Highly Polarized Photoluminescence and Photodetection from Single Indium Phosphide Nanowires

See allHide authors and affiliations

Science  24 Aug 2001:
Vol. 293, Issue 5534, pp. 1455-1457
DOI: 10.1126/science.1062340


We have characterized the fundamental photoluminescence (PL) properties of individual, isolated indium phosphide (InP) nanowires to define their potential for optoelectronics. Polarization-sensitive measurements reveal a striking anisotropy in the PL intensity recorded parallel and perpendicular to the long axis of a nanowire. The order-of-magnitude polarization anisotropy was quantitatively explained in terms of the large dielectric contrast between these free-standing nanowires and surrounding environment, as opposed to quantum confinement effects. This intrinsic anisotropy was used to create polarization-sensitive nanoscale photodetectors that may prove useful in integrated photonic circuits, optical switches and interconnects, near-field imaging, and high-resolution detectors.

Optical studies of one-dimensional (1D) nanostructures have focused primarily on lithographically and epitaxially defined quantum wires (1–5) embedded in a semiconductor medium. Free-standing nanowires have several attractive differences from these systems, including a large variation in the dielectric constant of the surrounding media and a cylindrical, strongly confining potential for both electrons and holes. Here we report optical studies of individual free-standing InP nanowires that demonstrate giant polarization anisotropy in PL measurements and the use of these InP nanowires as photoconductivity (PC)-based photodetectors. We synthesized single-crystal InP nanowires via a laser-assisted catalytic growth (LCG) method described previously (6–8). Monodisperse nanowire samples with diameters of 10, 15, 20, 30, and 50 nm were suspended in ethanol solution and later were dispersed onto quartz substrates for PL measurements. Atomic force microscopy images show that individual and well-isolated nanowires are readily produced by this method (Fig. 1A). To probe the PL and PC properties of a single InP nanowire, we used a home-built, far-field epifluorescence microscope equipped with a charge-coupled device (CCD) and spectrometer to image and obtain luminescence spectra (9). In this way, we avoid the averaging inherent in ensemble experiments (5, 10).

Figure 1

PL characterization of InP nanowires. (A) Atomic force microscopy image of nanowires dispersed on a substrate for PL measurements, showing that the individual nanowires are monodisperse and well separated from one another on the surface. Scale bar, 5 μm. (B) PL image of a single 20-nm InP nanowire with the exciting laser polarized along the wire axis. Scale bar, 3 μm. (C) PL image of the same nanowire as in (B) under perpendicular excitation. Intensity scale is identical to (B). Inset, variation of overall photoluminescence intensity as a function of excitation polarization angle with respect to the nanowire axis. The PL images were recorded at room temperature with integration times of 2 s.

PL images of single InP nanowires recorded at room temperature with the polarization of the exciting laser parallel (Fig. 1B) and perpendicular (Fig. 1C) to the nanowire show a giant polarization anisotropy. Essentially, the observed PL turns from “on” to “off” as the excitation polarization is rotated from parallel to perpendicular. Integration of the emission as a function of excitation angle shows that the intensity exhibits a periodic (cos2θ) dependence on angle (Fig. 1, inset). From the nanowire PL image (Fig. 1B), it is also evident that the emission intensity is relatively uniform along the wire axis.

We recorded PL spectra (11, 12) from a number of individual wires as a function of excitation (Fig. 2A) or emission (Fig. 2B) polarization. In both cases, the ratio of parallel to perpendicular emission is greater than an order of magnitude. The order-of-magnitude polarization anisotropy is exhibited over most of the energy range of the PL peak (insets of Fig. 2, A and B) and for excitation with both 488- and 514-nm laser wavelengths. On average, the measured excitation and emission polarization ratios, ρ = (I // − I )/(I //+ I ), of the intensities parallel (I //) and perpendicular (I ) to the wire axis are 0.91 ± 0.07. Many nanowires exhibited a polarization ratio of 0.96. The polarization ratio was independent of nanowire diameter between 10 and 50 nm, but radial quantum confinement effects were observed for diameters ≤20 nm.

Figure 2

Polarized excitation and emission spectra of nanowires. (A) Excitation spectra of a 15-nm-diameter InP nanowire. These spectra were recorded with the polarization of the exciting laser aligned parallel (solid line) and perpendicular (dashed line) to the wire axis. The polarization ratio, ρ, is 0.96. Inset, plot of the polarization ratio as a function of energy. (B) Emission spectra of the same wire as in (A). These spectra were taken with the excitation parallel to the wire, while a polarizer was placed in the detection optics. The polarization ratio of the parallel (solid line) to perpendicular (dashed line) emission is 0.92. The spectra were taken with integration times of 10 s. Inset, plot of the polarization ratio as a function of energy. (C) Dielectric contrast model of polarization anisotropy. The nanowire is treated as an infinite dielectric cylinder in a vacuum while the laser polarizations are considered as electrostatic fields oriented as depicted. Field intensities ( E 2 ) calculated from Maxwell's equations clearly show that the field is strongly attenuated inside the nanowire for the perpendicular polarization, E, whereas the field inside the nanowire is unaffected for the parallel polarization,E //.

The ratio of PL intensities (I ///I ) is at least 10 times greater than the ratio for previously reported quantum wire samples (2–4). Polarization anisotropy in quantum wires has been attributed to the mixing of valence bands due to quantum confinement. This quantum mechanical effect yields substantially smaller polarization ratios (ρ < 0.60) than we observed. In our case, the large polarization response can be accounted for in terms of the large dielectric contrast between the nanowire and its air or vacuum surroundings. We have modeled this effect quantitatively by treating the nanowire as an infinite dielectric cylinder in a vacuum, because the wavelength of the exciting light is much greater than the wire diameter (Fig. 2C). When the incident field is polarized parallel to the cylinder, the electric field inside the cylinder is not reduced. But when polarized perpendicular to the cylinder, the electric field amplitude is attenuated according toEmbedded Image(1)where E i is the electric field inside the cylinder, E e the excitation field, and ɛ (ɛ0) is the dielectric constant of the cylinder (vacuum) (13). Using the dielectric constant for bulk InP of 12.4, we calculate a theoretical polarization ratio, ρ = 0.96, which is in excellent agreement with the maximum values determined in our experiments.

These calculations show that classical electromagnetic theory accounts well for the observed polarization anisotropy and suggest that quantum effects do not contribute substantially. Moreover, these results suggest that by tailoring the environment around a nanowire, for example, by adsorption of molecules or growth of inorganic layers with varying dielectric properties, it will be possible to modify systematically the polarization response in a way not possible for surface-grown quantum wires.

The extreme PL polarization anisotropy of these InP nanowires suggests that they could serve as photodetectors, optically gated switches, and light sources, and we have fabricated polarization-sensitive photodetectors in which an individual nanowire serves as the device element (Fig. 3A). Nanowires were dispersed onto silicon substrates, and electrical contacts were defined at the nanowire ends with the use of electron beam lithography and followed by thermal evaporation of the metal electrodes. The nanowire devices were then placed in the epifluorescence microscope used for PL imaging, and the change in conductance (G) of the nanowires was measured via lock-in technique as a function of the laser intensity and polarization.

Figure 3

Polarized photodetection using individual InP nanowires. (A) Schematic depicting the use of a nanowire as a photodetector by measuring the change in PC as a function of incident light intensity and polarization. Inset, field-emission scanning electron microscopy image of a 20-nm-diameter nanowire and contact electrodes for PC measurements. Scale bar, 2 μm. Nanowires were first dispersed in ethanol and then deposited onto silicon substrates (600-nm oxide, 1 to 10 ohm-cm resistivity). Electrical contacts to the wires were defined using electron beam lithography, and Ni/In/Au contact electrodes were thermally evaporated. (B) Conductance,G, versus excitation power density. Shown is the PC response when the illumination is polarized parallel (black) and perpendicular (red) to the wire. Inset, PC anisotropy, σ, versus excitation power calculated from (B). The measured anisotropy for the shown device is 0.96. (C) Conductance versus polarization angle as the polarization was manually rotated while measuring the PC. All PC measurements were done at room temperature. Current collected at drain electrode was measured using standard lock-in techniques, with an excitation voltage of 50 mV at 31 Hz. No gate voltage was applied. An excitation wavelength of 514.5 nm was used for these measurements.

In general, the conductance of individual nanowires increased by two to three orders of magnitude (Fig. 3B) with increasing excitation power density. The PC is reproducible and reversible with respect to changes in the excitation power. This response suggests that the increases are due to direct carrier collection at the nanowire–metal contact interface versus population of surface traps, which would effectively gate the nanowire. The PC also shows a striking polarization anisotropy with parallel excitation producing Gthat is over an order of magnitude larger than perpendicular excitation (Fig. 3B). The photoconductivity anisotropy ratio, σ = (G // − G )/(G // + G ), where G //(G ) is the conductance with parallel (perpendicular) excitation, is 0.96 for the shown device (Fig. 3B, inset), in excellent agreement with the polarization ratio measured from PL. The reproducibility of the PC polarization response is explicitly seen in plots of conductance recorded as the excitation polarization vector is continuously rotated (Fig. 3C). We also note that the PC polarization anisotropy is expected to be nearly wavelength-independent for energies larger than the nanowire band gap. By making a cross of two nanowires (14–16) and independently measuring their PCs, one could make a device to simultaneously measure intensity and polarization. Moreover, the active device element in our nano-photodetector is substantially smaller than other polarization-sensitive quantum well-based detectors (17,18), which are not smaller than 50 μm by 50 μm and are often sensitive to only one wavelength of light.

To gauge the sensitivity of these photodetectors, we determined that the responsivities are as high as 3000 amperes/watt (A/W) as a figure of merit. This represents a high value for our unoptimized device (19), and we believe that investigations of nanowire composition and contacts should lead to further improvements. For example, InP nanowire detectors incorporating Ge traps exhibit responsivities up to 10,000 A/W, although this comes at a cost of reduced detector speed. In addition, these extremely small devices open the possibility of creating high-speed detectors (20,21).

Nanowire photodetectors could be exploited as optically gated switches, used to create very high-density optical interconnects, and incorporated into photonic-based circuits, where polarization detection could vastly increase the information bandwidth. Combined with the ability to synthesize nanowires out of virtually any group IV, III-V, or II-VI semiconductor material (6), we believe that this work opens up exciting opportunities for the creation of a wide range of detectors and high-resolution detector arrays for different spectral regimes, including the 1.5-μm regime important in current optical communications.

  • * These authors contributed equally to this work.

  • To whom correspondence should be addressed. E-mail: cml{at}


View Abstract

Navigate This Article