A simplified synthesis of N-doped zeolite-templated carbons, the control of the level of zeolite-like ordering and its effect on hydrogen storage properties

CARBON

4 9 ( 2 0 1 1 ) 8 4 4 –8 5 3

available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/carbon

A simplified synthesis of N-doped zeolite-templated carbons, the control of the level of zeolite-like ordering and its effect on hydrogen storage properties Yongde Xia a, Robert Mokaya b, David M. Grant a, Gavin S. Walker

a,*

a Energy and Sustainability Research Division, Faculty of Engineering, University of Nottingham, University Park, Nottingham NG7 2RD, United Kingdom b School of Chemistry, University of Nottingham, University Park, Nottingham NG7 2RD, United Kingdom

A R T I C L E I N F O

A B S T R A C T

Article history:

Nitrogen-doped, microporous carbon materials have been prepared using zeolite EMC-2 as

Received 22 April 2010

a hard template and acetonitrile as the carbon source via chemical vapour deposition (CVD)

Accepted 19 October 2010

in the temperature range 700–950 C. The carbon products exhibited high surface areas (up

Available online 23 October 2010

to 3360 m2/g), high pore volumes (up to 1.71 cm3/g) and had zeolite-like structural ordering derived from the template. The carbons had XRD patterns that exhibited two well resolved peaks and TEM images that showed well ordered pore channels. A high proportion of porosity (up to 85% of surface area and 73% of pore volume) for the best ordered carbon ˚ . The cararose from micropores that exhibited narrow size distribution in the range 5–15 A bons generally retained the morphology of the template with solid-core particles at CVD temperatures up to 900 C and hollow shells at 950 C. The carbons had total hydrogen storage capacities up to 6.0 wt.% at

196 C and 20 bar. The hydrogen uptake was found to be

dependent on the level of zeolite-like ordering and the resulting textural properties. Partic˚ which are ularly, high levels of zeolite-like ordering favoured micropores of size <15 A favourable for higher hydrogen uptake capacities.  2010 Elsevier Ltd. All rights reserved.

1.

Introduction

Carbon materials, characterised by chemical inertness, high thermal stability and good mechanical stability, have been widely used in catalysis, adsorption and separation. The past two decades have witnessed intensive studies on two classes of carbons; zero-dimensional fullerenes and one-dimensional carbon nanotubes [1,2]. Another important class of carbons is porous carbon materials. Due to their high surface area and large pore volume, there has been increasing interest over the past decade in novel applications of porous carbon materials as templates, electrode materials and gas stores [3]. Of particular interest is hydrogen storage, which is relevant for

the anticipated hydrogen economy where hydrogen may be used as an energy carrier [3]. Nanostructured carbon materials, including nanotubes, nanofibers and a variety of nanoporous carbons have been extensively studied for hydrogen storage [4–9]. A number of studies on microporous and mesoporous carbons have suggested a link between micropore surface area and hydrogen uptake [5,6,10–12], and indicated the importance of micropores with diameter below 1 nm (in particular pore size of ca. 0.7 nm) for hydrogen uptake [5,10–13]. For example, Rzepka et al. reported that hydrogen uptake is proportional to the specific surface area of carbons, but that optimal uptake occurs for slit pores with a diameter of ca. 0.7 nm [12].

* Corresponding author: Fax: +44 115 9513800. E-mail address: [email protected] (G.S. Walker). 0008-6223/$ - see front matter  2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.carbon.2010.10.028

CARBON

4 9 (2 0 1 1) 8 4 4–85 3

Texier-Mandoki et al. reported on the hydrogen adsorption capacity of various commercial activated carbons and demonstrated that hydrogen uptake is directly related to micropore volume (determined using CO2 sorption) and the presence of a homogeneous narrow microporosity of ca. 0.7 nm [5]. Since the hydrogen uptake capacity generally increases with surface area and is enhanced for carbons with high microporosity, it is desirable to prepare high surface area carbon materials with controlled microporosity for use as hydrogen stores. In recent years, much work has been carried out to prepare carbon materials with ordered porous structures [14–19] and among various synthesis strategies, the hard template route has attracted considerable attention [16,20,21]. A variety of inorganic substrates including microporous zeolites [22–28], mesoporous silicas [16,17,29–31] and colloidal silica [32] have been explored as hard templates and various carbon precursors including sucrose [16], furfuryl alcohol [23,24,26,27,33], acrylonitrile [23,26], propylene [24,25], pyrene [26], vinyl acetate [26] and acetonitrile [17,31,34,35] have been investigated to prepare porous carbon materials. The ability to control the porosity of the templated carbons depends on the extent of structural replication from the inorganic template to the carbon. In general, the pore channel ordering of mesoporous silicas and colloidal silicas can be faithfully replicated to produce mesoporous carbons because these templates have relatively large pore sizes which enable easy diffusion of the carbon precursor molecules into the template pores. In fact, the X-ray diffraction (XRD) patterns for mesoporous carbons usually exhibit several diffraction peaks arising from mesostructural ordering replicated from the inorganic template [16]. However, although microporous carbons templated by zeolites are much more attractive as hydrogen stores, it is much more difficult to replicate the pore channel structure of zeolites due to the smaller pore diameter of the templates which hinders diffusion of carbon precursor molecules. As a result, the XRD patterns of zeolite-templated carbons usually display either none or at most one diffraction peak [14,15]. Structural replication in zeolite-templated carbons may be improved by using larger pore zeolites with a two- or threedimensional non-cubic pore system as the template [36–38]. We have recently reported the synthesis of zeolite-like (with respect to structural pore ordering) carbon materials with well-resolved XRD patterns using zeolite b as a template [36]. These carbon materials possess very high surface areas and have enhanced hydrogen uptake capacities [36]. Gaslain and co-workers have prepared a carbon replica with a wellresolved XRD pattern that displays up to three diffraction peaks using zeolite EMC-2 as a template [37]. However, the preparation protocol involves several steps, requiring both liquid impregnation and CVD processes, to deposit the carbon. The group did not measure the hydrogen uptake properties for these carbons [37]. Utilising a single step deposition of the carbon would greatly simplify the synthesis procedure and through the use of a heteroatom containing feedstock, produce doped microporous carbons. It is also of interest to ascertain whether a higher level of zeolite-like ordering translates to enhanced textural properties, optimised microporosity and thus resulting in high hydrogen uptakes. In this

845

report, a simple CVD method with short processing time was used to synthesize N-doped microporous carbons with zeolite-like pore channel ordering using zeolite EMC-2 as template. The CVD temperature was used to vary the structural ordering, textural properties and levels of graphitization in the carbon materials. The effect of the level of structural ordering, textural properties and extent of graphitization, on hydrogen uptake capacity is evaluated.

2.

Experimental section

2.1.

Material synthesis

The zeolite EMC-2 template was synthesized following established procedures and the zeolite was used as Na+ type and the SiO2/Al2O3 ratio is 10 [39]. The porous carbon materials were prepared as follows: an alumina boat with 0.6 g of calcined zeolite EMC-2 was placed in a flow-through tube furnace. The furnace was heated to the required temperature (between 700 and 950 C) under an argon flow and then maintained at the target temperature for 3 h under a flow of argon saturated with acetonitrile at room temperature, followed by further heat treatment in an argon atmosphere at 900 C for 3 h, and cooling down to room temperature under an argon flow. The resulting carbon/zeolite composites were recovered and washed with 10% hydrofluoric (HF) acid several times, followed by refluxing at 60 C in concentrated hydrochloric acid (HCl) to remove the zeolite framework. The resulting carbon materials were dried in an oven at 120 C. (Thermogravimetric analysis of the HF-treated and HCl refluxed carbons indicated a residual mass typically lower than 1% at 800 C, which confirmed that the carbon materials were virtually zeolite free.) The carbon samples were denoted as CEMC700, CEMC750, CEMC800, CEMC850, CEMC900, and CEMC950, corresponding to CVD temperatures of 700, 750, 800, 850, 900, and 950 C, respectively.

2.2.

Material characterization

Powder XRD analysis was performed using a Siemens D500 powder diffractometer with Cu Ka radiation (40 kV, 25 mA), 0.02 step size, and 2 s step time. Textural properties were determined via nitrogen sorption at 196 C using a conventional volumetric technique on a Quantachrome Autosorb-1 sorptometer. Before analysis, the samples were evacuated for 12 h at 300 C under vacuum. The surface area was calculated using the Brunauer–Emmett–Teller (BET) method based on adsorption data in the partial pressure (p/po) range 0.02– 0.22. The total pore volume was determined from the amount of nitrogen adsorbed at p/po of ca. 0.99. Micropore surface area and micropore volume were obtained via t-plot analysis. The pore size distributions (PSD) were given with the non-local density functional theory (NLDFT) method for slit/cylinder pores using the software provided by Quantachrome. The fitting error was typically in the range of 2.5–3.0%. Elemental analysis was carried out using a CHNS analyzer (Fishons EA 1108). Thermogravimetric analysis (TGA) was performed using a TA SDT Q600 instrument with a heating rate of 2 C/min under air flow of 100 mL/min. Scanning electron microscopy

CARBON

4 9 ( 2 0 1 1 ) 8 4 4 –8 5 3

(SEM) images were recorded using a Philips XL-30 scanning electron microscope. Samples were mounted using a conductive carbon double-sided sticky tape. A thin (ca. 10 nm) coating of gold was sputtered onto the samples to reduce the effects of charging. Transmission electron microscopy (TEM) images were recorded on a JEOL 2000-FX electron microscope operating at 200 kV. Samples for analysis were prepared by spreading them on a holey carbon film supported on a grid. X-ray photoelectron spectroscopy (XPS) was performed using a Kratos AXIS ULTRA spectrometer with a mono-chromated Al X-ray source (1486.6 eV) operated at 10 mA emission current and 15 kV anode potential. The analysis chamber pressure was better than 1.3 · 10 12 bar. The take-off angle for the photoelectron analyzer was 90, and the acceptance angle was 30 (in magnetic lens modes).

2.3.

g

f e Intensity (a. u.)

846

d

Hydrogen uptake measurements 5

Gravimetric analysis of hydrogen uptake capacity was performed using high-purity hydrogen (99.9999%) over the pressure range 0–20 bar with an Intelligent Gravimetric Analyzer (IGA-003, Hiden), which incorporates a microbalance capable of measuring weights with a resolution of ±0.2 lg. The samples in the analysis chamber of the IGA-003 were evacuated up to 10 10 bar and kept at 250 C overnight before measurement. The hydrogen uptake measurements were carried out at 196 C in a liquid nitrogen bath. The high-purity hydrogen (99.9999%) was additionally purified by a molecular sieve filter and a liquid nitrogen trap before use for the hydrogen uptake measurements. The hydrogen uptake data was corrected for the buoyancy of the system and samples. The excess hydrogen uptake was calculated on the basis of a density for the carbon sample measured by He adsorption at 25 C on IGA [36]. The total hydrogen uptake was calculated from the measured excess capacity plus the amount of hydrogen that would occupy the pore volume of the sample at the prevailing temperature and pressure [40]. In all cases the hydrogen uptake capacities (excess or total) are reported as a weight percent of the loaded materials (capacity with respect to the bare materials are given for comparison). The hydrogen uptake capacity for loaded materials is quoted as a wt.% in terms of 100 · (weight of adsorbed H2)/(weight of host material + weight of H2 adsorbed) and hydrogen uptake capacity for bare materials is 100 · (weight of adsorbed H2)/ (weight of host material).

3.

Results and discussions

3.1.

Structural ordering, textural properties and porosity

Fig. 1 shows the powder XRD patterns of the zeolite EMC-2 templated carbons (after removal of the zeolite) prepared via CVD at various temperatures. For comparison the XRD pattern for the zeolite EMC-2 template is also shown. The XRD patterns of all the carbon samples showed a peak at 2h = 6.1 (hereinafter referred to as the basal peak). This ‘basal peak’ was at the same position as the (1 0 0) diffraction of the zeolite C-2 template and therefore indicates that the carbon materials exhibited structural pore ordering similar to that

c b a 10 15 20 25 30 35 40 45 50 2θ (degree)

Fig. 1 – Powder XRD patterns of zeolite EMC-2 (a) and carbon materials prepared via CVD using acetonitrile as the carbon source and zeolite EMC-2 as the template at (b) 700 C (CEMC700), (c) 750 C (CEMC750), (d) 800 C (CEMC800), (e) 850 C (CEMC850), (f) 900 C (CEMC900) and (g) 950 C (CEMC950).

of the zeolite with a ‘basal spacing’ of ca. 1.4 nm. The intensity of the ‘basal peak’ increased with CVD temperature from 700 to 750 C, and decreased at 800 C and above. The XRD patterns suggest that zeolite-type pore ordering was best replicated in the carbon prepared at 750 C. Besides the basal peak, all the carbon materials exhibited a further peak at 2h = 6.8 (although only a shoulder for samples CEMC700 and CEMC950), at a position similar to the (1 0 1) diffraction of the zeolite C-2 template. The second diffraction peak was consistent with a high level of replication of zeolite-type structural ordering in the carbon materials [37]. The intensity of the second diffraction peak was highest for CEMC750, further highlighting the high zeolite-like ordering of this sample, which is comparable to that reported by Gaslain and co-workers [37]. However, we noticed that the carbons only show two diffraction peaks replicated from zeolite rather than 3 diffraction peaks in the literature report [37]. This is maybe due to the fact that acetonitrile is a heavier carbon precursor than ethylene, which can result in a slower diffusion rate for acetonitrile under similar conditions and form carbon which does not replicate as faithfully the zeolite structure. For carbon materials prepared at a CVD temperature of 800 C and above, a further peak was observed at 2h = 26, the (0 0 2) diffraction line for graphitic carbon. The intensity and sharpness of this peak increased at higher carbonization temperature, implying the presence of greater proportions of graphitic carbon domains. At a CVD temperature of 950 C (sample CEMC950) the basal peak at 2h = 6.1 was negligible compared with a sharper graphitic (0 0 2) reflection, implying high graphitisation but poor zeolite-like ordering. The XRD patterns clearly show that changes in CVD temperature provide an avenue to varying the level of zeolite-like ordering in the carbons. At CVD temperatures below 900 C,

CARBON

carbonization was relatively slow and therefore the carbon precursor (acetonitrile) permeates into and fills the zeolite pores, which results in the replication of zeolite-like pore channel regularity in the carbons. Most of the deposited carbon was spatially constrained within the zeolite pores, which limits graphitisation resulting in amorphous carbons with few graphitic domains. The poor replication of the zeolite-like ordering and significant graphitisation at high CVD temperature (950 C), was due to faster carbonization of the acetonitrile which deposited carbon on the surface and/or near surface region of the zeolite particles thus blocking pores and hindering access into the interior of the particles. Furthermore, the zeolite template framework suffered from thermal degradation as evidenced by the XRD patterns of carbon/zeolite composites (Supplementary materials Fig. S1). At CVD temperatures lower than 900 C, the composites exhibited structures similar to the parent EMC-2 zeolite,

1400

b

Volume adsorbed (cm3/g STP)

1200 c d

1000

e

800

f

600 400 a

200 g

0 0.0

0.2

0.4 0.6 0.8 Pressure (P/Po)

847

4 9 (2 0 1 1) 8 4 4–85 3

1.0

Fig. 2 – Nitrogen sorption isotherms of zeolite EMC-2 (a) and carbon materials prepared via CVD using acetonitrile as the carbon source and zeolite EMC-2 as the template at (b) 700 C (CEMC700), (c) 750 C (CEMC750), (d) 800 C (CEMC800), (e) 850 C (CEMC850), (f) 900 C (CEMC900) and (g) 950 C (CEMC950). For clarity, isotherm b is offset (y-axis) by 200.

while at 950 C, zeolite peaks were absent or much reduced in intensity and the pattern was dominated by a strong graphitic carbon peak at a 2h value of ca. 26. Faster carbonization, pore blocking and collapse of the zeolite template at higher CVD temperatures hindered replication, with most of the carbon deposited outside the pores and therefore more readily graphitized. To assess whether a high level of zeolite-like ordering results in higher surface area and microporosity, porosity analysis of the carbons was performed and Fig. 2 shows the nitrogen sorption isotherms of the carbons and the zeolite EMC-2 template. The textural properties are summarised in Table 1. All the carbons displayed type I isotherms with very high adsorption below p/po = 0.1, due to micropore filling. This suggests that a large proportion of the pore channels were micropores, which is consistent with zeolite-like structural ordering. Some samples (CEMC700 and CEMC950) also exhibit significant nitrogen uptake at p/po > 0.6, which may be attributed to adsorption into larger pores arising from inter-particle voids. Carbons obtained at CVD temperatures from 700 C up to 900 C had high total surface areas (2160–3360 m2/g) and pore volumes (1.11–1.71 cm3/g). These samples being the ones that XRD showed had high levels of zeolite-like structural ordering and we can therefore conclude that zeolite-like ordering does enhance the textural properties. Furthermore, sample CEMC750, which exhibited the highest level of zeolite-like ordering (according to XRD patterns in Fig. 1) also had the highest surface area and pore volume and attains textural values that are comparable to those reported by Gaslain and co-workers [37]. On the other hand, the surface area and pore volume of sample CEMC950 was much lower due to poorer zeolite-like ordering and the high level of graphitization. For microporosity, the carbons exhibited high micropore surface areas up to 2838 m2/g and micropore volumes up to 1.24 cm3/g. It is noteworthy that sample CEMC750 has both the highest micropore surface area (2838 m2/g) and proportion of surface area arising from micropores (85%), and also the highest micropore volume (1.24 cm3/g) and proportion of volume arising from micropores (73%). Clearly, the high zeolite-like structural ordering of this sample enhanced the overall textural properties and microporosity. The porosity data showed that by changing the CVD temperature it was possible

Table 1 – Textural properties, N content, and hydrogen uptake capacity of N-doped carbon materials prepared via CVD using zeolite EMC-2 as template and acetonitrile as carbon source. Sample EMC-2 CEMC700 CEMC750 CEMC800 CEMC850 CEMC900 CEMC950 a b c d

CVD temperature (C) 700 750 800 850 900 950

N content (wt.%)a 3.65 4.70 6.51 6.33 6.19 7.66

(3.7) (5.2) (7.6) (6.5) (5.9) (6.4)

Surface area (m2/g)b 819 2878 (1913) 3360 (2838) 2762 (2302) 2658 (2167) 2168 (1761) 565 (401)

Pore volume (cm3/g)b 0.40 1.67 1.71 1.41 1.42 1.11 0.41

(0.77) (1.24) (0.99) (0.98) (0.75) (0.17)

Values in parentheses are obtained from XPS analysis. Values in parentheses are micropore surface area/pore volume. Hydrogen uptake capacity (with respect to loaded materials) at 196 C and 20 bar. Values in parentheses are hydrogen storage capacities with respect to the bare material at

Excess H2 uptake (wt.%)c,d 4.8 5.0 4.6 4.3 3.7 1.0

(5.0) (5.3) (4.8) (4.5) (3.8) (1.0)

196 C and 20 bar.

Total H2 uptake (wt.%)c,d 5.5 6.0 5.4 5.1 4.3 1.2

(5.8) (6.4) (5.7) (5.4) (4.5) (1.2)

CARBON

4 9 ( 2 0 1 1 ) 8 4 4 –8 5 3

to vary (and indeed optimise) the level and proportion of microporosity in the zeolite-templated carbons. The micropore/small mesopore size distribution was measured from nitrogen sorption for the optimised sample (CEMC750) in comparison with two other carbons as shown in Fig. 3. The pore size distribution (PSD) was determined using the NLDFT model applied to nitrogen adsorption data. The PSD of the carbons appeared to be dominated by pores ˚ , with sample CEMC750 displaying the in the range 10–15 A sharpest distribution. Furthermore, sample CEMC750 has ˚ size range while samples hardly any pores in the 15–30 A CEMC700 and CEMC800 showed a broad distribution of pores in this size range. These findings are consistent with a higher level of zeolite-like ordering in sample CEMC750 as discussed above. Ultramicropores in the obtained carbon materials are not accessible by N2, other adsorption gases such as CO2 are needed to measure these smaller pores [41,42]. The measurement of micropore size can be obtained from CO2 adsorption at 273 K, as shown in Fig. 4, which gave pore sizes in the range ˚ . The formation of pores of size between 5 and 10 A ˚ of 5–10 A for zeolite-templated carbons is reasonable given that the thickness of the zeolite wall framework (which becomes the pores in the carbons) is ca. 0.6 nm [22]. Obviously any incomplete filling of the zeolite pores during CVD would lead to slightly larger pores remaining in the resultant carbon. There does appear to be a bimodal distribution of sites which may result from either incomplete filling of the pores or from the varying wall thickness of the zeolite cage structure, but given the coincident nature of the minima, a modelling artefact cannot be ruled out. There is however a discrepancy between the N2 and CO2 PSDs as both molecules probe pores in the ˚ and therefore should be giving consistent data range of 6–15 A over these ranges. TEM was therefore used to investigate the pore structure, in addition to probing the localised crystallinity of the sample. Fig. 5 (and Supplementary materials Fig. S2) shows transmission electron microscopy (TEM) images of samples CEMC750, CEMC850 and CEMC950. The TEM results showed ordered pore channel structures for samples CEMC750 and

dV/dD (cm 3 g -1 -1)

CEMC800

CEMC750

CEMC700

5

10

15 20 Pore size (Å)

25

30

Fig. 3 – Representative pore size distribution of zeolitetemplated carbons determined using NLDFT method applied to nitrogen sorption (at 196 C) data.

CEMC800

dV/dD (cm3 g-1 Å -1)

848

CEMC750

CEMC700

2

4

6

8 10 12 Pore size (Å)

14

16

Fig. 4 – Representative pore size distribution of zeolitetemplated carbons determined using NLDFT methods applied to CO2 sorption (at 0 C) data.

CEMC850, while sample CEMC950 exhibited no ordered channels. Image analysis of sample CEMC750, gave an aver˚ , which age peak–peak distance for the striations of 9–10 A ˚ (as the pore walls indicates that the pore size must be <10 A have a finite thickness). This is in agreement with the pore ˚ ) as shown size determined from CO2 sorption data (5–10 A ˚ in Fig. 4. A PSD of 5–10 A also agrees with the results for an undoped EMC templated carbon [38]. The combination of supporting data for the CO2 derived PSD gives one more confidence in this result. If this assumption is correct, the discrepancy for the N2 PSD is most likely due to the NLDFT model not adequately representing the cage-derived morphology of the pores and highlights the need for more appropriate models for this interesting class of carbons. The arcs in the selected area electron diffraction (SAED) patterns (inset Fig. 5) show that CEMC950 had some graphitic crystallinity, but the lack of any diffraction rings for CEMC750 shows that this sample was largely amorphous, in agreement with the XRD data (Fig. 1). SAED of sample CEMC850 (Fig. S2 in Supplementary material) showed that in addition to the amorphous carbon within the particle, there was a thin (6 nm) layer of more graphitic carbon (not evident for carbons deposited at 750 C and below). This shows that as the CVD temperature increased so the amount of surface deposition increased, and that the surface carbon was the more graphitic of the two carbon forms. This is reflected in the XRD data and an increase in intensity of the (0 0 2) reflection for graphite with increasing CVD temperature. It is also evident in the XRD data that there is a loss in intensity of the low 2h peaks. Given that the amount of surface carbon for sample CEMC850 accounts of ca. 7% of the carbon this will lead to a corresponding attenuation of the low 2h peaks. As these peaks show a significantly greater decrease in intensity than just 7%, this supports the argument that the loss in intensity of these reflections is the result of poorer replication of the ordered zeolite structure. This is also supported by the textural properties which are lower for the samples prepared at temperatures above 750 C.

CARBON

849

4 9 (2 0 1 1) 8 4 4–85 3

Fig. 5 – TEM images of carbon materials prepared via CVD using acetonitrile as the carbon source and zeolite EMC-2 as the template at (a) 750 C (CEMC750) (b) 850 C (CEMC850) and (c) 950 C (CEMC950).

3.2.

Carbon yield and particle morphology

The yield of carbon has been identified as an important factor in determining the replication of zeolite-like structural ordering in zeolite-templated carbons. This is because the replication of zeolite-like ordering in the carbons requires extensive infiltration of the carbon precursor into the pores of the zeolite template. Thermogravimetric analysis (TGA) of carbon/ zeolite composites was performed to assess the carbon yield (Supplementary materials Fig. S3 shows a representative TGA curve for the CEMC750/EMC-2 composite). After excluding the removal of water below 300 C, the mass loss by 700 C was 25.2% with a residue of 65.2%. This suggests that the carbon/zeolite composite contains 25.2 wt.% carbon, and therefore that the carbon yield for sample CEMC750 was 0.39 g of carbon per g of EMC-2 zeolite (i.e., 0.39 g of C/g of zeolite). The carbon yield (g of C/g of zeolite) for the other samples was 0.27 (CEMC700), 0.4 (CEMC800), 0.43 (CEMC850),

0.51 (CEMC900) and 0.53 (CEMC950). The carbon yield generally increased with CVD temperature. Given that the density of the deposited carbons by this CVD method is ca. 1.5 g cm 3 [28], this carbon deposition trend may be explained by considering that at 700 C the carbon precursor did not completely infiltrate the zeolite EMC-2 pore channel system leading to a low carbon yield, large pores due to incomplete precursor filling, and a lower level of zeolite-like ordering (Fig. 1). There was very good infiltration of the pores at a CVD temperature of 750 C, but at higher temperatures the carbon yield increases primarily due to greater surface deposition of graphitic carbon (Fig. 1) outside the zeolite pore system. The formation of increasing amounts of graphitic carbon was confirmed by differential thermal-analysis (DTA) of the carbons (Supplementary materials Fig. S4). All the carbons exhibit one mass loss event (combustion of carbon) in the temperature range 450–650 C (Supplementary materials Fig. S3). The DTA curves show that samples synthesized at 700 and

a

b

CPS

CPS

CEMC950

CEMC850

CEMC950

CEMC850

CEMC750

CEMC750

396

398

400 402 404 Binding energy (eV)

406

290

288

286 284 282 Binding energy (eV)

280

Fig. 6 – XPS of N 1s (a) and C 1s (b) of carbon materials prepared via CVD using acetonitrile as the carbon source and zeolite EMC-2 as the template at 750 C (CEMC750), 850 C (CEMC850), and 950 C (CEMC950).

850

CARBON

4 9 ( 2 0 1 1 ) 8 4 4 –8 5 3

750 C (CEMC700, CEMC750) combust at ca. 540 C, reflecting their amorphous (i.e. thermally less stable) nature. Samples synthesized at 800 and 850 C (CEMC800 and CEMC850) burn off at ca. 555 and 570 C, respectively, while samples synthesized at 900 and 950 C combust at ca. 590 C due to their higher levels of graphitization [19,43]. The nitrogen content of the carbons (reported as wt.% in Table 1) increased from 3.65 wt.% to a maximum of 7.66 wt.% at high CVD temperature. Information on the nature of the nitrogen and its binding to carbon was obtained from X-ray photoelectron spectroscopy (XPS). As shown in Table 1, the surface nitrogen content calculated from the XPS data is very close to the bulk nitrogen content calculated

from elemental analysis due to the use of a single molecule precursor (i.e. acetonitrile) that ensures that the nitrogen is homogeneously distributed in the carbons [34]. The XPS spectra in Fig. 6 shows that the N 1s signal (Fig. 6a) is split into peaks at 401.2, 398.5 and 403.1 eV. The N 1s binding energies may be ascribed to highly coordinated (quaternary) nitrogen atoms (401.2 eV) incorporated into graphene sheets [44], and pyridine-like nitrogen atoms (398.5 eV) in graphene sheets (i.e. sp2 pyridine-like nitrogen bonded to two C atoms) [45,46]. The peak at ca. 403.1 eV may be ascribed to graphitic N atoms that are coordinated to O atoms [41,42,46]. The O atoms may arise from the replication process which involves contact with various oxygen containing reagents (zeolite tem-

Fig. 7 – Representative SEM images of zeolite EMC-2 (a) and zeolite-templated carbons prepared at (b) 700 C (CEMC700), (c) 750 C (CEMC750), (d) 800 C (CEMC800), (e) 850 C (CEMC850), (f) 900 C (CEMC900) and (g) 950 C (CEMC950).

CARBON

4 9 (2 0 1 1) 8 4 4–85 3

plate and aqueous acid). Fig. 6b shows that the C 1s peak for the carbon samples was centred at ca. 284.5 eV, which is consistent with sp2 graphitic carbon. Fig. 7 shows representative scanning electron microscopy (SEM) images of zeolite EMC-2 and the templated carbons. Fig. 7a shows sheet or disc-like particles, of diameter no more than 3 lm and thickness of less than 0.3 lm, for zeolite EMC2. Carbons prepared at 700–900 C display solid core disc-like particles, while the sample prepared at 950 C exhibits broken hollow sheet-like particles (Supplementary materials Fig. S2c). The morphology of the zeolite EMC-2 template was clearly transferred to the carbon materials, which is consistent with a templating mechanism whereby the carbon is predominantly nanocast either within the pore channels of the zeolite EMC-2 (700–900 C) or on the surface of the zeolite particles (CEMC950).

3.3.

Hydrogen storage properties

Hydrogen uptake isotherms of the zeolite-templated carbons, measured gravimetrically at 196 C over the pressure range 0–20 bar are shown in Fig. 8. The hydrogen uptake isotherms show no hysteresis, which indicates total reversibility in the take-up of hydrogen. The hydrogen uptake does not approach saturation even at a pressure of 20 bar implying that even greater uptake capacities are possible at even higher pressures. The excess and total hydrogen uptake capacities of the carbons (at 20 bar), reported in Table 1 varies between 1.0–5.0 and 1.2–6.0 wt.%, respectively (where 6.0 wt.% is equivalent to 6.4 wt.% when calculating the hydrogen capacity relative to the bare material). The sample prepared at 750 C (CEMC750) exhibits the highest hydrogen uptake. The trend in hydrogen uptake capacity is clearly related to the level of zeolite-like ordering, the improved uptake capacities being a result of the improved textural properties that these carbons exhibit. The total and excess hydrogen uptake capacb a

6

c

5

d

H2 uptake (wt%)

e

851

ities of up to 6.0 and 5.0 wt.% for the zeolite-like carbon material prepared at 750 C (CEMC750) is greater than that of most other carbon and non-carbon porous materials [6,8,47]. An analysis of the hydrogen uptake data in comparison with the textural properties (Table 1) clearly reveals that samples that exhibit higher levels of zeolite-like structural ordering, have higher textural properties (including both higher surface area and higher microporosity) and have enhanced hydrogen uptake capacities. The hydrogen uptake capacity is therefore dependent on the textural properties of the carbons and consequently also dependent on the level of zeolite-like ordering. Clearly, a linear relationship between hydrogen uptake and surface area can be expected (as shown in Fig. S5) for those porous carbon materials. The link between zeolite-like ordering, textural properties and hydrogen uptake is emphasised by the fact that sample CEMC750 with the highest hydrogen uptake possesses not only the highest micropore surface area and pore volume, but also the highest proportion of micropore surface area (85%) and micropore volume (73%). Our data shows that a high level of zeolite-like ordering in zeolite-templated carbons is an excellent way of optimising microporosity and thereby maximising surface area. Recent studies have shown that it is not simply the overall surface area that determines hydrogen uptake in porous carbons but the surface area associated with ‘optimal pores’ of a specific size ca. 0.7 nm [5,11,47]. According to the pore size distribution data (Figs. 3 and 4), sample CEMC750 possesses ˚ . Some microporosity that is dominated by pores below 15 A other samples appear to have broader micropore size distri˚ . Howbution that includes supermicropores larger than 15 A ever, there was no marked difference between these samples in the hydrogen capacity per m2 of surface area. Thus, the most important textural property in determining hydrogen capacity was surface area. As detailed in our recent work [28], the effect of N-doping of porous carbons on hydrogen uptake only has a minor effect in comparison to the textural properties. A challenge in the area is to increase the isosteric heat of adsorption of hydrogen to such porous materials which could be achieved by doping the carbons with other elements. The CVD methodology used to make these N-doped carbons shows great potential to be modified to produce a range of other types of doped carbons.

4

4.

3

2 f

1

0 0

2

4

6 8 10 12 14 16 18 20 Pressure (bar)

Fig. 8 – Hydrogen uptake at 196 C for zeolite-templated carbon temperatures prepared at: (a) 700 C (CEMC700), (b) 750 C (CEMC750), (c) 800 C (CEMC800), (d) 850 C (CEMC850), (e) 900 C (CEMC900) and (f) 950 C (CEMC950). Solid symbols for adsorption and empty symbols for desorption.

Conclusions

In summary, carbon materials that exhibit high surface area and large pore volume have been prepared via a simple CVD route using zeolite EMC-2 as a solid template and acetonitrile as the precursor for forming N-doped carbons. The carbon materials obtained at CVD temperatures of 700–900 C are essentially amorphous and have high surface areas (up to 3360 m2/g), high pore volume (up to 1.79 cm3/g), and exhibit varying levels of zeolite-like structural ordering derived from the zeolite template. The zeolite-like structural ordering of the best ordered carbon materials is indicated by powder XRD patterns with at least two well resolved diffraction peaks and TEM images that show well ordered pore channels. Carbons with high levels of zeolite-like ordering exhibit enhanced total hydrogen storage capacity of up to 6.0 wt.% at

852

CARBON

4 9 ( 2 0 1 1 ) 8 4 4 –8 5 3

20 bar and 196 C. Hydrogen uptake capacity was therefore found to be strongly dependent on the level of zeolite-like ordering and on the textural properties. In particular micro˚ plays an imporporosity associated with pores of size <15 A tant role in hydrogen uptake. Zeolite templating can therefore be used to generate carbons and doped carbons with the required properties for enhanced hydrogen storage.

Acknowledgement The authors are grateful to the EPSRC and EU for financial support.

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.carbon.2010.10.028.

R E F E R E N C E S

[1] Kroto HW, Heath JR, O’Brien SC, Curl RF, Smalley RE. C60: buckminsterfullerene. Nature 1985;318(6042):162–3. [2] Iijima S. Helical microtubules of graphitic carbon. Nature 1991;354(6348):56–8. [3] Thomas KM. Hydrogen adsorption and storage on porous materials. Catal Today 2007;120(3–4):389–98. [4] de la Casa-Lillo MA, Lamari-Darkrim F, Cazorla-Amoros D, Linares-Solano A. Hydrogen storage in activated carbons and activated carbon fibers. J Phys Chem B 2002;106(42):10930–4. [5] Texier-Mandoki N, Dentzer J, Piquero T, Saadallah S, David P, Vix-Guterl C. Hydrogen storage in activated carbon materials: role of the nanoporous texture. Carbon 2004;42(12–13):2744–7. [6] Zhao XB, Xiao B, Fletcher AJ, Thomas KM. Hydrogen adsorption on functionalized nanoporous activated carbons. J Phys Chem B 2005;109(18):8880–8. [7] Xia Y, Mokaya R. Ordered mesoporous carbon monoliths: CVD nanocasting and hydrogen storage properties. J Phys Chem C 2007;111(27):10035–9. [8] Yang ZX, Xia YD, Sun XZ, Mokaya R. Preparation and hydrogen storage properties of zeolite-templated carbon materials nanocast via chemical vapor deposition: effect of the zeolite template and nitrogen doping. J Phys Chem B 2006;110(37):18424–31. [9] Nishihara H, Hou PX, Li LX, Ito M, Uchiyama M, Kaburagi T, et al. High-pressure hydrogen storage in zeolite-templated carbon. J Phys Chem C 2009;113(8):3189–96. [10] Gogotsi Y, Dash RK, Yushin G, Yildirim T, Laudisio G, Fischer JE. Tailoring of nanoscale porosity in carbide-derived carbons for hydrogen storage. J Am Chem Soc 2005;127(46):16006–7. [11] Vix-Guterl C, Frackowiak E, Jurewicz K, Friebe M, Parmentier J, Beguin F. Electrochemical energy storage in ordered porous carbon materials. Carbon 2005;43(6):1293–302. [12] Rzepka M, Lamp P, De la Casa-Lillo MA. Physisorption of hydrogen on microporous carbon and carbon nanotubes. J Phys Chem B 1998;102(52):10894–8. [13] Gadiou R, Saadallah S-E, Piquero T, David P, Parmentier J, VixGuterl C. The influence of textural properties on the adsorption of hydrogen on ordered nanostructured carbons. Microporous Mesoporous Mater 2005;79(1–3):121–8. [14] Ma ZX, Kyotani T, Tomita A. Preparation of a high surface area microporous carbon having the structural regularity of Y zeolite. Chem Commun 2000:2365–6.

[15] Ma ZX, Kyotani T, Liu Z, Terasaki O, Tomita A. Very high surface area microporous carbon with a three-dimensional nano-array structure: synthesis and its molecular structure. Chem Mater 2001;13(12):4413–5. [16] Ryoo R, Joo SH, Jun S. Synthesis of highly ordered carbon molecular sieves via template-mediated structural transformation. J Phys Chem B 1999;103(37):7743–6. [17] Xia YD, Mokaya R. Ordered mesoporous carbon hollow spheres nanocast using mesoporous silica via chemical vapor deposition. Adv Mater 2004;16(11):886–91. [18] Kyotani T. Control of pore structure in carbon. Carbon 2000;38(2):269–86. [19] Xia YD, Mokaya R. Synthesis of ordered mesoporous carbon and nitrogen-doped carbon materials with graphitic pore walls via a simple chemical vapor deposition method. Adv Mater 2004;16(17):1553–8. [20] Vix-Guterl C, Saadallah S, Vidal L, Reda M, Parmentier J, Patarin J. Template synthesis of a new type of ordered carbon structure from pitch. J Mater Chem 2003;13(10):2535–9. [21] Xia Y, Yang Z, Mokaya R. Templated nanoscale porous carbons. Nanoscale 2010;2(5):639–59. [22] Johnson SA, Brigham ES, Ollivier PJ, Mallouk TE. Effect of micropore topology on the structure and properties of zeolite polymer replicas. Chem Mater 1997;9(11):2448–58. [23] Kyotani T, Nagai T, Inoue S, Tomita A. Formation of new type of porous carbon by carbonization in zeolite nanochannels. Chem Mater 1997;9(2):609–15. [24] Kyotani T, Ma Z, Tomita A. Template synthesis of novel porous carbons using various types of zeolites. Carbon 2003;41(7):1451–9. [25] Rodriguez-Mirasol J, Cordero T, Radovic LR, Rodriguez JJ. Structural and textural properties of pyrolytic carbon formed within a microporous zeolite template. Chem Mater 1998;10(2):550–8. [26] Meyers CJ, Shah SD, Patel SC, Sneeringer RM, Bessel CA, Dollahon NR, et al. Templated synthesis of carbon materials from zeolites (Y, beta, and ZSM-5) and a montmorillonite clay (K10): physical and electrochemical characterization. J Phys Chem B 2001;105(11):2143–52. [27] Su FB, Zhao XS, Lu L, Zhou ZC. Synthesis and characterization of microporous carbons templated by ammonium-form zeolite Y. Carbon 2004;42(14):2821–31. [28] Xia YD, Walker GS, Grant DM, Mokaya R. Hydrogen storage in high surface area carbons: experimental demonstration of the effects of nitrogen doping. J Am Chem Soc 2009;131(45):16493–9. [29] Xia YD, Yang ZX, Mokaya R. Mesostructured hollow spheres of graphitic N-doped carbon nanocast from spherical mesoporous silica. J Phys Chem B 2004;108(50): 19293–8. [30] Xia YD, Mokaya R. Generalized and facile synthesis approach to N-doped highly graphitic mesoporous carbon materials. Chem Mater 2005;17(6):1553–60. [31] Xia YD, Yang ZX, Mokaya R. Simultaneous control of morphology and porosity in nanoporous carbon: graphitic mesoporous carbon nanorods and nanotubules with tunable pore size. Chem Mater 2006;18(1):140–8. [32] Gierszal KP, Jaroniec M. Carbons with extremely large volume of uniform mesopores synthesized by carbonization of phenolic resin film formed on colloidal silica template. J Am Chem Soc 2006;128(31):10026–7. [33] Ma ZX, Kyotani T, Tomita A. Synthesis methods for preparing microporous carbons with a structural regularity of zeolite Y. Carbon 2002;40(13):2367–74. [34] Yang Z, Xia Y, Mokaya R. Hollow shells of high surface area graphitic N-doped carbon composites nanocast using zeolite templates. Microporous Mesoporous Mater 2005; 86(1–3):69–80.

CARBON

4 9 (2 0 1 1) 8 4 4–85 3

[35] Hou P, Orikasa H, Yamazaki T, Matsuoka K, Tomita A, Setoyama N, et al. Synthesis of nitrogen-containing microporous carbon with a highly ordered structure and effect of nitrogen doping on H2O adsorption. Chem Mater 2005;17(20):5187–93. [36] Yang ZX, Xia YD, Mokaya R. Enhanced hydrogen storage capacity of high surface area zeolite-like carbon materials. J Am Chem Soc 2007;129(6):1673–9. [37] Gaslain FOM, Parmentier J, Valtchev VP, Patarin J. First zeolite carbon replica with a well resolved X-ray diffraction pattern. Chem Commun 2006(6):991–3. [38] Ducrot-Boisgontier C, Parmentier J, Patarin J. Influence of the carbon precursors on the structural properties of EMT-type nanocasted-carbon replicas. Microporous Mesoporous Mater 2009;126(1–2):101–6. [39] Delprato F, Delmotte L, Guth JL, Huve L. Synthesis of new silica-rich cubic and hexagonal faujasites using crown-etherbased supramolecules as templates. Zeolites 1990;10(6):546–52. [40] Yan Y, Lin X, Yang SH, Blake AJ, Dailly A, Champness NR, et al. Exceptionally high H-2 storage by a metalorganic polyhedral framework. Chem Commun 2009:1025–7.

853

[41] Guan C, Wang K, Yang C, Zhao XS. Characterization of a zeolite-templated carbon for H-2 storage application. Microporous Mesoporous Mater 2009;118(1–3):503–7. [42] Guan C, Zhang X, Wang K, Yang C. Investigation of H-2 storage in a templated carbon derived from zeolite Y and PFA. Sep Purif Technol 2009;66(3):565–9. [43] Kim T, Park I, Ryoo R. A synthetic route to ordered mesoporous carbon materials with graphitic pore walls. Angew Chem Int Ed 2003;42(36):4375–9. [44] Nishihara H, Yang QH, Hou PX, Unno M, Yamauchi S, Saito R, et al. A possible buckybowl-like structure of zeolite templated carbon. Carbon 2009;47(5):1220–30. [45] Xu WH, Kyotani T, Pradhan BK, Nakajima T, Tomita A. Synthesis of aligned carbon nanotubes with double coaxial structure of nitrogen-doped and undoped multiwalls. Adv Mater 2003;15(13):1087–90. [46] Casanovas J, Ricart JM, Rubio J, Illas F, Jimenez-Mateos JM. Origin of the large N 1s binding energy in X-ray photoelectron spectra of calcined carbonaceous materials. J Am Chem Soc 1996;118(34):8071–6. [47] Yushin G, Dash R, Jagiello J, Fischer JE, Gogotsi Y. Carbidederived carbons: effect of pore size on hydrogen uptake and heat of adsorption. Adv Funct Mater 2006;16:2288–93.