You are on page 1of 3

APPLIED PHYSICS LETTERS 92, 152108 2008

Quantum transport in In0.75Ga0.25As quantum wires


P. J. Simmonds, F. Sgakis,a H. E. Beere, D. A. Ritchie, M. Pepper, D. Anderson, and G. A. C. Jones
Cavendish Laboratory, J.J. Thomson Avenue, Cambridge CB3 OHE, United Kingdom

Received 17 January 2008; accepted 27 March 2008; published online 16 April 2008 In addition to quantized conductance plateaus at integer multiples of 2e2 / h, the differential conductance G = dI / dV shows plateaus at 0.25 2e2 / h and 0.75 2e2 / h under applied source-drain bias in In0.75Ga0.25As quantum wires dened by insulated split gates. This observation is consistent with a spin-gap model for the 0.7 structure. Using a tilted magnetic eld to induce Landau level crossings, the g factor was measured to be 9 by the coincidence method. This material, with a mobility of 1.8 105 cm2 / V s at a carrier density of 1.4 1011 cm2, may prove useful for further study of electron-electron interaction effects in quantum wires. 2008 American Institute of Physics. DOI: 10.1063/1.2911730 The use of split-gates devices1 to dene quantum wires in the GaAs/ AlGaAs material system led to the rst observations of quantized differential conductance G = dI / dV in units of 2e2 / h Refs. 2 and 3 in the ballistic transport of electrons, at zero magnetic eld. In a nite magnetic eld B, Zeeman splitting lifts the electron spin degeneracy and conductance plateaus form at half-integer multiples of 2e2 / h . Due to the promising electron mobility of 2 105 cm2 / V s with a carrier density of 3.1 1011 cm2, which was recently achieved in In0.75Ga0.25As/ In0.75Al0.25As at T = 4.2 K,4 interest in this material system for the study of electron transport in quantum wires is growing. Conductance quantization and AharonovBohm oscillations have already been observed in etched devices.5 In0.75Ga0.25As has a much larger Land g factor g* 9 than GaAs g* 0.44 , which is benecial for the study of electron-electron interactions in magnetic elds such as the 0.7 analogue6 and the magnetic splitting of the zero-bias anomaly7 . Quantum wires with large onedimensional 1D energy level subband spacing can be fabricated because of its smaller effective mass m* = 0.040me versus m* = 0.067me in GaAs . Lastly, spin-orbit coupling is much more pronounced in In0.75Ga0.25As than in GaAs,8 which could enable the observation of spin polarization in electron focusing experiments.9 Here, we report quantized conductance observed in quantum wires dened by insulated split gates above a In0.75Ga0.25As quantum well. We observe 0.75 2e2 / h and 0.25 2e2 / h conductance plateaus under source-drain bias for an otherwise unpolarized quantum wire, which is an indication of spontaneous spin polarization. Strikingly, these plateaus exactly occur at the conductance values of quarterinteger plateaus, as normally only observed in the presence of a large magnetic eld or due to the effects of strong electron-electron interactions.1012 For G 2e2 / h, when a source-drain bias Vsd is applied and drain between the source s = + eVsd / 2 d= eVsd / 2 of the device in zero magnetic eld being the chemical potential at equilibrium , plateaus at half-integer multiples of 2e2 / h appear.13 If the minimum of a 1D subband lies between the two electrochemical potentials s and d, then there is one fewer propagating subband in one dia

Electronic mail: fs228@cam.ac.uk.

rection in the channel than in the other direction. It is important to note that these half-integer plateaus at high Vsd are still spin degenerate. Application of a magnetic eld results in the formation of quarter-integer plateaus14 at 1.25 2e2 / h , 1.75 2e2 / h , 2.25 2e2 / h , 2.75 2e2 / h , 3.25 2e2 / h , . The situation for G 2e2 / h is less well understood. The 0.7 structure is a commonly observed weak conductance plateau near 0.7 2e2 / h in zero magnetic eld.15 It evolves into a spin-polarized plateau at 0.5 2e2 / h with increasing magnetic eld, and becomes more pronounced as temperature increases. The 0.7 structure is thought to indicate partial spin polarization at B = 0 due to many-body spin interactions.16 To date, the origin of these effects has not been fully explained. When Vsd is applied in the absence of magnetic elds for G 2e2 / h, strong conductance plateaus are observed at 0.8 0.85 2e2 / h and 0.2 0.3 2e2 / h ,13 which are sometimes assigned the values of 0.85 2e2 / h and 0.5 2e2 / h in literature.1719 The 0.85 2e2 / h plateau has been shown to evolve from the 0.7 structure.20 These plateaus are not altered by the application of a magnetic eld,20 which may imply that these could be quarter-integer plateaus, nominally at 0.75 2e2 / h and 0.25 2e2 / h .1012 Undoped In0.75Ga0.25As two-dimensional electron gases 2DEGs were grown in a Veeco MOD GEN II solid-source molecular beam epitaxy MBE chamber on a 2 in. semiinsulating SI 001 InP substrates. The growth temperature was 440 C, which was calculated from the substrate bandedge wavelength using a KSA BANDIT system.21 Growth temperature was not intentionally altered during epitaxy. The V/III ratio of beam-equivalent pressures was maintained at 30, with a maximum growth rate of 1.3 m / h. Similar growth conditions for this material system have been previously optimized.22 A 700 nm InxAl1xAs buffer, in which the indium mole fraction was linearly graded from x = 0.52 to x = 0.85, was grown to achieve maximal relaxation of the device heterostructure. A 5 min growth interruption under As4 allowed stabilization of the In effusion cell and interface smoothing. A 50 nm In0.75Al0.25As layer was then grown, followed by a 30 nm In0.75Ga0.25As quantum well to reduce the effect of interface roughness scattering and to ensure single subband

0003-6951/2008/92 15 /152108/3/$23.00 92, 152108-1 2008 American Institute of Physics Downloaded 18 Apr 2011 to 121.240.208.65. Redistribution subject to AIP license or copyright; see http://apl.aip.org/about/rights_and_permissions

152108-2

Simmonds et al.

Appl. Phys. Lett. 92, 152108 2008

FIG. 1. Color online Conduction band prole and electron density for the InGaAs quantum well 2DEGs, calculated using Sniders 1D POISSON program Ref. 25 .

population in this range of electron densities23 , a 120 nm In0.75Al0.25As top barrier, and nally, a 10 nm In0.75Ga0.25As cap. No layers were intentionally doped. The resulting bandstructure, calculated from a self-consistent solution of the Poisson and Schrodinger equations, is shown in Fig. 1. The simulation predicted a carrier density of 1.4 1011 cm2 assuming a background of 1 1016 cm3 charged impurities in the InAlAs layers.23,24 Figure 2 shows magnetotransport measurements performed on Hall bars. Following illumination with a light emitting diode, the electron mobility for this 2DEG was measured to be 1.8 105 cm2 / V s at a carrier density ns of 1.4 1011 cm2. Although spin polarization only becomes apparent for B 1 T where the = 5 valley can be resolved , the Landau levels are most likely disorder broadened, preventing the resolution of spin splitting at lower magnetic elds. Using a top gate under forward bias, a peak mobility of 2.2 105 cm2 / V s was measured at a carrier density of 1.6 1011 cm2. Transport anisotropy was assessed along orthogonal crystal axes. Mobility was shown to be higher for currents owing along 110 , which was previously shown to be a direct function of surface morphology.24,26 In future, intentionally doping this MBE structure could further reduce disorder by suitably increasing electron-electron charge screening. Using a tilted-eld approach to identify coincidences between the Zeeman and cyclotron energies,27 the exchangeenhanced electron g factor was measured to be g* = 9.15

FIG. 3. a Conductance of an In0.75Ga0.25As quantum wire at T = 1.5 K, B = 0.6 T, and Vg = 0. b The feature near 0.5 2e2 / h evolves into a resonance as Vg = Vleft Vright increases from Vg = 0 leftmost trace to Vg g g = 0.4 V rightmost trace in steps of 0.1 V T = 1.5 K, B = 0.4 T, laterally traces offset . c From left to right, Vsd is increased from 0.1 to + 4.0 mV in steps of 0.050 mV traces offset laterally . Note the appearance of a 0.25 2e2 / h and 0.75 2e2 / h plateaus for the rst 1D subband, while only a 1.5 2e2 / h plateau appears for the second subband B = 0.6 T .

FIG. 2. Color online Magnetic eld dependence of longitudinal and transverse resistivities in an In0.75Ga0.25As 2DEG. The lling factors are shown for each Hall plateau. Downloaded 18 Apr 2011 to 121.240.208.65. Redistribution subject to AIP license or copyright; see http://apl.aip.org/about/rights_and_permissions

0.14 at lling factor = 4, and g* = 8.1 0.3 at = 6. These values for g* closely correlate with those found 9.2 for = 4 and 8.1 for = 6 in a previous study of In0.75Ga0.25As.4 A reduction in electron-electron exchange interactions is believed to be responsible for the lower values of g* at higher lling factors.28 Split gates were used to dene a quantum wire in the 2DEG and fabricated using e-beam lithography with a 1D channel length of 400 nm and a width of 500 nm. A 50 nm thick polyimide lm was used to insulate the split gates from the semiconductor due to the low metal-In0.75Ga0.25As Schottky barrier height 0.031 eV Ref. 29 . Maximum leakage current through the polyimide was measured to be 1.5 nA. A magnetic eld B perpendicular to the 2DEG plane was applied to reduce backscattering from impurities close to the quantum wire. From Fig. 2, this eld places the 2DEG in the regime where the rst Shubnikovde Haas oscillations can be seen = 12 . Figure 3 a shows the differential conductance G = dI / dVsd of our quantum wire at equilibrium Vsd = 0 . The rst and second integer plateaus are well resolved. The feature near 0.5 2e2 / h evolves into a resonant peak when the 1D channel is laterally moved by asymmetrically biasing each split gate, as shown in Fig. 3 b . A clean 0.7 structure should be insensitive to lateral displacement e.g., see Fig. 4 of Ref. 30 . The feature near 0.5 2e2 / h is most likely to be a 0.7 structure convolved with a resonance. In Fig. 3, for Vsd = 0, there is no half-integer plateau at 1.5 2e2 / h . The B = 0.6 T eld is therefore insufcient to spin polarize the system. This conclusion is further supported by the fact that, as Vsd is increased, quarter-integer plateaus at 1.25 2e2 / h or 1.75 2e2 / h are not observed. Instead, we only observe a 1.5 2e2 / h plateau: the expected behavior for an unpolarized 1D system under source-drain bias.13 As Vsd is increased for G 2e2 / h, plateaus at 0.75 2e2 / h and 0.25 2e2 / h also appear. These two plateaus

152108-3

Simmonds et al.
9

Appl. Phys. Lett. 92, 152108 2008 L. P. Rokhinson, L. N. Pfeiffer, and K. West, Phys. Rev. Lett. 96, 156602 2006 . 10 F. Sgakis, Ph.D. thesis, University of Cambridge, 2005. 11 A. C. Graham, M. Pepper, M. Y. Simmons, and D. Ritchie, Physica E Amsterdam 34, 588 2006 . 12 F. Sgakis, A. Graham, K. J. Thomas, and M. Pepper, J. Phys. Condens. Matter 20, 164213 2008 . 13 N. K. Patel, J. T. Nicholls, L. Martin-Moreno, M. Pepper, J. E. F. Frost, D. A. Ritchie, and G. A. C. Jones, Phys. Rev. B 44, 13549 1991 . 14 N. K. Patel, J. T. Nicholls, L. Martin-Moreno, M. Pepper, J. E. F. Frost, D. A. Ritchie, and G. A. C. Jones, Phys. Rev. B 44, 10973 1991 . 15 K. J. Thomas, J. T. Nicholls, M. Y. Simmons, M. Pepper, D. R. Mace, and D. A. Ritchie, Phys. Rev. Lett. 77, 135 1996 . 16 K. F. Berggren, P. Jaksch, and I. Yakimenko, Phys. Rev. B 71, 115303 2005 . 17 A. Kristensen, H. Bruus, A. Hansen, J. B. Jensen, P. E. Lindelof, C. J. Marckmann, J. Nygard, and C. B. Sorensen, Phys. Rev. B 62, 10950 2000 . 18 S. M. Cronenwett, H. J. Lynch, D. Goldhaber-Gordon, L. P. Kouwenhoven, C. M. Marcus, K. Hirose, N. S. Wingreen, and V. Umansky, Phys. Rev. Lett. 88, 226805 2002 . 19 D. J. Reilly, Y. Zhang, and L. DiCarlo, Physica E Amsterdam 34, 27 2006 . 20 K. J. Thomas, J. T. Nicholls, M. Y. Simmons, M. Pepper, D. R. Mace, and D. A. Ritchie, Philos. Mag. B 77, 1213 1998 . 21 I. Farrer, J. J. Harris, R. Thomson, D. Bartlett, C. A. Taylor, and D. A. Ritchie, J. Cryst. Growth 301, 88 2007 . 22 V. Drouot, M. Gendry, C. Santinelli, P. Viktorovitch, G. Hollinger, S. Elleuch, and J. L. Pelouard, J. Appl. Phys. 77, 1810 1995 . 23 F. Capotondi, G. Biasiol, D. Ercolani, and L. Sorba, J. Cryst. Growth 278, 538 2005 . 24 F. Capotondi, G. Biasiol, I. Vobornik, L. Sorba, F. Giazotto, A. Cavallini, and B. Fraboni, J. Vac. Sci. Technol. B 22, 702 2004 . 25 G. Snider, 1D POISSON/SCHRODINGER. This program can be downloaded from http://www.nd.edu/~gsnider/. 26 A. Richter, M. Koch, T. Matsuyama, C. Heyn, and U. Merkt, Appl. Phys. Lett. 77, 3227 2000 . 27 F. F. Fang and P. J. Stiles, Phys. Rev. 174, 823 1968 . 28 T. Ando and Y. Uemura, J. Phys. Soc. Jpn. 37, 1044 1974 . 29 K. Kajiyama, Y. Mizushim, and S. Sakata, Appl. Phys. Lett. 23, 458 1973 . 30 K. J. Thomas, J. T. Nicholls, N. J. Appleyard, M. Y. Simmons, M. Pepper, D. R. Mace, W. R. Tribe, and D. A. Ritchie, Phys. Rev. B 58, 4846 1998 .

are nearly identical to the 0.8 0.85 2e2 / h and 0.2 0.3 2e2 / h plateaus normally observed in GaAs quantum wires. These are believed to be zero-eld spin-polarized quarter-integer plateaus.1012 The late formation in sourcedrain bias voltage of the 0.25 2e2 / h plateau with respect to the formation of the 0.75 2e2 / h plateau is consistent with the predictions from a spin-gap model.10 The 0.25 2e2 / h plateau indicates that spin-polarized electrons ow in only one direction. This relation between momentum and spin may be of consequence in applications of nanostructures. In conclusion, we have fabricated quantum wires on a In0.75Ga0.25As/ In0.75Al0.25As heterostructure using split gates. The differential conductance shows quantized plateaus at 0.25 2e2 / h and 0.75 2e2 / h in nonequilibrium transport. This high g-factor material may prove useful for probing spin-related phenomena. The authors would like to thank A. C. Graham, K. Das Gupta, D. Ercolani, and K. J. Thomas for useful discussions. This work was supported by the EPSRC. P.J.S. acknowledges the support from the Semiconductor Physics Group.
1

T. J. Thornton, M. Pepper, H. Ahmed, D. Andrews, and G. Davies, Phys. Rev. Lett. 56, 1198 1986 . 2 D. A. Wharam, T. J. Thornton, R. Newbury, M. Pepper, H. Ahmed, J. E. F. Frost, D. G. Hasko, D. C. Peacock, D. A. Ritchie, and G. A. C. Jones, J. Phys. C 21, L209 1988 . 3 B. J. van Wees, H. van Houten, C. W. J. Beenaker, J. G. Williamson, L. P. Kouwenhoven, D. van der Marel, and C. T. Foxon, Phys. Rev. Lett. 60, 848 1988 . 4 W. Desrat, F. Giazotto, V. Pellegrini, F. Beltram, F. Capotondi, G. Biasiol, L. Sorba, and D. K. Maude, Phys. Rev. B 69, 245324 2004 . 5 F. Carillo, G. Biasiol, D. Frustaglia, F. Giazotto, L. Sorba, and F. Beltram, Physica E Amsterdam 32, 53 2006 . 6 A. C. Graham, K. J. Thomas, M. Pepper, N. R. Cooper, M. Y. Simmons, and D. A. Ritchie, Phys. Rev. Lett. 91, 136404 2003 . 7 F. Sgakis, C. J. B. Ford, M. Pepper, M. Kataoka, D. A. Ritchie, and M. Y. Simmons, Phys. Rev. Lett. 100, 026807 2008 . 8 P. J. Simmonds, S. N. Holmes, H. E. Beere, and D. A. Ritchie, J. Appl. Phys. to be published .

Downloaded 18 Apr 2011 to 121.240.208.65. Redistribution subject to AIP license or copyright; see http://apl.aip.org/about/rights_and_permissions