Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Stepped Graphene-based Aharonov-Bohm Interferometers

Stepped Graphene-based Aharonov-Bohm Interferometers V. Hung Nguyen and J.-C. Charlier Institute of Condensed Matter and Nanosciences, Universit e catholique de Louvain, Chemin des  etoiles 8, B-1348 Louvain-la-Neuve, Belgium Aharonov-Bohm interferences in the quantum Hall regime are observed when electrons are trans- mitted between two edge channels. Such a phenomenon has been realized in 2D systems such as quantum point contacts, anti-dots and p-n junctions. Based on a theoretical investigation of the magnetotransport in stepped graphene, a new kind of Aharonov-Bohm interferometers is proposed herewith. Indeed, when a strong magnetic eld is applied in a proper direction, oppositely propagat- ing edge states can be achieved in both terrace and facet zones of the step, leading to the interedge scatterings and hence strong Aharonov-Bohm oscillations in the conductance in the quantum Hall regime. Taking place in the unipolar regime, this interference is also predicted in stepped systems of other 2D layered materials. The fascinating properties of quantum hall devices graphene p-n junctions can also work as AB interferom- arise from their ideal 1D edge states formed in a 2D eters [32{35] in the quantum Hall regime. In particular, electron system when a high magnetic eld is applied the oppositely propagating edge states are formed in two [1, 2]. These edge states are particularly attractive due di erent doped zones and their interaction at the p-n to their large coherence lengths, which is mandatory for interface acquires conductance oscillations of the AB pe- constructing electron interferometers. However, since the riodicity at high B- elds. edge channels are spatially separated, a mechanism for Motivated by such scienti c context, a new kind creating the electron transmission between them is re- of Aharonov-Bohm interferometers based on stepped quired to achieve the interference e ects. In this regard, graphene channels is proposed herewith. These non- one explored technique consists in building constrictions planar systems have been actually achieved in several (quantum point contacts) in a sample, where the in- experimental situations. For example, the step bunch- teredge tunneling paths can occur [3{17]. Setups consist- ing on the SiC surface is often observed in epitaxial ing of a pair of quantum point contacts with an internal graphene growth by thermal decomposition of SiC [37{ cavity has been demonstrated to work well as quantum Hall, electronic Fabry-P erot, and Aharonov-Bohm inter- ferometers. Another mechanism has also been suggested in systems consisting of an antidot introduced between y-axis (a) their edges [18{25]. Electronic currents encircling the antidot can be achieved and a similar Aharonov-Bohm (AB) interference is hence observed. Graphene, a truly 2D material, is an ideal platform z-axis for investigating quantum Hall and interference e ects. x-axis Remarkably, owning to an unique linear dispersion and normal B Dirac-like fermions [26], Landau levels and a half-integer quantum Hall e ect with an unusual quantized sequence compared to the conventional systems have been ob- (b) served in graphene when a strong magnetic eld (B- eld) is applied [27, 28]. In addition, with its semimetal charac- normal B ter, quantum Hall systems of graphene can work in both the unipolar and bipolar regimes that can be generated normal B and controlled by gate voltages [28{35]. Interestingly, in applied B the bipolar regime the chiral edge states equilibration and interedge scatterings at the p-n interfaces in graphene have been observed, resulting in fractional conductance plateaus [31]. FIG. 1. Stepped graphene (a) with the ribbon width W and With its typically high carrier mobilities, graphene is the length of facet zone L . (b) Schematic of the side view also an ideal material to perform the investigation on illustrating the applied magnetic eld (red dashed lines of interference e ects, including the Aharonov-Bohm one. arrows) and its normal components (green and blue arrows, respectively) in both terrace and facet zones.  and  are Several experimental and theoretical observations of the B S the angles of the eld and of the facet zone relative to the AB e ect in graphene nanorings have been reported terrace one (Ox axis), respectively. (i.e., see Ref. [36] and references therein). Remarkably, arXiv:1812.02845v2 [cond-mat.mes-hall] 20 Jun 2019 2 47], a promising method for the production of large-area same direction are obtained, thus inducing the same high-quality graphene. These stepped graphene channels propagating edge states in the two zones as illustrated with terrace size of several m and step height of tens in Fig.2b. Therefore, when a large B- eld is applied, a nm can be controllably produced by varying the heating conventional Landau quantization is still obtained, i.e., rate [38]. Non-planar graphene systems have been also the conductance represents quantized values as in Fig.2a synthesized in an even better controllable way by draping for  = 90 . For    , the B -component in the B B S N graphene on pre-structured substrates [48{55]. facet zone is canceled and hence the B- eld has no e ect on the in-plane transport in this zone. Consequently, the In this work, a B- eld is found to induce di erent ef- system behaves as a heterojunction consisting of nite- fects on the electron motion in the terrace and facet zones and zero-magnetic eld zones and the scatterings at their of stepped graphene, essentially resulting from the cre- interface basically explain the reduction of conductance ation of di erent normal components of the eld in these obtained for  = 60   displayed in Fig.2a. zones. This feature has been also observed in several B S non-planar systems of graphene [41{47] and 2DEG [56{ Most interestingly, when 0 <  <  , two opposite B S 61], leading to anisotropic pictures when the transport B -components alternate in the terrace and facet zones, takes place in the directions aligned parallel and per- as discussed above. Similarly to the e ects of B- eld in pendicular to the step edge. Here, a novel phenomenon di erent doped zones of graphene p-n junctions [31{33], is predicted when tuning the direction of B- eld ap- opposite edge states in the terrace and facet zones are plied to stepped graphene. In particular, an inhomo- geneous pro le containing alternatively opposite normal B-components along the channel can be created, thus in- (a) ducing accordingly opposite edge states. The interaction between these edge states nally results in strong AB os- cillations in the quantum Hall regime as presented in this  = 90° >  B S article.  = 60°   B S The considered systems consist in graphene nanorib-  = 30° <  B S bons (GNRs) in the step geometry as illustrated in Fig.1. In general, in-plane local strains can occur in the bent zones of the step, however, have been demonstrated to be small (i.e., < 1%) and negligible even in epitaxial graphene [55, 62]. Moreover, such small local strains are shown not to strongly a ect the predicted AB interfer- ence picture [63]. Therefore, these local strains are ne- glected and the p tight-binding Hamiltonian [26] is em- ployed to compute the magneto-transport in this work. In particular, when a B- eld is applied, (b)  >  B S X X y i y nm H = U c c + t e c c (1) n n 0 m n n hn;mi    th where U represents the potential energy at the n site, t = 2:7 eV corresponds the nearest-neighbor hopping e m energies, and  = A(r)dr is the Peierls phase (c) nm ~ r  <  describing the e ects of the B - eld. Here, the magnetic B S eld B = B(cos  ; 0; sin  ) is considered by introduc- B B ing the vector potential A(r) = B(y sin  ; z cos  ; 0).   B B The above Hamiltonian is solved using the Green's- function technique [63{65], allowing for the calculation of the transport quantities perpendicular to the step edges, i.e., along the Ox axis shown in Fig.1. FIG. 2. (a) Conductance as a function of B- eld in a stepped Fig.2a displays the dependence of conductance on B- armchair GNR for di erent angles  and with  = 60 , B S eld applied in di erent directions in a step constituted E = 75 meV , L ' 75 nm, and W ' 40 nm (i.e., number F F by an armchair GNR. As mentioned, the applied B- eld of dimer lines N = 324, a semiconducting GNR). (b,c) Dia- induces di erent normal components (B ) in the terrace grams illustrating the interedge scatterings for  >  and N B S <  , when the normal B-components in the facet and and facet zones, i.e., B = B sin  (green arrows) and B S N B terrace zones (see Fig.1b) are pointing out in the same and B sin(  ) (blue arrow), respectively (see Fig.1b). S B opposite directions, respectively. First, for  >  , B -components pointing out in the B S N 3 are formed inside the facet zone (see Fig.3) and hence the value of S in Eq.(2) to estimate B is basically pro- portional to but smaller than the area of the facet zone. Thus, the origin of observed conductance oscillation is re- ally the AB interference due to the interaction between edge states in both terrace and facet zones. The observed AB interference is also found to be sen- sitive to some other structural parameters. First, per- (a) fect armchair GNRs can be divided in two main classes with di erent electronic properties, either quasi-metallic with negligible bandgap or semiconducting, depending on high the number of zigzag lines N across the ribbon width: N = 3p + 2 and N 6= 3p + 2 [67], respectively. More- z z over, in contrast to semiconducting ribbons, the rst sub- band of metallic GNRs is linear, thus inducing massless fermions at low energies that contribute mainly to the transport at high B- elds. As a consequence, a signi cant zero (b) Ox axis (nm) (a) semiconducting GNRs L  75.0 nm FIG. 3. Left-injected local density of states at B = 25.8 T L  75.0 nm (a) and 27.4 T (b), corresponding to conductance peak and L  150.0 nm valley, respectively (see Fig. 2a for  = 30 ). created, thus inducing the strong interedge scatterings at their interface as illustrated in Fig.2c. As an important consequence, the conductance as a function of B- eld represents a strong AB oscillation in the quantum Hall regime (see the case of  = 30 in Fig.2a). This result is essentially due to the interedge backscatterings diagram- matically described in Fig.2c and is further demonstrated by analyzing the computed left-injected local density of (b) metallic GNRs perfect edge states in Fig.3, that illustrates the left-to-right electron- disorder 1 wave propagation. Indeed, backscatterings are almost disorder 2 absent (Fig.3a) when the phase coherence condition is disorder 3 satis ed, leading to conductance peaks. In the phase incoherence condition, strong interedge backscatterings (Fig.3b) and hence a low conductance are achieved. AB oscillation period observed in quantum rings with area S is known to be given by B = h=eS [66]. To examine this property in the considered systems (for <  ), the above formula should be rewritten as B S h 1 B = (2) eS jsin(  )j S B where S is the area of the surface enclosed by the edge FIG. 4. (a,b) Conductance as a function of B- eld at E = channel in the facet zone. Actually, the oscillation peri- 75 meV computed for semiconducting (N = 324 [67]) and ods B ' 3.61 T and 1.84 T are obtained in the high eld metallic (N = 326) GNR systems, respectively, with  = z S 2 2 regime with the facet zones of  3000 nm and 6000 nm , 60 ,  = 30 , and W ' 40 nm. L ' 75 nm and 150 nm are B F respectively (see Fig.4a and additionally Figs.2a and 4b). studied in (a) while only L ' 75 nm in (b). Except for the Indeed, the formula (2) predicts quite well these results perfect edge in (b), di erent disordered con gurations with 2 2 the variation of ribbon width W modeled by a Gaussian auto- of B if S ' 2332 nm and 4640 nm are considered, correlation function [63], particularly, with the rsm W = rsm which are about 22.5 % smaller than the area of the cor- 0.6 nm and the correlation length  = 4.8 nm are considered. responding facet zones. Note that here, the edge states Oy axis (nm) Oy axis (nm) LDOS (a.u) 4 energy E and B- eld is presented. Basically, two typ- 150 F ical zones, E  E and > E , are speci ed where F 1 1 T F T E = minfE ; E g with the rst Landau levels E = 1 1 1 p p 2 F 2 2e~v Bj sin  j and E = 2e~v Bj sin(  )j B S B F 1 F [27] formed in the terrace and facet zones, respectively, 100 and v is the Fermi velocity in graphene. In particular, strong AB oscillations are predicted for E  E when F 1 only a single energy band is presented in both terrace and facet zones whereas the interference is blurred at higher energies. This can be explained by an inherent property of AB interference, similarly observed and demonstrated in nanoring systems [79, 80], that the strong oscillations can be observed when only a single energy band con- tributes to the transport, otherwise the e ect can be sig- ni cantly disturbed by the multi-bands contribution [63]. 0 5 10 15 20 25 30 35 Magnetic Field (T) FIG. 5. Conductance map with respect to B- eld and Fermi energy obtained in the system presented in Fig.2a phosphorene for  = 30 . The dashed line indicates the energy level T F E = minfE ; E g (see text). 1 1 di erence in the interference e ect in the stepped systems made of these perfect GNRs is predicted. In particular, even though the conductance oscillation is similarly ob- served in both cases, the e ect in the metallic systems (see Fig.4b for the perfect edge case) is relatively weaker than the one observed in the semiconducting GNRs (see Fig.2a for  = 30 ). Next, the e ects of edge disorder, which are practically inevitable and known to degrade strongly the transport properties of GNRs [68{78], have to be evaluated. In FIG. 6. Conductance as a function of B- eld obtained in a stepped phosphorene system with  = 60 and  = 30 . this work, disorder is modeled (see the details in [63]) S B W ' 40 nm, L ' 110 nm and the Fermi level is at 10 meV either by a Gaussian autocorrelation function (presented F above the bottom of conduction bands. in Fig.4) or by randomly removing the edge atoms. The edge disorder indeed degrades signi cantly the transport It is very worth noting that the interedge scatterings at low elds. However, as shown in Figs.4a-b and in the in the considered structures is achieved in the unipolar Supplementary Material [63], the AB oscillations at high regime. Therefore, di erent from the interference phe- elds are found to be much more robust under the ef- nomena reported in graphene p-n junctions, our predic- fect of the considered disorders than the zero- eld trans- tion can be achieved in both cases of semi-metallic and port. This can be explained by a fascinating feature that semiconducting materials. Indeed, a similar AB interfer- in the quantum Hall regime, the forward and backward ence picture is obtained in monolayer phosphorene sys- edge channels are spatially separated while the scatter- tems by tight-binding calculations [81] and is presented ings at these disordered edges do not allow electrons to in Fig. 6. Moreover, structural engineering for creat- transmit across the sample [63], thus not inducing the ing stepped structures can also allow for avoiding the strong backscatterings as at low elds. More interest- junction smoothness issues, that has been shown to of- ingly, the edge disorder even eliminates the di erence ten perturb dramatically similar quantum phenomena in between the metallic and semiconducting GNR systems graphene p-n junctions [86]. Finally, this predicted mech- discussed above (i.e., comparison of the results obtained anism for achieving AB interferences, in principle, can for perfect and disordered edges in Figs.2a and 4). Note be also applied to systems using ferromagnetic strips to that a picture, similar to the metallic armchair GNR case, create inhomogeneous B - elds [87], however, obtaining is also observed in zigzag GNR systems, i.e., the AB os- sharp junctions could be a practical challenge. cillation is almost invisible for perfect edges but much more pronounced in edge disordered ones [63]. To conclude, the magnetotransport through stepped In Fig.5, the conductance as a function of both Fermi graphene was investigated using atomistic tight-binding Fermi Energy (meV) Conductance (2e /h) 2 5 calculations. By applying the B- eld in a proper direc- [18] M. Kataoka, C. J. B. Ford, G. Faini, D. Mailly, M. Y. Simmons, and D. A. Ritchie, Phys. Rev. B 62, R4817- tion, opposite normal components of the eld can be R4820 (2000). created, thus inducing opposite edge states, in the ter- [19] H.-S. Sim, M. Kataoka, Hangmo Yi, N. Y. Hwang, M.-S. race and facet zones. The interedge scatterings were ob- Choi, and S.-R. Eric Yang, Phys. Rev. Lett. 91, 266801 served, leading to strong Aharonov-Bohm oscillations in (2003). the quantum Hall regime. The properties of this interfer- [20] S. Ihnatsenka, I. V. Zozoulenko, and G. Kirczenow, Phys. ence, depending on the carrier energy and structural pa- Rev. B 80, 115303 (2009). rameters, were systematically clari ed. Moreover, since [21] H.-S. Sim, M. Kataoka, and C.J.B. Ford, Phys. Rep. 456, 127-165 (2008). it is observed in the unipolar regime, our prediction can [22] B. Hackens, F. Martins, S. Faniel, C.A. Dutu, H. Sellier, be also achieved in stepped systems made of other (both S. Huant, M. Pala, L. Desplanque, X. Wallart, and V. semimetallic and semiconducting) 2D layered materials. Bayot, Nat. Commun. 1, 39 (2010). Acknowledgments - We acknowledge nancial sup- [23] N. Paradiso, S. Heun, S. Roddaro, G. Biasiol, L. Sorba, port from the F.R.S.-FNRS of Belgium through the re- D. Venturelli, F. Taddei, V. Giovannetti, and F. Beltram, Phys. Rev. B 86, 085326 (2012). search project (N T.1077.15), from the Flag-Era JTC [24] F. Martins, S. Faniel, B. Rosenow, M. G. Pala, H. Sellier, 2017 project "MECHANIC" (N R.50.07.18.F), from the S. Huant, L. Desplanque, X. Wallart, V. Bayot, and B. F ed eration Wallonie-Bruxelles through the ARC on 3D Hackens, New J. Phys. 15, 013049 (2013). nanoarchitecturing of 2D crystals (N 16/21-077) and [25] L. S. Sarah, S. Ady, R. Bernd, and I. H. Bertrand, Phys- from the European Union's Horizon 2020 research and ica E 76, 82-87 (2016). innovation program (N 696656). [26] A. H. Castro Neto, F. Guinea, N. M. R. Peres, K. S. Novoselov, and A. K. Geim, Rev. Mod. Phys. 81, 109- 162 (2009). [27] L.-J. Yin, K.-K. Bai, W.-X. Wang, S.-Y. Li, Y. Zhang, and L. He, Front. Phys. 12, 127208 (2017). [1] K. von Klitzing, Rev. Mod. Phys. 58, 519-531 (1986). [28] T. Machida, S. Morikawa, S. Masubuchi, R. Moriya, M. [2] David K. Ferry (IOP Publishing, 2015) p. 6.1 to 6.26. Arai, K. Watanabe, and T. Taniguchi, J. Phys. Soc. Jpn. 84, 121007 (2015). [3] B. W. Alphenaar, A. A. M. Staring, H. van Houten, M. A. A. Mabesoone, O. J. A. Buyk, and C. T. Foxon, Phys. [29] D. A. Abanin and L. S. Levitov, Science 317, 641-643 (2007). Rev. B 46, 7236-7239 (1992). [4] C. de C. Chamon, D. E. Freed, S. A. Kivelson, S. L. [30] J. R. Williams, L. DiCarlo, and C. M. Marcus, Science 317, 638-641 (2007). Sondhi, and X. G. Wen, Phys. Rev. B 55, 2331-2343 (1997). [31] B. Ozyilmaz, P. Jarillo-Herrero, D. Efetov, D. A. Abanin, [5] B. I. Halperin, A. Stern, I. Neder, and B. Rosenow, Phys. L. S. Levitov, and P. Kim, Phys. Rev. Lett. 99, 166804 Rev. B 83, 155440 (2011). (2007). [6] P. Bonderson, A. Kitaev, and K. Shtengel, Phys. Rev. [32] S. Morikawa, S. Masubuchi, R. Moriya, K. Watanabe, Lett. 96, 016803 (2006). T. Taniguchi, and T. Machida, Appl. Phys. Lett. 106, [7] P. Bonderson, K. Shtengel, and J. K. Slingerland, Phys. 183101 (2015). Rev. Lett. 97, 016401 (2006). [33] A. Mrenca-Kolasi  nsk  a, S. Heun, and B. Szafran, Phys. [8] A. Stern and B. I. Halperin, Phys. Rev. Lett. 96, 016802 Rev. B 93, 125411 (2016). (2006). [34] D. S. Wei, T. van der Sar, J. D. Sanchez-Yamagishi, [9] R. Ilan, E. Grosfeld, K. Schoutens, and A. Stern, Phys. K. Watanabe, T. Taniguchi, P. Jarillo-Herrero, B. I. Rev. B 79, 245305 (2009). Halperin, and A. Yacoby, Sci. Adv. 3, e1700600 (2017). [10] F. E. Camino, W. Zhou, and V. J. Goldman, Phys. Rev. [35] P. Makk, C. Handschin, E. T ov ari, K. Watanabe, T. B 76, 155305 (2007). Taniguchi, K. Richter, M.-H. Liu, and C. Sch onenberger, [11] F. E. Camino, W. Zhou, and V. J. Goldman, Phys. Rev. Phys. Rev. B 98, 035413 (2018). Lett. 98, 076805 (2007). [36] S. Jorg, R. Patrik, and T. Bjorn, Sol. State Commun. [12] Y. Zhang, D. T. McClure, E. M. Levenson-Falk, C. M. 152, 1411-1419 (2012). Marcus, L. N. Pfei er, and K. W. West, Phys. Rev. B [37] F. Giannazzo, I. Deretzis, G.Nicotra, G.Fisichella, 79, 241304 (2009). C.Spinella, F.Roccaforte, and A.La Magna, Appl. Surf. [13] D. T. McClure, Y. Zhang, B. Rosenow, E. M. Levenson- Sci. 291, 53-57 (2014). Falk, C. M. Marcus, L. N. Pfei er, and K. W. West, Phys. [38] J. Bao, O. Yasui, W. Norimatsu, K. Matsuda, and M. Rev. Lett. 103, 206806 (2009). Kusunoki, Appl. Phys. Lett. 109, 081602 (2016). [14] P. V. Lin, F. E. Camino, and V. J. Goldman, Phys. Rev. [39] F. Speck, M. Ostler, S. Besendrfer, J. Krone, M. Wanke, B 80, 125310 (2009). and T. Seyller, Ann. Phys. 529, 1700046 (2017). [15] N. Ofek, A. Bid, M. Heiblum, A. Stern, V. Umansky, and [40] A. St ohr et al., Ann. Phys. 529, 1700052 (2017). D. Mahalu, Proc. Natl. Acad. Sci. USA 107, 5276-5281 [41] S. Odaka, H. Miyazaki, S.-L. Li, A. Kanda, K. Morita, (2010). S. Tanaka, Y. Miyata, H. Kataura, K. Tsukagoshi, and [16] D. T. McClure, W. Chang, C. M. Marcus, L. N. Pfei er, Y. Aoyagi, Appl. Phys. Lett. 96, 062111 (2010). and K. W. West, Phys. Rev. Lett. 108, 256804 (2012). [42] H. Kuramochi, S. Odaka, K. Morita, S. Tanaka, H. [17] R. Sabo, I. Gurman, A. Rosenblatt, F. Lafont, D. Banitt, Miyazaki, M. V. Lee, S.-L. Li, H. Hiura, and K. Tsuk- J. Park, M. Heiblum, Y. Gefen, V. Umansky, and D. agoshi, AIP Advances 2, 012115 (2012). Mahalu, Nat. Phys. 13, 491-496 (2017). 6 [43] T. Schumann, K.-J. Friedland, M. H. Oliveira, A. [69] M. Y. Han, J. C. Brant, and P. Kim, Phys. Rev. Lett. Tahraoui, J. M. J. Lopes, and H. Riechert, Phys. Rev. 104, 056801 (2010). B 85, 235402 (2012). [70] M. V. Fischetti and S. Narayanan, J. Appl. Phys. 110, [44] E. Pallecchi, E. Lafont, V. Cavaliere, F. Schopfer, D. 083713 (2011). Mailly, W. Poirier, and A. Ouerghia, Sci. Rep. 4, 4558 [71] N. Djavid et al., IEEE Trans. Electron Devices 61, 23 - (2014). 29 (2014). [45] T. Ciuk, S. Cakmakyapan, E. Ozbay, P. Caban, K. [72] T. Misawa et al., Jpn. J. Appl. Phys. 54, 05EB01 (2015). Grodecki, A. Krajewska, I. Pasternak, J. Szmidt, and [73] T. Fang, A. Konar, H. Xing, and D. Jena, Phys. Rev. B W. Strupinski, J. Appl. Phys. 116, 123708 (2014). 78, 205403 (2008). [46] E. Akira, K. Fumio, M. Kouhei, K. Takashi, and T. [74] A. Y. Goharrizi et al., IEEE Trans. Electron Devices 58, Satoru, J. Low Temp. Phys. 179, 237-250 (2015). 3725 - 3735 (2011). [47] D. Momeni Pakdehi et al., ACS Appl. Mater. Interfaces [75] M. Evaldsson et al., Phys. Rev. B 78, 161407(R) (2008). 10, 6039-6045 (2018). [76] D. Querlioz et al., Appl. Phys. Lett. 92, 042108 (2008). [48] T. Takahiro and O. Toshio, Carbon 50, 674-679 (2012). [77] M. Poljak and T. Suligoj, IEEE Trans. Electron Devices [49] H. Kenjiro, S. Shintaro, and Y. Naoki, Nanotechnol. 24, 63, 537-543 (2016). 025603 (2012). [78] A. Cresti and S. Roche, New J. Phys. 11, 095004 (2009). [50] J.-K. Lee et al., Nano Lett. 13, 3494-3500 (2013). [79] V. Hung Nguyen, Y. M. Niquet, and P. Dollfus, Phys. [51] C.-H. Lee et al., Adv. Mater. 26, 2812-2817 (2014). Rev. B 88, 035408 (2013). [52] M. E. Ayhan, G. Kalita, M. Kondo, and M. Tanemura, [80] R. Zhang, Z. Wu, X. J. Li, and K. Chang, Phys. Rev. B RSC Adv. 4, 26866-26871 (2014). 95, 125418 (2017). [53] S. Dai et al., Nat. Nanotechnol. 10, 682-686 (2015). [81] A. N. Rudenko and M. I. Katsnelson, Phys. Rev. B 89, [54] K.-K. Bai, J.-B. Qiao, H. Jiang, H. Liu, and L. He, Phys. 201408 (2014). Rev. B 95, 201406 (2017). [82] Y. Xu, H. Gao, M. Li, Z. Guo, H. Chen, Z. Jin and B. [55] B. D. Briggs et al., Appl. Phys. Lett. 97, 223102 (2010). Yu, Nanotechnol. 22, 365202 (2011). [56] M. L. Leadbeater, C. L. Foden, J. H. Burroughes, M. [83] T. Low, V. Perebeinos, J. Terso , and Ph. Avouris, Phys. Pepper, T. M. Burke, L. L. Wang, M. P. Grimshaw, and Rev. Lett. 108, 096601 (2012). D. A. Ritchie, Phys. Rev. B 52, R8629-R8632 (1995). [84] http://www.openmx-square.org. [57] M. L. Leadbeater, C. L. Foden, J. H. Burroughes, T. M. [85] V. M. Pereira, A. H. Castro Neto, and N. M. R. Peres, Burke, L. L. Wang, M. P. Grimshaw, D. A. Ritchie, and Phys. Rev. B 80, 045401 (2009). M. Pepper, Surf. Sci. 361/362, 587-590 (1996). [86] V. V. Cheianov and V. I. Fal'ko, Phys. Rev. B 74, [58] I. S. Ibrahim, V. A. Schweigert, and F. M. Peeters, Phys. 041403(R) (2006). Rev. B 56, 7508 (1997). [87] A. Nogaret, J. Phys.: Condens. Matter 22, 253201 [59] S. Cin a, D. M. Whittaker, D. D. Arnone, T. Burke, H. P. (2010). Hughes, M. Leadbeater, M. Pepper, and D. A. Ritchie, Phys. Rev. Lett. 83, 4425-4428 (1999). [60] A. Nauen, U. Zeitler, R. J. Haug, A. G. M. Jansen, M. Dilger, and K. Eberl, Physica E 13, 732-735 (2002). [61] M. Grayson, D. Schuh, M. Bichler, M. Huber, G. Abstre- iter, L. Hoeppel, J. Smet, and K. von Klitzing, Physica E 22, 181-184 (2004). [62] J. A. Robinson, C. P. Puls, N. E. Staley, J. P. Stitt, M. A. Fanton, K. V. Emtsev, T. Seyller and Y. Liu, Nano Lett. 9, 964-968 (2009). [63] See Supplemental Material, which includes Refs. [27, 33, 55, 62, 64, 65, 68{80, 82{85]. In this material, computa- tional methodologies, e ects of local strains, edge disor- ders, and multi-bands contribution are presented in de- tail. [64] V. Hung Nguyen, J. Saint-Martin, D. Querlioz, F. Maz- zamuto, A. Bournel, Y.-M. Niquet, and P. Dollfus, J. Comput. Electron. 12, 85-93 (2013). [65] C. H. Lewenkopf and E. R. Mucciolo, J. Comput. Elec- tron. 12, 203-231 (2013). [66] Y. Aharonov and D. Bohm, Phys. Rev. 115, 485-491 (1959). [67] Armchair GNRs with perfect edges can be basically di- vided in two main classes, depending on the number of zigzag lines N across their width, i.e., GNRs with N = 3p + 2 are metallic while they are semiconducting in the other cases. The details can be found in Y.-W. Son et al., Phys. Rev. Lett. 97, 216803 (2016) or in S. M.-M. Dubois et al., Eur. Phys. J. B 72, 1-24 (2009). [68] M. Y. Han, B. zyilmaz, Y. Zhang, and P. Kim, Phys. Rev. Lett. 98, 206805 (2007). 7 Supplemental material for V. Hung Nguyen and J.-C. Charlier Institute of Condensed Matter and Nanosciences, Université catholique de Louvain, Chemin des étoiles 8, B-1348 Louvain-la-Neuve, Belgium Contents: 1. Computational methodologies 2. Curvature-induced local strains in stepped graphene Local strains induced by curvature and simulated models are discussed. 3. Tight Binding versus Density Functional Theory calculations Validity of tight binding model to investigate the local strains is considered by the fit to the density functional theory calculations 4. Effects of curvature-induced local strains on Aharonov-Bohm interference 5. Effects of edge disorder in graphene nanoribbons 6. Multi sub-bands contribution References 1. Computational methodologies In order to compute the electronic transport through stepped graphene systems, we employed the Green’s function technique [1] to solve the tight-binding Hamiltonian presented in the main text. In particular, the retarded Green’s function is determined as −1 R +. G ( E)= E+i 0 − H −Σ −Σ (S.1) [ ] D L R where H is the device Hamiltonian and  are self-energies describing the left and right device-to- D L,R lead couplings, respectively. This equation was solved using the recursive method [2]. The transport quantities such as transmission probability T(E), conductance G(E ) and local density of left- and right- injected states D (E,r) are then computed using the Landauer formalism as follows: L,R R R † T(E) = (S.2) Tr [Γ G Γ G ] L R +∞ ∂ f 2 e G(E ) = d E T ( E) − (S.3) F ∫ ( ) h −∞ ∂ E R R † G Γ G L , R D (E,r) = (S.4) L,R 2π Here, Γ =i Σ −Σ and f (E) is the Fermi-Dirac distribution function with the Fermi level E . [ ] F F L , R L , R L , R The total local density of states can be computed either by D(E,r) = D (E,r)+D (E,r) or L R D( E , r)=−ℑ( G )/π . 8 2. Curvature-induced local strains in stepped graphene In general, the non-planar geometry can induce in-plane strain inhomogeneities (i.e., local strains). Different from the uniform strains, local strains can result in electron scatterings and can be effectively described as the effects of strain-induced gauge fields in stepped graphene systems. These local strains have been shown to be often observed in the detached regions, i.e., around the step edges [3-5]. However, it has been experimentally demonstrated that if no external stress is applied, only small local strains can be observed in the stepped graphene systems, because of the mechanical robustness of graphene layers. For instance, it has been investigated and reported in refs. [3,4] that graphene at the step could experience small uniaxial strains (i.e., < 1%) relative to the rest of the sheet and nearly strain-free graphene is possible in epitaxial graphene. Extremely small strains of ~ 0.025% were also demonstrated in ref. [5] by Raman spectroscopy measurements. To investigate the effects of such local strains induced by graphene curvature, we assume a simple model that the strain is maximum at the step edges and gradually release in two sides. Here, we use the following simple formula to model these strains: max ε(r)= (S.5) 1+(d / d ) where  is the maximum value of strain, d is the in-plane distance from the position r to the max considered step edge, and d characterizes the strain release distance. Fig. S1 shows a typical picture of possible strain gradients induced by graphene curvature, that can be modelled by the above formula. Fig. S1: Possible local strains induced by curvature in stepped graphene systems. In next sections 3 and 4, using such the simple model we present an investigation to clarify the effects of possible local strains mentioned above on the Aharonov-Bohm interference predicted in this work. 9 3. Tight Binding versus Density Functional Theory calculations In this section, we present some calculations demonstrating the validity of the tight binding (TB) Hamiltonian to investigate the curvature-induced local strains in stepped graphene systems by the fit to the ab initio quantum transport data [6]. Actually, it has been shown by several studies in the literature that a simple p TB Hamiltonian with only nearest neighbor interactions can be used to compute very accurately the electronic properties and electron transport in planar graphene systems. However, the curvature in the stepped graphene systems can alter their electronic properties, i.e., can induce electron scatterings, especially, when a local strain occur around the step edges as discussed above. Hence, the validity of the p TB model is needed to be examined. To this aim, we performed quantum transport calculations based on density functional theory (DFT) [6] for stepped graphene systems, which are assumed very large so as to neglect the finite-width effects and hence the periodic boundary condition can be applied along Oy axis. The transmission coefficients through stepped graphene systems obtained for three different k - momentum modes around the Dirac point are computed and presented in Fig. S2. Two cases without and with a local strain ( = 1%) are considered. In agreement with the study in [7] with different max curvature radius, step heights and step angles, the results obtained without the local in-plane strain show that the effects of curvature on the electronic transport through the system is negligible. In particular, for k = K , the transmission coefficient exhibits only negligibly small reduction around the y 0 zero energy point, compared to the planar case where it is unity and constant in the considered energy range. Thus, the curvature does not induce significant electron scatterings and hence, similar to the planar cases, the p TB Hamiltonian still works well for these considered stepped graphene systems. Fig. S2: Electronic transport through stepped 2D graphene systems using DFT calculations: strained (with  = 1%) system as described in Fig.S1 versus unstrained one. max Fig. S3: Electronic transport through stepped 2D graphene system with local strain  = 1%: max Tight-Binding vs DFT calculations. When the curvature-induced local strains are introduced, significant electron scatterings at the step edges can be observed. Indeed, the transmission coefficient exhibits significant reduction around the zero energy point for k = K mode and close to the edges of energy gap for other modes K . In order y 0 1,2 to compute these effects, the p TB Hamiltonian must be adjusted. In particular, a model where the hopping term is determined a function of C-C bond length r as t =t exp β(1−r / r ) with  =3.37 ij [ ] ij 0 ij 0 and r = 0.142 nm has been demonstrated [8] to compute well the strain effects in graphene. Fig. S3 demonstrates a quite good agreement between the DFT data and results obtained by such strained TB model. Our calculations show that this strained TB Hamiltonian without any other adjustment is still a good model for the considered stepped graphene systems with local strains of  ≲ 4%. max 4. Effects of curvature-induced local strains on Aharonov-Bohm interference In this section, we employed the strained TB model presented above to investigate the effects of curvature-induced local strains in the Aharonov-Bohm interferometers predicted in this work. The conductance as a function of Fermi energy obtained at zero magnetic field is presented in Fig. S4. Similar to the results presented in section 3, the considered local strain can induce electron scatterings and affect significantly the transport through the system, leading to the conductance reduction. The effects of such strains on the predicted Aharonov-Bohm interference are investigated and presented in Fig. S5. It is however shown that the effects on the conductance oscillations in the quantum Hall regime are relatively weaker than those observed in the low field one. This can be explained as follows. In the low field regime, the electron transport is essentially due to “bulk states”, which transmit across the step edges (i.e., zones of local strains) and hence undergo strong back-scatterings as described in the top-right image of Fig.S5. In the quantum Hall regime, the presence of these locally strained zones however affects the transport picture differently. In particular, when a high magnetic field with  <  B S is applied, opposite edge states are formed in terrace and facet zones (in two sides of the step edges) and electrons transmitting through the system have to follow trajectories as described in the bottom- 11 right image of Fig.S5, i.e., when reaching the step edges, electrons transmit along (not directly across, as in the zero-field case) these edges. This can eliminate the scatterings induced by the considered local strains and explain their weak effects in the quantum Hall regime, compared to those in the low field one. Hence, within the range of experimentally reported strains discussed above, the considered local strains even though alter but does not strongly perturb the predicted Aharonov-Bohm interference. Fig. S4: Conductance as a function of energy at zero magnetic field with different curvature-induced strains. The strain release distance d ≈ 11 nm while other parameters are as in Fig.2 of the main text. Fig. S5: Conductance as a function of magnetic field at E = 75 meV and with different curvature-induced strains. Parameters are as in Fig.S4. Right panels illustrate the electron transport pictures at low and high magnetic fields. 12 Certainly, the effects of these local strains on the Aharonov-Bohm interference can be enlarged and significant when much larger strain gradients occur, for instance, when external stresses are additionally applied. 5. Effects of edge disorder in graphene nanoribbons Similar to the surface roughness in many nanoscale systems of conventional semiconductors, the edge disorder (i.e., edge roughness) is often a practical issue for graphene ribbons [9-19]. The edge disorder has been shown to affect strongly the electronic transport through graphene ribbons, especially, when their width reaches the nanoscale regime. Fig. S6: Graphene ribbon with edge disorder. To investigate the effects of edge disorder, there were two models widely used in the literature. In particular, the disordered edges can either be generated by randomly removing the edge atoms with a certain probability [11-14] or are modeled using auto-correlation functions [15-19]. These models have been demonstrated to interpret well the electronic properties of GNRs in experiments with a wide range of disorder level [9-19]. Here, the effects of edge disorder are examined using both two models. In the latter case, the edge disorder is generated by a Gaussian autocorrelation function, particularly, the variation of ribbon width  W is effectively described as 2 Δ x ⟨δ W ( x)δ W ( x+Δ x)⟩=W exp − (S.6) rms ( ) 2ξ where W is the rms value charactering the disorder strength and  represents the correlation length. rms A typical picture of edge disordered graphene ribbons is illustrated in Fig. S6. The effects of edge disorder generated by a Gaussian autocorrelation function are presented in Fig. 4 of the main text and the results obtained in the systems by randomly removing edge atoms are displayed in Fig. S7. Actually, two main features are found with both disorder models. First, in the case of semiconducting GNRs, the Aharonov-Bohm oscillations obtained at high magnetic fields (B-fields) are shown to be much more robust under the effect of the considered disorders than the zero-field transport. This can be explained by a fascinating feature that different from the zero-field case (see Figs.S8 (a,d)), the forward and backward edge channels are spatially separated in the quantum Hall regime (see Figs.S8 (b-c, e-f)) while the scatterings at the disordered edges do not allow electrons 13 to transmit across the sample. As a consequence, the edge disorder effects do not contribute significantly to backscatterings in the considered stepped systems, i.e., the dominant mechanism of backscattering is still interaction between edge states in zones of opposite normal B-fields and hence strong Aharonov-Bohm oscillations are still achieved. Fig. S7: (a,b) Conductance as a function of B-field at E = 75 meV computed for semiconducting (N = F z 324) and metallic (N = 326) GNR systems, respectively, with θ = 60°, θ =30°, L  nm and z S B F W  40nm. In the edge disordered systems, 15% of edge atoms are randomly removed. Second, as presented in Fig.4 of the main text and in Fig. S7, there is a significant difference between semiconducting and metallic GNR systems with perfect edges, i.e., the Aharonov-Bohm oscillation is relatively weak in the metallic cases, compared to the results obtained in semiconducting ones. The 14 edge disorder however strengthens the Aharonov-Bohm oscillation in metallic GNR systems, thus eliminating the difference mentioned. Fig. S8: Edge disorder effects on electron propagation in planar graphene ribbons: left-injected LDOS (a,b,d,e) and total LDOS (c,f). Magnetic fields B = 0 in (a,d) and 25 T in (b,c,e,f) while carrier energy E = 75 meV. Perfect and disordered edges are considered in (a,b,c) and (d,e,f), respectively. Fig. S9: Conductance as a function of B-field at E = 75 meV obtained in stepped systems of a zigzag GNR with θ = 60°, θ =30°, L  72 nm and W  40 nm. The disordered edge systems are generated S B F by the Gaussian autocorrelation function (S.6) with W  0.6 nm and  nm. rms 15 Note additionally that a picture, similar to those obtained in the metallic armchair GNRs, is also observed in zigzag GNR systems, i.e., the Aharonov-Bohm oscillation is relatively weak for perfect edges but the effect is much more pronounced when edge disorder is introduced (see Fig. S9). 6. Multi sub-bands contribution In this section, we analyze the effects of multi sub-bands contribution on the predicted Aharonov-Bohm oscillation in more detail. As it has been demonstrated in nanorings [20,21] and also in graphene p-n junctions [22], the Aharonov-Bohm interference has an inherent property that the strong oscillation of conductance can be observed in the low energy regime where only a single band is obtained. In the regime of high energies when multi energy bands can contribute to the transport, the interference picture can be significantly blurred. A similar feature is also observed in the Aharonov-Bohm interferometers considered here. Fig. S10: Conductance as a function of Fermi energy at B = 20 T obtained in stepped systems of an armchair GNR with L  110 nm and W   0 nm. The stepped angle  = 60° while three directions of F S T(F) B-field are considered. E indicates the first Landau level formed in the terrace (facet) zone. 1 16 Indeed, as shown in Fig. S10 and Fig. 5 of the main text, strong conductance oscillations in all T F T ,F considered cases are observed in the regime where are the first Landau E ≤ min( E , E ) E F 1 1 1 levels formed in the terrace and facet zones, respectively, at high magnetic fields . In Fig. S11, the energy bands of an armchair GNR under the effect of magnetic field are presented. When a high magnetic field is applied, the energy bands of GNR are strongly modified and the Landau quantization is observed as seen in the right panel of Fig.S11. This Landau quantization (i.e., energy levels E ) 0, 1, 2,...   obeys the well-known formula [23] E = sign(n) 2 e ℏ v |n|B (S.7) n F ⟂ T F where v is the Fermi velocity in graphene. The energy levels E and E mentioned above are hence F 1 1 T 2 F 2 determined as E = 2 eℏ v B|sin θ | and E = 2 e ℏ v B|sin (θ −θ )|. √ √ 1 F B 1 F S B B = 0.0 T B = 6.5 T B = 13 T -1 -2 Momentum k Momentum k Momentum k Fig. S11: Energy bands of the armchair GNR of W   0 nm under the effect of magnetic field. Energy levels E (n=0, 1, 2,...) in the right panel represent the Landau quantization at a high B-field [23]. In the high energy regime, the effects of multi sub-bands contribution, similar to those observed in graphene and phosphorene nanorings [20,21] and mentioned above, are clearly demonstrated in Fig.S10 (see the zones highlighted in yellow). In general, the Aharonov-Bohm oscillations are T F significantly blurred by such contribution when E >min ( E , E ), as illustrated in both Fig. S10 here F 1 1 T F and Fig.5 of the main text. In the cases if E  E (e.g.,  = 15° and 45°), some oscillations can still 1 1 B be well defined in the range between these two values (see Fig. S10 for  = 15°). The features presented here explain clearly the results presented and discussed in Fig.5 of the main text. Energy (eV) 17 References: [1] V. Hung Nguyen et al., J. Comput. Electron. 12, 85-93 (2013). [2] C. H. Lewenkopf and E. R. Mucciolo, J. Comput. Electron. 12, 203-231 (2013). [3] J. A. Robinson et al., Nano Lett. 9, 964-968 (2009). [4] T. Low, V. Perebeinos, J. Tersoff, and Ph. Avouris, Phys. Rev. Lett. 108, 096601 (2012). [5] B. D. Briggs et al., Appl. Phys. Lett. 97, 223102 (2010). [6] http://www.openmx-square.org [7] Y. Xu, H. Gao, M. Li, Z. Guo, H. Chen, Z. Jin and B. Yu, Nanotechnol. 22, 365202 (2011). [8] V. M. Pereira, A. H. Castro Neto, and N. M. R. Peres, Phys. Rev. B 80, 045401 (2009). [9] M. Y. Han, B. Özyilmaz, Y. Zhang, and P. Kim, Phys. Rev. Lett. 98, 206805 (2007). [10] M. Y. Han, J. C. Brant, and P. Kim, Phys. Rev. Lett. 104, 056801 (2010). [11] M. Evaldsson et al., Phys. Rev. B 78, 161407(R) (2008). [12] D. Querlioz et al., Appl. Phys. Lett. 92, 042108 (2008). [13] A. Cresti and S. Roche, New J. Phys. 11, 095004 (2009). [14] M. Poljak and T. Suligoj, IEEE Trans. Electron Devices 63, 537-543 (2016). [15] M. V. Fischetti and S. Narayanan, J. Appl. Phys. 110, 083713 (2011). [16] N. Djavid et al., IEEE Trans. Electron Devices 61, 23 - 29 (2014). [17] T. Misawa et al., Jpn. J. Appl. Phys. 54, 05EB01 (2015). [18] T. Fang, A. Konar, H. Xing, and D. Jena, Phys. Rev. B 78, 205403 (2008). [19] A. Y. Goharrizi et al., IEEE Trans. Electron Devices 58, 3725 - 3735 (2011). [20] V. Hung Nguyen, Y. M. Niquet, and P. Dollfus, Phys. Rev. B 88, 035408 (2013). [21] R. Zhang, Z. Wu, X. J. Li, and K. Chang, Phys. Rev. B 95, 125418 (2017). [22] A. Mreńca-Kolasińska, S. Heun, and B. Szafran, Phys. Rev. B 93, 125411 (2016). [23] L.-J. Yin, K.-K. Bai, W.-X. Wang, S.-Y. Li, Y. Zhang, and L. He, Front. Phys. 12, 127208 (2017). http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Condensed Matter arXiv (Cornell University)

Stepped Graphene-based Aharonov-Bohm Interferometers

Condensed Matter , Volume 2018 (1812) – Dec 6, 2018

Loading next page...
 
/lp/arxiv-cornell-university/stepped-graphene-based-aharonov-bohm-interferometers-PUz3zx6GSe
ISSN
2053-1583
eISSN
ARCH-3331
DOI
10.1088/2053-1583/ab3500
Publisher site
See Article on Publisher Site

Abstract

V. Hung Nguyen and J.-C. Charlier Institute of Condensed Matter and Nanosciences, Universit e catholique de Louvain, Chemin des  etoiles 8, B-1348 Louvain-la-Neuve, Belgium Aharonov-Bohm interferences in the quantum Hall regime are observed when electrons are trans- mitted between two edge channels. Such a phenomenon has been realized in 2D systems such as quantum point contacts, anti-dots and p-n junctions. Based on a theoretical investigation of the magnetotransport in stepped graphene, a new kind of Aharonov-Bohm interferometers is proposed herewith. Indeed, when a strong magnetic eld is applied in a proper direction, oppositely propagat- ing edge states can be achieved in both terrace and facet zones of the step, leading to the interedge scatterings and hence strong Aharonov-Bohm oscillations in the conductance in the quantum Hall regime. Taking place in the unipolar regime, this interference is also predicted in stepped systems of other 2D layered materials. The fascinating properties of quantum hall devices graphene p-n junctions can also work as AB interferom- arise from their ideal 1D edge states formed in a 2D eters [32{35] in the quantum Hall regime. In particular, electron system when a high magnetic eld is applied the oppositely propagating edge states are formed in two [1, 2]. These edge states are particularly attractive due di erent doped zones and their interaction at the p-n to their large coherence lengths, which is mandatory for interface acquires conductance oscillations of the AB pe- constructing electron interferometers. However, since the riodicity at high B- elds. edge channels are spatially separated, a mechanism for Motivated by such scienti c context, a new kind creating the electron transmission between them is re- of Aharonov-Bohm interferometers based on stepped quired to achieve the interference e ects. In this regard, graphene channels is proposed herewith. These non- one explored technique consists in building constrictions planar systems have been actually achieved in several (quantum point contacts) in a sample, where the in- experimental situations. For example, the step bunch- teredge tunneling paths can occur [3{17]. Setups consist- ing on the SiC surface is often observed in epitaxial ing of a pair of quantum point contacts with an internal graphene growth by thermal decomposition of SiC [37{ cavity has been demonstrated to work well as quantum Hall, electronic Fabry-P erot, and Aharonov-Bohm inter- ferometers. Another mechanism has also been suggested in systems consisting of an antidot introduced between y-axis (a) their edges [18{25]. Electronic currents encircling the antidot can be achieved and a similar Aharonov-Bohm (AB) interference is hence observed. Graphene, a truly 2D material, is an ideal platform z-axis for investigating quantum Hall and interference e ects. x-axis Remarkably, owning to an unique linear dispersion and normal B Dirac-like fermions [26], Landau levels and a half-integer quantum Hall e ect with an unusual quantized sequence compared to the conventional systems have been ob- (b) served in graphene when a strong magnetic eld (B- eld) is applied [27, 28]. In addition, with its semimetal charac- normal B ter, quantum Hall systems of graphene can work in both the unipolar and bipolar regimes that can be generated normal B and controlled by gate voltages [28{35]. Interestingly, in applied B the bipolar regime the chiral edge states equilibration and interedge scatterings at the p-n interfaces in graphene have been observed, resulting in fractional conductance plateaus [31]. FIG. 1. Stepped graphene (a) with the ribbon width W and With its typically high carrier mobilities, graphene is the length of facet zone L . (b) Schematic of the side view also an ideal material to perform the investigation on illustrating the applied magnetic eld (red dashed lines of interference e ects, including the Aharonov-Bohm one. arrows) and its normal components (green and blue arrows, respectively) in both terrace and facet zones.  and  are Several experimental and theoretical observations of the B S the angles of the eld and of the facet zone relative to the AB e ect in graphene nanorings have been reported terrace one (Ox axis), respectively. (i.e., see Ref. [36] and references therein). Remarkably, arXiv:1812.02845v2 [cond-mat.mes-hall] 20 Jun 2019 2 47], a promising method for the production of large-area same direction are obtained, thus inducing the same high-quality graphene. These stepped graphene channels propagating edge states in the two zones as illustrated with terrace size of several m and step height of tens in Fig.2b. Therefore, when a large B- eld is applied, a nm can be controllably produced by varying the heating conventional Landau quantization is still obtained, i.e., rate [38]. Non-planar graphene systems have been also the conductance represents quantized values as in Fig.2a synthesized in an even better controllable way by draping for  = 90 . For    , the B -component in the B B S N graphene on pre-structured substrates [48{55]. facet zone is canceled and hence the B- eld has no e ect on the in-plane transport in this zone. Consequently, the In this work, a B- eld is found to induce di erent ef- system behaves as a heterojunction consisting of nite- fects on the electron motion in the terrace and facet zones and zero-magnetic eld zones and the scatterings at their of stepped graphene, essentially resulting from the cre- interface basically explain the reduction of conductance ation of di erent normal components of the eld in these obtained for  = 60   displayed in Fig.2a. zones. This feature has been also observed in several B S non-planar systems of graphene [41{47] and 2DEG [56{ Most interestingly, when 0 <  <  , two opposite B S 61], leading to anisotropic pictures when the transport B -components alternate in the terrace and facet zones, takes place in the directions aligned parallel and per- as discussed above. Similarly to the e ects of B- eld in pendicular to the step edge. Here, a novel phenomenon di erent doped zones of graphene p-n junctions [31{33], is predicted when tuning the direction of B- eld ap- opposite edge states in the terrace and facet zones are plied to stepped graphene. In particular, an inhomo- geneous pro le containing alternatively opposite normal B-components along the channel can be created, thus in- (a) ducing accordingly opposite edge states. The interaction between these edge states nally results in strong AB os- cillations in the quantum Hall regime as presented in this  = 90° >  B S article.  = 60°   B S The considered systems consist in graphene nanorib-  = 30° <  B S bons (GNRs) in the step geometry as illustrated in Fig.1. In general, in-plane local strains can occur in the bent zones of the step, however, have been demonstrated to be small (i.e., < 1%) and negligible even in epitaxial graphene [55, 62]. Moreover, such small local strains are shown not to strongly a ect the predicted AB interfer- ence picture [63]. Therefore, these local strains are ne- glected and the p tight-binding Hamiltonian [26] is em- ployed to compute the magneto-transport in this work. In particular, when a B- eld is applied, (b)  >  B S X X y i y nm H = U c c + t e c c (1) n n 0 m n n hn;mi    th where U represents the potential energy at the n site, t = 2:7 eV corresponds the nearest-neighbor hopping e m energies, and  = A(r)dr is the Peierls phase (c) nm ~ r  <  describing the e ects of the B - eld. Here, the magnetic B S eld B = B(cos  ; 0; sin  ) is considered by introduc- B B ing the vector potential A(r) = B(y sin  ; z cos  ; 0).   B B The above Hamiltonian is solved using the Green's- function technique [63{65], allowing for the calculation of the transport quantities perpendicular to the step edges, i.e., along the Ox axis shown in Fig.1. FIG. 2. (a) Conductance as a function of B- eld in a stepped Fig.2a displays the dependence of conductance on B- armchair GNR for di erent angles  and with  = 60 , B S eld applied in di erent directions in a step constituted E = 75 meV , L ' 75 nm, and W ' 40 nm (i.e., number F F by an armchair GNR. As mentioned, the applied B- eld of dimer lines N = 324, a semiconducting GNR). (b,c) Dia- induces di erent normal components (B ) in the terrace grams illustrating the interedge scatterings for  >  and N B S <  , when the normal B-components in the facet and and facet zones, i.e., B = B sin  (green arrows) and B S N B terrace zones (see Fig.1b) are pointing out in the same and B sin(  ) (blue arrow), respectively (see Fig.1b). S B opposite directions, respectively. First, for  >  , B -components pointing out in the B S N 3 are formed inside the facet zone (see Fig.3) and hence the value of S in Eq.(2) to estimate B is basically pro- portional to but smaller than the area of the facet zone. Thus, the origin of observed conductance oscillation is re- ally the AB interference due to the interaction between edge states in both terrace and facet zones. The observed AB interference is also found to be sen- sitive to some other structural parameters. First, per- (a) fect armchair GNRs can be divided in two main classes with di erent electronic properties, either quasi-metallic with negligible bandgap or semiconducting, depending on high the number of zigzag lines N across the ribbon width: N = 3p + 2 and N 6= 3p + 2 [67], respectively. More- z z over, in contrast to semiconducting ribbons, the rst sub- band of metallic GNRs is linear, thus inducing massless fermions at low energies that contribute mainly to the transport at high B- elds. As a consequence, a signi cant zero (b) Ox axis (nm) (a) semiconducting GNRs L  75.0 nm FIG. 3. Left-injected local density of states at B = 25.8 T L  75.0 nm (a) and 27.4 T (b), corresponding to conductance peak and L  150.0 nm valley, respectively (see Fig. 2a for  = 30 ). created, thus inducing the strong interedge scatterings at their interface as illustrated in Fig.2c. As an important consequence, the conductance as a function of B- eld represents a strong AB oscillation in the quantum Hall regime (see the case of  = 30 in Fig.2a). This result is essentially due to the interedge backscatterings diagram- matically described in Fig.2c and is further demonstrated by analyzing the computed left-injected local density of (b) metallic GNRs perfect edge states in Fig.3, that illustrates the left-to-right electron- disorder 1 wave propagation. Indeed, backscatterings are almost disorder 2 absent (Fig.3a) when the phase coherence condition is disorder 3 satis ed, leading to conductance peaks. In the phase incoherence condition, strong interedge backscatterings (Fig.3b) and hence a low conductance are achieved. AB oscillation period observed in quantum rings with area S is known to be given by B = h=eS [66]. To examine this property in the considered systems (for <  ), the above formula should be rewritten as B S h 1 B = (2) eS jsin(  )j S B where S is the area of the surface enclosed by the edge FIG. 4. (a,b) Conductance as a function of B- eld at E = channel in the facet zone. Actually, the oscillation peri- 75 meV computed for semiconducting (N = 324 [67]) and ods B ' 3.61 T and 1.84 T are obtained in the high eld metallic (N = 326) GNR systems, respectively, with  = z S 2 2 regime with the facet zones of  3000 nm and 6000 nm , 60 ,  = 30 , and W ' 40 nm. L ' 75 nm and 150 nm are B F respectively (see Fig.4a and additionally Figs.2a and 4b). studied in (a) while only L ' 75 nm in (b). Except for the Indeed, the formula (2) predicts quite well these results perfect edge in (b), di erent disordered con gurations with 2 2 the variation of ribbon width W modeled by a Gaussian auto- of B if S ' 2332 nm and 4640 nm are considered, correlation function [63], particularly, with the rsm W = rsm which are about 22.5 % smaller than the area of the cor- 0.6 nm and the correlation length  = 4.8 nm are considered. responding facet zones. Note that here, the edge states Oy axis (nm) Oy axis (nm) LDOS (a.u) 4 energy E and B- eld is presented. Basically, two typ- 150 F ical zones, E  E and > E , are speci ed where F 1 1 T F T E = minfE ; E g with the rst Landau levels E = 1 1 1 p p 2 F 2 2e~v Bj sin  j and E = 2e~v Bj sin(  )j B S B F 1 F [27] formed in the terrace and facet zones, respectively, 100 and v is the Fermi velocity in graphene. In particular, strong AB oscillations are predicted for E  E when F 1 only a single energy band is presented in both terrace and facet zones whereas the interference is blurred at higher energies. This can be explained by an inherent property of AB interference, similarly observed and demonstrated in nanoring systems [79, 80], that the strong oscillations can be observed when only a single energy band con- tributes to the transport, otherwise the e ect can be sig- ni cantly disturbed by the multi-bands contribution [63]. 0 5 10 15 20 25 30 35 Magnetic Field (T) FIG. 5. Conductance map with respect to B- eld and Fermi energy obtained in the system presented in Fig.2a phosphorene for  = 30 . The dashed line indicates the energy level T F E = minfE ; E g (see text). 1 1 di erence in the interference e ect in the stepped systems made of these perfect GNRs is predicted. In particular, even though the conductance oscillation is similarly ob- served in both cases, the e ect in the metallic systems (see Fig.4b for the perfect edge case) is relatively weaker than the one observed in the semiconducting GNRs (see Fig.2a for  = 30 ). Next, the e ects of edge disorder, which are practically inevitable and known to degrade strongly the transport properties of GNRs [68{78], have to be evaluated. In FIG. 6. Conductance as a function of B- eld obtained in a stepped phosphorene system with  = 60 and  = 30 . this work, disorder is modeled (see the details in [63]) S B W ' 40 nm, L ' 110 nm and the Fermi level is at 10 meV either by a Gaussian autocorrelation function (presented F above the bottom of conduction bands. in Fig.4) or by randomly removing the edge atoms. The edge disorder indeed degrades signi cantly the transport It is very worth noting that the interedge scatterings at low elds. However, as shown in Figs.4a-b and in the in the considered structures is achieved in the unipolar Supplementary Material [63], the AB oscillations at high regime. Therefore, di erent from the interference phe- elds are found to be much more robust under the ef- nomena reported in graphene p-n junctions, our predic- fect of the considered disorders than the zero- eld trans- tion can be achieved in both cases of semi-metallic and port. This can be explained by a fascinating feature that semiconducting materials. Indeed, a similar AB interfer- in the quantum Hall regime, the forward and backward ence picture is obtained in monolayer phosphorene sys- edge channels are spatially separated while the scatter- tems by tight-binding calculations [81] and is presented ings at these disordered edges do not allow electrons to in Fig. 6. Moreover, structural engineering for creat- transmit across the sample [63], thus not inducing the ing stepped structures can also allow for avoiding the strong backscatterings as at low elds. More interest- junction smoothness issues, that has been shown to of- ingly, the edge disorder even eliminates the di erence ten perturb dramatically similar quantum phenomena in between the metallic and semiconducting GNR systems graphene p-n junctions [86]. Finally, this predicted mech- discussed above (i.e., comparison of the results obtained anism for achieving AB interferences, in principle, can for perfect and disordered edges in Figs.2a and 4). Note be also applied to systems using ferromagnetic strips to that a picture, similar to the metallic armchair GNR case, create inhomogeneous B - elds [87], however, obtaining is also observed in zigzag GNR systems, i.e., the AB os- sharp junctions could be a practical challenge. cillation is almost invisible for perfect edges but much more pronounced in edge disordered ones [63]. To conclude, the magnetotransport through stepped In Fig.5, the conductance as a function of both Fermi graphene was investigated using atomistic tight-binding Fermi Energy (meV) Conductance (2e /h) 2 5 calculations. By applying the B- eld in a proper direc- [18] M. Kataoka, C. J. B. Ford, G. Faini, D. Mailly, M. Y. Simmons, and D. A. Ritchie, Phys. Rev. B 62, R4817- tion, opposite normal components of the eld can be R4820 (2000). created, thus inducing opposite edge states, in the ter- [19] H.-S. Sim, M. Kataoka, Hangmo Yi, N. Y. Hwang, M.-S. race and facet zones. The interedge scatterings were ob- Choi, and S.-R. Eric Yang, Phys. Rev. Lett. 91, 266801 served, leading to strong Aharonov-Bohm oscillations in (2003). the quantum Hall regime. The properties of this interfer- [20] S. Ihnatsenka, I. V. Zozoulenko, and G. Kirczenow, Phys. ence, depending on the carrier energy and structural pa- Rev. B 80, 115303 (2009). rameters, were systematically clari ed. Moreover, since [21] H.-S. Sim, M. Kataoka, and C.J.B. Ford, Phys. Rep. 456, 127-165 (2008). it is observed in the unipolar regime, our prediction can [22] B. Hackens, F. Martins, S. Faniel, C.A. Dutu, H. Sellier, be also achieved in stepped systems made of other (both S. Huant, M. Pala, L. Desplanque, X. Wallart, and V. semimetallic and semiconducting) 2D layered materials. Bayot, Nat. Commun. 1, 39 (2010). Acknowledgments - We acknowledge nancial sup- [23] N. Paradiso, S. Heun, S. Roddaro, G. Biasiol, L. Sorba, port from the F.R.S.-FNRS of Belgium through the re- D. Venturelli, F. Taddei, V. Giovannetti, and F. Beltram, Phys. Rev. B 86, 085326 (2012). search project (N T.1077.15), from the Flag-Era JTC [24] F. Martins, S. Faniel, B. Rosenow, M. G. Pala, H. Sellier, 2017 project "MECHANIC" (N R.50.07.18.F), from the S. Huant, L. Desplanque, X. Wallart, V. Bayot, and B. F ed eration Wallonie-Bruxelles through the ARC on 3D Hackens, New J. Phys. 15, 013049 (2013). nanoarchitecturing of 2D crystals (N 16/21-077) and [25] L. S. Sarah, S. Ady, R. Bernd, and I. H. Bertrand, Phys- from the European Union's Horizon 2020 research and ica E 76, 82-87 (2016). innovation program (N 696656). [26] A. H. Castro Neto, F. Guinea, N. M. R. Peres, K. S. Novoselov, and A. K. Geim, Rev. Mod. Phys. 81, 109- 162 (2009). [27] L.-J. Yin, K.-K. Bai, W.-X. Wang, S.-Y. Li, Y. Zhang, and L. He, Front. Phys. 12, 127208 (2017). [1] K. von Klitzing, Rev. Mod. Phys. 58, 519-531 (1986). [28] T. Machida, S. Morikawa, S. Masubuchi, R. Moriya, M. [2] David K. Ferry (IOP Publishing, 2015) p. 6.1 to 6.26. Arai, K. Watanabe, and T. Taniguchi, J. Phys. Soc. Jpn. 84, 121007 (2015). [3] B. W. Alphenaar, A. A. M. Staring, H. van Houten, M. A. A. Mabesoone, O. J. A. Buyk, and C. T. Foxon, Phys. [29] D. A. Abanin and L. S. Levitov, Science 317, 641-643 (2007). Rev. B 46, 7236-7239 (1992). [4] C. de C. Chamon, D. E. Freed, S. A. Kivelson, S. L. [30] J. R. Williams, L. DiCarlo, and C. M. Marcus, Science 317, 638-641 (2007). Sondhi, and X. G. Wen, Phys. Rev. B 55, 2331-2343 (1997). [31] B. Ozyilmaz, P. Jarillo-Herrero, D. Efetov, D. A. Abanin, [5] B. I. Halperin, A. Stern, I. Neder, and B. Rosenow, Phys. L. S. Levitov, and P. Kim, Phys. Rev. Lett. 99, 166804 Rev. B 83, 155440 (2011). (2007). [6] P. Bonderson, A. Kitaev, and K. Shtengel, Phys. Rev. [32] S. Morikawa, S. Masubuchi, R. Moriya, K. Watanabe, Lett. 96, 016803 (2006). T. Taniguchi, and T. Machida, Appl. Phys. Lett. 106, [7] P. Bonderson, K. Shtengel, and J. K. Slingerland, Phys. 183101 (2015). Rev. Lett. 97, 016401 (2006). [33] A. Mrenca-Kolasi  nsk  a, S. Heun, and B. Szafran, Phys. [8] A. Stern and B. I. Halperin, Phys. Rev. Lett. 96, 016802 Rev. B 93, 125411 (2016). (2006). [34] D. S. Wei, T. van der Sar, J. D. Sanchez-Yamagishi, [9] R. Ilan, E. Grosfeld, K. Schoutens, and A. Stern, Phys. K. Watanabe, T. Taniguchi, P. Jarillo-Herrero, B. I. Rev. B 79, 245305 (2009). Halperin, and A. Yacoby, Sci. Adv. 3, e1700600 (2017). [10] F. E. Camino, W. Zhou, and V. J. Goldman, Phys. Rev. [35] P. Makk, C. Handschin, E. T ov ari, K. Watanabe, T. B 76, 155305 (2007). Taniguchi, K. Richter, M.-H. Liu, and C. Sch onenberger, [11] F. E. Camino, W. Zhou, and V. J. Goldman, Phys. Rev. Phys. Rev. B 98, 035413 (2018). Lett. 98, 076805 (2007). [36] S. Jorg, R. Patrik, and T. Bjorn, Sol. State Commun. [12] Y. Zhang, D. T. McClure, E. M. Levenson-Falk, C. M. 152, 1411-1419 (2012). Marcus, L. N. Pfei er, and K. W. West, Phys. Rev. B [37] F. Giannazzo, I. Deretzis, G.Nicotra, G.Fisichella, 79, 241304 (2009). C.Spinella, F.Roccaforte, and A.La Magna, Appl. Surf. [13] D. T. McClure, Y. Zhang, B. Rosenow, E. M. Levenson- Sci. 291, 53-57 (2014). Falk, C. M. Marcus, L. N. Pfei er, and K. W. West, Phys. [38] J. Bao, O. Yasui, W. Norimatsu, K. Matsuda, and M. Rev. Lett. 103, 206806 (2009). Kusunoki, Appl. Phys. Lett. 109, 081602 (2016). [14] P. V. Lin, F. E. Camino, and V. J. Goldman, Phys. Rev. [39] F. Speck, M. Ostler, S. Besendrfer, J. Krone, M. Wanke, B 80, 125310 (2009). and T. Seyller, Ann. Phys. 529, 1700046 (2017). [15] N. Ofek, A. Bid, M. Heiblum, A. Stern, V. Umansky, and [40] A. St ohr et al., Ann. Phys. 529, 1700052 (2017). D. Mahalu, Proc. Natl. Acad. Sci. USA 107, 5276-5281 [41] S. Odaka, H. Miyazaki, S.-L. Li, A. Kanda, K. Morita, (2010). S. Tanaka, Y. Miyata, H. Kataura, K. Tsukagoshi, and [16] D. T. McClure, W. Chang, C. M. Marcus, L. N. Pfei er, Y. Aoyagi, Appl. Phys. Lett. 96, 062111 (2010). and K. W. West, Phys. Rev. Lett. 108, 256804 (2012). [42] H. Kuramochi, S. Odaka, K. Morita, S. Tanaka, H. [17] R. Sabo, I. Gurman, A. Rosenblatt, F. Lafont, D. Banitt, Miyazaki, M. V. Lee, S.-L. Li, H. Hiura, and K. Tsuk- J. Park, M. Heiblum, Y. Gefen, V. Umansky, and D. agoshi, AIP Advances 2, 012115 (2012). Mahalu, Nat. Phys. 13, 491-496 (2017). 6 [43] T. Schumann, K.-J. Friedland, M. H. Oliveira, A. [69] M. Y. Han, J. C. Brant, and P. Kim, Phys. Rev. Lett. Tahraoui, J. M. J. Lopes, and H. Riechert, Phys. Rev. 104, 056801 (2010). B 85, 235402 (2012). [70] M. V. Fischetti and S. Narayanan, J. Appl. Phys. 110, [44] E. Pallecchi, E. Lafont, V. Cavaliere, F. Schopfer, D. 083713 (2011). Mailly, W. Poirier, and A. Ouerghia, Sci. Rep. 4, 4558 [71] N. Djavid et al., IEEE Trans. Electron Devices 61, 23 - (2014). 29 (2014). [45] T. Ciuk, S. Cakmakyapan, E. Ozbay, P. Caban, K. [72] T. Misawa et al., Jpn. J. Appl. Phys. 54, 05EB01 (2015). Grodecki, A. Krajewska, I. Pasternak, J. Szmidt, and [73] T. Fang, A. Konar, H. Xing, and D. Jena, Phys. Rev. B W. Strupinski, J. Appl. Phys. 116, 123708 (2014). 78, 205403 (2008). [46] E. Akira, K. Fumio, M. Kouhei, K. Takashi, and T. [74] A. Y. Goharrizi et al., IEEE Trans. Electron Devices 58, Satoru, J. Low Temp. Phys. 179, 237-250 (2015). 3725 - 3735 (2011). [47] D. Momeni Pakdehi et al., ACS Appl. Mater. Interfaces [75] M. Evaldsson et al., Phys. Rev. B 78, 161407(R) (2008). 10, 6039-6045 (2018). [76] D. Querlioz et al., Appl. Phys. Lett. 92, 042108 (2008). [48] T. Takahiro and O. Toshio, Carbon 50, 674-679 (2012). [77] M. Poljak and T. Suligoj, IEEE Trans. Electron Devices [49] H. Kenjiro, S. Shintaro, and Y. Naoki, Nanotechnol. 24, 63, 537-543 (2016). 025603 (2012). [78] A. Cresti and S. Roche, New J. Phys. 11, 095004 (2009). [50] J.-K. Lee et al., Nano Lett. 13, 3494-3500 (2013). [79] V. Hung Nguyen, Y. M. Niquet, and P. Dollfus, Phys. [51] C.-H. Lee et al., Adv. Mater. 26, 2812-2817 (2014). Rev. B 88, 035408 (2013). [52] M. E. Ayhan, G. Kalita, M. Kondo, and M. Tanemura, [80] R. Zhang, Z. Wu, X. J. Li, and K. Chang, Phys. Rev. B RSC Adv. 4, 26866-26871 (2014). 95, 125418 (2017). [53] S. Dai et al., Nat. Nanotechnol. 10, 682-686 (2015). [81] A. N. Rudenko and M. I. Katsnelson, Phys. Rev. B 89, [54] K.-K. Bai, J.-B. Qiao, H. Jiang, H. Liu, and L. He, Phys. 201408 (2014). Rev. B 95, 201406 (2017). [82] Y. Xu, H. Gao, M. Li, Z. Guo, H. Chen, Z. Jin and B. [55] B. D. Briggs et al., Appl. Phys. Lett. 97, 223102 (2010). Yu, Nanotechnol. 22, 365202 (2011). [56] M. L. Leadbeater, C. L. Foden, J. H. Burroughes, M. [83] T. Low, V. Perebeinos, J. Terso , and Ph. Avouris, Phys. Pepper, T. M. Burke, L. L. Wang, M. P. Grimshaw, and Rev. Lett. 108, 096601 (2012). D. A. Ritchie, Phys. Rev. B 52, R8629-R8632 (1995). [84] http://www.openmx-square.org. [57] M. L. Leadbeater, C. L. Foden, J. H. Burroughes, T. M. [85] V. M. Pereira, A. H. Castro Neto, and N. M. R. Peres, Burke, L. L. Wang, M. P. Grimshaw, D. A. Ritchie, and Phys. Rev. B 80, 045401 (2009). M. Pepper, Surf. Sci. 361/362, 587-590 (1996). [86] V. V. Cheianov and V. I. Fal'ko, Phys. Rev. B 74, [58] I. S. Ibrahim, V. A. Schweigert, and F. M. Peeters, Phys. 041403(R) (2006). Rev. B 56, 7508 (1997). [87] A. Nogaret, J. Phys.: Condens. Matter 22, 253201 [59] S. Cin a, D. M. Whittaker, D. D. Arnone, T. Burke, H. P. (2010). Hughes, M. Leadbeater, M. Pepper, and D. A. Ritchie, Phys. Rev. Lett. 83, 4425-4428 (1999). [60] A. Nauen, U. Zeitler, R. J. Haug, A. G. M. Jansen, M. Dilger, and K. Eberl, Physica E 13, 732-735 (2002). [61] M. Grayson, D. Schuh, M. Bichler, M. Huber, G. Abstre- iter, L. Hoeppel, J. Smet, and K. von Klitzing, Physica E 22, 181-184 (2004). [62] J. A. Robinson, C. P. Puls, N. E. Staley, J. P. Stitt, M. A. Fanton, K. V. Emtsev, T. Seyller and Y. Liu, Nano Lett. 9, 964-968 (2009). [63] See Supplemental Material, which includes Refs. [27, 33, 55, 62, 64, 65, 68{80, 82{85]. In this material, computa- tional methodologies, e ects of local strains, edge disor- ders, and multi-bands contribution are presented in de- tail. [64] V. Hung Nguyen, J. Saint-Martin, D. Querlioz, F. Maz- zamuto, A. Bournel, Y.-M. Niquet, and P. Dollfus, J. Comput. Electron. 12, 85-93 (2013). [65] C. H. Lewenkopf and E. R. Mucciolo, J. Comput. Elec- tron. 12, 203-231 (2013). [66] Y. Aharonov and D. Bohm, Phys. Rev. 115, 485-491 (1959). [67] Armchair GNRs with perfect edges can be basically di- vided in two main classes, depending on the number of zigzag lines N across their width, i.e., GNRs with N = 3p + 2 are metallic while they are semiconducting in the other cases. The details can be found in Y.-W. Son et al., Phys. Rev. Lett. 97, 216803 (2016) or in S. M.-M. Dubois et al., Eur. Phys. J. B 72, 1-24 (2009). [68] M. Y. Han, B. zyilmaz, Y. Zhang, and P. Kim, Phys. Rev. Lett. 98, 206805 (2007). 7 Supplemental material for V. Hung Nguyen and J.-C. Charlier Institute of Condensed Matter and Nanosciences, Université catholique de Louvain, Chemin des étoiles 8, B-1348 Louvain-la-Neuve, Belgium Contents: 1. Computational methodologies 2. Curvature-induced local strains in stepped graphene Local strains induced by curvature and simulated models are discussed. 3. Tight Binding versus Density Functional Theory calculations Validity of tight binding model to investigate the local strains is considered by the fit to the density functional theory calculations 4. Effects of curvature-induced local strains on Aharonov-Bohm interference 5. Effects of edge disorder in graphene nanoribbons 6. Multi sub-bands contribution References 1. Computational methodologies In order to compute the electronic transport through stepped graphene systems, we employed the Green’s function technique [1] to solve the tight-binding Hamiltonian presented in the main text. In particular, the retarded Green’s function is determined as −1 R +. G ( E)= E+i 0 − H −Σ −Σ (S.1) [ ] D L R where H is the device Hamiltonian and  are self-energies describing the left and right device-to- D L,R lead couplings, respectively. This equation was solved using the recursive method [2]. The transport quantities such as transmission probability T(E), conductance G(E ) and local density of left- and right- injected states D (E,r) are then computed using the Landauer formalism as follows: L,R R R † T(E) = (S.2) Tr [Γ G Γ G ] L R +∞ ∂ f 2 e G(E ) = d E T ( E) − (S.3) F ∫ ( ) h −∞ ∂ E R R † G Γ G L , R D (E,r) = (S.4) L,R 2π Here, Γ =i Σ −Σ and f (E) is the Fermi-Dirac distribution function with the Fermi level E . [ ] F F L , R L , R L , R The total local density of states can be computed either by D(E,r) = D (E,r)+D (E,r) or L R D( E , r)=−ℑ( G )/π . 8 2. Curvature-induced local strains in stepped graphene In general, the non-planar geometry can induce in-plane strain inhomogeneities (i.e., local strains). Different from the uniform strains, local strains can result in electron scatterings and can be effectively described as the effects of strain-induced gauge fields in stepped graphene systems. These local strains have been shown to be often observed in the detached regions, i.e., around the step edges [3-5]. However, it has been experimentally demonstrated that if no external stress is applied, only small local strains can be observed in the stepped graphene systems, because of the mechanical robustness of graphene layers. For instance, it has been investigated and reported in refs. [3,4] that graphene at the step could experience small uniaxial strains (i.e., < 1%) relative to the rest of the sheet and nearly strain-free graphene is possible in epitaxial graphene. Extremely small strains of ~ 0.025% were also demonstrated in ref. [5] by Raman spectroscopy measurements. To investigate the effects of such local strains induced by graphene curvature, we assume a simple model that the strain is maximum at the step edges and gradually release in two sides. Here, we use the following simple formula to model these strains: max ε(r)= (S.5) 1+(d / d ) where  is the maximum value of strain, d is the in-plane distance from the position r to the max considered step edge, and d characterizes the strain release distance. Fig. S1 shows a typical picture of possible strain gradients induced by graphene curvature, that can be modelled by the above formula. Fig. S1: Possible local strains induced by curvature in stepped graphene systems. In next sections 3 and 4, using such the simple model we present an investigation to clarify the effects of possible local strains mentioned above on the Aharonov-Bohm interference predicted in this work. 9 3. Tight Binding versus Density Functional Theory calculations In this section, we present some calculations demonstrating the validity of the tight binding (TB) Hamiltonian to investigate the curvature-induced local strains in stepped graphene systems by the fit to the ab initio quantum transport data [6]. Actually, it has been shown by several studies in the literature that a simple p TB Hamiltonian with only nearest neighbor interactions can be used to compute very accurately the electronic properties and electron transport in planar graphene systems. However, the curvature in the stepped graphene systems can alter their electronic properties, i.e., can induce electron scatterings, especially, when a local strain occur around the step edges as discussed above. Hence, the validity of the p TB model is needed to be examined. To this aim, we performed quantum transport calculations based on density functional theory (DFT) [6] for stepped graphene systems, which are assumed very large so as to neglect the finite-width effects and hence the periodic boundary condition can be applied along Oy axis. The transmission coefficients through stepped graphene systems obtained for three different k - momentum modes around the Dirac point are computed and presented in Fig. S2. Two cases without and with a local strain ( = 1%) are considered. In agreement with the study in [7] with different max curvature radius, step heights and step angles, the results obtained without the local in-plane strain show that the effects of curvature on the electronic transport through the system is negligible. In particular, for k = K , the transmission coefficient exhibits only negligibly small reduction around the y 0 zero energy point, compared to the planar case where it is unity and constant in the considered energy range. Thus, the curvature does not induce significant electron scatterings and hence, similar to the planar cases, the p TB Hamiltonian still works well for these considered stepped graphene systems. Fig. S2: Electronic transport through stepped 2D graphene systems using DFT calculations: strained (with  = 1%) system as described in Fig.S1 versus unstrained one. max Fig. S3: Electronic transport through stepped 2D graphene system with local strain  = 1%: max Tight-Binding vs DFT calculations. When the curvature-induced local strains are introduced, significant electron scatterings at the step edges can be observed. Indeed, the transmission coefficient exhibits significant reduction around the zero energy point for k = K mode and close to the edges of energy gap for other modes K . In order y 0 1,2 to compute these effects, the p TB Hamiltonian must be adjusted. In particular, a model where the hopping term is determined a function of C-C bond length r as t =t exp β(1−r / r ) with  =3.37 ij [ ] ij 0 ij 0 and r = 0.142 nm has been demonstrated [8] to compute well the strain effects in graphene. Fig. S3 demonstrates a quite good agreement between the DFT data and results obtained by such strained TB model. Our calculations show that this strained TB Hamiltonian without any other adjustment is still a good model for the considered stepped graphene systems with local strains of  ≲ 4%. max 4. Effects of curvature-induced local strains on Aharonov-Bohm interference In this section, we employed the strained TB model presented above to investigate the effects of curvature-induced local strains in the Aharonov-Bohm interferometers predicted in this work. The conductance as a function of Fermi energy obtained at zero magnetic field is presented in Fig. S4. Similar to the results presented in section 3, the considered local strain can induce electron scatterings and affect significantly the transport through the system, leading to the conductance reduction. The effects of such strains on the predicted Aharonov-Bohm interference are investigated and presented in Fig. S5. It is however shown that the effects on the conductance oscillations in the quantum Hall regime are relatively weaker than those observed in the low field one. This can be explained as follows. In the low field regime, the electron transport is essentially due to “bulk states”, which transmit across the step edges (i.e., zones of local strains) and hence undergo strong back-scatterings as described in the top-right image of Fig.S5. In the quantum Hall regime, the presence of these locally strained zones however affects the transport picture differently. In particular, when a high magnetic field with  <  B S is applied, opposite edge states are formed in terrace and facet zones (in two sides of the step edges) and electrons transmitting through the system have to follow trajectories as described in the bottom- 11 right image of Fig.S5, i.e., when reaching the step edges, electrons transmit along (not directly across, as in the zero-field case) these edges. This can eliminate the scatterings induced by the considered local strains and explain their weak effects in the quantum Hall regime, compared to those in the low field one. Hence, within the range of experimentally reported strains discussed above, the considered local strains even though alter but does not strongly perturb the predicted Aharonov-Bohm interference. Fig. S4: Conductance as a function of energy at zero magnetic field with different curvature-induced strains. The strain release distance d ≈ 11 nm while other parameters are as in Fig.2 of the main text. Fig. S5: Conductance as a function of magnetic field at E = 75 meV and with different curvature-induced strains. Parameters are as in Fig.S4. Right panels illustrate the electron transport pictures at low and high magnetic fields. 12 Certainly, the effects of these local strains on the Aharonov-Bohm interference can be enlarged and significant when much larger strain gradients occur, for instance, when external stresses are additionally applied. 5. Effects of edge disorder in graphene nanoribbons Similar to the surface roughness in many nanoscale systems of conventional semiconductors, the edge disorder (i.e., edge roughness) is often a practical issue for graphene ribbons [9-19]. The edge disorder has been shown to affect strongly the electronic transport through graphene ribbons, especially, when their width reaches the nanoscale regime. Fig. S6: Graphene ribbon with edge disorder. To investigate the effects of edge disorder, there were two models widely used in the literature. In particular, the disordered edges can either be generated by randomly removing the edge atoms with a certain probability [11-14] or are modeled using auto-correlation functions [15-19]. These models have been demonstrated to interpret well the electronic properties of GNRs in experiments with a wide range of disorder level [9-19]. Here, the effects of edge disorder are examined using both two models. In the latter case, the edge disorder is generated by a Gaussian autocorrelation function, particularly, the variation of ribbon width  W is effectively described as 2 Δ x ⟨δ W ( x)δ W ( x+Δ x)⟩=W exp − (S.6) rms ( ) 2ξ where W is the rms value charactering the disorder strength and  represents the correlation length. rms A typical picture of edge disordered graphene ribbons is illustrated in Fig. S6. The effects of edge disorder generated by a Gaussian autocorrelation function are presented in Fig. 4 of the main text and the results obtained in the systems by randomly removing edge atoms are displayed in Fig. S7. Actually, two main features are found with both disorder models. First, in the case of semiconducting GNRs, the Aharonov-Bohm oscillations obtained at high magnetic fields (B-fields) are shown to be much more robust under the effect of the considered disorders than the zero-field transport. This can be explained by a fascinating feature that different from the zero-field case (see Figs.S8 (a,d)), the forward and backward edge channels are spatially separated in the quantum Hall regime (see Figs.S8 (b-c, e-f)) while the scatterings at the disordered edges do not allow electrons 13 to transmit across the sample. As a consequence, the edge disorder effects do not contribute significantly to backscatterings in the considered stepped systems, i.e., the dominant mechanism of backscattering is still interaction between edge states in zones of opposite normal B-fields and hence strong Aharonov-Bohm oscillations are still achieved. Fig. S7: (a,b) Conductance as a function of B-field at E = 75 meV computed for semiconducting (N = F z 324) and metallic (N = 326) GNR systems, respectively, with θ = 60°, θ =30°, L  nm and z S B F W  40nm. In the edge disordered systems, 15% of edge atoms are randomly removed. Second, as presented in Fig.4 of the main text and in Fig. S7, there is a significant difference between semiconducting and metallic GNR systems with perfect edges, i.e., the Aharonov-Bohm oscillation is relatively weak in the metallic cases, compared to the results obtained in semiconducting ones. The 14 edge disorder however strengthens the Aharonov-Bohm oscillation in metallic GNR systems, thus eliminating the difference mentioned. Fig. S8: Edge disorder effects on electron propagation in planar graphene ribbons: left-injected LDOS (a,b,d,e) and total LDOS (c,f). Magnetic fields B = 0 in (a,d) and 25 T in (b,c,e,f) while carrier energy E = 75 meV. Perfect and disordered edges are considered in (a,b,c) and (d,e,f), respectively. Fig. S9: Conductance as a function of B-field at E = 75 meV obtained in stepped systems of a zigzag GNR with θ = 60°, θ =30°, L  72 nm and W  40 nm. The disordered edge systems are generated S B F by the Gaussian autocorrelation function (S.6) with W  0.6 nm and  nm. rms 15 Note additionally that a picture, similar to those obtained in the metallic armchair GNRs, is also observed in zigzag GNR systems, i.e., the Aharonov-Bohm oscillation is relatively weak for perfect edges but the effect is much more pronounced when edge disorder is introduced (see Fig. S9). 6. Multi sub-bands contribution In this section, we analyze the effects of multi sub-bands contribution on the predicted Aharonov-Bohm oscillation in more detail. As it has been demonstrated in nanorings [20,21] and also in graphene p-n junctions [22], the Aharonov-Bohm interference has an inherent property that the strong oscillation of conductance can be observed in the low energy regime where only a single band is obtained. In the regime of high energies when multi energy bands can contribute to the transport, the interference picture can be significantly blurred. A similar feature is also observed in the Aharonov-Bohm interferometers considered here. Fig. S10: Conductance as a function of Fermi energy at B = 20 T obtained in stepped systems of an armchair GNR with L  110 nm and W   0 nm. The stepped angle  = 60° while three directions of F S T(F) B-field are considered. E indicates the first Landau level formed in the terrace (facet) zone. 1 16 Indeed, as shown in Fig. S10 and Fig. 5 of the main text, strong conductance oscillations in all T F T ,F considered cases are observed in the regime where are the first Landau E ≤ min( E , E ) E F 1 1 1 levels formed in the terrace and facet zones, respectively, at high magnetic fields . In Fig. S11, the energy bands of an armchair GNR under the effect of magnetic field are presented. When a high magnetic field is applied, the energy bands of GNR are strongly modified and the Landau quantization is observed as seen in the right panel of Fig.S11. This Landau quantization (i.e., energy levels E ) 0, 1, 2,...   obeys the well-known formula [23] E = sign(n) 2 e ℏ v |n|B (S.7) n F ⟂ T F where v is the Fermi velocity in graphene. The energy levels E and E mentioned above are hence F 1 1 T 2 F 2 determined as E = 2 eℏ v B|sin θ | and E = 2 e ℏ v B|sin (θ −θ )|. √ √ 1 F B 1 F S B B = 0.0 T B = 6.5 T B = 13 T -1 -2 Momentum k Momentum k Momentum k Fig. S11: Energy bands of the armchair GNR of W   0 nm under the effect of magnetic field. Energy levels E (n=0, 1, 2,...) in the right panel represent the Landau quantization at a high B-field [23]. In the high energy regime, the effects of multi sub-bands contribution, similar to those observed in graphene and phosphorene nanorings [20,21] and mentioned above, are clearly demonstrated in Fig.S10 (see the zones highlighted in yellow). In general, the Aharonov-Bohm oscillations are T F significantly blurred by such contribution when E >min ( E , E ), as illustrated in both Fig. S10 here F 1 1 T F and Fig.5 of the main text. In the cases if E  E (e.g.,  = 15° and 45°), some oscillations can still 1 1 B be well defined in the range between these two values (see Fig. S10 for  = 15°). The features presented here explain clearly the results presented and discussed in Fig.5 of the main text. Energy (eV) 17 References: [1] V. Hung Nguyen et al., J. Comput. Electron. 12, 85-93 (2013). [2] C. H. Lewenkopf and E. R. Mucciolo, J. Comput. Electron. 12, 203-231 (2013). [3] J. A. Robinson et al., Nano Lett. 9, 964-968 (2009). [4] T. Low, V. Perebeinos, J. Tersoff, and Ph. Avouris, Phys. Rev. Lett. 108, 096601 (2012). [5] B. D. Briggs et al., Appl. Phys. Lett. 97, 223102 (2010). [6] http://www.openmx-square.org [7] Y. Xu, H. Gao, M. Li, Z. Guo, H. Chen, Z. Jin and B. Yu, Nanotechnol. 22, 365202 (2011). [8] V. M. Pereira, A. H. Castro Neto, and N. M. R. Peres, Phys. Rev. B 80, 045401 (2009). [9] M. Y. Han, B. Özyilmaz, Y. Zhang, and P. Kim, Phys. Rev. Lett. 98, 206805 (2007). [10] M. Y. Han, J. C. Brant, and P. Kim, Phys. Rev. Lett. 104, 056801 (2010). [11] M. Evaldsson et al., Phys. Rev. B 78, 161407(R) (2008). [12] D. Querlioz et al., Appl. Phys. Lett. 92, 042108 (2008). [13] A. Cresti and S. Roche, New J. Phys. 11, 095004 (2009). [14] M. Poljak and T. Suligoj, IEEE Trans. Electron Devices 63, 537-543 (2016). [15] M. V. Fischetti and S. Narayanan, J. Appl. Phys. 110, 083713 (2011). [16] N. Djavid et al., IEEE Trans. Electron Devices 61, 23 - 29 (2014). [17] T. Misawa et al., Jpn. J. Appl. Phys. 54, 05EB01 (2015). [18] T. Fang, A. Konar, H. Xing, and D. Jena, Phys. Rev. B 78, 205403 (2008). [19] A. Y. Goharrizi et al., IEEE Trans. Electron Devices 58, 3725 - 3735 (2011). [20] V. Hung Nguyen, Y. M. Niquet, and P. Dollfus, Phys. Rev. B 88, 035408 (2013). [21] R. Zhang, Z. Wu, X. J. Li, and K. Chang, Phys. Rev. B 95, 125418 (2017). [22] A. Mreńca-Kolasińska, S. Heun, and B. Szafran, Phys. Rev. B 93, 125411 (2016). [23] L.-J. Yin, K.-K. Bai, W.-X. Wang, S.-Y. Li, Y. Zhang, and L. He, Front. Phys. 12, 127208 (2017).

Journal

Condensed MatterarXiv (Cornell University)

Published: Dec 6, 2018

There are no references for this article.