научная статья по теме NEGATIVE DIFFERENTIAL CONDUCTANCE IN INAS WIRE BASED DOUBLE QUANTUM DOT INDUCED BY A CHARGED AFM TIP Физика

Текст научной статьи на тему «NEGATIVE DIFFERENTIAL CONDUCTANCE IN INAS WIRE BASED DOUBLE QUANTUM DOT INDUCED BY A CHARGED AFM TIP»

NEGATIVE DIFFERENTIAL CONDUCTANCE IN InAs WIRE BASED DOUBLE QUANTUM DOT INDUCED BY A CHARGED AFM TIP

A. A. Zhukova* Ch. Volkh'c, A. Windenh>c, H. Hardtdegenhc, Th. Schäpersh'c'd

" Institute of Solid State Physics, Russian Academy of Science 142432, Chernogolovka, Russia

bPeter Grimberg Institut (PGI-9), Forschungszentrum Jülich 52425, Jülich, Germany

c JARA-Fundamentals of Future Information Technology, Forschungszentrum Jülich

52425, Jülich, Germany

dII. Physikalisches Institut, B.WTH Aachen University 52056, Aachen, Germany

Received April 6, 2012

We investigate the conductance of an InAs nanowire in the nonlinear regime in the case of low electron density where the wire is split into quantum dots connected in series. The negative differential conductance in the wire is initiated by means of a charged atomic force microscope tip adjusting the transparency of the tunneling barrier between two adjoining quantum dots. We confirm that the negative differential conductance arises due to the resonant tunneling between these two adjoining quantum dots. The influence of the transparency of the blocking barriers and the relative position of energy states in the adjoining dots on the decrease in the negative differential conductance is investigated in detail.

Regarding the investigation of electronic transport of, one-dimensional systems such as InAs fl 5] and InN fC] nanowires, or carbon nanotubes [7, 8] became the focus of interest. One of the most effective methods to investigate such objects locally is a scanning gate measurement at room temperature [9] and helium temperatures [10, 11]. This technique was also used previously to investigate two-dimensional heterostruc-tures [12 14] and graphene [15]. Furthermore, in scanning gate measurements performed at helium temperatures, the separation of an InAs nanowire in interconnected quantum dots of different sizes was demonstrated [10, 11].

Negative differential conductance (NDC) has been found in quantum dot structures of different types of realizations [16 20,24]. It was observed in quantum dots defined by split gates 011 two-dimensional electron gases, where it has been regarded as a result of the presence of excited states weakly coupled to the main bath [16, 17]. A negative differential conductance has also been found in a linear array of metallic islands [18]. There, the NDC mechanism is based 011 the competi-

* E-mail: azhukovä'issp.ac.ru

tion between the forward rate of injecting charges into a system, which increases with the bias, and the tunneling rate across some junctions, which can be reduced with increasing the bias [18]. A device consisting of a single electron transistor and an electron box attached to it as a gate was measured in [19]. The authors reported a negative differential conductance for a range of conditions. Furthermore, a negative differential conductance was observed in an especially grown InAs p t n light-emitting diode [20].

The aim of this paper is to induce a negative differential conductance in an InAs nanowire. Furthermore, we want to identify the mechanism of negative differential conductance in a sequentially connected double dot structure in a nanowire or nanotube (see [24]) by adjusting the thickness of the tunneling barrier between two adjoining quantum dots by a charged atomic force microscope (AFM) tip. In these experiments, a thick tunneling barrier is necessary to eliminate the strong increase in the current while the energy levels of the quantum dots are tuned out of resonance with increasing the source-to-drain voltage [25]. In Ref. [24], the electrochemical potential could only be changed in both dots simultaneously, whereas in our experiments, we

are able to shift the mutual positions of the energy states of both dots by means of a charged AFM tip. Therefore, our approach might help confirm the theoretical explanation suggested by Frannson and Eriksson [25]. The subject of our work is therefore to realize a negative differential conductance in an InAs wire by using a charged AFM tip and to investigate how this effect is influenced by the tunneling barrier thickness and the relative positions of energy levels of adjoining quantum dots.

The nominally undoped InAs nanowire used in our experiment was grown by selective-area metal-organic vapor-phase epitaxy [22]. The diameter of the wires is 100 nm. For the transport measurements, the nanowi-res are separated from the growth template and subsequently placed on a Si n-type doped (100) wafer covered with an Si02 insulating layer 100 nm thick. The doped silicon substrate is used as a back-gate to vary the electron density in the nanowire by applying a back-gate voltage Vbg- The evaporated ohmic Ti/Au contacts to the wires and the markers of the search pattern are defined by electron beam lithography and lift-off. For the sample studied here, the distance between the contacts is 3.5 fini. A scanning electron microscopy image of the sample under investigation is shown in Fig. la.

All measurements are performed at T = 4.2 K. The charged tip of a home-built scanning probe microscope [23] for scanning gate imaging measurements is used as a relocatable gate. We keep the tip 300 nm above the Si02 surface to eliminate any mechanical or electrical contact of the tip to the nanowire. All scanning gate measurements are performed by keeping the potential of the scanning probe microscope tip (Vt.) and Vbg constant. The electrical scheme of the scanning gate imaging measurements is shown in Fig. 16.

The conductance of the wire during the scan and the conductance as a function of the source-to-drain voltage (Vsd) and Vbg measured in a two-probe scheme. A standard lock-in technique by applying a voltage and measuring the current with a current amplifier is used. The amplitude of the driving AC voltage is Vac =0.1 mV at a frequency of 231 Hz.

If | Vt — VsG | < 50 meV, no influence of the tip on the conductance of the wire is observed for Vbg >5V. At such high back-gate voltages, no dot in the wire is observed. Hence, the whole wire-to-back-gate capacitance is much lager than the wire-to-metallic-contacts ones and the potential of the wire is the same as Vbg in this case. The absence of the influence of the charged tip means that no additional impurity with its own charge is sited on the tip.

Vt

Fig. 1. a) Scanning electron microscope image of the InAs wire. Source and drain contact pads are marked by 'S' and 'D'. Positions of dots A and B are marked by arrows. The scale bar corresponds to 1 /im. The rectangle represents the area of scanning gate measurements at helium temperature, b) Electrical scheme of the measurements. The driving AC voltage is labeled as Vac; the respective tip, gate, and source-to-drain DC voltages are labeled with Vt, Vbg, and Vsd

To localize the closed dot in our wire, a preliminary scanning gate measurement was performed (see Fig. 2 a). This scan was done at Vbg = 1-725 V, Vsd = —8 mV, and Vt = 0 V. The presence of two small dots in the wire is clearly visible. The centers of the dots are labeled by 'A' and 4B' in Fig. 2a. Positions of the dots are defined from the shape of the equipo-tential line and the position of the wire itself. The accuracy of the definition of the dot positions is better than 100 nm and is comparable with the wire diameter. Both dots are formed by potential barriers created by defects of the wire crystal structure. Formation of dots of such a type was observed previously (see [10, 11]).

J Tip

200

- N.

| Source —

[Sc

Back gate

, |Drair

y Tip

b

f

[Sc

J Tip

a b

K,

Back gate

|Drain|

Back gate

|Drair

Fig. 2. a) Scanning gate measurement of the InAs wire made with v = 0, Vbg = 1.725 V, and Vsd = 8 mV. The position of the scanning area is marked with a square in Fig. la. The positions of the centers of dots A and B are marked by squares. The tip position during measurements of the conductance map as a function of Vbg and Vsd (see Fig. 3) is indicated by V. The dashed lines are guides to the eye and follow the equipotential lines of dot A (black) and B (white). The white horizontal bar corresponds to 1 pm. The side color bar represents the scale of the measured wire conductance. Figures b,c show the condition for the alignment of the energy levels of dots A and B for different tip voltages (V/) and source-to-drain voltages (Vsd). To maintain the alignment of the energy levels of both dots, a higher source-to-drain voltage Vsd (Fig. b) < Vsd (Fig. c) must be applied in the case where V (Fig. b) > V (Fig. c) and where the distance from the tip to dot B is smaller than the corresponding distance to dot A. Images d to /demonstrate the modification of the band profile along the InAs wire with increasing V. An increase in V not only results in an increasing opacity of the drain-side blocking barrier (Fig. e) but also opens an additional conductivity channel through the energy state of dot B (Fig. /, dashed arrows)

No additional small and closed dots beyond dots A and B were observed with SGM scanning all over the wire.

For the next set of experiments, we placed the tip in an asymmetric position between the two dots but outside the wire to eliminate any possibility of mechanical or electrical contact between the charged tip and the dot (cf. the asterisk in Fig. 2a). By varying the tip voltage, we can simultaneously alter the opacity of depletion regions of the tunneling barrier separating the dots and change the mutual positions of the energy-levels of the dots. The capacitance of the tip to the closer placed dot is larger than one of the tip to the dot at larger distance. The resulting wire conductance maps as functions of Vsd and Vbg measured at various tip voltages are presented in Fig. 3. A negative differential conductance is clearly visible in the experimental data measured at V¿ = 0,1 V, as can be seen in Fig. 3a.b. Remains of this effect are still observable in Fig. Be (V. = 2 V), while the negative differ

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