In our first “Behind the Paper” post, we cover a paper recently published in Nature Communications by Mallikarjuna Gurram et al., entitled “Bias induced up to 100% spin-injection and detection polarizations in ferromagnet/bilayer-hBN/graphene/hBN heterostructures”. We have interviewed Mallikarjuna Gurram, who is currently working towards a PhD at the Zernike Institute for Advanced Materials, University of Groningen, under the supervision of Prof Bart van Wees.
Could you briefly outline the key findings of your paper?
M. Gurram: “A poor injection and detection of spins in graphene has been a bottleneck in realizing graphene spintronic devices for practical purposes. We address this issue in a fully hexagonal boron nitride (hBN) encapsulated graphene spin valve device which demonstrated the possibility to inject and detect spins in graphene with differential spin injection and detection polarizations up to 100% by applying a bias across the cobalt/2L-hBN/graphene/hBN contacts at room temperature.
Surprisingly, we also found a unique sign inversion of the spin polarization, at a small voltage bias, which is unique only to two-layers of hBN tunnel barrier. Besides, we also demonstrate two-terminal spin valve with a record magnitude of the spin signal up to 800 Ohms and magnetoresistance ratio of 2.7%.”
What is your role in this work?
M. Gurram: “I, along with my colleague Siddhartha Omar and my PhD advisor Prof. Bart van Wees conceived the experiments. I carried out the sample fabrication and measurements. All authors carried out the analysis, discussed the results and the manuscript.”
What was the genesis of this paper? How did you come to this particular problem?
M. Gurram: “Nearly a decade ago, our group first reported the electrical spin injection and transport in graphene, published in Nature in 2007. Since then, most research on graphene spintronics is focused on finding a better substrate and better tunnel barrier in order to realize graphene’s full potential in a clean environment. In 2010, hBN was found to be a better substrate for studying spin transport in graphene. However, the conventional oxide tunnel barriers were still limiting the spin injection efficiencies and spin lifetimes due to the presence of unwanted spin sinks and spin scatterers at the interface. Later, it was discovered that hBN can also act as a tunnel barrier in its thin form with an atomically smooth surface and no signature of spin sinks.
In principle, the problems associated with the commonly used SiO2 substrate and the conventional oxide tunnel barriers can be overcome by using hBN material. Therefore, in our study, we prepared a fully hBN encapsulated device where the bottom thick-hBN acts as a substrate and the top thin monolayer-hBN acts as a tunnel barrier for electrical spin injection and detection, published in Phys. Rev. B in 2016. Herein, we reported a low spin injection efficiency due to the atomically thin nature of hBN tunnel barrier causing a well-known conductivity mismatch problem between the ferromagnet and graphene. To overcome this mismatch and to achieve large spin injection efficiency, a recent theoretical study, published by Q. Wu, et al. in Phys. Rev. Appl. in 2014, suggests to use bi or tri layers of hBN. In a follow up experiment in this paper, we used a similar device geometry as before but with bilayer-hBN as a tunnel barrier for spin injection and detection.”
What is the most empowering implication of your results?
M. Gurram: “Our results have some interesting implications. Firstly, large spin polarization achieved in this type of graphene spin valve devices can be used in spin-transfer torque memory devices (random access memory, RAM) technology wherein the magnetization orientation of various magnetic materials can be changed. Secondly, we found sign reversal of spin polarization just by applying electrical bias across the ferromagnet/bilayer-hBN/graphene/hBN interface. The implication of this finding is that it avoids the necessity of external magnetic field to change the orientation of spins in graphene. Other promising applications include magnetic sensors, spin logic gates, and spin transistors.”
How have 2D materials been uniquely instrumental to enabling these results?
M. Gurram: “Without 2D materials our results would not have been achieved. Firstly, graphene is the first 2D material to demonstrate spin transport at room temperature, and instrumental in enlightening 2D spintronics research field. Finding a good tunnel barrier for electrical spin injection and detection in graphene has been the basis for most of the graphene spintronics research in the past decade. For this, we need a very thin, few atoms think, ultra-smooth, insulating, and pinhole free material. hBN, an isomorph of graphene and belonging to 2D materials family, has been found to satisfy these requirements. Moreover, multilayer hBN acts as a flat and neutral dielectric substrate for achieving large electron mobility in graphene. Besides their individual properties of graphene and hBN, one unique characteristics of the 2D materials is the possibility of making hBN-graphene-hBN heterostructures by utilizing the van der Waals forces attraction at their interfaces.”
Can you describe the main challenges associated to the preparation of this manuscript? Any anecdotes you’d like to share with us?
M. Gurram: “It may sound trivial but finding a thin layer of hBN tunnel barrier flake was challenging for making the device. We regularly use a SiO2/Si substrate with oxide thickness of 300 nm to exfoliate and find the graphene flakes. However, it is not optimal for providing good optical contrast to find the hBN flakes down to monolayer. Now we started to use 90 nm SiO2/Si substrate which gives better colour contrast for “transparent” hBN. However, we still face the problem to find an optimal thin bilayer of hBN flake for tunnel barrier.”
Anything that stroke you as particularly surprising, unexpectedly pleasant/unpleasant during the peer review process?
M. Gurram: “During the review process, I was very glad that the reviewers were highly positive and enthusiastic about the results and the paper publishing went smoothly. One useful comment, pointed by a referee was the concern about relatively lower mobility and spin transport properties of our sample in spite of minimizing the chances of impurity inclusion during the fabrication process. We still tend to believe that thin hBN flakes such as mono-bi layers are not optimal for complete encapsulation of graphene and one needs to find alternate routes to optimize the cleanliness of these samples.”
What is your favourite 2D paper published in 2016/2017, and why?
M. Gurram: “The paper published by Huang B. et al., Nature 546, 270(2017) is quite interesting to me. The paper reported for the first time, an intrinsic ferromagnetism in a 2D material, CrI3. Different 2D materials with distinct physical properties have been found since the discovery of graphene. These can be broadly categorized into conductors, semiconductors, and insulators. Whereas, this paper becomes the first one to report a 2D material, CrI3 which is ferromagnetic down to monolayer. Magnetic materials are essential building blocks of spintronics devices. Especially graphene spin valves need a ferromagnetic material for electrical spin injection and detection. However, the conventional ferromagnetic materials used in graphene spinvalves include cobalt and permalloy, need to be grown in bulk amounts on top of the device. Our spin valve device, ferromagnet/bilayer-hBN/graphene/hBN consists of 2D materials except the cobalt ferromagnet. The use of 2D ferromagnet in place of the conventional ferromagnet creates a possibility of making ultra-thin graphene spin valve devices completely out of 2D materials. Of course, 2D ferromagnets hold promise for many more opportunities for spintronics as outlined in this paper.”
Which is the development in the field of 2D materials that you would like to see in the next 10 years?
M. Gurram: “Currently, academic research is done using micro scale flakes which is great to explore the interesting physics. But for the practical applications, these materials should be produced in bulk with high quality and with a possibility of growing heterostructures of desired layers.”
And now, what’s next?
M. Gurram: “From our results, we do not understand what is causing the unusual behaviour of spin polarization as a function of bias applied across ferromagnet/bilayer-hBN/graphene/graphene. Currently we are working on understanding the underlying mechanism.”