Reduction of metal-graphene contact resistance by direct growth of graphene over metal

  • cc icon
  • ABSTRACT

    The high quality contact between graphene and the metal electrode is a crucial factor in achieving the high performance of graphene transistors. However, there is not sufficient research about contact resistance reduction methods to improve the junction of metal-graphene. In this paper, we propose a new method to decrease the contact resistance between graphene and metal using directly grown graphene over a metal surface. The study found that the grown graphene over copper, as an intermediate layer between the copper and the transferred graphene, reduces contact resistance, and that the adhesion strength between graphene and metal becomes stronger. The results confirmed the contact resistance of the metalgraphene of the proposed structure is nearly half that of the conventional contact structure.

  • KEYWORD

    graphene , contact resistance , chemical vapor deposition

  • 1. Introduction

    Graphene is a zero-band gap semiconductor and has emerged as a promising candidate for device application due to its superior electrical performance [1-5]. Despite the good potential as channel material for field effect transistors (FETs), the contact resistance between metal and graphene can be a major limiting factor of device performance [6-8]. Therefore, understanding the metal-graphene contact is of great scientific and technological importance. Carrier transport from a three dimensional metal to two dimensional graphene has a different mechanism from the conventional metal-metal contact [9]. However, in spite of its importance, there have been few reports on understanding the dominant factors that control the contact resistance between graphene and metal. The detailed properties and methods to improve the contact between graphene and metal are not well studied. Here we demonstrate a new method to reduce the contact resistance between transferred graphene and metal using chemical vapor deposition (CVD) grown graphene as an intermediate layer. In a conventional graphene-metal contact, a transferred graphene layer contacts metal. In our approach, however, there is another synthesized graphene layer between the metal and graphene, which can reduce the contact resistance between them.

    2. Experiments

    In our new approach, a synthesized graphene layer on top of the metal was used underneath the transferred graphene for a FET channel. In a conventional graphene transistor, metal electrodes are deposited on top of the transferred graphene placed on a substrate, which is usually SiO2/Si. In this approach, however, an embedded metal electrode structure is used to insert the graphene intermediate layer between the metal-graphene contacts. The Cu is used as an embedded electrode, which is widely used as catalyst metal for the graphene synthesis by CVD growth [10,11]. The overall process of fabricating the embedded Cu electrode structure is summarized in Fig. 1a. To make space for the Cu electrode, 1-μm-thick SiO2 was etched by BOE after surface

    patterning. The Cu electrode was then deposited by thermal evaporation and lifted off to eliminate Cu on SiO2. A graphene intermediate layer was grown on the embedded Cu electrodes by induction coupled plasma enhanced CVD (ICP-CVD). The substrate temperature was ramped up to 850℃, and C2H2 gas was supplied to a chamber with plasma, which decomposed the C2H2 gas to carbon source for graphene deposition. Then the substrate was cooled to room temperature in vacuum conditions. Another graphene layer as a channel material for a FET, which was grown separately on another Cu/SiO2/Si wafer, was transferred onto the substrate, which is illustrated in Fig. 1b. To confirm the growth of the graphene intermediate layer, Raman spectroscopy was used and prominent G and 2D peaks showed that the carbon film on the metal was truly graphene (Fig. 2c).

    The metal-graphene contact resistance was measured using transfer length methods (TLM) [12-14] and a TLM device is shown in Fig. 2a. The graphene channel width was 10 μm and the channel length varied from 2 μm to 10 μm in steps of 2 μm and a top gate was fabricated in one of the TLM segments with a 20-nm-thick Al2O3 dielectric and Au as the top gate material (the top gate is not shown in the figure). The graphene intermediate layer was synthesized only on the Cu electrodes and bar-shaped graphene as a channel was placed on them. For comparison, not only graphene but also an amorphous carbon (a-C) film was synthesized as an intermediate layer and the Raman spectroscopy measurement result is shown in Fig. 2c.

    3. Results and Discussion

    We conducted three different tests, TLM measurement, Hall effect measurement, and a cantilever beam fracture test, to access the electronic and mechanical performances of the intermediate layer. First, we made TLM measurements at 300 K with a drain bias of 1 V on three different structures, Cu/transferred graphene contact (called direct contact), Cu/intermediate graphene/transferred graphene (indirect contact), and Cu/amorphous carbon/transferred graphene (amorphous contact). The drain current-gate voltage (I-V) characteristics of all of the top gate graphene FETs are plotted in Fig. 3a as a function of top-gate bias for all three different types of FETs. The highest source-drain current was from the indirect contact. The total resistance (Rtotal) between a source and a drain electrode includes contact and sheet resistances, and its dependence on the channel length is plotted in Fig. 3b. The Rtotal was measured at a gate bias of -1 V. The black (red, blue) squares denote the Rtotal of the FETs having direct (indirect, amorphous) contact and the lines are linear fitting curves. The resistance values increased with the channel length as expected. However, the contact resistance was constant regardless of the channel length, and the value was obtained at the intercept at the zero channel length.

    The measured contact resistance is plotted in Fig. 4. Note that the error bars are significantly larger near Vdirac due to the large resistance used in extrapolation around Vdirac, leading to larger absolute errors despite the high fit quality. Moreover, near Vdirac, Rtotal changes rapidly and a slight variation in Vdirac across all different structures leads to additional errors. We measured more than 20 FETs for each type of contact and consistent results were obtained. The smallest contact resistance was observed in the indirect contact. When the gate voltage was applied in the range of -2.0 V to -1.7 V far away from the Vdirac, the contact resistances did not vary significantly as shown in the inset of Fig. 4, and their values are similar to reported values [15]. The contact resistances of the proposed structure, indirect Cu-graphene contact by a graphene intermediate layer, had less

    than half the value of the direct Cu-graphene throughout most of the investigated gate voltages. That means that the intermediate layer assisted the charge transfer from Cu to the transferred graphene. In the amorphous contact, the contact resistance was over the 4 kΩ- μm because the charge transfer was harder than the other structures due to low conductivity of the amorphous carbon film.

    As another test method for the electrical performance enhancement and the adhesion property, we used Hall effect measurements and cantilever beam fracture measurement. For the electrical contact analysis, resistivity of Cu, CVD grown graphene on Cu, and transferred graphene on Cu had been measured by the Hall Effect measurement (Fig. 5). The resistivity of the grown graphene on Cu was lower than the transferred one, and it was closer to the pure Cu value. The adhesion energy between graphene and Cu was measured by double cantilever beam fracture mechanics testing with a high-precision micromechanical test system (Delaminator Adhesion Test System; DTS Company, MenloPark, CA, USA) [16]. The transferred graphene and Cu had 0.6 J/m2 adhesion energy, while the CVD graphene had 0.72 J/m2 adhesion energy in the Fig. 5 inset. The

    CVD grown graphene had harder contact with Cu than the transferred graphene, which can make the electric current easy to turn on. From these results, we conclude that CVD grown graphene has better surface contact than transferred graphene.

    4. Conclusions

    In summary, we proposed a new contact structure to reduce contact resistance by inserting a graphene intermediate layer between the metal and the transferred graphene. Using TLM, Halleffect, and adhesion tests, we verified that CVD grown graphene has lower resistivity and higher adhesion energy than transferred graphene on Cu. The proposed structure decreases nearly half the contact resistance. The results indicate that CVD grown graphene is more strongly contacted with Cu than the transferred graphene, which makes the charge transfer more effective between the electrode and the channel.

  • 1. Novoselov KS, Geim AK, Morozov SV, Jiang D, Zhang Y, Dubonos SV, Grigorieva IV, Firsov AA (2004) Electric field effect in atomically thin carbon films. [Science] Vol.306 P.666 google doi
  • 2. Schedin F, Geim AK, Morozov SV, Hill EW, Blake P, Katsnelson MI, Novoselov KS (2007) Detection of individual gas molecules adsorbed on graphene. [Nat Mater] Vol.6 P.652 google doi
  • 3. Murali R, Yang Y, Brenner K, Beck T, Meindl JD (2009) Breakdown current density of graphene nanoribbons. [Appl Phys Lett] Vol.94 P.243114 google doi
  • 4. Lee CH, Nishimura T, Saido N, Nagashio K, Kita K, Toriumi A (2009) Record-high electron mobility in Ge n-MOSFETs exceeding Si universality. [IEEE International Electron Devices Meeting] P.1 google doi
  • 5. Farmer DB, Chiu HY, Lin YM, Jenkins KA, Xia F, Avouris P (2009) Utilization of a buffered dielectric to achieve high field-effect carrier mobility in graphene transistors. [Nano Lett] Vol.9 P.4474 google doi
  • 6. Lin YM, Dimitrakopoulos C, Jenkins KA, Farmer DB, Chiu HY, Grill A, Avouris P (2010) 100-GHz transistors from wafer-scale epitaxial graphene. [Science] Vol.337 P.662 google doi
  • 7. Lin YM, Jenkins KA, Valdes-Garcia A, Small JP, Farmer DB, Avouris P (2008) Operation of graphene transistors at gigahertz frequencies. [Nano Lett] Vol.9 P.422 google doi
  • 8. Xia F, Mueller T, Golizadeh-Mojarad R, Freitag M, Lin YM, Tsang J, Perebeinos V, Avouris P (2009) Photocurrent imaging and efficient photon detection in a graphene transistor. [Nano Lett] Vol.9 P.1039 google doi
  • 9. Xia F, Perebeinos V, Lin YM, Wu Y, Avouris P (2011) The origins and limits of metal-graphene junction resistance. [Nat Nanotechnol] Vol.6 P.179 google doi
  • 10. Li X, Cai W, An J, Kim S, Nah J, Yang D, Piner R, Velamakanni A, Jung I, Tutuc E, Banerjee SK, Colombo L, Ruoff RS (2009) Large-area synthesis of high-quality and uniform graphene films on copper foils. [Science] Vol.324 P.1312 google doi
  • 11. Lee Y, Bae S, Jang H, Jang S, Zhu SE, Sim SH, Song YI, Hong BH, Ahn JH (2010) Wafer-scale synthesis and transfer of graphene films. [Nano Lett] Vol.10 P.490 google doi
  • 12. Danneau R, Wu F, Craciun MF, Russo S, Tomi MY, Salmilehto J, Morpurgo AF, Hakonen PJ (2008) Shot noise in ballistic graphene. [Phys Rev Lett] Vol.100 P.196802 google
  • 13. Blake P, Yang R, Morozov SV, Schedin F, Ponomarenko LA, Zhukov AA, Nair RR, Grigorieva IV, Novoselov KS, Geim AK (2009) Influence of metal contacts and charge inhomogeneity on transport properties of graphene near the neutrality point. [Solid State Commun] Vol.149 P.1068 google doi
  • 14. Venugopal A, Colombo L, Vogel EM (2010) Contact resistance in few and multilayer graphene devices. [Appl Phys Lett] Vol.96 P.013512 google doi
  • 15. Malec CE, Elkus B, Davidovi? D (2011) Vacuum-annealed Cu contacts for graphene electronics. [Solid State Commun] Vol.151 P.1791 google doi
  • 16. Yoon T, Shin WC, Kim TY, Mun JH, Kim TS, Cho BJ (2012) Direct measurement of adhesion energy of monolayer graphene as-grown on copper and its application to renewable transfer process. [Nano Lett] Vol.12 P.1448 google doi
  • [Fig. 1.] (a) The overall process of fabricating an embedded Cu electrode structure. (b) The structure to improve contact between the electrode and graphene using chemical vapor deposition (CVD) grown graphene as an intermediate layer.
    (a) The overall process of fabricating an embedded Cu electrode structure. (b) The structure to improve contact between the electrode and graphene
using chemical vapor deposition (CVD) grown graphene as an intermediate layer.
  • [Fig. 2.] (a) The transfer length methods (TLM) device for contact resistance measurement, and (b) transferred graphene as a channel in the TLM device. (c) The Raman spectroscopy of graphene and amorphous carbon (a-C) on an embedded Cu substrate using chemical vapor deposition growth.
    (a) The transfer length methods (TLM) device for contact resistance
measurement, and (b) transferred graphene as a channel in the TLM device.
(c) The Raman spectroscopy of graphene and amorphous carbon (a-C) on an
embedded Cu substrate using chemical vapor deposition growth.
  • [Fig. 3.] (a) Drain current-gate voltage curve of a graphene field effect transistor (FET) with a 2 μm channel length at 300 K. (b) Derived contact resistance of three different structures (transferred graphene-Cu, transferred graphene-synthesized graphene/Cu, and transferred graphene-a- C/Cu) using the top gate FET at gate biases of -1 V at 300 K.
    (a) Drain current-gate voltage curve of a graphene field effect
transistor (FET) with a 2 μm channel length at 300 K. (b) Derived contact
resistance of three different structures (transferred graphene-Cu, transferred
graphene-synthesized graphene/Cu, and transferred graphene-a-
C/Cu) using the top gate FET at gate biases of -1 V at 300 K.
  • [Fig. 4.] Contact resistance as a function of gate bias, showing the gate bias and contact structure dependence. The inset graph shows contact resistance results at gate biases from -2 V to -1.7 V.
    Contact resistance as a function of gate bias, showing the gate
bias and contact structure dependence. The inset graph shows contact
resistance results at gate biases from -2 V to -1.7 V.
  • [Fig. 5.] The resistivity of three different structures: Cu only, as-grown graphene on Cu (as grown G-Cu), and transferred graphene on Cu. The inset shows the adhesion energy between the as-grown graphene with Cu and the transferred graphene with Cu.
    The resistivity of three different structures: Cu only, as-grown
graphene on Cu (as grown G-Cu), and transferred graphene on Cu. The
inset shows the adhesion energy between the as-grown graphene with
Cu and the transferred graphene with Cu.