Tunable p-type doping of Si nanostructures for near infrared light photodetector application

Abstract

In this study, we present a simple oxide assisted p-type doping of Si nanostructures by evaporating a mixed powder composed of SiB₆ and SiO. It was found that Si nanoribbons (Si NRs) which can be obtained at high SiB₆ content, will give way to Si nanowires (Si NWs) when the content of SiB₆ in the mixed powder was reduced. According to our transport measurement of field effect transistors (FETs) assembled on individual Si nanostructures, the as-prepared Si nanostructures with different boron doping levels all exhibit typical p-type conduction characteristics. Additionally, the electrical conductivity of the Si nanostructures can be tuned over 7 orders of magnitude from 8.98 × 10² S cm⁻¹ for the highly doped sample to 3.36 × 10⁻⁵ S cm⁻¹ for the lightly doped sample. We also assembled a nano-photodetector based on monolayer graphene and the as-prepared Si nanostructures, which exhibits ultra-sensitivity to 850 nm near infrared light (NIR) illumination with a nanosecond response speed (τ_rise/τ_fall: 181/233 ns). The generality of the above results suggest that the Si nanostructures are promising building blocks for future electronic and optoelectronic device applications.

---

Introduction

Silicon (Si) with an indirect narrow band-gap (∼1.12 eV), has been central to numerous technological innovations for decades and remains an irreplaceable material for the semiconductor industry.^1,2 In comparison with their thin film and bulk counterparts, one dimensional (1-D) silicon nanostructures with large surface-to-volume ratio including nanoribbons (NR),³ nanowires (NWs),⁴ and nanotubes (NTs) have exhibited various unique properties in terms of light harvesting and carrier transport.^5,6 By this merit, Si nanostructures have been widely used lately as building blocks for constructing various electronic and optoelectronic devices, such as field effect transistors (FETs),^7,8 near infrared light (NIR) photodetectors,^9,10 solar cells,^11–13 lithium ion batteries,^14,15 thermoelectric devices,^16,17 chemical or biological sensors,^18,19 and so on.

Needless to say, the fabrication of 1-D Si nanostructures with controlled diameter, length and electronic properties is essential to the above nano-device applications. Thus far, significant efforts have been devoted to the development of facile and controlled methods for SiNW fabrication in recent years. In principle, there are two basic methodologies for Si nanostructures, i.e., the top-down (metal-catalysed chemical etching,²⁰ and reactive ion etching,²¹ etc.) and bottom-up approaches (oxide-assisted growth (OAG),²² vapour–liquid–solid (VLS) growth,²³ and solid–liquid–solid (SLS).²⁴). In order to tune the electrical properties of Si nanostructures, people normally employed conventional doping method, namely, intentionally incorporating impurity atoms into the host lattice of silicon crystal to control the density of the free charge carriers available in the semiconductor nanostructures, which determines the operation of the Si based nano-devices mentioned above. Take p-type doping for example, Cui et al. initiated the first p-type silicon NWs by developing a laser-assisted catalytic growth method.²⁵ During growth process, boron atoms were incorporated into the Si nanostructures by using diborane (B₂H₆) in the reactant flow. According to their electrical analysis, the boron doped Si NWs had good crystal structure, with tunable electrical property. What is more, Lew found that trimethylboron (TMB) can also be used as efficient p-type dopant source.²⁶ It was found that the as-fabricated Si NWs were predominantly single crystal, with controlled boron concentrations. In spite of these efforts, it is undeniable that these gaseous dopants either in the form of B₂H₆, and B(CH₃)₃ are extremely toxic and flammable, which is environmentally unfriendly. As a result, their further application was greatly restricted. Herein, we present a new synthetic approach to the synthesis of p-type silicon nanostructures by using SiB₆ powder as doping source. Electrical analysis revealed that the conductivity can be tuned over a large order of magnitude by adjusting the SiB₆ content in the mixed precursor. What is more, it was observed that the Schottky junction composed of monolayer graphene/individual Si nanostructures exhibited ultra-sensitivity to NIR irradiation with excellent spectral response, and fast response speed (τ_rise/τ_fall: 181/233 ns). The generality of this study suggests that the present Si nanostructures with controlled electrical property will have potential application for future electronic and optoelectronic devices.

Experimental

Fabrication of Si nanostructures and their microstructures analysis

The preparation of the Si nanostructures was performed on a horizontal tube furnace. In brief, a mixed powder composed of SiB₆ (Analytical grade, Sigma-Aldrich Co.) and SiO (Analytical grade, Aladdin Co.) was placed at the center of the alumina tube, silicon substrates were then placed at the downstream position of the source material. Prior to heating, the system was evacuated to a base pressure of 10⁻⁴ Torr, and back filled with an Ar and H₂ (5%) gas mixture at a constant flow rate of 30 sccm. Afterwards, the source was heated up to 1050 °C at a rate of 20 °C min⁻¹ and then was maintained at this temperature for ∼2 hours. The crystallinity, morphology and structure of the as-prepared Si nanostructures were characterized using X-ray diffractor (XRD, Siemens D-500 using Cu Kα radiation), field emission scanning electron microscopy (FESEM, Philips XL 30 FEG), transmission electron microscopy (TEM, Philips CM 20, operated at 200 kV), and high-resolution transmission electron microscopy (HRTEM, Philips CM 200 FEG, operated at 200 kV). The chemical composition of the product was analyzed by X-ray photoemission spectroscopy (XPS) which was performed on a VG ESCALAV 220i-XL surface analysis system equipped with a monochromatic Al Kα X-ray (1486.6 eV) source.

Devices fabrication and measurement

To fabricate back-gate nano Si-FET, p⁺ Si substrates with 300 nm thick thermally oxidized SiO₂ were cut into small pieces (1.5 × 1.5 cm²), and washed stepwise with absolute alcohol and distilled water under ultrasonication for 10 minutes, respectively. Then Si nanostructures in alcohol which was briefly treated with diluted HF solution, was dropped and dried on the Si substrate with desired density. Two parallel source and drain electrodes (2 nm Ti/58 nm Au) were then defined by photolithography, followed by electron-beam evaporation. Finally, the remaining photoresist was removed by acetone soaking. To fabricate Si nanostructure/monolayer graphene Schottky junction photodetector, the as-synthesized Si nanostructures were parallel dispersed on the substrate by contact print technique, followed by fabrication of single Ti/Au (2/58 nm) electrode on the Si nanostructure via photolithography, electron beam evaporation and lift-off process. Afterwards, monolayer graphene was transferred onto the one head. In order to form a good contact to the graphene, silver paste was finally placed onto the graphene film. A light system including a xenon lamp (150 W) and a monochromator (1/8, Spectra-physics 74000) was used to provide the monochromatic light, which was focused and guided perpendicularly onto the nanowires by a quartz lens. Electrical and optoelectronic measurements were carried out on the semiconductor characterization system (4200-SCS, Keithley Co.) with picoampere resolution. Unless stated, these measurements are all carried out at room temperature. To determine the spectral response and time response of the Schottky devices, a home-built system composed of an oscilloscope (Tektronix, TDS2012B), and a pulsed LED with frequency as high as 1 MHz were used.

Results and discussion

The controlled boron doping of Si nanostructures was realized by a conventional thermal evaporation of SiB₆ and SiO powders in different ratios, which was carried out in a three-temperature-zone tube furnace, as shown in Fig. 1(a). After synthesis, wool-like product (marked as Sample A, B, and C) can be obtained on the substrate by using a mixed powder composed of SiB₆ and SiO in an atomic ratio of 1 : 1, 1 : 2, 1 : 3, respectively (see Fig. S1†). Fig. 1(b) shows the XRD patterns acquired from the three samples, according to which, all five peaks in the range of 20–80° can be readily ascribed to diamond-like cubic phase Si (JCPDF card no. 27-1402),²⁷ and no obvious impurity peaks were observed. What is more, the diffraction signal of the Si nanostructures does not exhibit obvious peak shift or broadening in comparison with the standard Si, suggesting that the incorporation of boron atoms into the crystal lattice cannot alter the crystal quality of the NWs. Fig. 1(c)–(e) display the FESEM images of Sample A, B and C respectively. It is visible that for sample with high boron doping level (Sample A), the product is largely composed of long ribbon-like nanostructures, the length and width of the NR are in the range of 10–30 μm and 150–500 nm, respectively. Interestingly, when we gradually reduce the content of SiB₆ in the mixed power, the as-collected products with the decrease of boron doping level (Sample B and C), is mainly composed of long NWs. Although the three samples (Sample A, B and C) are different in morphology, their growth directions are highly identical to each other. Fig. 1(f) and (g) show the representative TEM and HRTEM analysis of Sample A. From the corresponding lattice resolved image, one can conclude that the growth direction is along [100], in good consistence with the corresponding Fast Fourier Transformation (FFT). To study the precursor dependent morphology of the product, other precursors with controlled compositions were taken into consideration as well. It is found that when only SiB₆ were used as evaporant, Si NRs with obvious contaminant was observed. However, when the contents of SiB₆ were further reduced to 1 : 4 and 1 : 6, tortuous Si NWs with lengths of several micrometers will be obtained (Fig. S2†).

Fig. 1 (a) Schematic illustration of the experimental setup for synthesis of silicon nanostructures. (b) XRD patterns of three samples. (c) SEM image of Sample A which is mainly composed of Si NRs, the inset shows a SEM image at large magnification. SEM images of Sample B (d) and Sample C (e). Typical TEM image of Sample A (f), Sample B (g), and Sample C (h). (i) HRTEM image of Si NRs, the inset shows the corresponding FFT.

Fig. 2 shows a typical XPS survey spectrum of Sample A, in which five peaks at 99.5, 150.7, 189.6, 284.8 and 533.2 eV due to Si 2s, Si 2p, B 1s, C 1s, and O 1s are observed. The signal C 1s is attributable to surface adsorption at ambient condition. What is more, the O 1s peak can be ascribed to SiO, SiO₂ and oxygen due to surface adsorption. According to the B 1s spectra of the three samples shown in the inset, the boron contents in the Si nanostructures (Sample A, and B) are estimated to be 1.3 and 0.037% (mol%), respectively. Moreover, the concentration of boron in the Sample C is too low to be detected. Table S1† shows the standard deviation of boron composition collected from five synthesis, from which, one can see that all samples exhibit small (±1.2%) standard deviation, suggesting that high reliability and reproducibility of such doping strategy. We speculate the modulation of the boron atoms concentration in the Si nanostructures was achieved by a SiB₆ decomposition assisted growth process: as observed in the growth of Si NW via OAG method,²⁸ at high temperature the SiO in the mixed powder will discompose to form SiO₂ and Si through the reaction: 2SiO → Si(g) + SiO₂(g). The Si gas will be transported to the deposition zone, where it will grow in one-dimensional way. In the meantime, the SiB₆ is unstable and it will decompose to form both silicon and boron atoms through the following chemical equation: SiB₆(g) → Si(g) + 6B(g). The resultant boron gas is highly reactive and it can readily incorporate into the crystal lattice of silicon, giving rise to the formation of boron doped Si nanostructures. Note that, in the course of growth, the doping level is largely dependent on the SiB₆ content in the mixed powder, the more the SiB₆, the higher the doping level of silicon nanostructure will be. The doped boron distribution was determined by the EDS elemental mapping. As shown in Fig. 3(c) and (d), the boron element is uniformly distributed in the Si NRs. Note that the mapping profile of boron atoms from Sample B can hardly be distinguished due to low doping level (see Fig. 3(h)).

Fig. 2 Survey XPS spectrum of the Sample A, the inset shows the B 1s spectra of Sample A, B, and C.

Fig. 3 (a) TEM image of a boron doped Si NR from Sample A. (b) Magnified TEM image of the Si NR in bright field. (c) and (d) are the corresponding EDS elemental mapping for Si and B, respectively. (e) TEM image of a boron doped Si NR from Sample B. (f) Magnified TEM image of the Si NR in bright field. (g and h) are the corresponding EDS elemental mapping for Si and boron, respectively.

Individual nanostructure based filed-effect transistors (FETs) were fabricated to study the electrical transport properties. Fig. 4(a) shows I_ds vs. V_ds curves of a FET assembled from a single highly doped Si NR (Sample A). The obvious linear curves insinuate that Ohmic contact was formed between the Ti/Au electrodes and the Si NR. In addition, the device exhibits a very weak p-type gating effect: that is, when V_g decreases (increases), the conductance increases (decreases) a lit bit (Fig. 4(b)). For such a NR based FET device, the hole mobility (μ_h) of the device can be calculated by using the equation of μ_h = g_mL/ZC₀V_ds (ref. 29 and 30) where g_m = dI_ds/dV_gs is the linear-region channel transconductance, which is actually the slope of the I_ds/V_gs shown in Fig. 4(b), Z/L is the width-to-length ratio of the channel (300 nm/3000 nm), C₀ is the capacitance of the back gate, and can be given by C₀ = ε_rε₀/h,³¹ where ε_r is the dielectric constant for SiO₂ layer (3.9), ε₀ the vacuum permittivity (8.85 × 10⁻¹⁴ F cm⁻¹), and h the thickness of the SiO₂ layer (300 nm). Based on the above constants and equations, the transconductance and hole mobility are estimated to be 3.0 × 10⁻⁸ S and 8.7 cm² V⁻¹ s⁻¹, respectively. In addition, the hole concentration (p) can be described by the relation of σ = pqμ_h, where q is the elementary charge (1.6 × 10⁻¹⁹ C), σ is conductivity of the Si NR. The conductivity can be estimated by the formula of IL/UA, where I is the current (1.616 × 10⁻⁶ A), L is the length of the NR (3 × 10⁻⁴ m), U is the applied voltage (3 V), A is the cross-section area of the NR (300 × 60 nm²). Based on these values, the conductivity and hole concentration are estimated to be 8.977 × 10² S cm⁻¹ and 6.45 × 10²⁰ cm⁻³, respectively.

Fig. 4 Electrical characteristics of Si nano-FETs. (a) I_ds–V_ds of a NR FET at different V_g. The inset shows a representative SEM image of the single NR FET. (b) Transfer characteristics of the FET at V_ds of 3 V. (c) I_ds–V_ds curves at different V_g of Sample B, inset in (c) is the SEM image of the NW FET. (d) The corresponding I_ds vs. V_g at V_ds of 3 V. (e) I_ds–V_ds curves at different V_g of Sample C, inset in (e) is the SEM image of the NW FET. (f) The corresponding I_ds vs. V_g at V_ds of 3 V.

Unlike the device made from Sample A, much stronger p-type gating effect was observed on nano-FET made from Sample B and C (cf. Fig. 4(c)–(f)). The semilogarithmic plots of I_ds vs. V_g at a constant V_ds = 3 V in the Fig. 4(d) and (f) reveal an on/off ratios as large as 6 × 10³ for Sample B, and 8 × 10² for Sample C, respectively. The carrier mobility of the NW from Sample B and C cannot be estimated by the above equation, considering the fact that the building blocks has changed from NRs to NWs in morphology.³² Assuming a cylinder on an infinite plate model for the Si NW FET (Sample B & C), the channel capacitance of the back gate can be estimated by the formula of C = 2πε₀ε_rL/ln(4h/d),³³ where ε₀ and ε_r are the vacuum permittivity (8.85 × 10⁻¹⁴ F cm⁻¹) and effective dielectric constant of SiO₂ (3.9), respectively, h the thickness of the dielectric layer, L the channel length, and d the radius of the Si NW. By fitting the linear part of the I_ds–V_g curve, the transconductance (g_m) is determined to be 1.37 × 10⁻⁷ S for Sample B, and 2.25 × 10⁻¹¹ S for Sample C. Moreover, the hole mobility is calculated to be 15.7 cm² V⁻¹ s⁻¹ (Sample B), and 2.59 × 10⁻³ cm² V⁻¹ s⁻¹ (Sample C), respectively. By using the relation of σ = pqμ_h mentioned above, the hole concentrations are estimated to be 3.34 × 10¹⁷ cm⁻³ and 8.11 × 10¹⁶ cm⁻³ (Sample C) from the relation: μ_h = g_mL²/CV_ds.

Fig. 5(a) compares the typical I–V curves of Si nanostructure from Sample A, B and C. It is revealed that the conductivity can be readily tuned over a large order of magnitude by adjusting the SiB₆ content in the precursor. To statistically evaluate the electrical conductivity of the boron doped Si nanostructures, we have in total fabricated and analyzed 30 FETs, with 10 devices from each sample (Fig. 5(b)). Obviously, the conductivity of highly doped Si NR (Sample A) is very high and in the range 500–1200 S cm⁻¹. But when the doping level was gradually reduced by decreasing the SiB₆ content in the mixed precursor, the conductivity will substantially decreases to 0.5–1 S cm⁻¹ for Sample B, and to 0.30–1.0 × 10⁻⁵ S cm⁻¹ for Sample C. This result signifies that such a doping method using SiB₆ as dopant could provide an alternative doping approach for tuning the electrical conductivity of Si nanostructures. Table 1 compares the SiB₆/SiO ratio dependent electrical property of the Si nanostructures. It can be seen that the hole concentration increases continuously with increasing SiB₆ content in the mixed precursor, which suggests that the carrier concentration in the Si nanostructures can be controlled by adjusting the ratio of the precursor. However, the hole mobility increases at first from 2.57 × 10⁻³, to 15.7 cm V⁻¹ s⁻¹, and then decreases to 8.7 cm V⁻¹ s⁻¹ at high doping level. Understandably, this phenomenon is due to enhanced carrier scattering of impurity atoms in heavily doped Si nanostructures.³⁴

Fig. 5 I–V characteristics of Si NWs (Sample B & C) in the dark. The below right inset shows the I–V characteristics of Si NR (Sample A) in the dark. The above left inset shows the distribution of electrical conductivity for 30 FETs from Sample A, B and C nanostructures, ten FETs for each sample.

Table 1 Summary of the key parameters of the FETs based on Si nanostructures with different doping levels

Sample no.	SiB6/SiO	gm (nS)	σ (S cm^−1)	μp (cm^2 V^−1 s^−1)	p (10^17 cm^−3)
Sample A	1 : 1	30	8.98 × 10^2	8.7	6450
Sample B	1 : 2	137	8.39 × 10^−1	15.7	3.34
Sample C	1 : 3	2.25 × 10^−2	3.36 × 10^−5	2.59 × 10^−3	0.81

---

Next, three kinds of silicon nanostructures/monolayer graphene heterojunction were assembled from Sample A, B, and C (cf. Fig. 6(a)) in order to explore the potential for optoelectronic devices application. The monolayer graphene (MLG) was confirmed by the Raman analysis in Fig. S3.† The inset in Fig. 7(a) shows the I–V characteristic of heterojunction composed of the Si NR (Sample A) and MLG. Obviously, the current increases linearly with the bias voltage in the range from −5 to 5 V. This characteristic is completely different from that of Schottky junctions made from Sample B and C, in which typical rectifying behavior was observed. Such a rectifying effect can be exclusively ascribed to the graphene/Si NW considering Ohmic contacts were formed at both the graphene/silver paste and Ti/Au electrode/Si NW interfaces (Fig. S4†). Remarkably, once illuminated by NIR with wavelength of 850 nm, both Schottky junctions made from Sample B and C displayed pronounced photoresponse. As shown in Fig. 7(b), the electrical current increased dramatically and stabilized at a low-resistance “on” state upon light irradiation, but decreased quickly to a high-resistance “off” state when the NIR was turned off, yielding an on/off ratio as high as 10⁵. To quantitatively assess the sensitivity of the present device to NIR illumination, the specific detectivity (D*) can be estimated by the formula of D* = (AR/2qI_d)^1/2,³⁵ where A is the active area of the device, q is the elementary charge, R is the responsivity of the device, which can be described by R = (I_p − I_d)/P_opt, where I_p, I_d and P_opt represent the photocurrent, the dark current and the incident light power, respectively. Based on the these values, the R and D* are estimated to be 8.7 A W⁻¹ and 7.94 × 10¹² cm Hz^1/2 W⁻¹ for Sample B, and 47.3 mA W⁻¹ and 1.97 × 10¹¹ cm Hz^1/2 W⁻¹ for Sample C at zero bias, respectively.

Fig. 6 (a) SEM image of the monolayer graphene/Si nanostructure Schottky junction device. (b) Schematic illustration of the photoresponse measurement setup.

Fig. 7 (a) Comparison of I–V curves of the graphene/NWs from both Sample B and C, the above left inset shows the I–V curves of Sample B in the dark and under light illumination, the below right inset shows the I–V curve of MLG/NW from Sample A. (b) Response of the hetero-junction device to the pulsed photo-irradiation at bias voltage of 0 V, the light wavelength was 850 nm, the powder intensity of the illumination was ∼9.8 × 10⁻³ W cm⁻². (c) Spectral response of the MLG/Sample B NW Schottky junction device.

The device performance of the present device and other photodetector with similar structures is summarized in Table 2. Although both detectivity does not show any obvious advantage, the on/off ratio and responsivity are as high as 10⁷ and 8.7 A W⁻¹, respectively, the highest values in comparison with other Si nanostructures based photodetector ever reported. Moreover, Fig. 7(c) shows the photoresponse of the Gr/Sample B Schottky junction photodetector to varied illumination with different wavelengths (to make this study more reliable, the light intensity was kept identical during test). It is visible that the device exhibits good spectral selectivity, with peak sensitivity at ∼960 nm.

Table 2 Comparison of the device performance of the current Si nanostructures/MLG Schottky junction with other Si bulk or nanostructures based NIRPDs

Device structure	Responsivity	τr/τf	Ilight/Idark	Detectivity	Ref.
Individual Si NW/MLG	8.7 A W^−1	181/233 ns	10^7	7.94 × 10^12	This work
Si NWs array/AuNPs@ graphene	1.5 A W^−1	73/96 μs	10^6	1.09 × 10^14	10
Si NWs/carbon QDs	1.73 mA W^−1	20/40 μs	3 × 10^3	3.79 × 10^9	36
Si bulk/graphene NIRPD	562 mA W^−1	93/110 μs	1.1	3.90 × 10^11	37
n-type Si/graphene	435 mA W^−1	—	3 × 10^3	7.69 × 10^9	38

---

In addition to the high sensitivity and good spectral selectivity, the present device enjoys a slow relative balance ((I_max − I_min)/I_max) decay when irradiated by pulsed illumination with frequency in the range from 1 to 1 MHz. In this study, the pulsed optical signal was generated by a light emitting diode driven by a signal generator with frequency as high as 20 MHz. Fig. 8(a) plots the relative balance as a function of switching frequency, from which one can see that the relative balance is less than 1% when the switching frequency reaches 1 × 10⁴ Hz. Notably, even the frequency was increased to 1 MHz, the relative balance decreases by less than 10%. This value, to the best of our knowledge is the highest ever reported, which suggests great potential for monitoring optical signal with ultra-high switching frequency.³⁹ It should be noted that, our Si NW/MLG Schottky junction can work properly after long-term storage at ambient condition. As shown in Fig. 8(b), the NIR photodetector can almost retain the same photocurrent after one month storage. In general, the response speed of a photodetector is often evaluated by the rise time of its response to a pulsed signal. From the magnified photoresponse curve in Fig. 8(c), the rise time (τ_rise) and fall time (τ_fall) can be estimated to be 181 and 233 ns, respectively, which are far faster than other Si nanostructures based NIR photodetectors (Table 2). Two factors are considered important to this ultra-fast response speed. (1) The specific geometry of the device. In this study, the three-dimensional core–shell configuration is capable of quickly separating and collecting of the carrier by increasing the Si NW/MLG interfacial area and allowing for the diffusion of photo-generated carriers toward the depletion layer within a small distance. (2) The high-quality of the hetero-junction. Due to the formation of good Schottky junction, a built-in electric field was formed. As a result, the photo-generated carriers will be effectively and rapidly separated. This fast response speed, along with the excellent device stability renders the current device promising candidate for future nano-optoelectronic devices application.

Fig. 8 (a) Relative balance of both MLG/Sample B, and MLG/Sample C photodetector. (b) Photoresponse of the MLG/Sample B to pulsed near infrared light with frequency of 1 MHz. (c) A single normalized cycle of the MLG/Sample B Schottky junction photodetector for estimating both rise time and fall time.

Conclusions

In this work, we reported on a synthetic approach to silicon nanostructures with controlled boron doping levels by a thermal evaporation of a mixed powder of SiB₆ and SiO. It is revealed that the morphology of the product is mainly determined by the content of the SiB₆ in the mixed powder. Field effect transistors (FETs) analysis shows that the as-prepared Si nanostructures all exhibit typical p-type conduction behaviour. The hole mobility and concentration can be readily tailored by the adjusting the content of SiB₆. It is also found that by tuning the content of the SiB₆ in the mixed powder, the electrical conductivity of the Si nanostructures can be tuned over 7 orders of magnitude from 8.98 × 10² S cm⁻¹ for the highly doped sample to 3.36 × 10⁻⁵ S cm⁻¹ to lightly doped sample. What is more, we observed that the Schottky junction composed of monolayer graphene/individual Si nanostructures exhibit ultra-sensitivity to 850 nm NIR irradiation with excellent spectral response, and fast response speed (τ_rise/τ_fall: 181/233 ns). We believe the present Si nanostructures with controlled electrical property will have potential application for future electronic and optoelectronic devices.

Acknowledgements

This work was supported by the Natural Science Foundation of China (NSFC, nos 51172151, 21101051), the Fundamental Research Funds for the Central Universities (2012HGCX0003, 2013HGCH0012, 2014HGCH0005), the China Postdoctoral Science Foundation (103471013), and the Natural Science Foundation of Anhui Province (Grant no. J2014AKZR0036).

Notes and references

1. S. M. Sze, Physics of Semiconductor Devices, John Wiley & Sons, New York, 1981.
2. Y. Cui and C. M. Lieber, Science, 2001, 291, 851–853.
3. W. Shi, H. Peng, N. Wang, C. P. Li, L. Xu, C. S. Lee, R. Kalish and S. T. Lee, J. Am. Chem. Soc., 2001, 123, 11095–11096.
4. J. S. Jie, W. J. Zhang, K. Q. Peng, G. D. Yuan, C. S. Lee and S. T. Lee, Adv. Funct. Mater., 2008, 18, 3251–3258.
5. E. Garnett and P. D. Yang, Nano Lett., 2010, 10, 1082–1087.
6. L. B. Luo, X. B. Yang, F. X. Liang, H. Xu, Y. Zhao, X. Xie, W. F. Zhang and S. T. Lee, J. Phys. Chem. C, 2011, 115, 18453–18458.
7. J. Goldberger, A. I. Hochbaum, R. Fan and P. D. Yang, Nano Lett., 2006, 6, 973–977.
8. C. Celle, A. Carella, D. Mariolle, N. Chevalier, E. Eouviere and J. P. Simonato, Nanoscale, 2010, 2, 677–680.
9. O. Hayden, R. Agarwal and C. M. Lieber, Nat. Mater., 2006, 5, 352–356.
10. L. B. Luo, L. H. Zeng, C. Xie, Y. Q. Yu, F. X. Liang, C. Y. Wu, L. Wang and J. G. Hu, Sci. Rep., 2014, 4, 3914.
11. K. Q. Peng and S. T. Lee, Adv. Mater., 2011, 23, 198–215.
12. L. Tsakalakos, J. Balch, J. Fronheiser, B. A. Jorevaar, O. Sulima and J. Rand, Appl. Phys. Lett., 2007, 91, 233117.
13. M. D. Kelzenberg, D. B. Turner-Evans, B. M. Kayers, M. A. Filler, M. C. Putnam, N. S. Lewis and H. A. Atwater, Nano Lett., 2008, 8, 710–714.
14. C. K. Chan, H. L. Peng, G. Liu, K. Mcllwrath, X. F. Zhang, R. A. Huggins and Y. Cui, Nat. Nanotechnol., 2008, 3, 31–35.
15. A. Magasinski, P. Dixon, B. Hertzberg, A. Kvit, J. Ayala and G. Yushin, Nat. Mater., 2010, 9, 353–358.
16. A. I. Boukai, Y. Bunimovich, J. Tahir-Kheli, J. K. Yu, W. A. Goddard and J. R. Heath, Nature, 2008, 451, 168–171.
17. A. Yusufu, K. Kurosaki, Y. Miyazaki, M. Ishimaru, A. Kosuga, Y. Ohishi, H. Muta and S. Yamanaka, Nanoscale, 2014, 22, 13921–13927.
18. L. B. Luo, L. M. Chen, M. L. Zhang, Z. B. He, W. F. Zhang, G. D. Yuan, W. J. Zhang and S. T. Lee, J. Phys. Chem. C, 2009, 113, 9191–9196.
19. L. B. Luo, J. S. Jie, W. F. Zhang, Z. B. He, J. X. Wang, G. D. Yuan, W. F. Zhang, L. C. M. Wu and S. T. Lee, Appl. Phys. Lett., 2009, 94, 193101.
20. C. Xie, L. B. Luo, L. H. Zeng, L. Zhu, J. J. Chen, B. Nie, J. G. Hu, Q. Li, C. Y. Wu, L. Wang and J. S. Jie, CrystEngComm, 2012, 14, 7222–7228.
21. L. A. Woldering, R. W. Tjerkstra, H. V. Jansen, I. D. Setija and W. L. Wos, Nanotechnology, 2008, 19, 145304.
22. R. Q. Zhang, Y. Lifshitz and S. T. Lee, Adv. Mater., 2003, 15, 635–640.
23. C. Xie, J. S. Jie, B. Nie, T. X. Yan, Q. Li, P. Lv, F. Z. Li, M. Z. Wang, C. Y. Wu, L. Wang and L. B. Luo, Appl. Phys. Lett., 2012, 100, 193103.
24. H. F. Yan, Y. J. Xing, Q. L. Hang, D. P. Yu, Y. P. Wang, J. Xu, Z. H. Xie and S. Q. Feng, Chem. Phys. Lett., 2000, 323, 224–227.
25. Y. Cui, X. F. Cui, J. T. Hu and C. M. Lieber, J. Phys. Chem. B, 2000, 104, 5213–5216.
26. K. K. Lew, L. Pan, T. Bogart, S. M. Dilts, E. C. Dickey, J. M. Redwing, Y. F. Wang, M. Cabassi, T. S. Mayer and S. W. Movak, Appl. Phys. Lett., 2004, 85, 3101–3103.
27. Z. H. Chen, J. S. Jie, L. B. Luo, H. Wang, C. S. Lee and S. T. Lee, Nanotechnology, 2007, 18, 345502.
28. N. Wang, Y. H. Tang, Y. F. Zhang, C. S. Lee, I. Bello and S. T. Lee, Chem. Phys. Lett., 1999, 299, 237–242.
29. J. S. Jie, W. J. Zhang, Y. Jiang and S. T. Lee, Appl. Phys. Lett., 2006, 89, 223117.
30. L. K. Li, Y. J. Yu, G. J. Ye, Q. Q. Ge, X. D. Ou, H. Wu, D. L. Feng, X. H. Chen and Y. B. Zhang, Nat. Nanotechnol., 2014, 9, 372.
31. M. S. Arnold, P. Avouris, Z. W. Pan and Z. L. Wang, J. Phys. Chem. B, 2003, 107, 659.
32. H. B. Huo, L. Dai, C. Liu, L. P. You, W. Q. Yang, R. M. Ma, G. Z. Ran and G. G. Qin, Nanotechnology, 2006, 17, 5912.
33. Z. Z. Hu, X. J. Zhang, C. Xie, C. Y. Wu, X. Z. Zhang, L. Bian, Y. M. Wu, L. Wang, Y. P. Zhang and J. S. Jie, Nanoscale, 2011, 3, 4798–4803.
34. G. D. Yuan, W. J. Zhang, J. S. Jie, X. Fan, J. X. Tang, I. Shafiq, Z. Z. Ye, C. S. Lee and S. T. Lee, Adv. Mater, 2008, 20, 168–173.
35. X. Liu, L. L. Gu, Q. P. Zhang, J. Y. Wu, Y. Z. Long and Z. Y. Fan, Nat. Commun., 2014, 5, 4007.
36. C. Xie, B. Nie, L. H. Zeng, F. X. Liang, M. Z. Wang, L. B. Luo, M. Feng, Y. Q. Yu, C. Y. Wu, Y. C. Wu and S. H. Yu, ACS Nano, 2014, 4, 4015–4022.
37. P. Lv, X. J. Zhang, X. W. Zhang, W. Deng and J. S. Jie, IEEE Electron Device Lett., 2013, 34, 1337–1339.
38. X. H. An, F. Z. Liu, Y. J. Jung and S. Kar, Nano Lett., 2013, 13, 909–916.
39. N. Guo, W. D. Hu, L. Liao, S. Yip, J. C. Ho, J. S. Miao, Z. Zhang, J. Zou, T. Jiang and S. W. Wu, Adv. Mater., 2014, 26, 8203–8209.

---

Footnote

† Electronic supplementary information (ESI) available: Digital camera pictures of the as prepared silicon nanostructures (Sample A, B and C), digital camera pictures and SEM images of other products obtained in the ratio of 1 : 0, 1 : 4, 1 : 6, the synthesis of monolayer graphene, Raman spectrum of the monolayer graphene. I–V curves of the monolayer graphene/silver paste contact, and Ti/Au electrode/Si NW (Sample B) contact. See DOI: 10.1039/c4ra16781c

---