научная статья по теме LOW-TEMPERATURE CONDUCTANCE OF THE WEAK JUNCTION IN INAS NANOWIRE IN THE FIELD OF AFM SCANNING GATE Физика

Текст научной статьи на тему «LOW-TEMPERATURE CONDUCTANCE OF THE WEAK JUNCTION IN INAS NANOWIRE IN THE FIELD OF AFM SCANNING GATE»

Pis'ma v ZhETF, vol.93, iss. 1, pp. 13-17

© 2011 January 10

Low-temperature conductance of the weak junction in In As nanowire

in the field of AFM scanning gate

A. A. Zhukov, Ch. Volk+, A. Winden*, H. Hardtdegen*, Th.Schäpers+ Institute of Solid State Physics RAS, 142432 Chernogolovka, Russia + Institute of Bio- and Nanosystems (IBN-1): Semiconductor Nanoelectronics, Juelich, Germany

*Institute of Bio- and Nanosystems (IBN-1) and JARA-Fundamentals of Future Information Technology, Research Centre Jülich,

52425 Jülich Germany

Submitted 1 November 2010 Resubmitted 19 November 2010

We investigate the conductance of the InAs nanowire in the presence of electrical potential created by AFM scanning gate. At helium temperature Coulomb blockade diamonds pattern give the same result for quantum dot sizes ratio as reveals scanning gate imaging. The essential influence of local electrical field direction on the tunneling rate through the weak junction in InAs wire is observed. To explain this behavior the redistribution of the electrons among conductive channels in the wire must be taken into account.

In the past decade an increasing number of publications were dedicated to investigations of electronic transport in semiconductor nanowires. Due to their small size these one-dimensional objects are very promising for future nanoelectronic and nanophotonic applications. InAs wires [1, 2], in particular, are in the focus of interest not only because these wires have surface charges accumulation layer similar to InN wires [3] but also because of the presence of the possibility to alter electronic density and conductivity by a voltage applied to a gate. This allows to construct the field-effect transistor based on InAs wires. Recently, detailed investigations of the dependence of electronic conductance of InAs wire on their diameter have been performed [1, 4]. In Ref. [2] electronic properties of quantum dot made of InAs wire between the InP barriers demonstrating Coulomb blockade is reported. No charging of the single-electron transistor like effect is observed in homogeneous InAs [2] showing absence of Schottky barrier in between InAs wire and metallic contacts [5]. Conductive scanning probe microscopy has been used to investigate distance-dependent electron transport behavior in InAs nanowires at room temperature [6]. Two regimes of transport behavior were observed. For distances less than 200 nm, the resistance was independent of the distance, thus ballistic electron transport was observed. For greater distances, the resistance was observed to increase linearly with distance, as expected for conventional drift transport.

Scanning probe techniques proved themselves as effective methods to study conductivity properties of low-dimensional structures such as 2D quantum dots (QD) [7-9], quantum point contacts [10, 11], QD made of car-

bon nanotubes [12, 13] and praphene [14]. However, so far there is only one work dedicated which investigated the influence of a charged scanning probe tip on conductivity of an InAs wire at cryogenic temperatures [15].

In this paper we study two field-effect transistors based on InAs wires. Both devices contain one weak junction dividing each InAs wire into two serial QD. Measurements of the conductance map of devices as function of source to drain voltage (Vsd) and back-gate voltage (Vbg) allow to estimate sizes of QD. These results will be compared to scanning gate imaging (SGI) which allow to reveal the exact positions of QD under investigation. We investigate the influence of local Coulomb potential altered with charged tip of the scanning probe microscope on tunneling rate through the weak junction as well.

The InAs undoped nanowires were fabricated using metal-organic vapor-phase epitaxy [16]. The diameter of the wires is 50 nm for device A and 60 nm for device B. The wires are placed on a Si n+ (100) wafer capped with a 100 nm thick Si02 insulating layer. The doped silicon substrate serves as the back gate to change electronic density in the wire with applying back gate voltage Vbg- The evaporated Ti/Au contacts to the wires and the markers of the search pattern are defined by electron-beam lithography. The distances in between contacts are 2 /¿m and 1.5 //111 for devices A and B, respectively. A scanning electron microscopy image of device A is presented in Fig. la.

All measurements are performed at 4.2 K. Scanning gate imaging measurements are made with home-built scanning probe microscope [17]. All SGI measurements are performed by keeping the potential of the scanning

ÜHCbMa b ?K3T<J> tom 93 Bbin.1-2 2011

13

14

A. A. Zhukov, Ch. Volk, A. Wînden et al.

V (mV)

1 um i_!_i

Fig.l. (a) Scanning electron microscope image of device A. The metallic stripe above InAs wire is a part of the search pattern marker, (b) SGI of devices A made at 4.2 K. Driving voltage of 10 mV is applied through 10 Mii resistor, source to drain voltage Vsd is measured during the charged tip scanning. A tip voltage of Vt = 1 V and back gate voltage of Vbg = 0 V is maintained. For both images the scale bar corresponds to 1 ¡im.. Dotted line denotes the wire position and dashed lines denote outline of metallic contacts

Vsd (mV)

3 4 VBG (V)

Fig.2. Conductance map of device A measured at 4.2 K as function of source to drain voltage and back gate voltage. Double Coulomb blockade pattern revealing two serial quantum dots in device A is presented at Vbg > 3.3 V

probe microscope tip (Vt) and Vbg constant. Electrical scheme of SGI experiment is presented in paper [13]. Scanning for SGI is performed by keeping the tip 200 nm above Si02 surface to eliminate any mechanical or electrical contact of the tip to the wire under investigation.

The conductance of the wire during the scanning is measured using two probe scheme with standard lock-in technique. In case of current-driven SGI measurement (device A) we measure Vsd while scanning by applying a current through 10 MO resistor. The bias voltage is Vac = 10 mV. In case of voltage-driven SGI measurement (device B) we measure the current while scanning applying voltage Vac = 0.1 mV.

In addition to the SGI measurements the differential conductance of the nanowires was measured as a function of Vsd and Vbg■ Here, a driving voltage of

Vac = 0.1 mV is used while the current is measured with the current amplifier.

In Fig.2 the conduction map as the function of Vsd and Vbg of the device A is presented. Larger diamonds of the Coulomb blockade are present at all values of back gate voltages come from the smaller dot in the wire. As closer look on Fig.2 reveals that at Vbg > 3.3 V an additional Coulomb blockade pattern with smaller period arises. We attribute this additional pattern to the presence of the larger dot in the wire.

Device A revealed a conduction map as function of Vsd and Vbg which is typical for two quantum dots with negligible mutual capacitance [18, 19]; a similar behavior was observed in InAs wires by Bleszynski et al. [15]. This behavior is the result of the small thickness (d = 100 nm) of the silicon oxide insulating layer

Письма в ЖЭТФ том 93 вып. 1-2 2011

Fig.3. (a) Conductance map of device B as function of source to drain voltage and back gate voltage at low back gate voltages. At Vbg < 1-1 V double Coulomb blockade pattern indicates two serial quantum dots formed in device B. (b) Conductance map of device B as function of Vsd and Vbg, both dots are completely open

(d -C compared to the lengths II and Is of the

larger and smaller dots, respectively. From the ratio of Coulomb blockade periods we can estimate the ratio of the QD sizes II/Is ~ 3 because the period of the oscillations of the conductance is inversely proportional to the wire to back gate capacitance and the capacitance itself is proportional to the length of the wire forming the quantum dot.

Fig. lb shows a scanning gate imaging of the device A. The value of the voltage applied to the back gate is Vbg = О V and to the tip Vt = IV. Equipotential lines mapped in the Fig. lb are slightly deformed because of the screening by metallic marker of the search pattern. The center dark region reveals the position of the center of the small quantum dot placed 0.35 /mi apart from the right contact plate. This allows us to estimate ratio of dots sizes from SGI as II/Is ~ 2.

In Fig.3a the conductance as the function of Vsd and Vbg at low back gate voltages ^0.2 < Vbg < 2.3 V of device В is presented. It is clearly visible that at Vbg < 1.1 V double Coulomb blockade diamond pattern is present while at higher back gate voltages only the single one survives. A calculation similar to the one performed for device A results for device В in a quantum dot sizes ratio of II/Is ~ 4. At even higher back gate

Письма в ЖЭТФ том 93 вып. 1-2 2011

voltages Vbg > 5.5 V (cf. Fig.3b) no Coulomb blockade oscillations are present because all dots are open and the nonlinearity in conductance comes from a weak junction splitting nanowire B into two quantum dots.

In Fig.4a series of scanning gate images of device B is presented. The back-gate voltage is maintained a same value for all images (Vbg = — 1 V) keeping the small QD near pinch off regime while the tip voltage is varied from scan to scan. Vt is of 3 V, 4 V and 5 V for scans (a), (b) and (c), respectively. The coincident centers of the equipotential curves reveals the position of the center of the small QD in the nanowire. The ratio of the dot sizes we estimated from the SGI experiment is Il/Is ~ 3. It is worth noting that the conductance along the equipotential lines of the device B strongly depends on the position of the charged tip.

In both devices no special preparation of additional tunneling barriers was made to form the quantum dots. We attribute the presence of the barrier between the InAs wire and the metallic contact to the mediocre quality of the contacts themselves. This barrier is presented according to the SGI measurements in the ot

Для дальнейшего прочтения статьи необходимо приобрести полный текст. Статьи высылаются в формате PDF на указанную при оплате почту. Время доставки составляет менее 10 минут. Стоимость одной статьи — 150 рублей.

Показать целиком