Selective Low-Energy Carbon Dioxide Adsorption Using

---

Subscriber access provided by University of Florida | Smathers Libraries

Article

Selective Low-energy Carbon Dioxide Adsorption using Monodisperse Nitrogen-rich Hollow Carbon Submicron Spheres Aditya Farhan Arif, Yuma Kobayashi, Elia Michael Schneider, Samuel C Hess, Ratna Balgis, Takafumi Izawa, Hideharu Iwasaki, Shuto Taniguchi, Takashi Ogi, Kikuo Okuyama, and Wendelin J. Stark Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.7b01353 • Publication Date (Web): 07 Dec 2017 Downloaded from http://pubs.acs.org on December 8, 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.

Langmuir 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 20 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

Langmuir

1

Selective Low-energy Carbon Dioxide Adsorption

2

using Monodisperse Nitrogen-rich Hollow Carbon

3

Submicron Spheres

4

Aditya F. Arif1, Yuma Kobayashi1, Elia M. Schneider2, Samuel C. Hess2, Ratna Balgis1,

5

Takafumi Izawa3, Hideharu Iwasaki3, Shuto Taniguchi1, Takashi Ogi1,*, Kikuo Okuyama1, and

6

Wendelin J. Stark2

7

1

8

Japan

9

2

Department of Chemical Engineering, Graduate School of Engineering, Hiroshima University,

Institute for Chemical and Bioengineering, ETH Zurich, Vladimir-Prelog-Weg 1, 8093 Zurich,

10

Switzerland

11

3

12

Sakazu, Kurashiki, Okayama 710-0801, Japan

Battery Materials Research Laboratory, Kurashiki Research Center, Kuraray Co., Ltd., 2045-1,

13 14 15

16

ACS Paragon Plus Environment

1

Langmuir 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 20

1

ABSTRACT

2

Monodisperse, nitrogen-doped hollow carbon spheres of submicron size were synthesized using

3

hexamethoxymethyl melamine as both a carbon and nitrogen source in a short (1 h) microwave-

4

assisted synthesis. After carbonization at 550 °C, porous carbon spheres with a remarkably high

5

nitrogen content of 37.1% were obtained, consisting mainly of highly basic pyridinic moieties.

6

The synthesized hollow spheres exhibited high selectivity for carbon dioxide (CO2) over nitrogen

7

and oxygen gases, with a capture capacity up to 1.56 mmol CO2 g−1. The low adsorption

8

enthalpy of the synthesized hollow carbon spheres permits good adsorbent regeneration.

9

Evaluation on the feasibility for scaling-up shows their potential for large-scale applications.

10 11

KEYWORDS microwave, polystyrene latex, pyridinic nitrogen, Van der Waals, physisorption

12 13 14 15 16 17 18 19 20

ACS Paragon Plus Environment

2

Page 3 of 20 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

Langmuir

1.

Introduction

2

Hollow nanostructures1 have attracted an enormous amount of attention lately because of

3

their tunable and often remarkable properties (e.g., ultrahigh surface area, shell permeability to

4

allow improved internal access, high surface-to-volume ratio, and low density). Such

5

nanostructures are appealing for various applications, such as drug delivery,2 catalysis,3 and

6

energy conversion.4 Carbon-based materials have been viewed as promising inexpensive sources

7

as the scaffold for hollow nanostructures, mainly because they exhibit the well-known features of

8

carbon; namely, high chemical stability and electrical conductivity. Combining the

9

aforementioned properties of carbon-based materials with hollow spherical structures has led to a

10

novel kind of material known as hollow carbon spheres (HCS). Various synthetic pathways to

11

obtain HCS that exploit either a soft or hard templating approach have been investigated.5 The

12

former method involves the self-assembly of carbon precursors using a variety of different

13

surfactants or organic additives,6,

14

nanoparticles,9, 10, 11 polystyrene spheres,12, 13, 14, 15 or nanoparticles with a metal core.16, 17 HCS

15

have been applied in catalysis, carriers, and drug delivery.18 However, the intrinsic chemical

16

inertness of pure carbon structures renders them ineffective for numerous interesting applications

17

like CO2 adsorption.

7, 8

while the latter uses hard templates such as silica

18

Among the current commercial CO2 capture technologies, such as chemical absorption by

19

monoethanolamine (MEA) or monodiethanolamine (MDEA), physical absorption by Selexol,

20

and chemical adsorption, adsorption by carbon generally has the lowest CO2 capture ability.

21

However, carbon exhibits interesting advantages in CO2 adsorption, including low desorption

22

energy because of its weak adsorption interaction, easy adsorbent regeneration, and low cost.19, 20

23

Unlike most absorption chemicals, carbon is non-corrosive. Thus, the need for corrosion

ACS Paragon Plus Environment

3

Langmuir 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 4 of 20

1

inhibitors and corrosion-resistant alloys for the construction materials is suppressed or even

2

eliminated in CO2 adsorption by carbon. The low CO2 uptake can be overcome through

3

functionalization to create basic sites,21, 22 primarily through functionalization with nitrogen. A

4

general way to dope HCS with nitrogen is by polymerization of nitrogen-containing monomers

5

and subsequent carbonization. Usually, 3-aminophenol or aniline is used as the nitrogen donor.23

6

It was recently shown that using a melamine resin as the N-dopant increased the nitrogen-doping

7

content of carbon.24 However, this synthesis was conducted using ultrasonic spray pyrolysis,

8

which gave a low yield of polydisperse particles.25 A general, scalable process for the large-scale

9

production of N-doped HCS is needed. There are several reports of hydrothermal syntheses of N-

10

doped HCS with high monodispersity, but the synthesis time is generally long (up to 24 h).26, 27,

11

28

12

Herein, we report the synthesis of monodisperse N-doped HCS of submicron size via a short

13

(1 h) microwave-assisted synthesis and subsequent carbonization. Because of the high content of

14

nitrogen atoms in the aromatic rings of hexamethoxymethylmelamine (HMMM), which is used

15

as a nitrogen source, HCS with nitrogen contents as high as 37.1% after carbonization at 550 °C

16

are achieved. A preliminary evaluation of CO2 adsorption ability is also conducted at ambient

17

temperature and pressure. We evaluate feasibility of scaling up this process from the standpoints

18

of energy consumption and equipment size through comparison with a state-of-the-art

19

commercial CO2 removal process.

20 21 22 23

2.

Experimental

2.1. Synthesis of HCS

ACS Paragon Plus Environment

4

Page 5 of 20 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

Langmuir

1

In a standard procedure, positively charged PSL particles with a diameter of 220 nm (ζ-

2

potential = 46 mV, 0.9 mL, 1.1 wt%) were diluted in ultrapure water (125 mL). HMMM (1 g,

3

2.56 mmol) was added and then the solution was stirred for 1 hat ambient temperature. AIBA

4

radical starter solution (15 mL, 0.01 g mL−1, 0.55 mmol) was added to the nanoparticle

5

dispersion, which was subsequently heated using a microwave-assisted organic initiator

6

(Biotage, Uppsala, Sweden) for 1 h at 90 °C. The obtained particles were washed with water

7

twice using centrifugation at 10000 rpm and then dried at 40 °C in a vacuum oven over night.

8

The particles (~0.1 g) were heated in an electric ceramic furnace at 550 °C under nitrogen

9

atmosphere (0.5 L min−1) for 4 h for carbonization, yielding the HCS product (0.03 g).

10 11

2.2. Physical Characterization

12

The ζ-potential of the PSL particles was measured using a zetasizer (Zetasizer Nano ZSP,

13

Malvern Instruments Ltd., Malvern, UK). The morphology of the synthesized HCS was observed

14

using field-emission SEM (S-5000, 20 kV, Hitachi High-Tech., Tokyo, Japan) and TEM (JEM-

15

2010, 200 kV, JEOL Ltd., Tokyo, Japan). Fourier transform infrared (FT-IR) spectroscopy was

16

used to investigate the chemical reactions that occurred during the microwave irradiation process

17

(Spectrum One, Perkin Elmer Inc., Waltham, MA, USA). Elemental analysis was carried out

18

using XPS (PHI Quantera II, Physical Electronics, Chanhassen, MN, USA). The thermal

19

behavior of the materials before and after carbonization was investigated using TGA (TGA-

20

50/51 Shimadzu, Kyoto, Japan). The nitrogen adsorption–desorption properties of the HCS were

21

evaluated using gas adsorption measurement equipment (BELSORP-max, MicrotracBEL, Osaka,

22

Japan), and their surface area was calculated using the BET method based on N2 isotherm

23

curves. CO2 adsorption measurements were conducted using the same equipment as the BET

ACS Paragon Plus Environment

5

Langmuir 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 6 of 20

1

analysis. Prior to the adsorption analysis, each sample was degassed at 300 °C for 3 h. CO2

2

capture experiments were performed at 25 °C in a thermostatic bath.

3 46

3.

Results and discussion

47

To obtain monodisperse hollow carbon spheres with a high nitrogen content, the following

48

two-step synthesis pathway was pursued (Scheme 1). In the first step, HMMM was cross-linked

49

and condensed as both a carbon and nitrogen source onto monodisperse polystyrene latex (PSL)

50

nanoparticles via microwave-assisted polymerization at 90 °C for 1 h.

51

ACS Paragon Plus Environment

6

Page 7 of 20 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

Langmuir

Scheme 1. Illustration of the microwave-assisted process used to synthesize HCS.

HMMM is usually used as a thermally activated curing agent for polymers bearing hydroxyl

6

groups.29,

30

7

dinonylnaphthalene disulfonic acid as a catalyst at high temperatures (>150 °C).31 However in

8

this case, the use of this kind of acid catalyst did not result in efficient crosslinking. Therefore,

9

the ability of other reactants to crosslink HMMM and PSL was examined. Surprisingly, the

10

crosslinking worked well in the presence of 2,2′-azobis(2-methylpropionamidine) (AIBA), a

11

common water-soluble radical initiator. A plausible explanation for this finding is the radical-

12

induced oxidation of polystyrene by an AIBA/H2O/O2 system,32,

13

binding sites (i.e., hydroxyl and carboxyl groups) for the covalent attachment of HMMM via

14

ether bond formation (Fig. 1). Formation of the ether bond is confirmed by the peak at 1150 cm-1

15

in the FT-IR spectrum of the polycondensated HMMM (Fig. S1 of supplementary information).34

16

This novel reaction is currently the subject of further investigation in our laboratory. An

17

important factor for the successful synthesis of monodisperse polymer-coated nanoparticles was

18

the amount of HMMM used. We found that 0.7 wt% of HMMM was the ideal amount; higher

19

and lower amounts resulted in agglomerated particles that were not monodisperse (Fig. S2). In

20

the second step, the melamine-polymer coated particles (1) were carbonized in a furnace under

21

nitrogen atmosphere at 550 °C. As expected from the size distribution before carbonization, only

22

the sample synthesized using 0.7 wt% melamine had spherical geometry and a monodisperse size

23

distribution (Fig. S3).

For example, HMMM and substituted phenols were condensed using

33

thus generating possible

24

ACS Paragon Plus Environment

7

Langmuir 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 8 of 20

1 2

Figure 1. Oxidation of polystyrene nanoparticles and subsequent polycondensation of HMMM

3

as a possible reaction pathway to obtain HCS.

4 5

To investigate the effect of the template size on the HCS formed, two different sizes of

6

PSL nanoparticles (120 and 220 nm) were coated with polymer and subsequently calcined (Fig.

7

2). The larger PSL template resulted in monodisperse, N-doped HCS 2, while with the smaller

8

template gave a mixture of hollow and dense particles. A possible explanation for this

9

phenomenon is the intrinsic shrinking of the polymer shell by 50% to 60% depending on the PSL

10

size,12 which occurs when the PSL template is thermally removed at high temperature during

11

carbonization.

12

ACS Paragon Plus Environment

8

Page 9 of 20 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

Langmuir

1 2

Figure 2. Figure 2. (a, b) SEM images of polymer-coated PSL particles and (c, d) TEM images

3

of HCS. (a) and (c) were synthesized from a PSL diameter of 120 nm while (b) and (d) were

4

prepared from a PSL diameter of 220 nm. TEM image of polymer-coated PSL particles and SEM

5

image of HCS are provided in the supplementary information.

6 7

The shrinkage factor should be considered when choosing the size of the template

8

nanoparticles to obtain optimal results. Therefore, the PSL template with a diameter of 220 nm

9

was used in the following experiments. The thickness of the resulting carbon shell was ~49 nm.

10

It should be noted that the diameters of the core (110.3±3.6 nm according to the TEM image in

11

Fig. 2) and shell (207.8±3.9 nm) are quite uniform; i.e., they exhibit quite small standard

ACS Paragon Plus Environment

9

Langmuir

1

deviations. Other TEM images of the HSC showed similar distributions, thus we can consider the

2

HCS monodisperse.

3

Like the previously reported carbonization of melamine resin,24 thermogravimetric

4

analysis (TGA) of 1 in N2 showed four carbonization stages (Fig. 3(a)). We therefore expect that

5

these materials possess a similar carbonization mechanism. Similar with the previous research on

6

the melamine resin, the first heating stage may be assigned to water evaporation.24 Cleavage of

7

hydrocarbon and nitrogen bonds occurs up to 400 °C (stage II), while the considerable weight

8

loss in stage III is caused by thermal decomposition of PSL.12, 24, 35 Decomposition of PSL led to

9

the formation of hollow structure inside the particles, while the gas released by PSL

10

decomposition promoted the formation of micro- and mesopores within the shell. Weight loss in

11

stage IV is consistent with continuous carbonization.12 Meanwhile, 2 was stable in air after

12

carbonization at temperatures up to 440 °C (Fig. 3(b)), indicating these HCS should tolerate gas

13

adsorption/desorption cycles. Melamine@PSL in N2

80 60 40 I

20

II

III

IV

(b) 100 Remaining weight (%)

(a) 100 Remaining 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

Page 10 of 20

0

80 60 40 20 0

0

14 15 16

HCS in air

200

400 600 Temperature (ºC)

800

0

200

400 600 Temperature (ºC)

800

Figure 3. TGA of sample (a) 1 in N2 and (b) 2 in air at a heating rate of 5 °C min−1.

17 18

An important feature of this synthesis using HMMM as a nitrogen source (HMMM has a

19

nitrogen content of 21.5%) is the resulting amount of nitrogen in the final HCS 2, which can be

ACS Paragon Plus Environment

10

Page 11 of 20 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

Langmuir

1

as high as 37.1% (Table S1). In a similar approach to synthesize HCS using 3-aminophenol as a

2

nitrogen source, a nitrogen content of 15.6% was obtained;12 this value is high compared to that

3

reported in other literature,36, 37 but lower than that achieved with this protocol. A high nitrogen

4

content is important for substantial CO2 adsorption, because it increases the number of basic sites

5

on the carbon surface, thus promoting the adsorption of acidic CO2.38, 39 Additionally, the nature

6

of the nitrogen incorporated in the HCS is important. Therefore, in-depth X-ray photoelectron

7

spectroscopy (XPS) analysis of HCS 1 was performed (Fig. S6). The nature of nitrogen species

8

present in 2 was determined to be 60.1% pyridinic nitrogen moieties, 31.9% pyrrolic and amino

9

nitrogen, and only 7.9% graphitic nitrogen species. Pyridinic moieties on graphene exhibit a

10

positive pKa of around 5, so are highly basic compared with graphitic nitrogen (pkA = −29).40 It

11

is assumed that a high content of pyridinic nitrogen in HCS will translate to a high CO2 uptake.

12

However, another factor influencing CO2 adsorption is porosity; i.e., accessible surface area.

13

Brunauer–Emmett–Teller (BET) measurements revealed that 2 possessed a specific surface area

14

of 323 m2 g−1.

15

The N2 adsorption–desorption isotherms of 2 were measured (Fig. 4a) to further

16

investigate the nature of its porosity. A substantial monolayer and micropore adsorption occurred

17

below p/po = 0.05, followed by a linear type-II isotherm between p/po = 0.05–1. The hysteresis

18

loop can be described as H4-type hysteresis, with a clear drop around p/po = 0.5. The observed

19

isotherms are typical for a hollow structure with a mesoporous shell.23 The pore size distribution

20

is multimodal in the micro- and mesoporous domains (Fig. 4b). Based on the N2 adsorption

21

analysis, most pores have a diameter of less than 2 nm, which implies they are in the micropore

22

range. However, it was difficult to precisely estimate the micropore size of 2 based on its N2

23

adsorption data because of the large molecule size. Therefore, another non-linear density

ACS Paragon Plus Environment

11

Langmuir

1

functional theory (NLDFT) analysis based on CO2 adsorption was performed, which gave a

2

micropore size of 1.9 nm (Fig. S8).

3

(b)

500

Incremental volume (cm3 g -1)

(a) Adsorbed value (cm3 g -1)

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 12 of 20

400

300

200

100

0 0.0

0.2

0.4

0.6

0.8

1.0

0.04

0.03

0.02

0.01

0.00

1

p/p 0

4

10

100

Pore diameter [nm]

Figure 4. a) N2 sorption isotherms of HCS 2 and b) NLDFT pore size distribution of HCS 2.

5 6

It was recently shown that the performance of a material in CO2 capture strongly depends

7

on its micropore content.41 Micropores with diameters of less than 1 nm are particularly desired

8

for CO2 adsorption at ambient temperature and pressure.42 Thus, the CO2 capture capacity of the

9

HCS was measured at room temperature (Fig. 5). Through a preliminary evaluation, a CO2

10

capture capacity up to 1.56 mmol g−1 at room temperature and 1 bar was obtained. This is subpar

11

when compared with recent literature This is subpar when compared with recent literature

12

although this material contains a significant amount of nitrogen.27, 36 It is important to note that

13

the depth of XPS analysis is 4 nm from the outer surface of the particles, hence the XPS results

14

cannot reflect the accessibility of the nitrogen sites. We therefore analyzed the surface acidity of

15

2 by measuring the ζ-potential. The average ζ-potential of 2 was -33.7 mV, indicating the

ACS Paragon Plus Environment

12

Page 13 of 20 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

Langmuir

1

presence of surface functional groups containing free electrons to be donated to acidic carbon

2

atoms of CO2 molecules. However, the adsorption enthalpy of CO2 on HCS was calculated to be

3

31.5 kJ mol−1 based on the adsorption kinetics, indicating that CO2 and 2 displayed van der

4

Waals interactions (see the Supporting Information). This is a typical value for the physisorption

5

of CO2, which also implies that the CO2 is dominantly physisorbed on the carbon instead of

6

chemisorbed on the nitrogen sites, i.e. the accessibility of nitrogen sites should be improved. In

7

addition, the presence of the highly electronegative C=O, shown by the strong peak at 1670 cm-1

8

in the FT-IR spectrum of 2 (Figure S1), potentially weakens the attractiveness of N to C from

9

CO2.43 While these issues give good insight for the future development of this material, it is

10

noteworthy that the current CO2 uptake was achieved in a structure without micropores of