Spatially Resolved Doping Concentration and Nonradiative Lifetime

Apr 1, 2015 - Hark Hoe Tan,. † and Chennupati Jagadish. †. †. Department of Electronic Materials Engineering, and. ‡. Department of Applied Ma...
0 downloads 0 Views 3MB Size
Letter pubs.acs.org/NanoLett

Spatially Resolved Doping Concentration and Nonradiative Lifetime Profiles in Single Si-Doped InP Nanowires Using Photoluminescence Mapping Fan Wang,*,† Qian Gao,† Kun Peng,† Zhe Li,‡ Ziyuan Li,† Yanan Guo,† Lan Fu,† Leigh Morris Smith,§ Hark Hoe Tan,† and Chennupati Jagadish† †

Department of Electronic Materials Engineering, and ‡Department of Applied Mathematics, Research School of Physics and EngineeringThe Australian National University, Canberra, ACT 2601, Australia § Department of Physics, University of Cincinnati, Cincinnati, Ohio 45221-0011, United States S Supporting Information *

ABSTRACT: We report an analysis method that combines microphotoluminescence mapping and lifetime mapping data of single semiconductor nanowires to extract the doping concentration, nonradiative lifetime, and internal quantum efficiency along the length of the nanowires. Using this method, the doping concentration of single Si-doped wurtzite InP nanowires are mapped out and confirmed by the electrical measurements of single nanowire devices. Our method has important implication for single nanowire detectors and LEDs and nanowire solar cells applications. KEYWORDS: Semiconductor nanowires, photoluminescence, doping, nonradiative lifetime, internal quantum efficiency

B

Currently, the most common method for measuring the doping level of individual nanowire is through four-probe measurements and gate measurements with nanowire fieldeffect transistor (NW-FET),17,22,23 where the doping concentration can be estimated from the measured conductivity and field-effect mobility of NWs. However, this technique requires complex nanofabrication processes to make the electrical contacts. Salehzadeh et al.24 used nanoprobe inside a scanning electron microscope for electrical measurements (J−V curve) of individual nanowires in order to extract the doping concentration, but this technique requires a metal-catalyzed nanowire to provide a reliable ohmic contacts. Parkinson et al.25,26 developed a terahertz photoconductivity spectroscopy technique to estimate the doping concentration but it can only be used on NW ensembles and is only sensitive to electron concentrations. Indeed, all these methods cannot be used to measure the doping distribution along an individual NW. Storm et al.27 employed both spatially resolved Hall effect and cathodoluminescence to measure the doping distribution; however this method is time-consuming which requires complicated device fabrication steps. Using secondary ion mass spectrometry (SIMS), Chia et al.28 were able to extract the doping profiles along nanowire but this technique is both time-consuming and complex in sample preparation. More advanced optical methods such as infrared near-field nano-

ecause of their inherent structural characteristics such as high aspect ratio, large surface area-to-volume ratio, and cylindrical shape, which result in unique optical and electrical properties, nanowires (NWs) are of considerable interest for novel nanophotonic applications. Their unique shape allows NWs to be well manipulated by optical tweezers,1,2 used in scanning probe3 microscopy to resolve small features or as an optical cavity to achieve lasing.4,5 Because of increased light trapping by optical scattering,6,7 NWs are currently being investigated vigorously as the basis for next-generation photovoltaic devices. On the basis of efficient strain relaxation as a result of high surface-to-volume ratio, high quantum efficiency, controllable band structure, semiconductor III−V NWs have become promising building blocks for nanoscale devices such as solar cells,8,9 lasers,10 photodetectors,11 waveguides,12 field effect transistors,13 single-electron transistors,14 spin−orbit qubits,15 and light-emitting diodes.16 The ability to precisely dope these nanowires to control their electrical properties is crucial for the operation of these nanoscale devices. Although doping in NWs has been extensively investigated17−20 there are still many challenges in controlling the doping of NWs due to the complexity of NW growth, especially in advanced axial and radial heterostructures.21 Compounding these challenges is the lack of a simple and efficient way of measuring the doping concentration in NWs such as conventional capacitance− voltage spectroscopy and Hall effect measurements, mainly due to the small size of NWs and the difficulty in interpreting the results. © XXXX American Chemical Society

Received: December 22, 2014 Revised: February 28, 2015

A

DOI: 10.1021/nl504929n Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

Figure 1. (a) Schematic diagram of the measurement technique for measuring spatially resolved doping concentration on single NWs. The setup consists of a 2D scanning stage, a 522 nm (frequency doubled) pulsed laser source with pulse width 300 fs and repetition rate 20.8 MHz, external power control, PL system, time-correlated single photon counting system (TCSPC), and a high NA focusing objective. A NW sample is laid horizontally on the scanning surface. Bright-field microscopy is used to visualize the NW sample. TCSPC is used to detect PL decay with a grating spectrometer and a CCD array used to record the spectrum. (b) Mapping of the NW is carried out by moving the stage along the nanowire axis, and for each data point the power dependent PL and PL intensity decay measurements are taken. (c) Simulated laser spot on sample plane with a spot size of 0.305 μm along the x-axis and 0.233 μm along the y-axis.

scopy29 where scattering-type scanning near-field optical microscopy (s-SNOM) is used to achieve visualization of free-carrier concentration distribution with 20 nm spatial resolution. However, these are specialized methods and do not allow a routine measurement of the doping profile across the NW. Internal quantum efficiency (IQE) is a figure of merit for the optical quality of a material and an important parameter for predicting the performance of NW-based optoelectronic devices. For instance, higher IQE will substantially decrease the lasing threshold of a semiconductor laser5 and increase the efficiency of solar cells.30 The popular methods to quantify IQE include power-dependent photoluminescence (PL) at different temperatures 31,32 and time-resolved photoluminescence (TRPL).33 However, the ambiguous assumption, complex setup, and modeling make these methods less likely to be used more generally. Recently Yoo et al.34 have developed a method that, by analyzing the power dependence PL data, the IQE of the sample can be extracted without determining A, B, and C coefficients from the simplified ABC rate equation,35 which provides a simple analysis process and more accurate result. However, this method is limited to undoped samples, which substantially limits its usage. In this work, by combining power-dependent PL and timeresolved PL measurements, we demonstrate a simple method to evaluate the internal quantum efficiency, nonradiative lifetime, and doping concentration of InP NWs. In addition, by using these techniques in a μ-PL setup we are able to spatially resolve these three properties simultaneously along the length of the nanowire. For doping concentration measurements, this contact-free method provides a significant advantage compared to conventional electrical characterization methods where contact pads have to be fabricated onto the NWs. Another advantage of this method is the ability to measure these parameters at room temperature without a high vacuum environment. Finally compared with other optical methods such as PL spectroscopy36 and Raman spectroscopy,37 this method can directly estimate the doping density rather than the relative value and can determine low doping density that may be too small to give a large enough peak shift for PL

spectroscopy and phonon frequency shift for Raman spectroscopy. Using this method, in this work we measured the doping concentration distribution in nominally undoped and Si-doped wurtzite (WZ) InP NWs, as well as NWs doped with variable doping levels during growth. The averaged values were then correlated with electrical measurements of single NWs. The IQE and nonradiative lifetime results provide us with a measure of the crystal quality of the NWs. Experimental Methods. WZ InP NW arrays were grown by selective-area metalorganic vapor phase epitaxy (MOVPE), which has been shown to produce high crystal quality NWs.38,39 A SiO2 mask layer was first deposited on (111)A InP substrates, then patterned by electron beam lithography (EBL) to create hexagonal arrays of circles, followed by chemical etching to selectively remove the SiO2 mask. After etching, the diameter of the circles was 200 nm with a pitch of 800 nm. A thin layer of the exposed InP within the circles was removed by a trim etching step using H3PO4 solution. The patterned substrates were then loaded into an horizontal flow low pressure MOVPE system for NW growth. The nanowires (NW1) were grown at 730 °C for 20 min with trimethylindium and phosphine at a flow rate of 6.1 × 10−6 and 4.9 × 10−4 mol/min, respectively. For Si-doped nanowires (NW2), the dopant SiH4 was introduced during the growth with flow rate of 3.1 × 10−7 mol/min with all other parameters kept constant. For spatially modulated Si-doped nanowires (NW3), the flow rates of SiH4 for undoped, lightly doped, and highly doped regions were 0, 5.1 × 10−8 and 3.1 × 10−7 mol/min, respectively, and the growth time was 7.5 min for each section. Figure 1a shows the experimental setup which incorporates microphotoluminescence spectroscopy and a time-correlated single photon counting (TCSPC) system. In this arrangement, a linearly polarized pulsed laser (frequency doubled to 522 nm with 300 fs pulse width and 20.8 MHz repetition rate) is directed to the back aperture of a high numerical aperture (NA = 0.9) objective (Nikon LU Plan 100×). The pump laser is sampled using a beam splitter and detected by a fast photodiode to determine the arrival time of the laser pulse on the NW. The photoluminescence (PL) signal is collected by a grating spectrometer, thereby either recording the PL B

DOI: 10.1021/nl504929n Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

where n(t) is the time-dependent carrier density, G is the carrier generation rate, τnr is the nonradiative recombination lifetime, B is the radiative recombination rate constant, ND is the doping concentration, and S and d are the surface recombination velocity and diameter of InP nanowires, respectively. Here the Auger recombination rate is ignored due to the small excitation power used in our measurements. For a pulsed laser, it is assumed that carriers are generated as a delta function at t = 0; therefore G is equal to zero (the initial carriers are only induced at t = 0). As shown by Gao et al.,38 the surface recombination velocity is about 161 cm/s, which indicates that the surface-related term 4Sn/d is far less than nonradiative recombination term n/τnr. As a result, the final term in eq 1 can be ignored. Setting τnr and N1= 1/(Bτnr) as the system time and carrier density unit respectively, the unitless rate equation can be rewritten as

spectrum through a charge coupled device (CCD) (Princeton Instruments, PIXIS) or detecting the PL intensity decay using a Si single photon avalanche diode (SPAD) with TCSPC provided by a PicoHarp 300 system with a system response (fwhm) of 45 ps. A neutral density filter wheel is used to control the laser excitation power over four orders-ofmagnitude. NWs dispersed on a Si substrate are then mounted on a two-dimensional (2D) motorized motion stage (Newport) with 200 nm minimum incremental motion. At the laser repetition rate, 1 mW average laser power is equivalent to an excitation pulse energy of 48 pJ. The time between pulses is 48 ns, which is sufficient to ensure that all the photoexcited carriers for one pulse decay completely before the arrival of a second pulse. Figure 1b shows the schematic for mapping the optical properties along the NW, where we use a 2D stage to move the NW with respect to the excitation spot along its main axis. For each excitation point on the NW, power dependence PL measurement and the PL intensity decay at peak wavelength are carried out to extract the doping density and nonradiative lifetime. The initial photoexcited distribution is estimated by the spot size of excitation laser, which is 0.305 μm for the polarized direction and 0.233 μm for the other, according to Figure 1c calculated by vector diffraction theory.40 Figure 2 shows the data used in our method to estimate the doping concentration, IQE, and nonradiative lifetime. This

dn′ = −(1 + nD)n′ − (n′)2 dt ′

(2)

where n′ is the carrier density in units of N1; nD is the doping density in unit of N1; t′ is the time in unit of τnr. Solving this rate equation, the power dependence curve can be expressed as ⎛ 1 + nD ⎞ ⎛ 1 + n0 + nD ⎞ I(P) ∝ nrad = log⎜ ⎟ − log⎜ ⎟ + n0 n0 ⎝ n0 ⎠ ⎝ ⎠ (3)

where nrad represents the carrier density involved in radiative recombination, n0 is the initial carrier density that is proportional to incident laser power. Figure 2a is the powerdependent PL intensity curve from the undoped (blue) and Sidoped (red) nanowires. By fitting these power-dependent curves, the doping concentration in unit of N1 can be extracted. The corresponding IQE can be calculated by applying the fitted doping density, according to IQE = nrad/n0, which is shown in Figure 2b. It is observed that the doped NW gives a higher IQE, even though its nonradiative lifetime lower as indicated in Figure 2c, resulting from increased radiative recombination rate. Using the resultant doping concentration in units of N1, the total carrier decay curve can be calculated from the rate equation as n′(t ′) =

e(1 + nD)t ′ −

+ nD)

n0 1 + nD + n0

(4)

which is proportional to the intensity decay. By fitting this equation to emission intensity decay that is shown in Figure 2c, the nonradiative lifetime can be extracted. For these two samples, τnr is 0.28 and 1.1 ns for the Si-doped and undoped samples, respectively. Here the shorter nonradiative lifetime for doped sample is indication of an inferior crystal quality such as higher point defect concentration. Considering the value of τnr and B, the doping concentration can be calculated. This yields (1.01 ± 0.23) × 1017 and (3.41 ± 0.55) × 1018 /cm3 for undoped and Si-doped samples, respectively. Here B is 1.2 × 10−10 cm3/s according to Levinshtein et al.41 The uncertainty in doping level is estimated from 98% confidence level of the fitting analyses. From these results, we can conclude that doping increases IQE and decreases the nonradiative lifetime as shown in Figure 2b,c. Applying the above optical method within a μ-PL setup, it is possible to probe the doping concentration, nonradiative lifetime and IQE along the NW axis. The doping concentration

Figure 2. (a) PL integrated intensity as a function of power. Fits to these curves are provided by a simplified rate equation from two spatially selected points from one undoped NW and one Si-doped NW. (b) The power-dependent internal quantum efficiency extracted from the fits of the carrier rate equation. (c) The PL intensity decay of these two samples is fitted by a single exponential decay with lifetime of 1.1 and 0.28 ns for undoped and Si-doped NWs, respectively.

method is based on the analysis of carrier rate equation, which is written as dn 4S n =G− − B(ND + n)n − n dt τnr d

n0 (1 1 + nD + n0

(1) C

DOI: 10.1021/nl504929n Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

Figure 3. False colored SEM images of NW1 (undoped) and NW2 (Si-doped) are shown in (a) and (d) respectively, with the color representing the doping concentration. The doping profiles are shown above NWs. The measured nonradiative lifetime along the axis of NW1 and NW2 are shown in (b) and (e), respectively. The measured peak position of the PL spectrum along the axis of NW1 and NW2 are shown in (c) and (f), respectively. The scale bar in (a) and (d) represent a length of 5 μm.

Table 1. Growth Condition and Measured Parameters for Undoped and Si-Doped WZ InP NWsa NW

SiH4 flow rate (mol/min)

averaged doping density1 (1017 cm−3)

averaged doping density2 (1017 cm−3)

averaged τnr (ns)

peak energy (eV)

peak wavelength (nm)

NW1 NW2

0 3.1 × 10−7

1.59 10.91

0.5 5.8

1 0.35

1.416 1.421

875.71 872.61

a

Note: Average doping density1 is measured by optical method claimed by this paper while average doping density2 is measured by the electrical method as shown in Figure 5.

ments. The power dependent PL results of NW2 in Supporting Information Figure S3 show a valence band splitting of the A and B bands. Therefore, it is essential to use spectrum time decay to check whether both A and B bands can be described by single rate equation. The dynamics of carriers in the A and B bands are investigated through time-resolved PL (TRPL) measurements, which are able to capture how the emission changes as carrier density changes with time. Figure 4 shows the time-resolved spectrum from one measurement point on a Si-doped NW (NW2) for low (0.0314 μW), medium (0.524 μW), and high (3.145 μW) excitation power. The two lines on the figures are at 1.423 and 1.459 eV, indicating the transition from the conduction band to the A and B bands, respectively. For low excitation power, as shown in Figure 4a,b, the peak position of spectra lies between these two energy levels and does not shift much within the first 3 ns. This is also observed for medium excitation power, as shown in Figure 4c,d. The result indicates that at room temperature the emission from A and B bands cannot be distinguished. Hence they can be described by single rate equation. This may because thermal energy at room temperature (kBT = 0.0256 eV) is close to the energy difference between band A and B (0.036 eV), which decreases the degree of carrier confinement of these two bands. For higher excitation power, which is shown in Figure 4e,f, the emission peak position is pushed toward the continuum state (1.51 eV) just after the excitation pulse due to the high density of photogenerated carriers, which lifts the Fermi level. After 1 ns, the peak energy shifts back to the energy range between 1.423 and 1.459 eV as a result of carrier renormalization. On the basis of the above TRPL results, within the lifetime of the nanowires, typically 1−1.5 ns for the undoped sample and 0.2 to 0.4 ns for the highly doped sample, the emission can be treated as originating from a single state, which is the mixture of

distribution of an undoped InP NW (NW1) and Si-doped InP NW (NW2) are shown in Figure 3a,d, respectively, where the SEM images of these two NWs are filled with false color to represent the doping concentration according to the measured doping profile results obtained by above method. The scale bar in Figure 3a,d is 5 μm. The results indicate average doping values are 1.59 × 1017 cm−3 and 1.09 × 1018 cm−3 for the undoped and Si-doped NWs as listed in Table1. The doping level for undoped NWs may result from both the surface states that behave as donors25 and impurities during growth.38 The spatially resolved nonradiative lifetimes determined from the fits are shown in Figure 3b,e for the undoped and Si-doped NWs, respectively, where it decreases with doping. The PL peak positions for different position of NW1 and NW2 are shown in Figure 3c,f, respectively. The PL peak position distributions are affected by excitation condition, intrinsic doping distribution, and valence band splitting; there is no linear relationship between PL peak position and doping density. Therefore, the PL peak position here cannot be used to estimate doping density. On the basis of the above measurement, our method is more sensitive to measure the doping density, particularly at low doping level or when the PL spectrum has overlapping peaks. Furthermore, it does not require precise spectrum calibration compared with conventional PL peak fitting and Raman methods. We will now discuss the validity of our method. The premise of this method is that the dynamics of photoexcited carrier can be described using a single rate equation as shown in eq 1. As one of the most studied III−V nanowires, the band structure of WZ InP NWs has been shown to split into three valence bands, namely the A heavy hole band, B light hole band, and C split off band based on many experimental observations such as lowtemperature photoluminescence excitation (PLE),42 photocurrent,43 and photomodulated Rayleigh scattering44 measureD

DOI: 10.1021/nl504929n Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

(NW1) NWs are shown in Figure 5a,b respectively, where the current for NW2 is substantially larger than that for NW1. The extremely low current of NW1 results from large surface depletion region for NWs with low doping density (120 nm for doping 8 × 1016 cm−3) according to45,46 −

⎛ r ⎞ 2εε0Φ ω2 + r0ω − (r0 − ω)2 ln⎜ 0 ⎟ = 2 eND ⎝ r0 − ω ⎠

(5)

where ω is the depletion region, r0 is the radius of the NW, Φ is the surface potential (0.45 V in n-doped InP47), and ND is the doping density. For NW2, the I−V curve can be well fitted with doping concentration of 5.8 × 1017 cm−3 using the COMSOL Multiphysics software. During the fitting, the NW was treated as a cylinder with its two ends covered by ideal Ohmic contacts. Details of the fitting parameters can be found in the Supporting Information. However, this doping concentration is estimated using the mobility value of 120 cm2 V−1 s−1, which is taken from the literature for undoped InP NWs.25 It is known that an increase in doping concentration will result in lower mobility, which indicates that the doping concentration obtained from our electrical measurement is slightly underestimated. The slightly difference between average doping densities obtained from the electrical and optical methods may also be due to the uncertainty of the mobility data and radiative recombination rate constant (B). The value of B used here is 1.2 × 10−10 cm3/ s, which is the reported value for bulk zincblende phase InP41 because the B value for WZ InP could not be found in the literature. We further demonstrate the powerfulness of our optical technique to profile the carrier concentration along the length of nanowires using specially designed samples. The nanowires in this case were grown with different doping profile as shown in Figure 6a. Starting from the bottom of the nanowires, it contains three different sections with decreasing SiH4 flow during growth. The final (fourth) section has a high doping level as with the bottommost section in order to facilitate contact fabrication. The length of each section is nominally the same. The flow rates of SiH4 for undoped, lightly doped and highly doped regions are 0, 5.1 × 10−8, and 3.1 × 10−7 mol/ min, respectively. Figure 6b shows the doping profile on a color scale superimposed on the SEM image of the nanowire. The zero position indicates the middle point of nanowire. The negative position values are in the direction toward the tip of the nanowire, which is flatter, while the positive values are toward the base of the nanowire, which is broken off the substrate during the transfer process. It can clearly be seen that the doping profile as determined by our optical technique follows closely that of the designed structured with the values ranging from 6.34 × 1016 to 9.63 × 1017 cm−3 from the undoped section to heavily doped section. However, rather than a steplike profile, the doping concentration along NW3 appears to have a somewhat smooth profile, most likely due to diffusion of the Si dopant during growth. It can also be noted that the highly doped region at the tip seems to be shorter than the designed value. Most likely this is due to the larger gradient of dopants between the highly doped region to the undoped region, which leads to a more pronounced diffusion. The PL peak energy profile is shown in Figure 6c, where the highly doped sections have larger peak energy, confirming that the doping is n-type36,48 and both of ends have higher doping. We also contacted the NWs on both ends (the highly doped sections) and performed photocurrent mapping measurements.

Figure 4. PL spectra at various times after laser excitation with a power of 0.0314 (a), 0.524 (c) and 3.145 μW (e). The 2D false color maps of the spectrum decay under these three excitation powers are shown in (b,d,f) respectively. The data is from the middle point of a Si-doped NW.

the two hole bands. This in turn allows us use single rate equation to describe the carrier dynamics. In order to verify this technique, we used an electrical method to compare the doping concentration measured in Figure 3 for both NWs. The single NW devices used for I−V measurements were fabricated by electron beam lithography (EBL) with four contact terminals as shown in Figure 5c. The measured I−V curves of the Si-doped (NW2) and undoped

Figure 5. (a) The I−V plot (data points) for (a) the Si-doped NW and (b) the undoped NW after fabricating a four-terminal contact as shown by the SEM image in (c). The curves are fitted to the data point using COMSOL to extract the doping concentration. E

DOI: 10.1021/nl504929n Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

Figure 6. (a) Schematic showing the NW with different doping levels (NW3). The length of each section is nominally the same. (b) False color SEM image of the NW, with color representing the doping concentration. The doping profile is shown above the NW. Note that the top of the NW, which has a flatter surface, is on the left-hand side of the figure. (c) The PL spectrum peak position at different points along the length of the NW. (d,e) The 2D photocurrent mapping results of NW3 under a bias of −1 and +1 V are overlaid on the SEM image of this device. (f) The dark (black) and illuminated (red) current−voltage (I−V) curves of the NW device.

Following the optical measurement of NW3, we use a photocurrent mapping method to estimate the position of the undoped region of NW3. In Figure 6d, the NW3 is contacted by direct laser writing lithography.49 The photocurrent mapping results for a bias voltage of −1 V (the electrical potential being higher at the top of the NW) and +1 V (the bottom of the NW having a higher potential) are shown in Figure 6d,e, respectively, overlaid on the SEM image of this device. The power of the incident laser is 3 μW. This 2D photocurrent mapping results indicate that reasonable ohmic contacts have been made to the NW and photocurrent is mainly generated from the NW.50,51 They also clearly indicate the undoped NW region, which exhibits a higher photocurrent response. More specifically, each doping segment in the NW can be treated as a separate resistor in series. The resistance of the undoped segment dominates the total conductivity without illumination. However, under illumination the local resistance of undoped segment will be substantially reduced while the doped regions will not show such a large effect because of the higher free carrier density. Besides, from Figure 6d,e, the reverse bias (−1 V) provides higher maximum current than that for forward bias (+1 V). This is due to the position of undoped segment in relation to the two contacts. For the undoped segment, holes limit the transport. Thus, for reverse bias the holes are closer to the target contact than that for forward bias, which results in a higher current flow for a fixed voltage. The high photocurrent in region surrounding the undoped region could be due to scattering of the laser beam. The dark (black) and illuminated (red) current−voltage (I−V) curves for the NW device are shown in Figure 6f. The linear and symmetric dark I−V curve confirms reasonable ohmic contacts have been made to the NW. The large photocurrents obtained at both biases can thus be attributed to carrier generated from the undoped NW region. In conclusion we have demonstrated a novel optical method for determining the doping concentration, nonradiative lifetime, and IQE of semiconductor nanowires without the need for making electrical contacts. In addition, we also show that by using a micro-PL setup, these values can be simultaneously profiled along the length of the NWs. The doping concentration obtain by this technique is verified by electrical

measurements. Our technique provides an easy approach to routinely measure the carrier concentration in NWs, facilitating advancement in NW doping calibration and characterizations, which has thus far been one of the main limitations for the true potential of NW devices. During the reviewing process for our work we have noticed a concepttually similar work by Lindgre et al.52



ASSOCIATED CONTENT

S Supporting Information *

Scanning electron microscopy, Section 1. Power-dependent PL measurement of Si-doped WZ InP NWs, Section 2. Electrical simulation parameters for single nanowire, Section 3. Doping measurements of an n-doped InP wafer, Section 4. This material is available free of charge via the Internet at http:// pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Author Contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The Australian National Fabrication Facility is acknowledged for the access to the growth and fabrication facilities used in this work. The Australian Research Council is acknowledged for the financial support. L.M.S. acknowledges the support of the National Science Foundation through Grants DMR-1105362, 1105121, and ECCS-1100489.



ABBREVIATIONS NW, nanowire; FET, field-effect transistor; SIMS, secondary ion mass spectrometry; s-SNOM, scattering-type scanning near-field optical microscopy; IQE, internal quantum efficiency; TRPL, time-resolved photoluminescence; MOVPE, metalorganic vapor phase epitaxy; TCSPC, time-correlated single F

DOI: 10.1021/nl504929n Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

(27) Storm, K.; Halvardsson, F.; Heurlin, M.; Lindgren, D.; Gustafsson, A.; Wu, P. M.; Monemar, B.; Samuelson, L. Nat. Nanotechnol. 2012, 7 (11), 718−722. (28) Chia, A. C. E.; Boulanger, J. P.; LaPierre, R. R. Nanotechnology 2013, 24 (4), 045701. (29) Stiegler, J. M.; Huber, A. J.; Diedenhofen, S. L.; Gómez Rivas, J.; Algra, R. E.; Bakkers, E. P. A. M.; Hillenbrand, R. Nano Lett. 2010, 10 (4), 1387−1392. (30) Brendel, R.; Hirsch, M.; Plieninger, R.; Werner, J. J. H. IEEE Trans. Electron Devices 1996, 43 (7), 1104−1113. (31) Li, X.; Ni, X.; Lee, J.; Wu, M.; Ö zgür, Ü .; Morkoç, H.; Paskova, T.; Mulholland, G.; Evans, K. R. Appl. Phys. Lett. 2009, 95 (12), 121107−121107-3. (32) Watanabe, S.; Yamada, N.; Nagashima, M.; Ueki, Y.; Sasaki, C.; Yamada, Y.; Taguchi, T.; Tadatomo, K.; Okagawa, H.; Kudo, H. Appl. Phys. Lett. 2003, 83 (24), 4906−4908. (33) Neal, T. D.; Okamoto, K.; Scherer, A.; Liu, M. S.; Jen, A. K. Y. Appl. Phys. Lett. 2006, 89 (22), 221106−221106-3. (34) Yoo, Y.-S.; Roh, T.-M.; Na, J.-H.; Son, S. J.; Cho, Y.-H. Appl. Phys. Lett. 2013, 102 (21), 211107. (35) Kioupakis, E.; Yan, Q.; Van de Walle, C. G. Appl. Phys. Lett. 2012, 101 (23), 231107. (36) Bugajski, M.; Lewandowski, W. J. Appl. Phys. 1985, 57 (2), 521− 530. (37) Kawashima, T.; Imamura, G.; Saitoh, T.; Komori, K.; Fujii, M.; Hayashi, S. J. Phys. Chem. C 2007, 111 (42), 15160−15165. (38) Gao, Q.; Saxena, D.; Wang, F.; Fu, L.; Mokkapati, S.; Guo, Y.; Li, L.; Wong-Leung, J.; Caroff, P.; Tan, H. H.; Jagadish, C. Nano Lett. 2014, 14 (9), 5206−5211. (39) Masatoshi, Y.; Eiji, N.; Katsuhiro, T.; Takashi, F. Appl. Phys. Express 2013, 6 (5), 052301. (40) Török, P.; Varga, P.; Laczik, Z.; Booker, G. R. J. Opt. Soc. Am. A 1995, 12 (2), 325−332. (41) Levinshtein, M. E.; Rumyantsev, S. L.; Shur, M. Handbook Series on Semiconductor Parameters. 1. Si, Ge, C (diamond), GaAs, GaP, GaSb, InAs, InP, InSb; World Scientific Publishing Company, Incorporated: New York, 1996. (42) Perera, S.; Shi, T.; Fickenscher, M. A.; Jackson, H. E.; Smith, L. M.; Yarrison-Rice, J. M.; Paiman, S.; Gao, Q.; Tan, H. H.; Jagadish, C. Nano Lett. 2013, 13 (11), 5367−5372. (43) Pemasiri, K. P. S.; Jackson, H. E.; Smith, L. M.; Yarrison-Rise, J. M.; Paiman, S.; Gao, Q.; Tan, H. H.; Jagadish, C., Photocurrent Spectroscopy of ZB and WZ InP Nanowire Ohmic devices. In APS March Meeting 2013. (44) Montazeri, M.; Wade, A.; Fickenscher, M.; Jackson, H. E.; Smith, L. M.; Yarrison-Rice, J. M.; Gao, Q.; Tan, H. H.; Jagadish, C. Nano Lett. 2011, 11 (10), 4329−4336. (45) Richter, T.; Meijers, H. L. R.; Calarco, R.; Marso, M. Nano Lett. 2008, 8 (9), 3056−3059. (46) Casadei, A.; Krogstrup, P.; Heiss, M.; Röhr, J. A.; Colombo, C.; Ruelle, T.; Upadhyay, S.; Sørensen, C. B.; Nygård, J.; Fontcuberta i Morral, A. Appl. Phys. Lett. 2013, 102 (1), 013117. (47) Bertness, K. A.; Kendelewicz, T.; List, R. S.; Williams, M. D.; Lindau, I.; Spicer, W. E. J. Vacuum Sci. Technol., A 1986, 4 (3), 1424− 1426. (48) van Weert, M. H. M.; Wunnicke, O.; Roest, A. L.; Eijkemans, T. J.; Yu Silov, A.; Haverkort, J. E. M.; ’t Hooft, G. W.; Bakkers, E. P. A. M. Appl. Phys. Lett. 2006, 88 (4), -. (49) Parkinson, P.; Jiang, N.; Gao, Q.; Tan, H. H.; Jagadish, C. Nanotechnology 2012, 23 (33), 335704. (50) Maharjan, A.; Pemasiri, K.; Kumar, P.; Wade, A.; Smith, L. M.; Jackson, H. E.; Yarrison-Rice, J. M.; Kogan, A.; Paiman, S.; Gao, Q.; Tan, H. H.; Jagadish, C. Appl. Phys. Lett. 2009, 94 (19), 193115. (51) Persano, A.; Taurino, A.; Prete, P.; Lovergine, N.; Nabet, B.; Cola, A. Nanotechnology 2012, 23 (46), 465701. (52) Lindgren, D.; Hultin, O.; Heurlin, M.; Strom, K.; Borgström, M. T.; Samuelson, L.; Gustafsson, A. Nanotechnology 2015, 26 (4), 045705.

photon counting system; SPAD, single photon avalanche diode; PL, photoluminescence; WZ, wurtzite; CCD, charge-coupled device; SEM, scanning electron microscope



REFERENCES

(1) Reece, P. J.; Toe, W. J.; Wang, F.; Paiman, S.; Gao, Q.; Tan, H. H.; Jagadish, C. Nano Lett. 2011, 11 (6), 2375−2381. (2) Pauzauskie, P. J.; Radenovic, A.; Trepagnier, E.; Shroff, H.; Yang, P.; Liphardt, J. Nat. Mater. 2006, 5 (2), 97−101. (3) Nakayama, Y.; Pauzauskie, P. J.; Radenovic, A.; Onorato, R. M.; Saykally, R. J.; Liphardt, J.; Yang, P. Nature 2007, 447 (7148), 1098− 1101. (4) Piccione, B.; Cho, C.-H.; van Vugt, L. K.; Agarwal, R. Nat. Nanotechnol. 2012, 7 (10), 640−645. (5) Saxena, D.; Mokkapati, S.; Parkinson, P.; Jiang, N.; Gao, Q.; Tan, H. H.; Jagadish, C. Nat. Photonics 2013, 7 (12), 963−968. (6) Hu, L.; Chen, G. Nano Lett. 2007, 7 (11), 3249−3252. (7) Polman, A.; Atwater, H. A. Nat. Mater. 2012, 11 (3), 174−177. (8) Wallentin, J.; Anttu, N.; Asoli, D.; Huffman, M.; Åberg, I.; Magnusson, M. H.; Siefer, G.; Fuss-Kailuweit, P.; Dimroth, F.; Witzigmann, B.; Xu, H. Q.; Samuelson, L.; Deppert, K.; Borgström, M. T. Science 2013, 339 (6123), 1057−1060. (9) Cui, Y.; Wang, J.; Plissard, S. R.; Cavalli, A.; Vu, T. T. T.; van Veldhoven, R. P. J.; Gao, L.; Trainor, M.; Verheijen, M. A.; Haverkort, J. E. M.; Bakkers, E. P. A. M. Nano Lett. 2013, 13 (9), 4113−4117. (10) Li, K.; Sun, H.; Ren, F.; Ng, K. W.; Tran, T.-T. D.; Chen, R.; Chang-Hasnain, C. J. Nano Lett. 2013, 14 (1), 183−190. (11) Wang, J.; Gudiksen, M. S.; Duan, X.; Cui, Y.; Lieber, C. M. Science 2001, 293 (5534), 1455−1457. (12) Ding, Y.; Motohisa, J.; Hua, B.; Hara, S.; Fukui, T. Nano Lett. 2007, 7 (12), 3598−3602. (13) De Franceschi, S.; van Dam, J. A.; Bakkers, E. P. A. M.; Feiner, L. F.; Gurevich, L.; Kouwenhoven, L. P. Appl. Phys. Lett. 2003, 83 (2), 344−346. (14) Thelander, C.; Mårtensson, T.; Björk, M. T.; Ohlsson, B. J.; Larsson, M. W.; Wallenberg, L. R.; Samuelson, L. Appl. Phys. Lett. 2003, 83 (10), 2052−2054. (15) Nadj-Perge, S.; Frolov, S. M.; Bakkers, E. P. A. M.; Kouwenhoven, L. P. Nature 2010, 468 (7327), 1084−1087. (16) Duan, X.; Huang, Y.; Cui, Y.; Wang, J.; Lieber, C. M. Nature 2001, 409 (6816), 66−69. (17) Cui, Y.; Duan, X.; Hu, J.; Lieber, C. M. J. Phys. Chem. B 2000, 104 (22), 5213−5216. (18) Kempa, T. J.; Tian, B.; Kim, D. R.; Hu, J.; Zheng, X.; Lieber, C. M. Nano Lett. 2008, 8 (10), 3456−3460. (19) Minot, E. D.; Kelkensberg, F.; van Kouwen, M.; van Dam, J. A.; Kouwenhoven, L. P.; Zwiller, V.; Borgström, M. T.; Wunnicke, O.; Verheijen, M. A.; Bakkers, E. P. A. M. Nano Lett. 2007, 7 (2), 367− 371. (20) Björk, M. T.; Knoch, J.; Schmid, H.; Riel, H.; Riess, W. Appl. Phys. Lett. 2008, 92 (19), -. (21) Connors, B.; Povolotskyi, M.; Hicks, R.; Klein, B. Simulation and Design of Core-Shell GaN Nanowire LEDs. Proc. SPIE 2010, 75970B−75970B-11. (22) Ford, A. C.; Ho, J. C.; Chueh, Y.-L.; Tseng, Y.-C.; Fan, Z.; Guo, J.; Bokor, J.; Javey, A. Nano Lett. 2008, 9 (1), 360−365. (23) Thathachary, A. V.; Agrawal, N.; Liu, L.; Datta, S. Nano Lett. 2014, 14 (2), 626−633. (24) Salehzadeh, O.; Zhang, X.; Gates, B. D.; Kavanagh, K. L.; Watkins, S. P. J. Appl. Phys. 2012, 112 (9), 094323. (25) Joyce, H. J.; Wong-Leung, J.; Yong, C.-K.; Docherty, C. J.; Paiman, S.; Gao, Q.; Tan, H. H.; Jagadish, C.; Lloyd-Hughes, J.; Herz, L. M.; Johnston, M. B. Nano Lett. 2012, 12 (10), 5325−5330. (26) Parkinson, P.; Joyce, H. J.; Gao, Q.; Tan, H. H.; Zhang, X.; Zou, J.; Jagadish, C.; Herz, L. M.; Johnston, M. B. Nano Lett. 2009, 9 (9), 3349−3353. G

DOI: 10.1021/nl504929n Nano Lett. XXXX, XXX, XXX−XXX