Spin relaxation in graphene with self-assembled cobalt porphyrin molecules

S. Omar, M. Gurram, I. J. Vera-Marun, X. Zhang, E. H. Huisman, A. Kaverzin, B. L. Feringa, and B. J. van Wees

Phys. Rev. B 92, 115442 – Published 25 September 2015

Abstract

In graphene spintronics, interaction of localized magnetic moments with the electron spins paves a new way to explore the underlying spin-relaxation mechanism. A self-assembled layer of organic cobalt porphyrin (CoPP) molecules on graphene provides a desired platform for such studies via the magnetic moments of porphyrin-bound cobalt atoms. In this work a study of spin-transport properties of graphene spin-valve devices functionalized with such CoPP molecules as a function of temperature via nonlocal spin-valve and Hanle spin-precession measurements is reported. For the functionalized (molecular) devices, we observe a decrease in the spin-relaxation time $τ_{s}$ even up to 50%, which could be an indication of enhanced spin-flip scattering of the electron spins in graphene in the presence of the molecular magnetic moments. The effect of the molecular layer is masked for low-quality samples (low mobility), possibly due to dominance of Elliot-Yafet-type spin relaxation mechanisms.

Received 20 April 2015
Revised 12 July 2015

DOI:https://doi.org/10.1103/PhysRevB.92.115442

Authors & Affiliations

S. Omar^1,*, M. Gurram¹, I. J. Vera-Marun^1,2, X. Zhang³, E. H. Huisman¹, A. Kaverzin¹, B. L. Feringa³, and B. J. van Wees¹

¹Physics of Nanodevices, Zernike Institute for Advanced Materials, University of Groningen, Nijenborgh 4, 9747 AG Groningen, Netherlands
²School of Physics and Astronomy, The University of Manchester, Manchester M13 9PL, United Kingdom
³Stratingh Institute for Chemistry, Zernike Institute for Advanced Materials, University of Groningen, Nijenborgh 4, 9747 AG Groningen, Netherlands

^*Corresponding author: s.omar@rug.nl

Article Text (Subscription Required)

Click to Expand

References (Subscription Required)

Click to Expand

Issue

Vol. 92, Iss. 11 — 15 September 2015

Reuse & Permissions

Access Options

Author publication services for translation and copyediting assistance advertisement

Images

Figure 1
(a) Molecular structure of a cobalt-bound porphyrin (CoPP) complex. ${Co}^{+ +}$ (red circle) is the central atom in the complex, surrounded by the porphyrin ligand. In the porphyrin ring R represents a long-chain alkyl group $(C_{10} H_{21})$ , which is responsible for making weak van der Waals interaction with graphene during the self-assembly. (b) Nonlocal measurement scheme for a graphene spin valve. Graphene (in gray) with a self-assembly of cobalt porphyrin molecules on top (cobalt magnetic moments in red) is probed with ferromagnetic tunnel contacts (in blue). (c) Scanning electron micrograph (SEM) of sample A. The distance between contacts 2 and 3 (transport channel) is $5 μ m$ . Outer contacts are chosen far enough from the inner ones to make sure that they do not affect the spin transport. (d) A scanning tunneling microscopy (STM) image of CVD graphene functionalized with cobalt porphyrin molecules on top (scan area $39 {nm}^{2})$ on Si/ ${SiO}_{2}$ substrate, which demonstrates an ordered self-assembly of the CoPP molecules on graphene. A bright spot in the image corresponds to the core of the porphyrin molecule.
Reuse & Permissions
Figure 2
Resistivity as a function of gate voltage for the CoPP device (sample A) at different temperatures. Solid (dashed) lines correspond to the forward (backward) sweeping direction of the back-gate voltage. The CoPP device shows hysteresis at room temperature (black curve), which disappears at low temperatures (blue curve). Hysteresis at RT indicates a charge-transfer process between the CoPP molecules and graphene, which disappears at low temperatures due to freezing of the charged states in the molecules [24]. A comparison between the Dirac measurement for the pristine state and the CoPP state of sample A is shown in the inset (at 4 K). After functionalization, the sheet resistance increases near the charge neutrality point, which is not significant at high carrier densities.
Reuse & Permissions
Figure 3
Spin-valve measurements for sample A (all the measurements in the electron-doped regime at $n \sim 10^{12} {cm}^{- 2})$ are shown in the positive $x$ axis for the pristine state, and those for the device after the functionalization are shown in the negative $x$ axis. A strongly reduced spin-valve signal is observed after the functionalization.
Reuse & Permissions
Figure 4
(a) Hanle measurements $[(R_{P} - R_{A P}) / 2]$ for the pristine state (black squares) and the CoPP state (red circles) at 4 K (sample B). The corresponding fittings are plotted in line. The curves are normalized with respect to the signal at $B = 0$ . After the functionalization Hanle line shape is broadened, indicating a reduced spin-relaxation time $τ_{s}$ . (b) Hanle measurements for sample B after self-assembly at RT and 4 K. The curves are normalized. Broadening of the black curve (with squares; RT) is dominated by the enhanced $D_{s}$ . The spin-relaxation time $τ_{s}$ only changes from 100 ps (RT) to 112 ps (4K). All the measurements were done at fixed carrier density $(n \sim 10^{12} {cm}^{- 2})$ .
Reuse & Permissions
Figure 5
A summary of (a) $τ_{s}$ and (b) $D_{s}$ , extracted from Hanle analysis, for samples A (squares plus line), B (triangle), and C (circle) before (black) and after (red) the functionalization. Black data correspond to the pristine state, and red data are for the CoPP state of the samples. Reduced $τ_{s}$ and $D_{s}$ were observed for the samples after the functionalization with a weak temperature dependence, which rules out any exchange coupling between the localized magnetic moments and the electron spins in graphene [28] and indicates an enhanced spin-flip process, where the present magnetic moments play only the role of spin-flip scatterers. The effect of the molecular layer is determined by the sample quality $(μ_{e}, D)$ in the pristine state. Since samples A and C have higher mobility and diffusion coefficient, $τ_{s}$ is highly reduced for these samples after the functionalization. Sample B, having lower mobility, did not show any significant change in $τ_{s}$ .
Reuse & Permissions

Physical Review B

covering condensed matter and materials physics