Elsevier

Carbon

Volume 58, July 2013, Pages 139-150
Carbon

Graphene as catalyst support: The influences of carbon additives and catalyst preparation methods on the performance of PEM fuel cells

https://doi.org/10.1016/j.carbon.2013.02.043Get rights and content

Abstract

The reduction of the platinum amount for efficient PEM (polymer electrolyte membrane) fuel cells was achieved by the use of graphene/carbon composites as catalyst support. The influences of the carbon support type and also of the catalyst preparation method on the fuel cell performance were investigated with electrochemical, spectroscopic and microscopic techniques. Using pure graphene supports the final catalyst layer consists of a dense and well orientated roof tile structure which causes strong mass transport limitations for fuels and products. Thus the catalysts efficiency and finally the fuel cell performance were reduced. The addition of different carbon additives like carbon black particles or multi-walled carbon nanotubes (MWCNT) destroys this structure and forms a porous layer which is very efficient for the mass transport. The network structure of the catalyst layer and therefore the performance depends on the amount and on the morphology of the carbon additives. Due to optimizing these parameters the platinum amount could be reduced by 37% compared to a commercial standard system.

Introduction

Graphene is known as monolayers of sp2 bonded carbon atoms into a two-dimensional structure and has become of rapidly expanding interest due to its extraordinary features such as high electronic and thermal conductivity [1], [2], [4], an ambipolar field-effect [3], intriguing (magneto-)transport characteristics [5], [7], and mechanical properties [6]. Graphene exhibits also a very high surface area (∼2630 m2 g−1) [8], much higher than graphite (∼10 m2 g−1) or carbon nanotubes (1300 m2 g−1). All these remarkable properties make graphene promising for applications like polymer-composite materials [9], photo-electronics [10], field-effect transistors [11], electromechanical systems [12], sensors [13], [15], [16], hydrogen storage [14], energy conversion and storage, batteries [17] and drug delivery systems [18].

PEM fuel cells are clean and environmentally-friendly power sources, which are essential for future energy solutions [19]. They exhibit good energy efficiency and high power density per volume, but the high costs of the noble metal catalysts, electrolyte membranes, bipolar plates and the control system constrain a widespread commercialization. The MEA-fabrication costs (membrane electrode assemblies) can be reduced through several approaches such as (i) reducing the platinum loading (ii) the use of high temperature-tolerant membranes and (iii) new alternative electrocatalyst materials [20], [21]. In the last 3 years some very promising efforts were made in electrode development and in membrane technology to reduce the price of fuel cell systems. Zirfon™ was sucsessfully tested to replace the expensive Nafion™-membranes [22] and activated carbon air-cathodes have a great potential as inexpensive alternative to Pt-catalysed electrodes in microbial fuel cells [23], [24]. For electrochemical cogeneration of chemicals and also for NO/hydrogen fuel cells some effective concepts for the preparation of gas diffusion electrodes in a cold rolling or casting process with carbonaceous compounds and binders were reported recently [25].

Our strategy to lower the MEA-costs is the improvement of the cell performance by developing more efficient catalyst support systems which has the following properties:

  • (i)

    high specific surface area, which improves the dispersion of the catalytic metals,

  • (ii)

    improved chemical and electrochemical stability at operation temperatures between 60 and 100 °C,

  • (iii)

    enhanced electronic conductivity due to strong percolation, and

  • (iv)

    high catalytic activity for hydrogen oxidation of the graphene itself [26], [27].

Currently, the most common electrocatalysts are platinum and platinum-based alloys supported on carbon black. A major problem of the catalysts is aggregation on the surface of the carbon black support [28]. A further problem is the formation of inactive catalysts during the manufacturing process that means about 30% of catalysts are not located at the triple-phase boundaries (TPBs) where the fuel cell reactions take place.

To reduce the amount of inactive catalysts, catalyst supports and also catalyst particles with a strong reduced size were developed [26] but these systems are very hard to handle in coating processes and additionally the very small catalyst particles are thermally instable. A third problem is the electronic conductivity because in such small carbon particles the graphite content is very low. Graphitized carbon black particles with good electronic conductivity have a larger specific surface of about 65 m2 g−1 (e.g. TIMREX™ HSAG400, Ensaco™ 350G). An alternative to conventional carbon black materials could be graphene flakes. The theoretical specific surface area of single-layer graphene is about 2600 m2 g−1, but the specific surface area of multilayer graphene is much smaller (70–130 m2 g−1). An advantage of these materials is the high graphite content. Only a few fundamental research results for the preparation of isolated metal particles on the surface of single-, double-, and few-layer graphene sheets [29], [30] are known.

In addition to the active surface area of the catalyst/carbon black system, the stability is very important. There are two mechanisms of degradation: coalescence sintering and Ostwald ripening [27]. Coalescence sintering is enhanced by carbon corrosion which occurs under a high potential in an acidic environment. During the oxidizing process surface-oxygen containing species are formed and the catalyst–carbon interactions are weakened. Graphene may improve catalyst stability by providing a more stable support material and by strengthening the interaction of the catalyst with the support [26]. These enhanced interactions between the metal particles and the graphene surface are described in [31].

Finally, in a high-performance PEM fuel cell electrode the supported catalysts should be located at the TPBs. The flake-like structure of the graphene should lead to a very porous and percolated system which is essential for the mass transport and the electronic conductivity [32]. This work shows two routes of platinum catalyst immobilization on the surface of multilayer graphene flakes: (i) thermally-induced chemical reduction of a platinum precursor and (ii) electrochemical reduction of a platinum precursor by potentiostatic electrodeposition using a hydrogen depolarized anode (HDA).

The advantage of the first method is the decrease of the degree of agglomeration of platinum precursors during the reduction process because of the high interaction between adsorbed precursor ions and the graphene surface. The catalyst loading can be controlled by the amount of hexachloroplatinic acid. The HDA-method [33] has the advantage that it enables a very precise control of the catalyst loading and leads to a very homogeneous platinum particle distribution. This method assures the formation of nearly 100% of active catalyst particles which are located at the TPBs. The samples were characterized by ICP-OES, SEM, TEM, XRD and Raman spectroscopy. The electrodes were used to prepare MEAs which were characterized in a modular fuel cell test station.

Section snippets

Materials

Many preparation methods for graphene and graphene oxide exist in the literature but most of them work in small laboratory scale [34], [35], [36], [37]. The evaluation of graphene as catalyst support requires relatively large sample amounts in the range of 10–20 g, each. This amount can only be prepared in several batches in a time consuming and extensive way. For reproducibility reasons commercial graphene samples which are available in the required amounts were used (Table 1).

Microscopic techniques

TEM pictures

Macrostructures of different graphene types

Fig. 2A and B show electron microscopic images of AO-2 graphene. The graphene flakes are well separated and consist of 44-52 monolayers, which can be clearly seen in the TEM image (Fig. 2B). The thickness of the multilayer flakes is between 11 and 13 nm.

Fig. 2C shows also well-dispersed graphene flakes of sample C1, which tend to exhibit a very rough surface structure. The average multilayer flake thickness is 3.7–4.0 nm (Fig. 2D), corresponding to 15–32 monolayers. As can be seen in the SEM

Conclusion

Different graphene types were tested for electrochemical energy conversion especially for fuel cell applications. It could be shown that pure graphene forms a very dense and ordered structure during the coating process of gas diffusion layers. For this reason the produced MEAs suffered from a very low performance. Carbon black and MWCNT were used to act as spacers between the graphene flakes, in order to increase the exposed surface area of the graphene, the electronic conductivity and the mass

Acknowledgments

This work was financially supported by the Bundesministerium für Bildung und Forschung within the framework of the WING programme (Einsatz von Graphenen in der Energietechnik – Lithiumbatterien und Brennstoffzellen – LiBZ). We wish to thank Sylvia Kuhn, Tanja Müller, Elfi Jungblut, Dieter Münch, Dr. Haibin Gao, Christoph Thome, Sabrina Fricke, Franco Cappuccio and Andreas Witzmann. We also thank Prof. Dr. Rolf Hempelmann, Bernd Oberschachtsiek and Prof. Dr. Axel Lorke for stimulating and

References (49)

  • J.B. Hou

    Electrochemical impedance investigation of proton exchange membrane fuel cells experienced subzerotemperature

    J Power Sources

    (2007)
  • J. Mitzel et al.

    Electrodeposition of PEM fuel cell catalysts by the use of a hydrogen depolarized anode

    Int J Hydrogen Energy

    (2012)
  • S. Stankovich et al.

    Synthesis of graphene-based nanosheets via chemical reduction of exfoliated graphite oxide

    Carbon

    (2007)
  • Y.S. Yun et al.

    Porous graphene/carbon nanotube composite cathode for proton exchange membrane fuel cell

    Synth Met

    (2011)
  • A.H. Castro Neto et al.

    The electronic properties of graphene

    Rev Mod Phys

    (2009)
  • A.A. Balandin et al.

    Superior thermal conductivity of single-layer graphene

    Nano Lett

    (2008)
  • K.S. Novoselov et al.

    Electric field effect in atomically thin carbon films

    Science

    (22 October 2004)
  • A.K. Geim et al.

    The rise of graphene

    Nat Mater

    (2007)
  • I.M. Katsnelson

    Graphene: carbon in two dimensions

    Mater Today

    (2007)
  • C. Lee et al.

    Measurement of the elastic properties and intrinsic strength of monolayer graphene

    Science

    (18 July 2008)
  • F. Guinea et al.

    Energy gaps and a zero-field quantum Hall effect in graphene by strain engineering

    Nat Phys

    (2010)
  • D.M. Stoller et al.

    Graphene-based ultracapacitors

    Nano Lett

    (2008)
  • A.D. Dikin et al.

    Preparation and characterization of graphene oxide paper

    Nature

    (2007)
  • H. Park et al.

    Doped graphene electrodes for organic solar cells

    Nanotechnology

    (2010)
  • Cited by (97)

    • Effect of reduced graphene oxide addition on cathode functional layer performance in solid oxide fuel cells

      2023, International Journal of Hydrogen Energy
      Citation Excerpt :

      There exist many studies in the related literature on the use of graphene, graphene oxide or doped graphene in fuel cells including even several review articles [5–20]. They generally center on the use of graphene or its derivatives as a catalyst support, electrolyte or catalyst for polymer electrolyte membrane fuel cell (PEMFC), direct methanol fuel cell (DMFC) or microbial fuel cell (MFC) [21–37]. On the other hand, it is seen that limited number of studies have been performed on the use of graphene or its derivatives in solid oxide fuel cells (SOFCs) compared to other fuel cells.

    View all citing articles on Scopus
    View full text