Microporous and Mesoporous Materials 117 (2009) 511–514
Contents lists available at ScienceDirect
Microporous and Mesoporous Materials journal homepage: www.elsevier.com/locate/micromeso
In situ synthesis and hydrogen storage properties of PdNi alloy nanoparticles in an ordered mesoporous carbon template R. Campesi a,1, F. Cuevas a,1, E. Leroy a,1, M. Hirscher c,3, R. Gadiou b,2, C. Vix-Guterl b,2, M. Latroche a,* a
CMTR-I.C.M.P.E-UMR 7182;2-8 rue Henri Dunant, 94320, Thiais, France (ICSI) UPR CNRS 906915,15 rue Jean Starcky, 68057, Mulhouse Cedex, France c MPI-MF, Heisenbergstrasse 3, 70569 Stuttgart, Germany b
a r t i c l e
i n f o
Article history: Received 6 May 2008 Received in revised form 9 July 2008 Accepted 11 July 2008 Available online 19 July 2008 Keywords: Porous carbons Carbon/metal composite Metal alloy Hydrogen storage
a b s t r a c t Organized mesoporous carbon has been used as a nanoreactor to prepare PdNi metallic particles using an incipient wetness method starting from Pd and Ni salts. The ﬁnal composite material consists of nanosized metallic particles of an alloy with composition Pd0.60Ni0.40 highly dispersed within the carbon host structure. The thermodynamic hydrogenation properties of both the PdNi-free OMC and the Pd0.60Ni0.40OMC composite have been determined by hydrogen isotherm sorption measurements. The introduction of the palladium–nickel alloy into the carbon matrix does not increase the hydrogen storage capacity at 77 K and 2 MPa, since the hydrogen uptake is mainly attributed to physisorption on the carbon surface. However, at room temperature and moderate pressure (0.5 MPa), the ﬁlling of the OMC with nanocrystalline Pd0.60Ni0.40 results in larger hydrogen uptake than that of the PdNi-free OMC. Ó 2008 Elsevier Inc. All rights reserved.
With the aim of ﬁnding new materials for hydrogen storage, we already investigated the possibility of using ordered mesoporous carbon compound (OMC), obtained by a replica technique [1–3], as a host for the insertion of Pd  or Ni nanoparticles. In order to tailor the thermodynamic properties for H-storage in metal–carbon composites, the feasibility of synthesizing intermetallic compounds in conﬁned carbon media is a key issue. In the literature, indeed, many works can be found dealing with the synthesis of metal alloys–carbon composites by mechanical milling [5–7]. Only few attempts have been done in order to synthesize in situ metal alloys (PtRu with Pt:Ru 3:1) in carbon support with well organized pore structure . In a recent work, we showed that nanosized palladium can be dispersed in an organized carbon porous network . However, instead of pure metallic elements, it is very challenging to prepare highly dispersed metallic alloys in nanoporous hosts. In the present work, we report on the in situ synthesis of a PdNi alloy–OMC composite. The composite was obtained using an incipient wetness method starting from an equal concentration of metal salt precursor of Ni [Ni(NO3)2] and Pd [PdCl2], acetone and the OMC.
The OMC has been prepared by a templating procedure using SBA-15 as silica template and propylene as carbon precursor. The carbon replica (CT) was obtained by carbon vapor deposition (CVD) of propylene at 750 °C as described elsewhere [1–3]. In order to disperse the PdNi nanoparticles in the carbon porosity, the CT was impregnated with a solution of tetrachloropalladous acid (Cl:Pd molar ratio equal to 48), nickel nitrate and acetone to get the desired metal content. Next, the impregnated carbon template was reduced by heating in an Ar/H2 ﬂow for 8 h at 673 K. The PdNi–OMC composite was analyzed by XRD diffraction (Bruker D8Advance Cu Ka, Bragg-Brentano geometry, backscattered rear graphite monochromator) and by Transmission Electron Microscopy TEM (Tecnai F20, Field Emission Gun FEG 200 KV, punctual resolution 0.24 nm, Energy Filtering GIF, EDX and EELS spectrometers). The textural properties were determined from nitrogen adsorption/desorption isotherms measured with a Quantachrome Autosorb A1-LP Instrument. The speciﬁc surface area was obtained by the BET method. The total pore volume was computed from the amount of gas adsorbed at P/P0 = 0.95 and the microporous volume was calculated using the DubininRadushkevich equation in the relative pressure range 10 4–10 2. Isotherm curves for hydrogen adsorption were performed both at 77 K and 298 K in the range 0–2 MPa with a PCTPro-2000 automatic volumetric device (Sievert’s method) equipped with calibrated and thermalized volumes and pressure gauges. Before each measurement, the sample has been outgassed overnight under
* Corresponding author. Tel.: +33 1 49 78 12 03; fax: +33 1 49 78 12 01. E-mail addresses: [email protected]
(M. Hirscher), [email protected]
(C. Vix-Guterl), [email protected]
(M. Latroche). 1 Tel.: +33 1 49 78 12 03; fax: +33 1 49 78 12 01. 2 Tel.: +33 3 89 60 87 45; fax: +33 3 89 60 87 99. 3 Tel.: 49 711 689 1808; fax: +49 711 689 1952. 1387-1811/$ - see front matter Ó 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.micromeso.2008.07.023
R. Campesi et al. / Microporous and Mesoporous Materials 117 (2009) 511–514
high vacuum (10 10 MPa) at 473 K. For sorption measurements at 77 K, the sample holder was immersed in a Dewar ﬁlled with liquid nitrogen.
3. Results The chemical composition of the PdNi–OMC composite was determined by Inductively Coupled Plasma Optical Emission Spectrometer (ICP-OES). The main constituents are: C (82.6 wt%), Pd (10.2 ± 0.3 wt%), Ni (6.48 ± 0.2 wt%). Traces, coming from the OMC synthesis, have also been detected (Si, Ca. . .) but in so small quantities that they can be considered as negligible. The structure of the PdNi hybrid composite has been investigated by XRD analysis. The pattern (top of Fig. 1) displays ﬁve main broad peaks centered at 2h =25°, 42°, 48°, 71° and 86°, respectively. The ﬁrst one can be attributed to the OMC whereas the remaining four peaks can be indexed with the (1 1 1), (2 0 0), (2 2 0) and (3 1 1) reﬂections of a fcc-structure with a cell parameter of a = 3.781 Å. This value lies between those of Pd (a = 3.8907 Å) and Ni (a = 3.524 Å) conﬁrming the formation of an alloy. Taking in account that the PdNi alloys respect the Vegard’s law, as reported by R. Bidwell , we have calculated that our PdNi alloy has an atomic composition close to Pd0.60Ni0.40 . The crystallite size was estimated from the FWHM of the (1 1 1) reﬂection, using the Scherrer’s equation. The value obtained is close to 5 nm. For comparison the patterns of the metal-free OMC, Pd-OMC and Ni-OMC composites are also shown. It can be seen that the main line of the OMC is unchanged but the lines belonging to the PdNi are located between those of Pd-and Ni in the OMC composites. The OMC structure and the dispersion of the PdNi alloy within the porosity of the carbon template were investigated by TEM/
CT Pd Ni PdNi
CT_PdNi CT_10wt% Ni
CT_10wt% Pd CT
2 Theta (deg.) Fig. 1. XRD patterns (CuKa radiation) of the OMC (line), OMC/10wt% Pd (open triangles), OMC/10wt% Ni (full triangles) and OMC/Pd0.60Ni0.40 composites (half full triangles) (from the bottom to the top).
c 1.0 Pd Ni
0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0
Particles Fig. 2. (a) bright ﬁeld and (b) dark ﬁeld STEM images of the Pd0.60Ni0.40 -OMC composite. (c) Atomic composition of several PdNi nanoparticles inside the circle of Fig. 2b (N Pd, j Ni).
R. Campesi et al. / Microporous and Mesoporous Materials 117 (2009) 511–514
Fig. 3. (a), (b) HRTEM images of Pd0.60Ni0.40 nanoparticles dispersed in the OMC graphene structure and FFT TEM image (c) of Fig. (b).
type IV following the IUPAC nomenclature. The total pore volume is 1.11 cm3/g. The low pressure adsorption values shows that this material has a small amount of micropores. The volume of micropores obtained with the Dubinin–Radushkevich method is 0.27
800 700 600
V N2 (cm STP/g)
STEM. The bright ﬁeld STEM image (Fig. 2a) displays the OMC channels structure (white fringes) and between them the PdNi nanoparticles (black spots) can be identiﬁed. The STEM dark ﬁeld image, in Fig. 2b, displays a region showing an OMC area where the white spots correspond to the PdNi nanoparticles. The atomic composition of several nanoparticles, inside the selected area of Fig. 2b, was determined by EDX (Fig. 2c). It corresponds to a Pd0.61±0.05Ni0.39±0.05 alloy. This result is in good agreement with those obtained by ICP and XRD analyses. In Fig. 3a, the HRTEM image shows some PdNi nanoparticles embedded into the graphene layers of the OMC. The particle size ranges between 2 and 5 nm, i.e. close to the coherent diffraction length determined by XRD. In particular the magniﬁcation of the particle indicated by the circle in Fig. 3a shows the periodic fringes corresponding to crystalline planes in the nanoparticle. In order to further investigate the crystal structure of the PdNi nanoparticles, we performed the fast Fourier transform (FFT) of the Fig. 3b showing one PdNi nanoparticle. The white spots in Fig. 3c correspond to the (-2 0 0) and (2 0 0) diffraction planes of a PdNi nanoparticle oriented perpendicular to [0 0 1] direction. The d-spacing of 1.894 ± 0.18 Å is in good agreement with the formation of a Pd1 xNix alloy. The textural properties of the OMC and OMC–Pd60Ni40 samples were analyzed by nitrogen adsorption at 77 K. The nitrogen adsorption/desorption isotherms of the samples are shown in Fig. 4. In this ﬁgure, the isotherm of the OMC–Pd60Ni40 composite has been normalized to take into account the mass of the Pd60Ni40 alloy, and all the textural parameters are also deﬁned on a mass of carbon basis. The isotherm measured for the pristine OMC is of
500 400 300 200 100 0 0.0
Relative pressure (P/P0) Fig. 4. Nitrogen sorption curves of the Pd-free carbon template (dashed line) and Pd0.60Ni0.40-OMC (full line) at 0.6 MPa and 298 K. The isotherm of the Pd0.60Ni0.40OMC composite has been normalized to take into account the mass of the palladium.
R. Campesi et al. / Microporous and Mesoporous Materials 117 (2009) 511–514
Several sorption/desorption cycles have been conducted at RT for this material and the capacity remains constant giving a proof that the hydrogen storage process in this composite is fully reversible.
H2 uptake (wt%)
0.6 0.4 0.2 0.04
298 K CT
0.01 0.00 0.0
Pressure (MPa) Fig. 5. Hydrogen isotherm sorption curves of the PdNi-free carbon template and Pd0.40Ni0.40-containing carbon template at 77 K (top) and 298 K (bottom). Empty and full symbols stand for PdNi-free carbon template and Pd0.40Ni0.40-containing carbon template, respectively.
cm3/g. This relatively small value is expected for an OMC obtained by chemical vapor deposition. The OMC exhibits a relatively wide distribution of mesopores with a size ranging between 3 and 8 nm. The shape of the isotherm is not modiﬁed by the introduction of the Pd60Ni40 alloy. The amount of nitrogen adsorbed in the low pressure region is exactly the same for the two samples. The micropore volume is unchanged at 0.24 cm3/g. The total surface area of the OMC sample is also nearly unchanged: 847 m2/g for the OMC and 874 m2/g for the OMC–Pd60Ni40. This shows that the alloy nanoparticles are inserted within the mesopores of the carbon material, and no impregnation is done in the micropores. Hydrogen isotherm sorption curves for PdNi-free and PdNi-containing OMCs have been obtained at liquid nitrogen and room temperature (Fig. 5). At 77 K (Fig. 5, top), the PdNi-free OMC sample (empty squares) adsorbs 1.4 wt.% at 1.9 MPa. At this temperature, the hydrogen storage capacity remains constant for the second and following cycles as already noted by Gadiou et al. [1,3]. The value measured for PdNi-free OMC material is in good agreement with the one obtained with similar nanostructured carbons in previous works [3,11]. For the PdNi-OMC sample (full squares), the hydrogen uptake at 77 K is lower and it achieve 0.9 wt.% at 2.0 MPa. This effect is attributed to both the increase of the overall sample weight by adding the Pd0.60Ni0.40 alloy and to the decrease of the micropore volume. At room temperature (Fig. 5, bottom), the hydrogen uptake for the PdNi-free OMC sample (empty circles) is lower than at 77 K. The capacity is about 0.01 wt.% at 0.5 MPa and 298 K. This value is similar to the hydrogen uptake measured at 298 K on other carbon materials such as carbon nanotubes [12,13] or activated carbons and carbon ﬁbers [14,15]. However, for the PdNi–OMC composite (full circles) the maximum storage capacity is almost 3 times larger than for the Pd-free material reaching a value of 0.027 wt.% at 0.5 MPa. Taking into account that palladium–nickel alloys absorb hydrogen in this pressure and temperature range [10,16–19], the larger hydrogen uptake for the PdNi–OMC composite compared to that of PdNi-free OMC is mainly attributed to the PdNi nanoparticles lying within the OMC pores, even if interactions with the OMC cannot be excluded.
We can conclude that the synthesis route we reported here allows the formation of metallic compounds in conﬁned media. In particular, it leads to a high dispersion of a Pd0.60Ni0.40 alloy through the meso/microporous carbon template. Pd0.60Ni0.40 nanoparticles are accessible for solid–gas reactions. Thus, the obtained carbon/PdNi composite shows an enhancement of the hydrogen storage capacity at RT. The reversible hydrogen capacity of the composite is 3 times larger than that of the PdNi-free carbon template. In contrast, the PdNi OMC composite shows lower hydrogen storage capacity compared to that of the PdNi-free carbon template at 77 K. Furthermore, these materials may have hydrogenation properties that can be tuned by modifying the alloy composition. Beside the interest of inserting metal alloys in carbon composites with well deﬁned pores structure for developing materials with improved hydrogen storage properties, such metal alloy-porous carbon composites could be very attractive either as catalysts in several hydrogenation reactions of organic compounds , for biological applications  or for hydrogen storage . Acknowledgments The present work has received funding from the European Community’s Sixth Framework Program trough a Marie Curie Research Training Network (MRTN-CT-2004-512443). The authors would acknowledgment Albrecht Meyer for chemical analysis and Ivana Krkljus for measuring the PCI isotherms. References  R. Gadiou, C. Vix-Guterl, Annales. des. Chimie. 30 (2005) 425.  C. Vix-Guterl, S. Boulard, J. Parmentier, J. Werckmann, J. Patarin, Chem. Lett. 10 (2002) 1062.  R. Gadiou, S.E. Saadallah, T. Piquero, P. David, J. Parmentier, C. Vix-Guterl, Micropor. Mesopor. Mater. 79 (2005) 121.  R. Campesi, F. Cuevas, R. Gadiou, E. Leroy, M. Hirscher, C. Vix-Guterl, M. Latroche, Carbon 46 (2008) 206.  L. Aymard, C. Lenain, L. Courvoisier, F. Salver-Disma, J.M. Tarascon, J. Electrchem. Chem. Soc. 146 (1999) 2015.  J.L. Bobet, E. Grigorova, M. Khrussanova, M. Khristov, P. Stefanov, P. Peshev, D. Radev, J. Alloy. Comp. 366 (2004) 298.  X.B. Yu, G.S. Walker, N. Bowering, D.M. Grant, J. Shen, Z. Wu, B.J. Xia, Electr. Sol. State. Lett. 8 (2005) A 596.  J. Ding, K.Y. Chan, J.W. Ren, F.S. Xiao, Electrochem. Acta 50 (2005) 3131.  L.R. Bidwell, Acta Cryst. 17 (1964) 1473.  S.S.M. Tavares, J.M. Pardal, T. Gurova, J.R.R. Bernardo, J.M. Neto, J. Alloy. Comp. 384 (2004) 152.  R. Gadiou, N. Texier, T. Piquero, S.E. Saadallah, J. Parmentier, J. Patarin, P. David, C. Vix-Guterl, Adsorption 11 (2005) 823.  A. Ansón, M.A. Callejas, A.M. Benito, W.K. Maser, M.T. Izquierdo, B. Rubio, J. Jagiello, M. Thommes, J.B. Parra, M.T. Martínez, Carbon 42 (2004) 1243.  A. Ansón, J. Jagiello, J.B. Parra, M.L. Sanjuán, A.M. Benito, W.K. Maser, M.T. Martínez, J. Phys. Chem. B 108 (2005) 15820.  B. Panella, M. Hirscher, S. Roth, Carbon 43 (2005) 2209.  H. Takagi, H. Hatori, Y. Yamada, S. Matsuo, M. Shiraishi, J. Alloy. Comp. 385 (2004) 257.  S. Luo, C.N. Park, T.B. Flanagan, J. Alloy. Comp. 384 (2004) 208.  T.B. Flanagan, H. Noh, J.D. Clewley, J.G. Barker, Scr. Mater. 39 (1998) 1607.  Z. Gavra, J.R. Johnson, J.J. Reilly, J. Less-Common. Met. 172–174 (1991) 107.  M. Hara, L. Wan, M. Matsuyama, K. Watanabe, J. Alloy. Comp. 428 (2007) 252.  S. Takenaka, Y. Shigeta, E. Tanabe, K. Otsuka, J. Phys. Chem. B 108 (2004) 7665.  T. Sen, A. Sebastianelli, I.J. Bruce, J. Am. Chem. Soc. 128 (2006) 7130.