than 75%, they found that the coexistence of large fraction of mesopores along with the micropores, increase the accessible surfaces and ultimately enhance the electrosorption process. In another study by Porada et al. the contribution of pore volume toward the final salt


5


Page 6 of 50

capacity depended on the pore size. They also found that designing a carbon structure with a multilevel of porosity (micro, meso, macro) is a strategic approach toward advancing the performance of the electrodes 28. Although most of these studies covered a wide range of surface areas, they focused on porous carbons that exhibit narrow pore volume range, mesopore fraction less than 80%, limited pore volume (less than 3 cm3 g-1) and mesopore size less than 4 nm. Hence, it is still unknown whether the reported correlations hold for higher pore volumes, higher mesopore fractions and higher mesopore sizes.

In addition, some of these studies employed different types of carbons with varied synthesis conditions and precursors leaving many other variables such as the surface chemistry and degree of graphitization untested. Therefore, in order to investigate and fully understand the impact of using mesopore containing carbons for electrosorption processes, it is crucial to evaluate HPCs that are synthesized using the same precursors and under similar conditions, and potentially allow tight control of mesopore size (i.e. carbons with a sharp peak in BJH plot) along with the ability to tune pore volume over a wide range. Here we report for the first time a systematic investigation using HPCs covering a wide range of primary mesopore size (6 - 17 nm), mesopore surface area (516 - 1640 m2 g-1), mesopore volume (1.1 - 4.3 cm3 g-1), mesopore fraction (85 – 99 %) and more importantly narrow pore size distribution. The HPCs were synthesized using the same precursors and enabled by the ice templation approach. The adapted synthesis approach eliminates other effects that might arise from using different precursors or different synthesis conditions. Due to the ease of precise and tight control over the mesopore size and volume that our approach provides, the resulting carbons represent excellent model systems to study the effect of varying the mesopore structure on


6

Page 7 of 50 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


performance. Understanding the role of the materials design parameters of porous carbon on the electrosorption is especially important to maximize the capacitive deionization electrode performance. We believe that both the chemical and physical characteristics of the CDI electrode material are critical in determining the electrode performance. For example, Gao et al. and others have focused on the chemical structure and charge of the carbon electrodes 41,42 while others have reported on the physical structure of the carbon materials28,43. In this work, we focus mainly on the physical characteristics of the porous carbons during the electrosorption experiments. To assure a systematic probing of the relationship between the pore matrices and the electrosorption performance, the synthesized porous carbons were categorized into two main groups; category I includes predominantly meso-macro porous carbons and category II includes carbon with a combination of porosities (micro-meso-macro ). For the category I, the target is to investigate how tuning the mesopore size, mesopore volume and the mesopore surface area affect the final performance. For category II, the aim is to understand how different mesopore structures behave as capacitive deionization electrodes after the introduction of micropores. The outcomes of this study (i.e. the existing correlations, broad trends, synergetic effects of micro-meso structure) can contribute toward advancing our understanding about the impact of some of the design parameters on the ions adsorption and ultimate performance.

Experimental Synthesis of hierarchical porous carbons, HPCs The synthesized HPCs are classified into two categories according to their level of porosities. Category I: meso-macro porous carbons All the carbons in this category are non-activated carbons (NA) with mainly meso-macro structure. The non-activated carbons were synthesized by dissolving a specific amount of sucrose in a


7


Page 8 of 50

colloidal silica suspension. The mixture solutions were prepared using different weight ratios (y:z) of silica to sucrose, and colloidal silica suspensions with different average colloidal silica size (T) (nm) (as provided by the supplier). The samples in this non-activated NA category were denoted TyzNA. To prepare 412NA, a weight ratio of (y:z= 1:2) was achieved by preparing a solution mixture composed of 15 g of a 15 wt.% colloidal silica suspension (2.25 g silica) with an average silica template size (T = 4 nm) and 4.5 g sucrose. Next, a centrifuge tube filled with the mixture solution was subjected to a freeze-casting condition in a liquid nitrogen bath before it was transferred to freeze-dryer machine (room temperature, 0.014 mBar) for at least 2 days in order to sublime the ice crystals from the sucrose-silica composite. For carbonization step, a crucible loaded with the resulted white mixture was placed in a tube furnace under a nitrogen environment and heated at 1050°C for period of 3 hours at 3 °C min-1 heating rate. Then, the produced carbonsilica composite was loaded in a bottle that contains 3 M sodium hydroxide and stirred overnight at 80 °C. Next, the resulted carbon powders were rinsed with deionized water DI (until the pH comes down from 11 to around 7), and dried in a vacuum oven for 24h at 80 °C.

Category II: micro- meso-macro porous carbons To introduce micropores, CO2 gas was flowed over the sample in a tubular furnace at a flow rate of 40 cm3 min-1 at 950 °C for different times (H). All the samples in this category are denoted by A which means an activated carbon. For instance, 2021A1.5H is an activated version of the 2021NA, and the duration of physical activation is 1.5 hour.

CDI electrode fabrication


8

Page 9 of 50 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The capacitive deionization electrodes (anode and cathode) were fabricated by preparing a slurry composed of HPCs powder (as an active material) polyvinyledene fluoride (PVDF) (as a binder) and SuperP carbon black (as a conducting agent) with following weight ratio 85: 10: 5, respectively. The PVDF was dissolved in N-methyl-2pyrrolidone (NMP) (as a solvent) with 1:50, PVDF:NMP weight ratio. Films with ~ 60 microns thickness was made by casting the slurry on square graphite sheets with total area of 36 cm2. Prior to slurry casting, a hole with an area of 1.8 cm2 was punched in the center of the graphite sheets. The films on graphite sheets were dried overnight at 80°C. Further heat treatment was carried out on the films in a vacuum oven overnight at 80 °C to ensure the removal of the solvent. The mass of the HPCs powder in each electrode was in the range between 20 to 30 mg

Capacitive deionization experiments Prior to the CDI tests, all the electrodes were immersed in a deionized water DI bath for at least 24 hours. In a typical CDI cell, two coated graphite sheets separated by 700 microns glass fiber spacers were placed parallel to each other and sandwiched in a plexiglass housing. The upper part of the plexiglass housing was drilled to have two water corner inlets and one center outlet concentric with graphite sheets hole. A batch mode capacitive deionization experiments was used to evaluate the electrosorption performance of HPCs. A peristaltic pump (with 0.416 mL s-1 flow rate) was used to circulate 48 mL volume of salt solution (5 mM NaCl) between an outside reservoir and the cell in a room temperature environment. The salt solution flow into the cell through two corners inlets and forced to pass between the parallel electrodes (through the middle spacers) before it leaves the cell from the middle hole. To monitor the pH of the salt solution, a custom-built flow-through cell equipped by pH electrode was placed between the cell and the reservoir. The salinity


9


Page 10 of 50

of the circulated water was monitored by placing a conductivity electrode in the reservoir. The conductivity measurements were later converted to concertation values to calculate the mass of salt that exist in the solution. In a typical CDI experiment, each cycle composed of two consecutive steps; adsorption and desorption. During the adsorption step the electrodes were polarized at 1.2 V for a course of 65 minutes. After that, a zero voltage was applied for 120 minutes during the desorption stage. All the reported measurements were taken after the first cycles. The calculation of the salt capacity (G, mg g-1) of each electrode was carried out according to Eq.(1): G=

"#$ %#& '(

(1)

)

Where 𝐶+ ( mg L-1 ) is the initial salt concertation and 𝐶, ( mg L-1 ) is the final salt concertation, 𝑉 ( L) is the total circulated volume of the solution, and m (g) is the total mass of the active material (HPCs powders) in all the electrodes that were used during the experiments. The average salt removal rate ASAR was calculated by the following Eq. (2):

𝐴𝑆𝐴𝑅12 24)567 =

89:; NO;P)

(2)

The charge efficiency was calculated by the following equation Eq.(3): Charge efficiency (%) = =

Q# R S ∫ UV2

(3)

Where the SC (mol g-1) is the salt capacity, F (96485 C mol-1) is Faraday’s constant and I (A) is the current supplied during the adsorption step.

Structural characterization


10

Page 11 of 50 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


Nitrogen adsorption-desorption isotherms were carried out by Micrometrics ASAP2020 instrument at 77 K. Specific mesosurface areas were calculated from Brunauer-EmmettTeller (BET) modal and the mesopore volume was determined using Barrett-JoynerHalenda (BJH) model. Renishaw InVia Confocal Raman microscope (with an excitation power of 10 mW at 514 nm) was used to record the Raman spectra. To analyse the surface chemistry of the hierarchal porous carbons, X-ray photon spectroscopy XPS was carried out using Surface Science Instruments (SSX-100) with monochromatic Al K-α x-rays (1486.6 eV).

Electrical conductivity A lab-made 4-point probe setup was used to measure the electrical conductivity of the synthesized carbon powders. Briefly, a pressure of 8000 psi was applied to press the carbon particles inside a mold which has two metal ends. Upon passing a known current, the potential across the sample was measured by using two metallic probes separated by a fixed distance44.

Results and discussion Morphology of hierarchical porous carbons HPCs A series of hierarchal porous carbons that demonstrates a wide range of a controllable specific mesosurface area (516- 1640 m2 g-1 ) and mesopore volume (1.1 to 4.3 cm3 g-1) was synthesized via the ice-templation approach. The adapted approach composed of the following main steps: 1) high rate freeze casting step (of the solution mixture) where the


11


Page 12 of 50

colloidal silica were locked by the fast grown ice crystals, 2) macropores generation through the sublimation of ice crystals, 3) carbonization, 4) mesopore creation by the removal of silica hard template silica used

45

. Figure 1(c-f) shows the high magnification

transmission electron microscopy TEM images of HPCs. The TEM images suggest that the synthesized HPCs are highly mesopores with a sponge-like morphology and interconnected channels. The mesopore size that replicate the size of silica template shown in the TEM images is in good agreement with the N2 adsorption data. BJH pore size distribution can provide significant information about the pore characteristics of hierarchal porous carbons. Fig. 2(a-c) shows the BJH plots for a series of non-activated carbons (category I). Initially all the carbons display a single and sharp peak centered around 6, 12 and 17nm that corresponds well with the size of the silica particles template that were selected during the synthesis. For all the carbons increasing the pore volume leads to slight broadening of the BJH peaks and slight shift in the main peak as in (412NA, 411NA) and (1211NA, 1221NA). Toward the left of the BJH plots a small shoulder appears for all the carbons, which corresponds to small mesopores (2 - 4 nm). This shoulder can be attributed to the lack of complete filling of the interparticles spaces by sucrose during the ice templation step. Theses spaces converted into pores after the etching and carbonization step.


12

Page 13 of 50 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


(b)

(a)

5 µm

10 µm (c)

(d)

50 nm (e)

20 nm (f)

50 nm

20 nm

Fig. 1 (a,b) representative SEM images of hierarchical porous carbon showing surface morphology and texture. TEM images of (c,d) 412NA, (e,f) 1211NA HPC illustrating Spongelike morphology with mesopores replicating the hard template size. ACS Paragon Plus Environment

13


Page 14 of 50

Fig. 2 (d-f) illustrates the BJH pore size distribution of the porous carbons before and after the introduction of micropores during the activation step. While all the carbons possess a single main peak, the BJH pore size distribution for the activated samples show relatively broader peaks compared with non-activated samples consistent with an increase in the BET surface area and pore volume. This broadening can be attributed to the creation of new pores (micropores and mesopores) as well as the widening of the existing mesopores 46. For some of the carbons (as in 411NA) there is a broad shoulder toward the right of the BJH plots, which corresponds to mesopores from 10 to 30 nm. This broadening can be attributed to a slight aggregation of the silica colloidal particles during the synthesis steps due to lack of enough sucrose to cover the silica particles. One way to minimize the right shoulder is to increase the fraction of sucrose prior to the freeze-casting. This effect can be clearly seen, when comparing 411NA with 412NA, which has double amount of sucrose (see Fig. 2a). Notably, this study aims to investigate the impact of a specific mesopore size and volume on the CDI performance of porous carbon electrodes. Hence, to achieve precise tuning of mesoporous structure, the ice-templating of the colloidal silica hard template particles approach was considered. During the solidification of water, the fast grown ice crystals expel and lock the colloidal silica within the sucrose matrix, which can largely minimize the potential aggregation of the silica particles. Interestingly, the resulting mesopore size corresponds well with the colloidal silica hard template size. In our approach, minimizing the aggregation of silica particles can be controlled by the speed of freezing and by the silica/sucrose ratio. Increasing the silica/sucrose ratio will decrease the amount of sucrose present around the silica particles, leading to aggregation of the latter. The impact of increasing silica/sucrose ratio increases as the colloidal silica size decreases. For instance, the ice templating of 4 nm colloidal


14

Page 15 of 50 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


silica particles using the ratios of 1:2, 1:1, 2:1 results in 6, 8, 21 nm mesopore size, respectively (see Fig. S1). Thus, in this study for 4 nm series we only considered 1:2 and 1:1 ratios (i.e 412NA, 411NA) as they demonstrate good agreement between the mesopore size and the colloidal silica size. Fig. 2f depicts the BJH pore size distribution of the 2021NA before and after the activation. Although activation leads to an increase in the small mesopores as it is shown from the left shoulder, all the carbons still possess one main peak corresponding to the same pore size around 17 nm. Besides the BJH, the isotherms can also reveal the nature of porosity. Fig. 3a displays the isotherms for three non-activated mesoporous carbons that have 6, 12, 17 nm mesopores size. Clearly all the carbons display one condensation step, which corresponds well to the single sharp peak in BJH plots in Fig. 2(a-c). The condensation step for 412NA, 1211NA and 2011NA starts at

P Po-1 = 0.54, 0.70, 0.74, respectively. The higher the pressure at the beginning of the

condensation step, the larger the size of the mesopore that leads to such condensation 47. Increasing the pore volume can be achieved by increasing the fraction of the silica template prior to the freeze-casting step. Accordingly, there is a high chance of slight aggregation as there are not enough sucrose molecules to wrap the silica template. Thus, increasing the pore volume leads to a second condensation step start at 0.89, 0.93 and 0.9 for 411NA, 1221NA and 2021NA as in Fig. 3b. This second condensation step coincides well with right shoulder that appears in the BJH plots (see Fig. 2(a-c)).


15


Page 16 of 50

(a)

(b)

(c)

(d)

(e)

(f)

Fig. 2 BJH pore size distributions plots of category I (a-c) and category II (d-f) HPCs, obtained via nitrogen adsorption measurement at 77 K.


16

Page 17 of 50 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 1 lists the textural characteristics for a series of hierarchical porous carbons, which can be classified into two main categories; non-activated HPCs (category I), and activated HPCs (category II). The non-activated carbons mainly possess two level of porosities: mesopores and macropores, with mesopore surface area from 516 up to 965 m2 g-1 and mesopore volume from 1.1 up to 2.9 cm3 g-1. This group corresponds to carbons with 3 primary mesopores; 6, 12 and 17 nm. For each mesopore size it shows the impact of increasing the fraction of silica template prior to freeze-casting on the surface area and pore volume. For example, 412NA has mesopore surface area 647 m2 g-1 and mesopore volume 1 cm3 g-1.

(a)

(b)

(c) (c)

(d)

Fig. 3 Adsorption-desorption isotherms and (b) BJH pore size distributions plots of category I (a,b) and category II (c,d) HPCs, obtained via nitrogen adsorption measurement at 77 K.


17


Page 18 of 50

However, doubling the fraction of the template leads to 411NA with almost twice the mesopore volume (2.176 cm3 g-1) and 37% higher mesopore surface area. Although 412NA and 411NA possess different pore volumes and surface area, both carbons display almost the same primary mesopore size thanks to the synthesis approach, which allow tight control over the mesopore structure. The activated HPCs (category II) contains carbons with three level of porosities, micro-meso- and macropores, and high mesopore surface area up to 1640 m2 g-1 and up to 4.3 cm3 g-1 mesopore volume. The carbons in this group represent the activated version of 412NA, 411NA and 2021NA. Typically, introducing micropores through the activation step results in an increase in the surface area and the micropores fraction

47

. Although the BET and BJH models are proper models to

consider, when dealing with mesoporous carbons, both fail to accurately analyze microporosity (less than 2nm). Thus, in this study, Non-Linear Density functional theory NLDFT method was employed to evaluate the micropore fraction of the activated hierarchical porous carbons HPCs (see Fig. S 2,3,8 & Table S1). For instance, activation of 2021NA for 90 minutes results in 2021A1.5H with 2793 m2g-1 micropore surface area, on the other hand shorter activation period (as in 2021A1H) leads to lower micropore surface area (2455 m2g-1).


18

Page 19 of 50 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 1 Surface area and porosity characteristics of the synthesized HPCs.

However longer activation of 2021NA also leads to an increase of the small mesopore fraction as it is confirmed by higher left shoulder in BJH plot of Fig. 2d. Due to the CO2 activation, the pore structure of the HPCs possess higher total pore volume and larger mesopore and micropore fractions. This impact in the pore structure is ascribed to the generation of extra pores and the enlargement of the pre-existed ones. It is worth mentioning that carful physical activation conditions (i.e temperature, duration) should be controlled to avoid the collapsing of the carbon pore structure. Notably, all the reported HPCs demonstrate average mesopore size that replicate the average template size. Remarkably, all the series of hierarchical porous carbons maintain higher mesopore fraction (above 85%) and the same mesopore size, which corresponds well to the size of the silica template. XPS analyses were carried out to evaluate the surface chemistry of the synthesized HPCs 48. Fig. 4 shows the XPS survey spectrums of all the HPCs in category I & II in a comparison with the


19


Page 20 of 50

non-etched 411NE (carbon-silica composite).). Notably, the non-etched sample 411NE shows peaks that corresponds to silicon (17 at. %), carbon (30.7 at. %) and oxygen (52.3 at. %).IN contrast, the XPS spectra of all the etched HPCs contain predominantly a C 1s peak (above 90 at. %) and a relatively small O 1s peak (below 10 at. %) and no silicon peak (see Table S2). Thus, this confirm that the removal of the silica hard template from the carbon-silica composite was successfully achieved during the etching step.

The HPCs were further analyzed using Raman spectroscopy to reveal more information about the local structure of the samples 49. Fig. 5(a & b) shows the Raman spectra for the hierarchical porous carbons with two main peaks centered around 1360 cm-1 and 1580 cm-1. Typically, the peak at around 1580 cm-1 is associated with G band which is related to the sp2 carbon bonds stretching, and the peak at around 1360 cm-1 corresponds to D bands which is associated with structure disorder (i.e. defects, imperfections) 50. Fig. 5c displays the effect of mesopore volumes in the Raman ratio (ID / IG). Doubling the mesopore volume for 412NA leads to an increase in the intensity ratios from 0.83 to 0.87. In contrast, doubling the mesopore volume of 1211NA and 2011NA shows no effect into the Raman ratio. Fig. 5d illustrates the difference between Raman ratio of two types of carbons 412 and 2021 as a function of mesopore surface area, before and after the activation. Clearly, both carbon types 412 and 2021 show consistent increase in the Raman ratio with the surface area. Obviously, the smaller the pore size (as in 412 type) the higher is the Raman ratio. Notably, the intensity of D-peak, which is related to the lattice distortion, increases with CO2 activated carbons which can be attributed to the introduction of small mesopore and micropores to the carbon matrix 51.


20

Page 21 of 50 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Fig. 4 XPS spectra of all the HPCs that belong in categories I & II. 411NE non-etched carbon-silica composite.


21


(a)

(b)

(c)

(d)

Page 22 of 50

Fig. 5 Raman spectra of HPCs in category I (a) and category II (b); (c) Raman intensity ratios for category I HPCs; (d) Raman intensity ratios of 412 and 2021 HPCs before and after the activation treatment.


22

Page 23 of 50 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


Capacitive deionization performance of HPCs This section presents a systematic study of the porous carbons and the effect on the capacitive deionization behavior of HPCs (see Fig.S4). More specifically, the impact of varying the mesopore size, mesopore volume, surface area and the introduction of micropores on the current profile, salt capacity and salt removal rate was evaluated and studied systemically.

Correlation between pore structure of category I and current profile Fig. S5 shows the conductivity and current density profile during the electrosorption step for a series of HPCs. Theoretically, each electrode in the capacitive deionization technique will act as a capacitor during the charging and discharging steps. Thus, upon polarization a high initial value appears and then the charging asymptotically approaches zero as the electrode becomes fully charged 52. Upon electrode polarization, the current with an initial value decays first before it levels out. Moreover, all the hierarchical porous carbons illustrate similar profile. When a potential of 1.2 V is applied all the HPCs show similar current density profile, where the current starts with an initial value, and then the current decays until it reaches to a certain level as the ions electrosorped to the electrode surface and screen the charges. The initial value of the current density could be used as an indication of the salt removal capability. Fig. 6(a-c) compares the current density profile for six mesoporous carbons with initial values ranging between 1.1 A g-1 and 5.15 A g-1 corresponding to 1211NA and 2021NA, respectively. For all the HPCs that belong to category I, increasing the amount of mesoporosity leads to an increase in the initial current density. For instance, 1211NA shows initial value around 1 A g-1. Increasing the mesopore volume by increasing the amount of silica template prior to the freeze-casting produced 1221NA with more than three times higher current density initial value (4.3 A g-1). Comparison between


23


Page 24 of 50

the initial current density values of different primary mesopore carbons indicates that the initial current density does not show a dependency on the total pore volume; instead the current density depends on the surface area and mesopore size. For instance, 411NA has higher initial current density (~5.1 A g-1) compared with 1221NA although the former has lower mesopore volume. In addition, by comparing 412NA and 1211NA, where the former has higher mesopore surface area and the later one has higher mesopore volume, we show that the 412NA with higher mesopore surface area shows higher initial current density. Moreover, when considering different HPCs with similar surface areas (HPCs with difference in surface areas less than 100 m2g-1) regardless of the pore volume, smaller mesopore size leads to higher initial current density values (compare 411NA vs. 1221NA & 1221NA vs. 2011NA). These correlations that exist between pore characteristics and current profile correlate well with HPCs CDI performances as it will be discussed later.

Correlation between pore structure of category II and current profile Fig. 6(c-d) compares the current density profiles between activated and non-activated carbons. Clearly, the activated samples always show higher initial current density. Thus, the introduction of micropores to the HPCs during the CO2 activation results in higher initial current density values, which could be indicative of better salt capability. In contrast to 412NA and 411NA, the current density profile for 2021NA show higher values than its activated versions 2021A1H and 2021A1.5H. Consequently, as it turns out these lower initial current density values have an impact on the capacitive deionization performance of the activated HPCs of 17 nm version as it will be discussed later.


24

Page 25 of 50 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


(a)

(b)

(c)

(d)

(e)

(f)

Fig. 6 Current density profile of category I (a-c) and category II (d-f) in a batch mode flow-by CDI cell with static film electrodes in 5 mM NaCl solution and at 1.2 V.


25


Page 26 of 50

Correlation between pore structure of category1 and salt capacity To investigate the impact of doubling the pore volume for a specific mesopore size on the final salt capacity, a series of 6 hierarchical porous carbons, which represent 3 different primary mesopore pore sizes 6, 12 and 17 nm were tested. For each mesopore size the increase in the mesopore volume leads to higher salt capacity (see Fig. S7). For instance, for the 6 nm primary mesoporous carbons increasing the mesopore volume from =1.1 to 2.2 cm3 g-1 results in in salt capacity of 6.6 and 11 mg g-1, respectively. Notably, the mesopore size plays a significant role in the relation between the mesopore volume and the final salt capacity. The percentages of the additional salt capacity as a result of doubling pore volume are 67, 58, 12 % for 6, 12 and 17 nm, respectively. Fig. 7a displays the salt capacity obtained during the electrosorption cycle as a function of mesopore surface area for a series of mesoporous carbons. For HPCs with similar mesopore surface areas (HPCs with difference in surface areas less than 100 m2g-1) regardless of mesopore volumes, smaller primary mesopore size leads to higher salt capacity (compare 411NA (888 m2 g-1, 2.2 cm3 g-1, 11 mg g-1), 1221NA (795 m2 g-1, 2.8 cm3 g-1, 9 mg g-1) and 2011NA (708 m2 g-1, 1.9 cm3 g-1, 8 mg g-1) and 2011NA (884 m2 g-1, 2 cm3 g-1, 8 mg g-1)). These results show that as the mesopore size decreases, increasing the mesopore volume leads to a higher salt capacity. This trend can be explained by correlating between the strength of the electric field inside the pore and the mesopore size. Typically, smaller mesopore size results in higher electric field, which in turn leads to higher salt adsorption53. These results suggest that hierarchical porous carbons that have smaller mesopore offer more room for improving the desalination performance. Interestingly, the salt capacity correlated positively with the mesopore surface area of porous carbons and it did show a steady increase. Notably, the increase in the surface area from 516 m2 g1

up to 965 m2 g-1 results in roughly two times higher salt capacity. These results suggest that for

mesoporous carbons, the higher the mesopore surface area the higher the salt capacity.


26

Page 27 of 50 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


Accordingly, improving the salt capacity of the hierarchical porous carbons can be achieved by tuning the mesopore surface area. Fig. 7b illustrate the dependency of salt capacity on the total mesopore volume in the region between 1.1 up to 2.2 cm3 g-1 (regardless of the mesopore size). While there exists some correlation between pore volume and salt capacity in a specific mesopore volume region (1.3 - 2.2 cm3 g-1), where the salt capacity value increases from an (5.5 up to 12 mg g-1), obviously there is no universal trend in the relation between salt capacity and mesopore volume. Therefore, for mesoporous carbons the salt capacity does not necessarily scale with mesopore volume. Fig. 7c plots the salt capacity against the primary mesopore pore size from 6 up to 17 nm. Clearly, there is no direct correlation with regard to the pore size. Thus, the mesopore pore size alone as a design parameter is not explicitly a dominant factor toward high salt capacity. To prove the significance of mesopore surface area over the mesopore volume, two samples were compared: 412NA with higher mesopore surface area but lower mesopore volume (647 m2 g-1, 1.1 cm3 g-1) and 1211NA with lower mesopore surface area but higher mesopore volume (516m2 g-1, 1.3 cm3 g-1). Interestingly, 412NA with higher mesopore surface area showed higher salt capacity (6.6 mg g-1) than the 1211NA with higher mesopore volume (5.5 mg g-1).


27


(a)

Page 28 of 50

(b)

(c)

(d)

(e)

(f)

Fig. 7 Salt capacity vs mesopore surface area (a), mesopore volume (b), and primary mesopore size (c). (d-f) The average salt adsorption rate vs. the salt adsorption capacity. All are for category I HPCs and were evaluated in 5 mM NaCl solution and at 1.2 V.


28

Page 29 of 50 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


While increasing the surface area leads to higher salt capacity, the mesopore volume and mesopore size fail to show such trend. These results suggest that for mesoporous HPCs, the salt capacity scales with mesopore surface area, which makes the mesopore surface area a more effective design parameter than mesopore volume and mesopore size to predict and boost the salt capacities of the carbon electrodes.

Correlation between pore structure of category I and removal rates Besides the salt capacity metric, the HPCs performance can be further assessed by considering the rate at which the salt adsorbed to the carbon surfaces. Thus, the average salt adsorption rate ASAR metric will be used to understand the role of pore structure on the kinetics of salt adsorption. Fig. 7(d-f) shows CDI Ragone plots (ASAR against the salt capacity) for meso- macro carbons (category I). The figure can be utilized to correlate between the carbon structure and electrosorption performance. For each mesopore size the smaller the mesopore volume the faster each carbon reaches its salt saturation or equilibrium salt capacity (see Fig. 7d for comparison between 412NA and 411NA). Increasing the mesopore volume for a specific mesopore size leads to simultaneously higher salt capacity and slightly higher removal rate shown. One way to analyze the plot in Fig. 7 is to multiply the removal rates value ASAR by the salt capacity to identify the optimum operational conditions, and, hence, the optimum value of each carbon can be used for comparison purposes. The mesoporous carbon 2011NA shows the highest optimum value at 3 mg g-1 min-1 and 6 mg g-1 after 2 minutes of polarizing the carbon electrodes at 1.2 V. These results imply that larger mesopore size leads to faster electrosorption of the ions compared with those possessing small mesopore size. Interestingly, doubling the mesopore volumes show no significant difference between the salt removal rates of different mesopore pore size carbons.


29


Page 30 of 50

According to these results we suggest that in order to achieve higher removal rates for mesoporous carbons, the carbons have to be designed such that the they exhibit large mesopore pore size and small mesopore volume. Typically, the removal rate is a combination of many factors, which are related to the electrode material, and other factors such as electrode thickness, charging time and cell architecture54. In our case, since all the carbons were synthesized in a similar way using similar precursors and were evaluated in similar test conditions (cell architecture, charging voltage, charging time, etc.) we believe that the obtained ASAR values are mainly the result of the textural characteristics of the porous carbon and/or the film quality. Just by considering the porous structure of the carbons, the ASAR values can be a combination of the three factors; (1) the electric field within the mesopore, which gets larger as the mesopore size decreases; (2) the electrical conductivity of the porous carbons particles, where increasing the porosity of the structure can lead to lower electric conductivity44, and (3) the ion transport kinetics, where the existence of large mesopores can act as a transport channel for the ions, and as the channel size increases the ion transport will be faster. According to Table S3, the electric conductivity of the particles seems not to be directly correlated with removal rates. Thus, from the material side, we believe that both the electric field inside the mesopore and the ion transport kinetics are the main competing factors contributing toward the observed removal rates. Additionally, we have noticed that increasing the carbon mesopore volume usually leads to less dense particles, which could affect the packing density of the film and ultimately affect the ASAR values54. Therefore, manipulating the packing density (such as by compressing the films) can improve the electrical conductivity and the capacity of the electrodes55.


30

Page 31 of 50 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


Correlation between pore structure of category II and salt capacity Fig. 8a shows the salt capacity against the surface area for category II, which includes HPCs with micro-meso porous structure. While the mesoporous carbons demonstrate a positive correlation between the mesopore surface area and the salt capacity, the salt capacity for activated samples shows no universal dependent neither on the mesopore surface area nor on the micropore area (see Fig.S8a). Each carbon responds differently after the activation. For 6 nm carbons, more than 50 % additional salt capacity is seen for the 412NA after the introduction of micropores through the CO2 activation step for 3 hours. Notably, 4 hours CO2 activation for 412NA (non-activated sample) produces 412A4H with two times higher salt capacity than 412NA (12.6 mg g-1). These results confirm that for 6 nm primary mesopore pore size, introducing micropores leads to higher salt capacity56. Two samples were compared (411A1.5H and 412A4H) to investigate the role of increasing the mesopore volume prior to activation step on the electrosorption performance. Notably, 411A1.5H with 1372 m2 g-1 mesopore surface area and 2000 m2 g-1 micropore area possesses high salt capacity (14 mg g-1) compared with 12.6 mg g-1 for 412A4H, which has 63% higher micropore surface area. Therefore, increasing the mesopore fraction for the activated samples can potentially improve the utilization of the accessible surfaces of the hierarchical carbons. For the 17 nm pore HPCs, the increase of micropore fraction through activation process did not demonstrate a significant impact on salt capacity. This behavior can be ascribed to the electric conductivity of the films, where introducing new micropores and small mesopores to 2021NA during the activation step can potentially lead to lower density carbons and ultimately affect the packing density and the electric conductivity of the film. Consequently, the electric field is not being effectively developed in all particles in the film. Thus, regardless of the high micropore surface areas (up to 2793 m2 g-1) that 2021A1H and 2021A1.5H possess, they present lower current profile than the 2021NA (see Fig. 6f). Similar to the mesoporous carbons, the electrical


31


Page 32 of 50

conductivity of the activated carbons seems not to be an effective parameter for the removal rate at least within the range of the samples used in this work (see Table S3 to compare between 2021NA and 2021A1.5H). Therefore, in order to control the salt capacity when dealing with large mesopores (as in the case with 2021versions) more attention should be directed toward tackling the issue of the film electric conductivity and the packing density. In other words, considering the totality of the electrode rather than only changing the materials parameters for the porous carbons. Fig. 8b illustrates the relation between salt capacity and pore volume for a series of activated category II HPCs. The salt capacity increases by 40%, when the mesopore volume increase from 1.5 to 2.7 cm3 g-1. However, further increase from 3.8 to 4.3 cm3 g-1 in the mesopore volume results in lower salt capacities (around 10 mg g-1). These results show that the activated HPCs exhibit similar behavior to the non-activated mesoporous carbons in which there is no general trend that correlate salt capacity with the mesopore volume as well as the micropore volume (see Fig. S8b). Furthermore, the impact of the primary mesopore size on the electrosorption of the activated hierarchical porous carbons were investigated. All the salt capacities for a series of five samples were plot against the average pore size (Fig. 8c). The salt capacity of the activated carbons does not show any general trends with the average mesopore pore size.


32

Page 33 of 50 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


(a)

(b)

(c)

(d)

(e)

(f)

Fig. 8 Salt capacity vs mesopore surface area (a), mesopore volume (b), and primary mesopore size (c) of category II HPCs. (d-f) The average salt adsorption rate vs. the salt adsorption capacity of the non-activated and activated versions of HPCs. All the reported values were evaluated in NaCl solution and at 1.2 V.


33


Page 34 of 50

It is worth mentioning that, although 411A3.5H and 2021A1H have similar mesopore surface area, 411A3.5H possess 40% higher salt capacity compared with 2021A1H (although the later has higher micropore surface area). These results indicate that, when considering similar mesopore surface areas micro-meso-macro pores carbons, smaller mesopore size is more preferable in order to achieve higher salt capacity.

Correlation between pore structure of category II and removal rates Fig. 8 (d-f) shows the removal rates ASAR against the salt capacities for non-activated HPCs 412NA, 411NA and 2021NA, and their activated versions. Notably, all of the activated samples display a shift to the upper right side of the graph (except for 411A3.5). Thus, introducing micropores results in higher removal rates compared with the non-activated samples. This can be ascribed to the strong electric field that is associated with micropores and, therefore, the introduction of micropores can result in higher salt capacity and higher removal rates. Interestingly, while the introduction of micropores to 2021NA does not result in a considerable increase in the salt capacity, the introduction of micropores shows a significant improvement in the removal rates (see Fig. 8f). These findings suggest that when dealing with large mesopore size (as in 17 nm) increasing removal rates can be achieved either by lowering the mesopore volume as in category I (see Fig. 7f to compare 2011NA with 2021NA) or introducing micropores as in category II (see Fig 8f to compare 2021NA with 2021A1H and 2021A1.5H the activated samples). To investigate the impact of increasing the mesopore volume prior to the introduction of micropores on the CDI removal rate, two HPCs were considered 412A3H and 411A3.5H, where the former has around 20 % higher micropore volume and the latter has two times higher mesopore volume. Notably, while doubling the mesopore volume prior to the introduction of micropores as in 411A3.5H leads to higher salt capacity and slower removal rate, 412A3H with higher amount


34

Page 35 of 50 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


of microporosity demonstrates higher removal rate and lower salt capacity compared with 411A3.5 (see Fig 9a). Fig. 9b shows the CDI performance of two HPCs; 411A3.5H and 2021A1H which have similar surface area and micropore area. Notably, 2021A1.5H with larger mesopore size and higher micropore area exhibits higher removal rate where 411A3.5H with the smaller mesopore size leads to higher salt capacity. These results lead to the following important question; what makes 2021A1H carbon demonstrates higher removal rate compared to 411A3.5H is it because the 2021A1H has higher micropore area (as it has been concluded from Fig 8 & Fig 9a)? or/and because it has larger mesopore size? To answer this question, the performance of 2021A1H that exhibit larger primary mesopore and 412A3H which possesses lower mesopore size and higher micropore area were compered. Fig 9c shows that both of these HPCs demonstrate similar CDI performance. These results suggest that carbon with larger mesopores (2021A1H compered to smaller mesopore size as in 411A3.5H) are preferable for higher removal rate as it can provide easier pathways for the ions to diffuse into the sample to establish the electric double layer. Hence, larger mesopore size and higher micropore fraction are two competing factors toward higher removal rates for category II.


35


Page 36 of 50

(a)

(b)

(c)

Fig. 9 Average salt adsorption rate vs. the salt adsorption capacity of the activated versions of HPCs (a) 411A3.5H vs. 2021A1H; (b) 412A3H vs. 411A3.5H; (c) 412A3H vs. 2021A1H. All the reported values were evaluated in NaCl solution and at 1.2 V. The plots are accompanied by schematics demonstrating the textural characteristics of the samples.

The above results can be explained as follows; 1) decorating the mesopore walls with micropores results in micropores experiencing a stronger electric field (compared with that in the mesopores) more accessible, which in turn increases the removal rate; 2) higher amount of micropores does not necessary lead to higher salt capacity as some of these micropores will not be accessible. 3) in


36

Page 37 of 50 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


general, mesopores are considered to be an important factor toward increasing the removal rate, however, as the electric field inside the mesopore is weaker compared with that in the micropores, increasing the amount of mesopore with smaller portion of micropores (411A3.5H vs. 412A3H) result in higher salt capacity through the electric double layer formation but over longer periods of time. The analysis of the removal rates for category II suggest that, for micro-meso-macro structures, higher removal rates can be achieved by using larger mesopore size and/or by the introduction of micropores. Focusing on the materials side of capacitive deionization technology, the ultimate goal is to optimize the designing parameters of the CDI electrode in order to achieve maximum performance (i.e. high salt capacity, faster removal rate). Theoretically, materials with high porosity can achieve high equilibrium salt capacity. However, for the removal rate, more porosity could imply more diffusion for the ions into the carbon matrix for the electric double layer formation. Thus, materials with high salt capacity but slower removal rate as well as materials with high removal rate but low salt capacity generally are viewed as not very good candidates for capacitive deionization. Therefore, delineating key materials parameters and intelligent design of the carbon porous structure can pave the way for simultaneously attaining high salt capacity and faster removal rate. The reported findings in this work can provide insights into the role of various design parameters on performance. The reported findings in this work can provide insights into the role and the impact of each design parameter of the mesoporous structure on the CDI performance of the carbon electrode. Moreover, by employing equation 3, all the HPCs demonstrate comparable charge efficiencies (see Table S4). Additionally, Table 2 includes a comparison between our HPCs and a wide range of reported hierarchical pristine carbons. Clearly, our HPCs salt capacities are among the best reported ones.


37


Page 38 of 50

Table 2 Comparison of salt capacity electrosorption of various types of hierarchical pristine carbon materials from the literature

Cell voltage (V)

NaCl Concentration (mg L-1)

Salt capacity (mg g-1)

Year of publication

Activated mesoporous carbon sheets16

1.4

1000

3.5

2011

Hierarchical ordered mesoporous

1.2

290

12.8

2013

Graphene sponge57

1.2

500

14.9

2015

Hierarchically Porous Carbon derived

1.2

300

17.5

2015

1.4

300

13.5

2016

1.2

500

16.98

2016

Micro/mesoporous sheets11

1.2

300

15.34

2017

3D porous graphene18

1.4

300

8.97

2017

Hierarchical porous carbon61

1.2

580

7.75

2017

2

270

16.9

2018

1.5

295

9.6

2019

1.2

292

12.63

1.2

292

14.0

CDI electrode material

silicon carbide-derived carbon28

from PolyHIPE58 3D foam-like carbon nanoarchitectures59 Carbon nanofibers reinforced 3D porous carbon polyhedra network60

Hierarchical hole-enhanced 3D graphene assembly62 3D Channel-structured graphene63 Hierarchical porous carbons HPCs


This Work

38

Page 39 of 50 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


Conclusions In summary, we have systematically investigated the electrosorption behavior of several hierarchical porous carbons with a wide range of mesopore surface area (516 - 1640 m2 g-1), highly tunable pore volume (1.1 - 4.3 cm3 g-1), and tightly controllable mesopore pore size (6 - 17 nm). The synthesized samples were used as model systems to delineate The impact of several design parameters (i.e. mesopore surface area, mesopore size, mesopore volume) on the capacitive deionization performance. For meso-macro HPCs we found that as the mesopore size decreases, increasing the mesopore volume for a specific mesopore size leads to a higher salt capacity. Therefore, hierarchical porous carbons that have smaller mesopore pore size offer more room toward improving and tuning the desalination performance. However, within the range of tested primary mesopore size (6 - 17 nm) there is no universal trend that directly correlate the mesopore size with the salt capacity. Hence, the mesopore pore size alone is not a dominant factor toward controlling the salt capacity. Additionally, we found that for mesoporous carbons the mesopore surface area is more effective design parameter than mesopore volume and pore size in order to achieve high salt capacity. In terms of removal rates for mesoporous carbons, we found that higher values can be achieved by designing porous carbons with large mesopore size but small mesopore volume. Contrary to mesomacro HPCs that demonstrate a positive correlation between the mesopore surface area and the salt capacity, the salt capacity of activated carbons shows no universal dependent neither on the total mesopore surface area nor on the micropore area. Interestingly, each mesopore size for the meso-macro HPCs responds differently during the electrosorption experiment after the physical activation. For all the micro-meso-macro HPCs, introducing micropores enhances the salt removal rates. However, increasing the mesopore fraction of activated samples can potentially increase the number of the accessible sites within the hierarchical carbons. When dealing with large mesopores


39


Page 40 of 50

(as in the case of carbons with 17 nm) more attention should be directed toward tackling the issue of the film characteristics (i.e. electric conductivity and the packing density). In other words, considering the totality of the electrode rather than only changing the materials parameters for the porous carbons. Besides the high salt capacity and fast removal rates that can be obtained by the synthesized HPCs, we believe that our approach can offer a new platform to improve the understanding of the effect of electrode materials on CDI performance. Accordingly, this understanding of the role of some of the design parameter of porous carbons can guide and provide a roadmap for further investigation, better design and development of porous carbons and ultimately pave the way for practical applications of capacitive deionization. Supporting Information Electronic supplementary information (ESI) available: BJH pore size distributions plots of 412NA, 411NA and 421NA HPCs; NLDFT differential pore size distribution and cumulative size distribution of category II; N2 adsorption-desorption isotherms of category II ; Surface characteristics of category II; atomic concentrations of various hierarchical porous carbons obtained via XPS; electrical conductivity of various hierarchical porous ; carbon films coated graphite sheets, SEM image of the dried films, optical image of single HPC particle, TEM images of HPCs, and BJH pore size distribution; conductivity and current density profile ; pH profile for 411NA and 2021NA HPCs at 1.2 V in 5 mM NaCl solution; the impact of doubling the mesopore volume on the salt capacity of category I; the relation between salt capacity and micropore strucure of category II; Charge efficiency of various hierarchical porous . Acknowledgements This work is supported by King Abdullah University of Science and Technology (KAUST), KAUST baseline fund (KUS-C1-018-02). This work made use of the Cornell Center for


40

Page 41 of 50 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


Materials Research Shared Facilities supported through the NSF MRSEC program (DMR1120296). The authors would like to thank Dr. R. Sahore for her help with the electrical conductivity of the carbon particles. Turki Baroud would like to thank King Fahd University of Petroleum and Minerals (Scholarship number: DS/2095), Dhahran, Saudi Arabia for a Ph.D. scholarship and for their support.

Reference (1)

The Global Risks Report 2016 11th Edition Insight Report. World Econ. Froum 2016.

(2)

Subramani, A.; Jacangelo, J. G. Emerging Desalination Technologies for Water Treatment: A Critical Review. Water Res. 2015, 75, 164–187, DOI 10.1016/j.watres.2015.02.032.

(3)

Welgemoed, T. J.; Schutte, C. F. Capacitive Deionization Technology???: An Alternative Desalination Solution. Desalination 2005, 183 (1–3), 327–340, DOI 10.1016/j.desal.2005.02.054.

(4)

Yang, C. M.; Choi, W. H.; Na, B. K.; Cho, B. W.; Cho, W. I. Capacitive Deionization of NaCl Solution with Carbon Aerogel-Silicagel Composite Electrodes. Desalination 2005, 174 (2), 125–133, DOI 10.1016/j.desal.2004.09.006.

(5)

Xu, P.; Drewes, J. E.; Heil, D.; Wang, G. Treatment of Brackish Produced Water Using Carbon Aerogel-Based Capacitive Deionization Technology. Water Res. 2008, 42 (10– 11), 2605–2617, DOI 10.1016/j.watres.2008.01.011.

(6)

Yang, K. L.; Ying, T. Y.; Yiacoumi, S.; Tsouris, C.; Vittoratos, E. S. Electrosorption of Ions from Aqueous Solutions by Carbon Aerogel: An Electrical Double-Layer Model. Langmuir 2001, 17 (6), 1961–1969, DOI 10.1021/la001527s.

(7)

Pugazhenthiran, N.; Sen Gupta, S.; Prabhath, A.; Manikandan, M.; Swathy, J. R.; Raman,


41


Page 42 of 50

V. K.; Pradeep, T. Cellulose Derived Graphenic Fibers for Capacitive Desalination of Brackish Water. ACS Appl. Mater. Interfaces 2015, 7 (36), 20156–20163, DOI 10.1021/acsami.5b05510. (8)

Wang, G.; Pan, C.; Wang, L.; Dong, Q.; Yu, C.; Zhao, Z.; Qiu, J. Activated Carbon Nanofiber Webs Made by Electrospinning for Capacitive Deionization. Electrochim. Acta 2012, 69, 65–70, DOI 10.1016/j.electacta.2012.02.066.

(9)

Luo, G.; Wang, Y.; Gao, L.; Zhang, D.; Lin, T. Graphene Bonded Carbon Nanofiber Aerogels with High Capacitive Deionization Capability. Electrochim. Acta 2018, 260, 656–663, DOI 10.1016/j.electacta.2017.12.012.

(10)

Wang, H.; Zhang, D.; Yan, T.; Wen, X.; Shi, L.; Zhang, J. Graphene Prepared via a Novel Pyridine-Thermal Strategy for Capacitive Deionization. J. Mater. Chem. 2012, 22 (45), 23745–23748, DOI 10.1039/c2jm35340g.

(11)

Jia, B.; Zou, L. Graphene Nanosheets Reduced by a Multi-Step Process as HighPerformance Electrode Material for Capacitive Deionisation. Carbon N. Y. 2012, 50 (6), 2315–2321, DOI 10.1016/j.carbon.2012.01.051.

(12)

Yeh, C. L.; Hsi, H. C.; Li, K. C.; Hou, C. H. Improved Performance in Capacitive Deionization of Activated Carbon Electrodes with a Tunable Mesopore and Micropore Ratio. Desalination 2015, 367, 60–68, DOI 10.1016/j.desal.2015.03.035.

(13)

Oh, H.; Lee, J.; Ahn, H.; Jeong, Y.; Kim, Y.; Chi, C. Nanoporous Activated Carbon Cloth for Capacitive Deionization of Aqueous Solution. Thin Solid Films 2006, 515, 220–225, DOI 10.1016/j.tsf.2005.12.146.

(14)

Chia-Hung Hou, Jing-Fang Huang, Hong-Ren Lin, B.-Y. W. Preparation of Activated Carbon Sheet Electrode Assisted Electrosorption Process. J. Taiwan Inst. Chem. Eng. 2012, 43 (3), 473–479, DOI 10.1016/j.jtice.2011.12.003.


42

Page 43 of 50 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


(15)

Zou, L.; Li, L.; Song, H.; Morris, G. Using Mesoporous Carbon Electrodes for Brackish Water Desalination. WATER Res. 2008, 42, 2340–2348, DOI 10.1016/j.watres.2007.12.022.

(16)

Tsouris, C.; Mayes, R.; Kiggans, J.; Sharma, K.; Yiacoumi, S.; Depaoli, D.; Dai, S. Mesoporous Carbon for Capacitive Deionization of Saline Water. Environ. Sci. Technol. 2011, 45 (23), 10243–10249, DOI 10.1021/es201551e.

(17)

Chang, L.; Hu, Y. H. Excellent Capacitive Deionization Performance of Meso-Carbon Microbeads. RSC Adv. 2016, 6 (53), 47285–47291, DOI 10.1039/C6RA08362E.

(18)

Hu, W.; Xie, F.; Li, Y.; Wu, Z.; Tian, K.; Wang, M.; Pan, L.; Li, L. Hierarchically Porous Carbon Derived from PolyHIPE for Supercapacitor and Deionization Applications. Langmuir 2017, 33 (46), 13364–13375, DOI 10.1021/acs.langmuir.7b03175.

(19)

Zhao, S.; Yan, T.; Wang, H.; Zhang, J.; Shi, L.; Zhang, D. Creating 3D Hierarchical Carbon Architectures with Micro ‑ , Meso ‑ , and Macropores via a Simple Self-Blowing Strategy for a Flow- through Deionization Capacitor. ACS Appl. Mater. Interfaces 2016, 8 (28), 18027–18035, DOI 10.1021/acsami.6b03704.

(20)

Wang, D.-W.; Li, F.; Liu, M.; Lu, G. Q.; Cheng, H.-M. 3D Aperiodic Hierarchical Porous Graphitic Carbon for High Rate - Suporting Informaton. Angew. Chemie 2008, 373–376, DOI 10.1002/anie.200702721.

(21)

Wen, X.; Zhang, D.; Yan, T.; Zhang, J.; Shi, L. Three-Dimensional Graphene-Based Hierarchically Porous Carbon Composites Prepared by a Dual-Template Strategy for Capacitive Deionization. J. Mater. Chem. A 2013, 1 (39), 12334–12344, DOI 10.1039/c3ta12683h.

(22)

Yin, H.; Zhao, S.; Wan, J.; Tang, H.; Chang, L.; He, L.; Zhao, H.; Gao, Y.; Tang, Z. Three-Dimensional Graphene/Metal Oxide Nanoparticle Hybrids for High-Performance


43


Page 44 of 50

Capacitive Deionization of Saline Water. Adv. Mater. 2013, 25 (43), 6270–6276, DOI 10.1002/adma.201302223. (23)

Liu, Y.; Nie, C.; Liu, X.; Xu, X.; Sun, Z.; Pan, L. Review on Carbon-Based Composite Materials for Capacitive Deionization. RSC Adv. 2015, 5 (20), 15205–15225, DOI 10.1039/C4RA14447C.

(24)

Porada, S.; Zhao, R.; van der Wal, a.; Presser, V.; Biesheuvel, P. M. Review on the Science and Technology of Water Desalination by Capacitive Deionization. Prog. Mater. Sci. 2013, 58 (8), 1388–1442, DOI 10.1016/j.pmatsci.2013.03.005.

(25)

Hou, C. H.; Liang, C.; Yiacoumi, S.; Dai, S.; Tsouris, C. Electrosorption Capacitance of Nanostructured Carbon-Based Materials. J. Colloid Interface Sci. 2006, 302 (1), 54–61, DOI 10.1016/j.jcis.2006.06.009.

(26)

Chmiola, J.; Yushin, G.; Gogotsi, Y.; Portet, C.; Simon, P.; Taberna, P. L. Anomalous Capacitance Less Than Increase in Carbon at Pore Sizes Less than 1 Nanometer. Science (80-. ). 2006, 313 (5794), 1760–1763.

(27)

Porada, S.; Weinstein, L.; Dash, R.; Van Der Wal, A.; Bryjak, M.; Gogotsi, Y.; Biesheuvel, P. M. Water Desalination Using Capacitive Deionization with Microporous Carbon Electrodes. ACS Appl. Mater. Interfaces 2012, 4 (3), 1194–1199, DOI 10.1021/am201683j.

(28)

Porada, S.; Borchardt, L.; Oschatz, M.; Bryjak, M.; Atchison, J. S.; Keesman, K. J.; Kaskel, S.; Biesheuvel, P. M.; Presser, V. Direct Prediction of the Desalination Performance of Porous Carbon Electrodes for Capacitive Deionization. Energy Environ. Sci. 2013, 6 (12), 3700–3712, DOI 10.1039/c3ee42209g.

(29)

Xu, X.; Tang, H.; Wang, M.; Liu, Y.; Li, Y.; Lu, T.; Pan, L. Carbon Spheres with Hierarchical Micro/Mesopores for Water Desalination by Capacitive Deionization. J.


44

Page 45 of 50 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


Mater. Chem. A 2016, 4 (41), 16094–16100, DOI 10.1039/c6ta07616e. (30)

Liu, Y.; Pan, L.; Chen, T.; Xu, X.; Lu, T.; Sun, Z.; Chua, D. H. C. Porous Carbon Spheres via Microwave-Assisted Synthesis for Capacitive Deionization. Electrochim. Acta 2015, 151, 489–496, DOI 10.1016/j.electacta.2014.11.086.

(31)

Wang, H.; Shi, L.; Yan, T.; Zhang, J.; Zhong, Q.; Zhang, D. Design of Graphene-Coated Hollow Mesoporous Carbon Spheres as High Performance Electrodes for Capacitive Deionization. J. Mater. Chem. A 2014, 2 (13), 4739–4750, DOI 10.1039/c3ta15152b.

(32)

Li, L.; Zou, L.; Song, H.; Morris, G. Ordered Mesoporous Carbons Synthesized by a Modified Sol–gel Process for Electrosorptive Removal of Sodium Chloride. Carbon N. Y. 2009, 47 (3), 775–781, DOI 10.1016/j.carbon.2008.11.012.

(33)

Xu, X.; Tang, H.; Wang, M.; Liu, Y.; Li, Y.; Lu, T.; Pan, L.; Zhang, D.; Presser, V. Carbon Spheres with Hierarchical Micro/Mesopores for Water Desalination by Capacitive Deionization. J. Mater. Chem. A 2016, 4 (41), 16094–16100, DOI 10.1039/C6TA07616E.

(34)

Dutta, S.; Huang, S.-Y.; Chen, C.; Chen, J. E.; Alothman, Z. A.; Yamauchi, Y.; Hou, C.H.; C-W Wu, K. Cellulose Framework Directed Construction of Hierarchically Porous Carbons Offering High-Performance Capacitive Deionization of Brackish Water. ACS Sustain. Chem. Eng. 2016, 4, 1885–1893, DOI 10.1021/acssuschemeng.5b01587.

(35)

Zhang, L. L.; Zhao, X. S. Carbon-Based Materials as Supercapacitor Electrodes. Chem. Soc. Rev. 2009, 38 (9), 2520–2531, DOI 10.1039/b813846j.

(36)

Wang, G.; Qian, B.; Wang, Y.; Dong, Q.; Zhan, F.; Qiu, J. Electrospun Porous Hierarchical Carbon Nanofibers with Tailored Structures for Supercapacitors and Capacitive Deionization. New J. Chem. New J. Chem 2016, 3786 (40), 3786–3792, DOI 10.1039/c5nj02963e.

(37)

Wang, G.; Qian, B.; Dong, Q.; Yang, J.; Zhao, Z.; Qiu, J. Highly Mesoporous Activated


45


Page 46 of 50

Carbon Electrode for Capacitive Deionization. Sep. Purif. Technol. 2013, 103, 216–221, DOI 10.1016/j.seppur.2012.10.041. (38)

Li, G.; Hou, P.; Zhao, S.; Liu, C.; Cheng, H. A Flexible Cotton-Derived Carbon Sponge for High-Performance Capacitive Deionization. Carbon N. Y. 2016, 101, 1–8, DOI 10.1016/j.carbon.2015.12.095.

(39)

Zhao, S.; Yan, T.; Wang, Z.; Zhang, J.; Shi, L.; Zhang, D. Removal of NaCl from Saltwater Solutions Using Micro/Mesoporous Carbon Sheets Derived from Watermelon Peel via Deionization Capacitors. RSC Adv. 2017, 7 (8), 4297–4305, DOI 10.1039/C6RA27127H.

(40)

Dehkhoda, A. M.; Ellis, N. Effect of Activated Biochar Porous Structure on the Capacitive Deionization of NaCl and ZnCl 2 Solutions. Microporous Mesoporous Mater. 2016, 224, DOI 10.1016/j.micromeso.2015.11.041.

(41)

Gao, X.; Porada, S.; Omosebi, A.; Liu, K. L.; Biesheuvel, P. M.; Landon, J. Complementary Surface Charge for Enhanced Capacitive Deionization. Water Res. 2016, 92, 275–282, DOI 10.1016/j.watres.2016.01.048.

(42)

Gao, X.; Omosebi, A.; Landon, J.; Liu, K. Dependence of the Capacitive Deionization Performance on Potential of Zero Charge Shifting of Carbon Xerogel Electrodes during Long-Term Operation. J. Electrochem. Soc. 2014, 161 (12), E159–E166, DOI 10.1149/2.0561412jes.

(43)

Porada, S.; Sales, B. B.; Hamelers, H. V. M.; Biesheuvel, P. M. Water Desalination with Wires. J. Phys. Chem. Lett. 2012, 3 (12), 1613–1618, DOI 10.1021/jz3005514.

(44)

Sahore, R.; Levin, B. D. A.; Pan, M.; Muller, D. A.; DiSalvo, F. J.; Giannelis, E. P. Design Principles for Optimum Performance of Porous Carbons in Lithium-Sulfur Batteries. Adv. Energy Mater. 2016, 6 (14), 1–9, DOI 10.1002/aenm.201600134.


46

Page 47 of 50 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


(45)

Estevez, L.; Dua, R.; Bhandari, N.; Ramanujapuram, A.; Wang, P.; Giannelis, E. P. A Facile Approach for the Synthesis of Monolithic Hierarchical Porous Carbons – High Performance Materials for Amine Based CO2 Capture and Supercapacitor Electrode. Energy Environ. Sci. 2013, 6 (6), 1785–1790, DOI 10.1039/c3ee40549d.

(46)

Ahmadpour, A.; Do, D. D. The Preparation of Activated Carbon from Macadamia Nutshell by Chemical Activation. Carbon N. Y. 1997, 35 (12), 1723–1732, DOI 10.1016/S0008-6223(97)00127-9.

(47)

Groen, J. C.; Peffer, L. A. A.; Pérez-Ramírez, J. Pore Size Determination in Modified Micro- and Mesoporous Materials. Pitfalls and Limitations in Gas Adsorption Data Analysis. Microporous Mesoporous Mater. 2003, 60 (1–3), 1–17, DOI 10.1016/S13871811(03)00339-1.

(48)

Zhu, S.; Li, J.; He, C.; Zhao, N.; Liu, E.; Shi, C.; Zhang, M. Soluble Salt Self-AssemblyAssisted Synthesis of Three-Dimensional Hierarchical Porous Carbon Networks for Supercapacitors. J. Mater. Chem. A Mater. energy Sustain. 2015, 3, 22266–22273, DOI 10.1039/C5TA04646G.

(49)

Ferrari, A. C. Raman Spectroscopy of Graphene and Graphite: Disorder, Electron-Phonon Coupling, Doping and Nonadiabatic Effects. Solid State Commun. 2007, 143 (1–2), 47– 57, DOI 10.1016/j.ssc.2007.03.052.

(50)

Nakamizo, M.; Kammereck, R.; Walker, P. L. LASER RAMAN STUDIES ON CARBONS. Carbon N. Y. 1974, 12, 25–267.

(51)

Roh, J. Structural Study of the Activated Carbon Fiber Using Laser Raman Spectroscopy. Carbon Lett. 2008, 9 (2), 127–130, DOI 10.5714/CL.2008.9.2.127.

(52)

Zhao, R. Theory and Operation of Capacitive Deionization Systems; Wageningen University: Wageningen, NL, 2013.


47


(53)

Page 48 of 50

Oldham, K. B. A Gouy–Chapman–Stern Model of the Double Layer at a (Metal)/(Ionic Liquid) Interface. J. Electroanal. Chem. 2008, 613 (2), 131–138, DOI 10.1016/j.jelechem.2007.10.017.

(54)

Suss, S. M. E.; Presser, V. Water Desalination via Capacitive Deionization : What Is It and What Can We Expect from It ? Energy Environ. Sci. 2015, 8 (CDI), 2296–2319, DOI 10.1039/C5EE00519A.

(55)

Bhandari, N.; Dua, R.; Giannelis, E. P. A Combined Salt – Hard Templating Approach for Synthesis of Multi-Modal Porous Carbons Used for Probing the Simultaneous Effects of Porosity and Electrode Engineering on EDLC Performance. Carbon N. Y. 2015, 87, 29– 43, DOI 10.1016/j.carbon.2015.01.039.

(56)

Baroud, T. N.; P. Giannelis, E. High Salt Capacity and High Removal Rate Capacitive Deionization Enabled by Hierarchical Porous Carbons. Carbon N. Y. 2018, DOI 10.1016/j.carbon.2018.05.053.

(57)

Xu, X.; Pan, L.; Liu, Y.; Lu, T.; Sun, Z.; Chua, D. H. C. Facile Synthesis of Novel Graphene Sponge for High Performance Capacitive Deionization. Sci. Rep. 2015, 5, 8458, DOI 10.1038/srep08458.

(58)

Xu, X.; Liu, Y.; Wang, M.; Yang, X.; Zhu, C.; Lu, T.; Zhao, R.; Pan, L. Design and Fabrication of Mesoporous Graphene via Carbothermal Reaction for Highly Efficient Capacitive Deionization. Electrochim. Acta 2016, 188, 406–413, DOI 10.1016/J.ELECTACTA.2015.12.028.

(59)

El-Deen, A. G.; Boom, R. M.; Kim, H. Y.; Duan, H.; Chan-Park, M. B.; Choi, J. H. Flexible 3D Nanoporous Graphene for Desalination and Bio-Decontamination of Brackish Water via Asymmetric Capacitive Deionization. ACS Appl. Mater. Interfaces 2016, 8 (38), 25313–25325, DOI 10.1021/acsami.6b08658.


48

Page 49 of 50 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


(60)

Liu, Y.; Ma, J.; Lu, T.; Pan, L. Electrospun Carbon Nanofibers Reinforced 3D Porous Carbon Polyhedra Network Derived from Metal-Organic Frameworks for Capacitive Deionization. Sci. Rep. 2016, 6 (August), 1–9, DOI 10.1038/srep32784.

(61)

Duan, H.; Yan, T.; An, Z.; Zhang, J.; Shi, L.; Zhang, D. Rapid Construction of 3D Foamlike Carbon Nanoarchitectures via a Simple Photochemical Strategy for Capacitive Deionization. RSC Adv. 2017, 7 (62), 39372–39382, DOI 10.1039/C7RA04613H.

(62)

Li, J.; Ji, B.; Jiang, R.; Zhang, P.; Chen, N.; Zhang, G.; Qu, L. Hierarchical HoleEnhanced 3D Graphene Assembly for Highly Efficient Capacitive Deionization. Carbon N. Y. 2018, 129, 95–103, DOI 10.1016/j.carbon.2017.11.095.

(63)

Chang, L.; Hang Hu, Y. 3D Channel-Structured Graphene as Efficient Electrodes for Capacitive Deionization. J. Colloid Interface Sci. 2018, 538, 420–425, DOI 10.1016/j.jcis.2018.11.105.


49


Page 50 of 50

For Table of Contents Use Only

Synopsis: The use of hierarchical porous carbons enabled by ice-templation as a model system to delineate the impact of mesoporous structure on capacitive deionization performance.


50

The Role of Mesopore Structure of Hierarchical Porous Carbons on

Recommend Documents