Open Access
Add to BibSonomyAbstract
We compare the optical response of wurtzite and zinc blende GaP nanowire arrays for varying geometry of the nanowires. We measure reflectance spectra of the arrays and extract from these measurements the absorption in the nanowires. To support our experimental findings and to allow for more detailed investigations of the optical response of the nanowire arrays than possible in experiments, we perform electromagnetic modeling. This modeling highlights the validity of the extraction of the absorptance from reflectance spectra, as well as limitations of the extraction due to anti-reflection properties of the nanowires. In our combined experimental and theoretical study, we find for both zinc blende and wurtzite nanowires an absorption resonance that can be tuned into the ultraviolet by decreasing the diameter of the nanowires. This peak stops blue-shifting with decreasing nanowire diameter at a wavelength of approximately 330 nm for zinc blende GaP. In contrast, for the wurtzite GaP nanowires, the resonance continues blue-shifting at 310 nm for the smallest diameters we succeeded in fabricating. We interpret this as a difference in refractive index between wurtzite and zinc blende GaP in this wavelength region. These results open up for optical applications through resonant absorption in the visible and ultraviolet wavelength regions with both zinc blende and wurtzite GaP nanowire arrays. Notably, zinc blende and wurtzite GaP support resonant absorption deeper into the ultraviolet region than previously found for zinc blende and wurtzite InP and InAs.
© 2015 Optical Society of America
1. Introduction
There is presently impressive progress in the fabrication of semiconductor nanowires [1–4 ] as well as in the control and characterization of their structural, electrical, and optical properties [5–13 ]. This progress enables the tailoring of nanowire length and diameter, as well as controlled placement of nanowires on the growth substrate [14–16 ]. Furthermore, it is possible to control the crystal phase to be either zinc blende and/or wurtzite for nanowires made from InAs, GaAs, GaP, and InP [9,17–20 ]. The possibility to tailor both the crystal phase and the geometry of nanowire arrays opens up a powerful method to control the optoelectronic characteristics of semiconductors. It is important to note that the wurtzite polytype of these materials does not exist in bulk form where they crystallize in the zinc blende crystal structure [21], with the corresponding refractive indices being well studied through for example ellipsometry on planar samples. By contrast, only very few studies on the linear optical response have been made on the wurtzite polytypes. Wurtzite InP [22] and wurtzite InAs [23] nanowires have been investigated, but no studies have been performed for wurtzite GaP nanowires which is an optoelectronically important material.
We have previously studied the absorption spectra of wurtzite and zinc blende InP [22] and InAs [23] nanowire arrays. For those studies, we used primarily the resonant excitation of the fundamental HE11 waveguide mode of individual nanowires as a probe to compare the refractive index behavior of the zinc blende and wurtzite polytypes. Resonant excitation of the waveguide mode leads to an absorption peak whose wavelength position depends on both the diameter of the nanowires and the wavelength-dependent refractive index of the nanowire material. In contrast, the wavelength for such resonant excitation in the individual nanowires is not expected to depend noticeably on the array period in sparse nanowire arrays [24,25 ]. By systematically designing and varying the diameter of the nanowires, we could show that the zinc blende and wurtzite polytypes of InP and InAs show differences in their refractive indices for short wavelengths. Specifically, the resonant absorption was not possible to tune with decreasing nanowire diameter to below 500 nm for zinc blende InAs, whereas it could be tuned down to 380 nm for wurtzite InAs. Similarly, the resonance could be tuned down to 390 nm for zinc blende InP and down to 345 nm for wurtzite InP. The wavelength at which the shifting stops indicates where the refractive index changes from increasing to decreasing with decreasing wavelength. Such anomalous dispersion in the refractive index on the short-wavelength side prohibits further blue-shifting of the resonance [23].
Due to the similarity between the different III-V semiconductors, we would anticipate a similar crystal-phase dependent resonant response also for GaP which we investigate in this study. Our fundamental results can have application for example in nanowire-based photodetectors and photovoltaics.
We note that the indirect band gap of zinc blende GaP limits its absorption performance and therefore its potential use for low-cost photovoltaics. However, band structure calculations have predicted that, in contrast to zinc blende GaP, wurtzite GaP has a direct band gap [26,27 ]. Such a direct band gap material is a promising candidate for photovoltaics due to its high absorption coefficient in the wavelength region close to the band gap region.
We report measurements and calculations of the absorption spectra of GaP nanowire arrays from 200 to 1100 nm in wavelength (1.1 to 6.2 eV in photon energy). Measurements have been done for both zinc blende and wurtzite GaP nanowires of similar dimensions to investigate differences that originate from differences in the crystal-phase dependent refractive index, without additional effects from varying geometry. We find for both zinc blende and wurtzite nanowires that the absorption peak due to the waveguide resonance can be tuned into the ultraviolet (UV) by decreasing the diameter of the nanowires. This peak stops blue-shifting with decreasing nanowire diameter at a wavelength of λ ≈330 nm for zinc blende GaP. In contrast, for the smallest diameter of 28 nm for wurtzite GaP we succeeded in fabricating, the peak continues to blue-shift at 310 nm. Thus, similarly as for InP [22] and InAs [23], the wurtzite polytype of GaP allows for shifting of the resonance to shorter wavelengths. However, both wurtzite and zinc blende GaP allow the tuning of the resonant absorption deeper into the UV than possible with wurtzite and zinc blende InP or InAs.
Next, we used the well-known refractive index of zinc blende GaP to perform electromagnetic modeling of the zinc blende nanowire arrays. We found good agreement between measured and modeled spectra for zinc blende GaP nanowires. The modeling showed that the nanowires themselves can lead to an anti-reflection property, which can manifest itself as absorption artifacts of increasing strength with increasing nanowire diameter. Thus, our results show how important it is to be aware of the assumptions made when extracting absorption values from reflection measurements. We note that in the present case of GaP, even weak absorption artifacts can draw considerable attention in the long-wavelength region in which absorption is not expected.
2. Experimental
We fabricated a large set of nanowire arrays with varying dimensions for the nanowires using metal organic vapor phase epitaxy (MOVPE) for the nanowire synthesis. We fixed the array period P to 500 nm in a square array and varied the nanowire diameter D and nanowire length L. An AIXTRON 200/4 horizontal setup with a total carrier gas flow of 13 slm was used for zinc blende growth and an Epiquip horizontal setup with a total carrier gas flow of 6 slm for wurtzite growth, both operated at a pressure of 100 mbar. Nanowire growth was carried out following the vapor-liquid-solid growth mode [28]. We tailored both D and L by using Au seeding material of a predefined size range and varied growth times (1 – 3 min for wurtzite and 5 min for zinc blende). The gold seeding material was defined on the growth substrates by an electron beam lithography (EBL) process in regular 100 μm × 100 μm square arrays. Wurtzite GaP nanowire growth was achieved by high temperature growth in the presence of HCl (χ = 1.1 × 10−4) [9,29 ] while GaP zinc blende formation was triggered by in situ use of diethylzinc (DEZn) (χ = 1.6 × 10−5) [30] and HCl (χ = 3.1 × 10−5), respectively. After a 10 min annealing step in a H2/PH3 atmosphere at set temperatures of 650 °C (wurtzite) and 600 °C (zinc blende) the temperature was lowered and set to growth temperatures of 600 °C (wurtzite) and 440 °C (zinc blende) with molar fractions for trimethylgallium of χ = 3.8 × 10−5 (wurtzite) or 2.0 × 10−5 (zinc blende) and PH3 of χ = 6.2 × 10−3 (wurtzite) or 6.9 × 10−3 (zinc blende). Additionally, nucleation times of 15 sec without and 30 sec with HCl before providing DEZn for zinc blende and 15 sec using lower PH3 molar fraction χ = 1.2 × 10−3 for wurtzite, respectively, were used to assure a high yield of vertical nanowire growth.
We determined the values of D and L experimentally for the arrays using scanning electron microscopy in a ZEISS Leo Gemini setup operated with a field emission gun (FEG) at 15 kV. The crystal structure of the nanowires was investigated using transmission electron microscopy in a JEOL 3000F setup operated with a FEG at 300 kV. For each D we fabricated one array of zinc blende GaP nanowires of constant L and three arrays of wurtzite GaP nanowires of varying L. Figure 1 shows SEM images of wurtzite and zinc blende nanowires with diameter and length equal to D/L = 39.8 ± 2.6/1270 ± 37 nm and 34 ± 1.2/1700 ± 42 nm, respectively. We used the NanoDim software [31] to determine the lengths and diameters of the grown nanowires with about 30 to 100 nanowires from each array included for statistics. The arrays are highly uniform, as demonstrated by the small standard deviations in the length and diameter of the nanowires (less than 8 nm for D and less than 178 nm for L). At the bottom of the nanowires we find some lateral growth, in particular for the wurtzite nanowires [9,29 ], giving the appearance of a pedestal for the nanowires [Fig. 1(b)]. This pedestal will affect the optical response of the arrays as detailed below through experiments and modeling. Furthermore, we notice that especially in the array shown in Fig. 1(b), it can be seen that nanowires did not nucleate from every single gold catalyst particle. Also, we see that some nanowires have shifted from their intended position due to slight movement of the gold particles during the annealing step. However, such missing or slightly shifted nanowires will not affect noticeably the resonant response of the HE11 waveguide mode of the individual nanowires.
Fig. 1 30° tilted SEM image of periodic (a) zinc blende and (b) wurtzite GaP nanowire arrays fabricated by metal organic vapor phase epitaxy (inset scale bar 200 nm). Overview and high resolution transmission electron microscopy images of (c) zinc blende and (d) wurtzite nanowires displaying details on the crystal structure.
We measured the reflectance of the arrays using a microscope with a 15X objective of 0.28 in numerical aperture. Light was transmitted through the microscope and the intensity of the reflected light was measured through the same objective as a function of wavelength from a measurement spot of 25 μm × 25 μm in size. The response of the system was calibrated using a planar Si sample.
In general, the absorptance A of the nanowires is related to the reflectance R and transmittance T into the substrate as A = 1 − R − T. However, we measured only R and need therefore an estimate for T. For this estimate, we used the so-called dual-pass approximation [32]. In this approximation, we assume that the nanowires cover a negligible fraction f of the substrate surface, that is, we assume that f << 1. Therefore, we assume that the reflectance of the air/nanowire-array top interface can be ignored. Instead, we assume that the reflectance R originates from light reflected at the nanowire-array/substrate bottom interface. Since f << 1, we assume that the reflectance at the nanowire-array/substrate bottom interface is given by R sub, the reflectance of a planar substrate without nanowires. Finally, we assume that the light is partially absorbed during the dual pass through the nanowire array before giving rise to R. Under these assumptions, T is given by [22]
We use measured values for R and R sub. For Rsub we measured the reflectance spectrum of a planar GaP substrate.First, we note that the fabricated nanowire arrays cover less than f = 0.03 of the substrate surface, as calculated from f = π(D/2)2/P 2. Thus, the assumption of f << 1 in the dual-pass approximation seems to be fulfilled. Furthermore, the substrate surface between the nanowires is very flat according to SEM images [Figs. 1(a) and 1(b)]. Therefore, we could expect that the use of the reflectance Rsub of a planar, nanowire-free substrate applies well in Eq. (1). However, the SEM images also reveal a pedestal, that is, a cone, around the foot of each nanowire [Figs. 1(a) and 1(b)]. The pedestal is noticeable especially in the wurtzite nanowires. Such a pedestal is expected to introduce an anti-reflection behavior to the nanowire arrays. In that case, the use of R sub for the reflectance at the nanowire-array/substrate interface can be questioned, despite f << 1. We show in the results section that such an anti-reflection effect leads to absorption artifacts, even in a wavelength region in which the nanowires are not expected to absorb any light.
3. Calculations
We modeled the absorption in zinc blende GaP nanowire arrays in order to compare with our experimental data. Maxwell’s equations were solved using a scattering matrix method [33]. In this way, we take into account the diffraction and scattering of light from the nanowires and the substrate. We used tabulated data for the refractive index of zinc blende GaP [34] (note that refractive index data is not available for the case of wurtzite GaP) and the experimentally determined geometry (P, L, and D) as inputs for the calculation. We modeled normally incident light since we do not expect strong dependence on the incidence angle within the incidence-cone of half angle of 16° in the experiments (numerical aperture of 0.28). We used periodic boundary conditions in the in-plane direction to model a periodic array of nanowires. The effect of the gold particle at the top of the nanowires was investigated by doing calculations with and without the particle, but we found only negligible differences, which will not affect the conclusions of this paper. We took into account also the pedestal around the foot of each nanowire to investigate the validity of the dual-pass approximation.
4. Results and discussion
Figure 2 shows experimentally obtained absorption spectra of arrays of zinc blende GaP and wurtzite GaP nanowires together with modeled spectra for zinc blende GaP nanowires. Notice that for clarity, each consecutive spectrum in Fig. 2 (a)-(c) has been vertically shifted down by an additional 15% with respect to the spectrum for the thickest diameter. The spectra for both zinc blende and wurtzite nanowires contain an absorption peak which is clearly visible for the larger diameter nanowires. This peak is due to resonant absorption through the HE11 waveguide mode in the individual nanowires of the array [25], and its shape and position is influenced, in addition to the diameter, also by the refractive index of the nanowire material. With decreasing diameter of the nanowires, this peak shifts towards shorter wavelengths and becomes less prominent.
Fig. 2 (a) Absorption spectra of GaP nanowire arrays obtained from electromagnetic modeling of the reflectance of zinc blende nanowires. (b)-(c) Experimentally determined absorption in zinc blende (b) and wurtzite (c) nanowires for varying nanowire diameter. The absorption spectra in (a)-(c) are obtained from modeled (a) or measured (b)-(c) reflection spectra through the dual-pass approximation [Eq. (1)]. In the electromagnetic modeling, we used the specular reflectance R 0 to represent the collected reflection signal (due to the small numerical aperture of 0.28 in the experiments). Thus, we used R 0 for R in Eq. (1) to calculate the estimate for T and in the consecutive calculation of A = 1 – R – T. Notice that for clarity, each consecutive spectrum in (a)-(c) has been vertically shifted down by an additional 15% with respect to the spectrum for the thickest diameter. (d) Wavelength-position of the HE11 absorption resonance, that is, the peak showing up at the longest wavelength (see (b) for an indication of this peak), as a function of nanowire diameter for a nominally constant nanowire length (L ≈1.2 μm for wurtzite and L ≈1.5 μm for zinc blende). The inset shows the real part of the refractive index of zinc blende GaP [34].
Figure 2(d) shows the wavelength-position of this peak as a function of the nanowire diameter for our measured and modelled arrays for a nominally constant nanowire length (L ≈1.2 μm for wurtzite and L ≈1.5 μm for zinc blende samples). From the measurements, we find that the blue-shift of the peak stops for the zinc blende nanowires at λ ≈330 nm. Also the modelling of the zinc blende nanowires shows this “stop” at λ ≈330 nm, in agreement with experiments [Fig. 2(d)]. In contrast, the shifting continues still at λ ≈310 nm for the smallest-diameter wurtzite nanowires. This behavior is similar to that of InP [22] and InAs [23] where the zinc blende and wurtzite nanowires showed the “stop” at very different wavelengths. Note that, as described in detail in the introduction above, the “stop” is expected to show up where the refractive index shows a peak, that is, changes from increasing to decreasing with decreasing wavelength. Indeed, the refractive index of zinc blende GaP shows a peak at λ ≈340 nm [inset of Fig. 2(d)], very close to where the “stop” is detected.
We have previously studied, for InAs nanowires [23], how the resonance wavelength λ peak, nanowire diameter D, and the refractive index Re(n(λ)) are connected to each other through:
Here, c is a parameter that depends on the geometrical shape of the resonant nanostructure. Since, we know the refractive index for the zinc blende III-V materials [inset in Fig. 2(d) for zinc blende GaP], we can determine c from the measurement of λ peak vs. D for the zinc blende nanowires. Previously [23], we determined for the InAs nanowires a mean value of c = 1.49 (with the six obtained values varying between 1.41 and 1.63). Here, we determined c = 1.55 from the 20 zinc blende GaP nanowire arrays. Based on this value for c, we used Eq. (2) to extract a value for Re(n(λ)) for the wurtzite GaP (Fig. 3 ). Notice that we here extract a representative isotropic refractive index for the wurtzite GaP, which is in more detailed studies expected to show anisotropy in Re(n(λ)). The results indicate how the refractive index differs between wurtzite and zinc blende GaP (Fig. 3). Specifically, we see a strong indication that the critical point where the refractive index changes from increasing to decreasing with decreasing wavelength is blue-shifted for the wurtzite polytype, similarly as for InAs [23].Fig. 3 Tabulated refractive index of zinc blende GaP (solid line) together with extracted refractive index for wurtzite GaP (circles). For this approximate extraction, we assumed for simplicity an isotropic optical response for the wurtzite GaP. After this, we used Eq. (2) to relate Re(n(λ peak)) to λ peak and D, with c determined from the zinc blende samples. Notice that each array shows a specific D dependent value for λ peak. Thus, each array gives rise to one extracted value (one circle).
In addition to the diameter dependence, we have also investigated the effect of the nanowire length L on the absorption spectra. Figure 4 shows spectra from three wurtzite GaP nanowire arrays having identical diameters but different lengths (D ≈76 nm and L ≈1.2, 2.0, and 2.7 μm). For increasing length we find that the wavelength position of the HE11 absorption peak does not shift, but the peak becomes broader and less distinct, due to saturated absorption around the peak (the nanowires absorb more light with increasing length since the light travels a longer distance in the absorbing nanowire array) [35]. When studying all of our arrays, there is no noticeable length dependence of the position of the main peak for any diameter as shown by Fig. 4(b).
Fig. 4 (a) Absorption spectra of wurtzite GaP nanowire arrays with different lengths and approximately identical diameter (D ≈76 nm) extracted from measured reflectance spectra through Eq. (1). (b) Wavelength position of the main absorption peak as a function of nanowire diameter for three different nanowire lengths L. (c) Modelled absorption via dual-pass approximation [Eq. (1)] for a zinc blende nanowire array with (dashed-dotted line) and without (dashed line) a pedestal for L = 1.2 µm. (d) Schematics of modeled nanowire arrays with and without a pedestal. Dped is the diameter of the pedestal and Lped is the height of the pedestal. θ ped indicates the angle of the side of the pedestal to the substrate.
However, we notice in Fig. 4(a) that there is apparent absorption in the wurtzite nanowires in the whole wavelength range we probe, even at λ = 1100 nm. Thus, we find apparent absorption in the nanowires far beyond the expected band gap wavelength of GaP (550 nm for zinc blende GaP [36]) at which the nanowires are expected to stop absorbing light. We attribute this apparent absorption at long wavelengths to the anti-reflection property of the pedestal around the foot of each wurtzite nanowire (see Fig. 1(b) for SEM images), which increases in size with increasing growth time, and therefore with nanowire length. We consider the pedestal to be of conical shape. The diameter of the bottom of the pedestal and the height of the pedestal (D ped/L ped) are determined from SEM cross-sectional images to be 156 ± 16/137 ± 16 nm, 178 ± 15/158 ± 23 nm, and 234 ± 27/237 ± 28 nm, respectively, for the nanowires of L = 1.2, 2.0, and 2.7 μm considered in Fig. 4(a). The angle of the side of the pedestal to the substrate (θ ped) is 116 ± 4° [see Fig. 4(d) for a schematic of the geometry with the pedestal].
To investigate the possible impact of the pedestal on the dual-pass assumption [Eq. (1)], we performed modeling with and without the pedestal for L = 1.2 µm [Fig. 4(c)]. Note that we calculated the specular reflectance R 0 in this modeling to represent the collected reflection signal (due to the small numerical aperture of 0.28 in the experiments). Thus, we used R 0 for R in Eq. (1) to calculate the estimate for T and in the consecutive calculation of A = 1 – R – T, assuming that R sub gives the reflectance at the nanowire-array/substrate interface. There is a clear increase in the apparent absorption at long wavelengths when the geometry includes the pedestal. The pedestal provides a smoother change of the refractive index at the nanowire/substrate interface than at the interface between air and the substrate which yields R sub. We notice that a similar reduction of the reflectance has been observed in specifically designed dual-diameter nanowires [37]. In the case of the pedestal, the reflectance at the nanowire/substrate interface R NW/sub is noticeably smaller than R sub, even if f << 1 as calculated from f = π(D/2)2/P 2. Thus, since R NW/sub < R sub, less light is reflected at the nanowire-array/substrate interface than assumed in Eq. (1). This reduced reflection leads to an over-estimation of A, and absorption artifacts can show up even at long wavelengths for which A = 0 is expected. However, such artifacts do not affect the resonant behavior of the HE11 waveguide mode, which we use to probe the refractive index behavior of the wurtzite GaP.
However, we note that even without the pedestal present, there is apparent absorption at long wavelengths [dashed line in Fig. 4(c)]. To understand the origin of this apparent absorption, we note that also the nanowires themselves can provide anti-reflection properties to the nanowire-array/substrate interface. To investigate this, we consider an increase of the diameter of the nanowires in the modeling, without including a pedestal. In Fig. 5 , absorption spectra for nanowire diameters of D = 50, 100, and 150 nm are shown. Here, we show two modeling results for each array. First, we show the dual-pass approximation (solid lines) where we used the modeled R 0 to represent R in Eq. (1) to estimate T and to calculate A = 1 − R – T. Second, we show the results using R and T directly from the modeling (dashed lines), without the use of the dual-pass approximation. Thus, in the second case of the fully modeled A, we do not expect absorption at long wavelengths beyond the band gap wavelength.
Fig. 5 Modeled absorption spectra of zinc blende GaP nanowires for varying diameter; (a) D = 50 nm, (b) D = 100 nm, and (c) D = 150 nm. We show results using the dual-pass approximation (solid line) as well as values for the fully modelled A (dashed lines). Here, for the dual-pass approximation, Eq. (1), we used the specular reflectance R 0 to represent the collected reflection signal (due to the small numerical aperture of 0.28 in the experiments). Thus, we used R 0 for R in Eq. (1) to calculate the estimate for T and in the consecutive calculation of A = 1 – R – T. In the fully modelled A, we take values for both R and T directly from the modelling, without using the dual-pass approximation to determine T from R.
There is a good agreement between the dual-pass approximation and the fully modeled A for D = 50 nm [Fig. 5(a)]. However, already for D = 100 nm [Fig. 5(b)] we find in the dual-pass approximation absorptance values on the order of 7% at long wavelengths, where the fully modeled A shows A = 0 as expected. Furthermore, we find a very noticeable peak around λ ≈500 nm. We attribute this peak to the resonant excitation of the HE11 mode at λ ≈500 nm for D = 100 nm and an efficient coupling of the light through this mode into the substrate. [Note that for D = 100 nm we expect a resonance at λ ≈500 nm when extrapolating the data in Fig. 2(d)]. Thus, we expect this excitation of the HE11 mode to work as an efficient anti-reflection mechanism since zinc blende GaP is expected to absorb weakly at this wavelength (zinc blende GaP shows weakly absorbing indirect transitions for 450 nm < λ < 550 nm [34]). We note that for wavelengths λ < 450 nm, the dual pass approximation gives good agreement with the fully modeled A. When D increases to 150 nm [Fig. 5(c)], we find in the dual-pass approximation higher values for A at long λ than at D = 100 nm, as well as interference oscillations. These oscillations, which increase in amplitude with increasing nanowire diameter, arise from increasing reflection of light at the air/nanowire-array top interface [32], which can interfere with the light reflected at the nanowire-array/substrate bottom interface. We note that also for D = 150 nm, the dual-pass approximation gives for λ < 450 nm good agreement with the fully modeled A (except for slightly stronger interference oscillations at λ < 350 nm). Thus, the dual-pass approximation shows only weak absorption artifacts for D < 100 nm when no pedestal is present. Furthermore, in the wavelength region of λ < 450 nm in which the nanowires can absorb light strongly, the dual-pass approximation shows good agreement with the fully modeled A even for D = 150 nm.
Thus, the nanowires themselves can lead to an anti-reflection property, which can manifest itself as absorption artifacts of increasing strength with increasing nanowire diameter. Thus, our results show how important it is to be aware of the assumptions made when extracting absorption values from reflection measurements. We note that our previous studies on wurtzite and zinc blende InAs [23] and InP [22] were performed on nanowires without considerable pedestal. Furthermore, those studies were performed in a wavelength range in which the nanowire material absorbed light. In those experiments, the absorption in the nanowires could therefore overshadow possible absorption artifacts. By contrast, in the present case of GaP, even weak absorption artifacts can draw considerable attention in the wavelength region in which absorption is not expected.
We note that some reports infer that wurtzite GaP shows a direct band gap [9,26,27 ] in contrast to zinc blende GaP which shows an indirect band gap. Such a difference in the type of the band gap should show up in the absorption properties of the nanowires. However, due to the absorption artifacts in our measurements, we are not able to extract quantitative information about the absorption properties in the vicinity of the band gap, or the band gap type, of wurtzite GaP. Instead, our study reveals that the refractive index of wurtzite and zinc blende GaP is dissimilar in the UV region when λ < 330 nm.
5. Conclusion
In summary, we have studied the optical response of wurtzite and zinc blende GaP nanowire arrays for varying diameter and length of the nanowires. We find that we can tune an absorption peak into the UV range by decreasing the diameter of the nanowires. An increase of the length of the nanowires does not affect the peak position but does increase the absorption. We find that the resonant absorption behavior is different for wurtzite and zinc blende GaP for λ < 330 nm. This behavior is similar to that of InP and InAs for which the two polytypes have very different refractive indices in the UV.
Acknowledgments
We thank Magnus T. Borgström for valuable discussions regarding the crystal phase tuning of GaP nanowires. This work was performed within NanoLund, and was supported by the Swedish Research Council (VR), the Swedish Foundation for Strategic Research (SSF), the EU project NWs4Light under Grant No. 280773 in the framework of the FP7 Program, and the Knut and Alice Wallenberg Foundation.
References and links
1. M. Heurlin, M. H. Magnusson, D. Lindgren, M. Ek, L. R. Wallenberg, K. Deppert, and L. Samuelson, “Continuous gas-phase synthesis of nanowires with tunable properties,” Nature 492(7427), 90–94 (2012). [CrossRef] [PubMed]
2. K. Hiruma, K. Tomioka, P. Mohan, L. Yang, J. Noborisaka, B. Hua, A. Hayashida, S. Fujisawa, S. Hara, J. Motohisa, and T. Fukui, “Fabrication of axial and radial heterostructures for semiconductor nanowires by using selective-area metal-organic vapor-phase epitaxy,” J. Nanotechnol. 2012, 169284 (2012). [CrossRef]
3. R. G. Hobbs, N. Petkov, and J. D. Holmes, “Semiconductor nanowire fabrication by bottom-up and top-down paradigms,” Chem. Mater. 24(11), 1975–1991 (2012). [CrossRef]
4. N. Wang, Y. Cai, and R. Q. Zhang, “Growth of nanowires,” Mater. Sci. Eng. 60(1-6), 1–51 (2008). [CrossRef]
5. N. Anttu, M. Heurlin, M. T. Borgström, M. E. Pistol, H. Q. Xu, and L. Samuelson, “Optical far-field method with subwavelength accuracy for the determination of nanostructure dimensions in large-area samples,” Nano Lett. 13(6), 2662–2667 (2013). [CrossRef] [PubMed]
6. M. H. Huang, S. Mao, H. Feick, H. Yan, Y. Wu, H. Kind, E. Weber, R. Russo, and P. Yang, “Room-temperature ultraviolet nanowire nanolasers,” Science 292(5523), 1897–1899 (2001). [CrossRef] [PubMed]
7. X. Duan, Y. Huang, Y. Cui, J. Wang, and C. M. Lieber, “Indium phosphide nanowires as building blocks for nanoscale electronic and optoelectronic devices,” Nature 409(6816), 66–69 (2001). [CrossRef] [PubMed]
8. J. Bao, D. C. Bell, F. Capasso, J. B. Wagner, T. Mårtensson, J. Trägårdh, and L. Samuelson, “Optical properties of rotationally twinned InP nanowire heterostructures,” Nano Lett. 8(3), 836–841 (2008). [CrossRef] [PubMed]
9. S. Assali, I. Zardo, S. Plissard, D. Kriegner, M. A. Verheijen, G. Bauer, A. Meijerink, A. Belabbes, F. Bechstedt, J. E. M. Haverkort, and E. P. A. M. Bakkers, “Direct band gap wurtzite gallium phosphide nanowires,” Nano Lett. 13(4), 1559–1563 (2013). [CrossRef] [PubMed]
10. K. Pemasiri, M. Montazeri, R. Gass, L. M. Smith, H. E. Jackson, J. Yarrison-Rice, S. Paiman, Q. Gao, H. H. Tan, C. Jagadish, X. Zhang, and J. Zou, “Carrier dynamics and quantum confinement in type II ZB-WZ InP nanowire homostructures,” Nano Lett. 9(2), 648–654 (2009). [CrossRef] [PubMed]
11. R. Agarwal and C. M. Lieber, “Semiconductor nanowires: Optics and optoelectronics,” Appl. Phys., A Mater. Sci. Process. 85(3), 209–215 (2006). [CrossRef]
12. S. L. Diedenhofen, O. T. A. Janssen, G. Grzela, E. P. A. M. Bakkers, and J. Gómez Rivas, “Strong geometrical dependence of the absorption of light in arrays of semiconductor nanowires,” ACS Nano 5(3), 2316–2323 (2011). [CrossRef] [PubMed]
13. Z. Ikonic, G. P. Srivastava, and J. C. Inkson, “Optical properties of twinning superlattices in diamond-type and zinc-blende-type semiconductors,” Phys. Rev. B 52(19), 14078–14085 (1995). [CrossRef] [PubMed]
14. T. Mårtensson, P. Carlberg, M. Borgström, L. Montelius, W. Seifert, and L. Samuelson, “Nanowire arrays defined by nanoimprint lithography,” Nano Lett. 4(4), 699–702 (2004). [CrossRef]
15. A. Dorn, P. M. Allen, and M. G. Bawendi, “Electrically controlling and monitoring InP nanowire growth from solution,” ACS Nano 3(10), 3260–3265 (2009). [CrossRef] [PubMed]
16. S. Koh, “Strategies for controlled placement of nanoscale building blocks,” Nanoscale Res. Lett. 2(11), 519–545 (2007). [CrossRef] [PubMed]
17. S. Lehmann, J. Wallentin, D. Jacobsson, K. Deppert, and K. A. Dick, “A general approach for sharp crystal phase switching in InAs, GaAs, InP, and GaP nanowires using only group V flow,” Nano Lett. 13(9), 4099–4105 (2013). [CrossRef] [PubMed]
18. P. Caroff, J. Bolinsson, and J. Johansson, “Crystal phases in III-V nanowires: From random toward engineered polytypism,” IEEE J. Sel. Top. Quantum Electron. 17(4), 829–846 (2011). [CrossRef]
19. H. J. Joyce, J. Wong-Leung, Q. Gao, H. H. Tan, and C. Jagadish, “Phase perfection in zinc blende and wurtzite III-V nanowires using basic growth parameters,” Nano Lett. 10(3), 908–915 (2010). [CrossRef] [PubMed]
20. T. T. T. Vu, T. Zehender, M. A. Verheijen, S. R. Plissard, G. W. G. Immink, J. E. M. Haverkort, and E. P. A. M. Bakkers, “High optical quality single crystal phase wurtzite and zincblende InP nanowires,” Nanotechnology 24(11), 115705 (2013). [CrossRef] [PubMed]
21. G. Chiarotti and G. Chiarotti, SpringerMaterials; (Springer-Verlag GmbH, Heidelberg, 1993), http://materials.springer.com/lb/docs/sm_lbs_978-3-540-47397-8_6.
22. M. Aghaeipour, N. Anttu, G. Nylund, L. Samuelson, S. Lehmann, and M. E. Pistol, “Tunable absorption resonances in the ultraviolet for InP nanowire arrays,” Opt. Express 22(23), 29204–29212 (2014). [CrossRef] [PubMed]
23. N. Anttu, S. Lehmann, K. Storm, K. A. Dick, L. Samuelson, P. M. Wu, and M. E. Pistol, “Crystal phase-dependent nanophotonic resonances in InAs nanowire arrays,” Nano Lett. 14(10), 5650–5655 (2014). [CrossRef] [PubMed]
24. K. Seo, M. Wober, P. Steinvurzel, E. Schonbrun, Y. Dan, T. Ellenbogen, and K. B. Crozier, “Multicolored vertical silicon nanowires,” Nano Lett. 11(4), 1851–1856 (2011). [CrossRef] [PubMed]
25. N. Anttu and H. Q. Xu, “Efficient light management in vertical nanowire arrays for photovoltaics,” Opt. Express 21(S3Suppl 3), A558–A575 (2013). [CrossRef] [PubMed]
26. A. De and C. E. Pryor, “Predicted band structures of III-V semiconductors in the wurtzite phase,” Phys. Rev. B 81(15), 155210 (2010). [CrossRef]
27. A. Belabbes, C. Panse, J. Furthmüller, and F. Bechstedt, “Electronic bands of III-V semiconductor polytypes and their alignment,” Phys. Rev. B 86(7), 075208 (2012). [CrossRef]
28. R. S. Wagner and W. C. Ellis, “Vapor-liquid-solid mechanism of single crystal growth,” Appl. Phys. Lett. 4(5), 89–90 (1964). [CrossRef]
29. A. Berg, S. Lehmann, N. Vainorius, A. Gustafsson, M. E. Pistol, L. R. Wallenberg, L. Samuelson, and M. T. Borgström, “Growth and characterization of wurtzite GaP nanowires with control over axial and radial growth by use of HCl in-situ etching,” J. Cryst. Growth 386, 47–51 (2014). [CrossRef]
30. R. E. Algra, M. A. Verheijen, L. F. Feiner, G. G. W. Immink, W. J. P. V. Enckevort, E. Vlieg, and E. P. A. M. Bakkers, “The role of surface energies and chemical potential during nanowire growth,” Nano Lett. 11(3), 1259–1264 (2011). [CrossRef] [PubMed]
31. More information available at the home page of NanoDim atwww.nanodim.net.
32. N. Anttu, A. Iqbal, M. Heurlin, L. Samuelson, M. T. Borgström, M. E. Pistol, and A. Yartsev, “Reflection measurements to reveal the absorption in nanowire arrays,” Opt. Lett. 38(9), 1449–1451 (2013). [CrossRef] [PubMed]
33. N. Anttu and H. Q. Xu, “Scattering matrix method for optical excitation of surface plasmons in metal films with periodic arrays of subwavelength holes,” Phys. Rev. B 83(16), 165431 (2011). [CrossRef]
34. A. Borghesi and G. Guizzetti, Gallium Phosphide (GaP), in Handbook of Optical Constants of Solids, E. D. Palik, ed. (Academic, 1997), pp. 445–464.
35. N. Anttu, A. Abrand, D. Asoli, M. Heurlin, I. Åberg, L. Samuelson, and M. Borgström, “Absorption of light in InP nanowire arrays,” Nano Res. 7(6), 816–823 (2014). [CrossRef]
36. I. Vurgaftman, J. R. Meyer, and L. R. Ram-Mohan, “Band parameters for III-V compound semiconductors and their alloys,” J. Appl. Phys. 89(11), 5815–5875 (2001). [CrossRef]
37. Z. Fan, R. Kapadia, P. W. Leu, X. Zhang, Y. L. Chueh, K. Takei, K. Yu, A. Jamshidi, A. A. Rathore, D. J. Ruebusch, M. Wu, and A. Javey, “Ordered arrays of dual-diameter nanopillars for maximized optical absorption,” Nano Lett. 10(10), 3823–3827 (2010). [CrossRef] [PubMed]
