Home

Search

Collections

Journals

About

Contact us

My IOPscience

II–VI nanowire radial heterostructures

This content has been downloaded from IOPscience. Please scroll down to see the full text. 2013 Nanotechnology 24 455603 (http://iopscience.iop.org/0957-4484/24/45/455603) View the table of contents for this issue, or go to the journal homepage for more

Download details: IP Address: 139.184.14.159 This content was downloaded on 27/03/2015 at 01:35

Please note that terms and conditions apply.

IOP PUBLISHING

NANOTECHNOLOGY

Nanotechnology 24 (2013) 455603 (8pp)

doi:10.1088/0957-4484/24/45/455603

II–VI nanowire radial heterostructures K B Kahen1 , Irene A Goldthorpe2 and M Holland3 1

Department of Chemical and Biological Engineering, State University of New York at Buffalo, Buffalo, NY 14260, USA 2 Department of Electrical and Computer Engineering, University of Waterloo, Waterloo, N2L 3G1, Canada 3 Eastman Kodak Company, Research Laboratories, Rochester, NY 14650, USA E-mail: [email protected]

Received 25 May 2013, in final form 28 August 2013 Published 18 October 2013 Online at stacks.iop.org/Nano/24/455603 Abstract There are many reports of ZnSe nanowire synthesis, but photoluminescence measurements on these nanowires indicate weak band-edge and high sub-bandgap defect emission. The two main contributors to the non-optimal photoluminescence are nanowire growth at high temperatures and unpassivated surface states. In this paper, the synthesis of II–VI core–shell nanowires by metal organic vapor phase epitaxy is reported. We demonstrate that larger bandgap shells that passivate the nanowire surface states can be deposited around the nanowires by increasing the partial pressures of the shell reactants without a large increase in growth temperature, allowing high quality material to be obtained. The deposition of nearly lattice-matched ZnMgSSe shells on the ZnSe nanowires increases the band-edge luminescent intensity of the ZnSe nanowires by more than four orders of magnitude and improves the band-edge to defect photoluminescence intensity ratio to 12 000:1. The corresponding full widths at half maximum of the band-edge exciton peaks of the core–shell nanowires can be as narrow as 2.8 nm. It is also shown that magnesium and chlorine can be incorporated into the ZnSe nanowire cores, which shortens the emission wavelength and is known to act as an n-type dopant, respectively.

1. Introduction

suitable for blue emission. ZnSe also has a relatively large exciton binding energy of 20 meV [6]. Several groups have synthesized ZnSe nanowires, however, photoluminescence (PL) measurements indicate weak band-edge and high sub-bandgap defect emission [7–10]. We previously reported the use of a novel metal alloy catalyst, AuSn, to permit the growth of ZnSe nanowires at reduced temperatures where low-defect material can be obtained [11]. However, we found that even these low-defect ZnSe nanowires emitted a broad sub-bandgap emission peak. The main contributors to the non-optimal PL spectra are surface states at which excitons can recombine non-radiatively and also emit sub-bandgap fluorescence. Because nanowires have large surface area-to-volume ratios, surface passivation is critical. The most effective route for passivating the surface states is to grow a shell around the nanowires with a higher band gap material, which results in confining the carriers away from the surface. There are many reports of the synthesis of group IV and III–V core–shell nanowires [2, 12–15]. Other than a brief summary of our work elsewhere [11], to our knowledge

Semiconductor nanowires have shown promise as the active material in optical and optoelectronic devices such as light emitting diodes (LEDs), phosphors, and diode lasers [1–4]. In bulk devices certain emission wavelengths are difficult to obtain due to the lack of appropriate latticematched substrates required for epitaxial thin-film growth. Single-crystalline nanowires, however, can be synthesized through the vapor–liquid–solid (VLS) mechanism without the requirement of a lattice-matched substrate [5]. As a result, single-crystal nanowires of arbitrary materials and alloys can be grown, allowing new wavelengths to be accessed. The wires can be grown dislocation free, enabling highly energy-efficient devices to be formed. Furthermore, inexpensive non-crystalline substrates can be used as well as silicon substrates, which may permit the integration of optical functions with microelectronics. ZnSe is a II–VI semiconductor which has a direct band gap of ∼2.75 eV at room temperature and is therefore 0957-4484/13/455603+08$33.00

1

c 2013 IOP Publishing Ltd Printed in the UK & the USA

Nanotechnology 24 (2013) 455603

K B Kahen et al

there is only one other report of VLS-grown II–VI core–shell nanowires [16]. However, in the latter paper the material quality and thus optical emission properties of the nanowires was poor. In this paper, shells from the ZnMgSSe material system are deposited around ZnSe and ZnMgSe nanowires. It is shown that the deposition of the shells greatly increases both the band-edge PL emission intensity and the peak-to-defect intensity ratio as compared to ZnSe nanowires without shells.

2. Synthesis 2.1. Core–shell nanowire synthesis The core–shell nanowires were grown in a home-built atmospheric pressure, metal organic vapor phase epitaxy (MOVPE) system. It is well known from the thin-film community that the optimal growth temperature for II–VI materials with high optoelectronic quality is between 300 and 350 ◦ C [17–19], with higher growth temperatures leading to unwanted point and extended defects. Gold-catalyzed ZnSe nanowires are generally grown at 550 ◦ C and above because of the high melting temperatures of Au–Zn–Se alloys. To be able to grow the ZnSe nanowire cores at low temperatures, Au0.35 Sn0.65 , which has a lower melting temperature than pure Au, was used as the catalyst [11]. To form the catalysts, a 4.5 nm thick Sn film, followed by a 1.5 nm thick Au film, was thermally evaporated onto the surface of a thermally oxidized Si(001) surface. The precursors of diethylzinc and ditertiarybutylselenide were flowed at 0.4 and 5.6 sccm, respectively. The carrier gas of 8% H2 in He was flowed at 1000 sccm. The growth temperature was 320 ◦ C. The axial growth rate was ∼40 nm min−1 and the average nanowire diameter was 21 nm. To deposit shells around the nanowires, the conditions must be changed such that uncatalyzed film deposition occurs on the nanowire sidewalls. In the literature this most often involves a large temperature increase [13–15]. This is undesirable because high deposition temperatures can [20] (i) promote interdiffusion of the core and shell materials, (ii) provide thermal energy for strain-induced dislocations to form more easily at a lattice-mismatched core/shell interface, and (iii) induce point defects in the core and shell, especially in II–VI materials. There are a few reports, however, where shell deposition was achieved, without a large temperature increase, through the use of different strategies. For example, pulsed-laser deposition was used to deposit polycrystalline shells on chemical vapor deposited silicon nanowires [21], and in another paper diborane was flowed along with a silicon-containing gas which made the latter less stable [12]. These strategies involve extra complexity in the synthesis process or limit the type of shell material that can be deposited. Here we used a strategy where shells can be obtained without a large temperature increase by increasing the flow rates of the shell precursor gases. In our process, the shell deposition temperature was 330 ◦ C, only ten degrees higher than the core growth temperature and still within the range where high optical and electrical quality II–VI material

Figure 1. Schematic illustrating the core–shell growth process. In (a), the precursors are able to diffuse to the catalyst when their partial pressure is low, resulting in lengthening of the nanowire core. In (b), the partial pressures are increased. Many of the column II and column VI precursors now encounter one another before they can move to the catalyst, resulting in uncatalyzed thin-film growth on the nanowire sidewalls and substrate surface.

can be obtained. A low shell deposition temperature is enabled through the use of precursors that permit films to be readily deposited at lower temperatures. Diethylzinc and ditertiarybutylselenide remained as the Zn and Se precursors, respectively, while bis(methylcyclopentadienyl)magnesium and ditertiarybutylsulphide were the Mg and S precursors, respectively. In fact, when bare SiO2 surfaces without metal catalysts were exposed to the same conditions as described above for the growth of the ZnSe cores, polycrystalline ZnSe films were deposited. As such, in the presence of the metal catalysts there is a competitive process between deposition through the catalysts and uncatalyzed thin-film deposition. VLS growth dominates when lower flow rates of the precursors are used. Low flow rates allow more time for the group II and VI precursors to diffuse to and dissolve in the catalysts before they react with one another on the nanowire sidewalls or on the substrate surface (figure 1(a)). The nanowire cores imaged in transmission electron microscopy (TEM) were untapered, indicating little to no deposition on their sidewalls [11]. There was also little deposition on the substrate surfaces as well; having a low-energy substrate surface, such as SiO2 , helps in this regard. Uncatalyzed film deposition can be ‘turned-on’ with a modest increase in temperature and an increase of flow rates. At higher temperatures the precursors are more reactive which increases their ability to deposit without a catalyst, while the higher flow rates of the precursors lead to an increase in their partial pressures. 2

Nanotechnology 24 (2013) 455603

K B Kahen et al

Table 1. Lattice constants and bandgap energies of relevant II–VI semiconductors. Material

Lattice constant (nm)

Bandgap energy (eV)

ZnSe [22] ZnS [22] MgSe [23]

0.568 0.541 0.589

2.7 3.7 3.6

Table 2. Shell deposition conditions that yielded the optimal photoluminescence upon shelling the ZnSe cores. The deposition temperature for all shells was 330 ◦ C and the carrier gas flow rate was 1500 sccm.

Figure 2. Comparison of the PL at 77 K resulting from varying the core Zn to Se flow ratio from 1:14 to 1:15.7, solid and dashed lines, respectively. ZnSeS shells were deposited at 330 ◦ C to passivate these nanowires. The gray dotted line shows the impact of lowering the shell deposition temperature to 320 ◦ C (for a core Zn:Se ratio of 1:15.7).

Precursor flow rates (sccm)

Shell material

Shell number

Zn

ZnSSe

First shell Second shell

1.1 1.1

ZnMgSe

First shell Second shell

0.7 0.6

20.4 30.6

0.64

37.8

ZnMgSSe One shell only

Mg

S

Se

Deposition time (min)

4.8 7.6

5.3 3.7

13 20

9.4 9.4

13 20

3.7

35

7.6

shell temperature of 330 ◦ C (dashed curve), while the defect emission (not shown) is higher by a factor of ∼2. We settled on using a shell deposition temperature of 330 ◦ C because of the lower defect emission, though reasonable shells can also be obtained at 320 ◦ C. Hence, without a temperature increase, higher Zn flows (about a factor of 3) are sufficient to switch from core growth (through the catalyst) to shell growth (which occurs on all surfaces).

More specifically with regard to growth rates, we found that the flow rates of the group II elements (Zn and Mg) control the growth rates of the cores and shells, while adjustments to the chalcogen flows have a minor impact on the growth rate while mainly controlling the overall PL quality (both intensity and degree of defect emission). With respect to the latter effect, figure 2 shows that the PL intensity increases by a factor of ∼3 by raising the Zn to Se ratio in the core nanowire from 1:14 (solid line) to 1:15.7 (dashed line), while keeping all other growth parameters constant. Accordingly, as will be shown in table 2, in order to boost the growth rate of the nanowires, the Zn flow for the shells (taking the case of ZnSeS shells) is a factor of 2.8 larger than that for the cores. It is this higher growth rate which changes the growth mechanism from core growth (figure 1(a)) to shell growth (figure 1(b)). As illustrated in figure 1(b), when the precursors are more concentrated the group II and VI atoms can react with one another on the nanowire sidewalls and substrate surface before a significant amount of material dissolves in the catalysts. The nanowires lengthen axially during shell deposition, but the growth rate is slow (

II-VI nanowire radial heterostructures.

There are many reports of ZnSe nanowire synthesis, but photoluminescence measurements on these nanowires indicate weak band-edge and high sub-bandgap ...
649KB Sizes 0 Downloads 0 Views