Enhanced Field Emission and Optical Properties of ... - ACS Publications

Mar 15, 2012 - ... Young , Shoou-Jinn Chang , Jia-Jyun Guo , Chung-Wei Liu , Tsung-Ying Yang. IEEE Transactions on Electron Devices 2013 60, 1905-1910...
0 downloads 0 Views 4MB Size
Download PDF
Article pubs.acs.org/JPCC

Enhanced Field Emission and Optical Properties of Controlled Tapered ZnS Nanostructures Muhammad Hafeez,†,‡ Tianyou Zhai,*,‡ Arshad S. Bhatti,*,† Yoshio Bando,‡ and Dmitri Golberg‡ †

Center for Micro and Nano Devices (CMND), Department of Physics, COMSATS Institute of Information Technology, Islamabad, 44000, Pakistan ‡ International Center for Materials Nanoarchitectonics (WPI-MANA), National Institute of Materials Science (NIMS), Namiki 1-1, Tsukuba, Ibaraki, 305-0044, Japan ABSTRACT: High-quality ZnS tapered 1D nanostructures were successfully synthesized by a simple thermal evaporation technique after having careful control on the size of the catalyst and growth temperature. The dynamics of varied gas flow on the extent of tapering was monitored with specific experimental settings. Systematic optical investigations revealed that the PL emission was due to the surface states. Surface optical phonons in tapered nanostructures were also observed, which showed variation in the surface potential. The dielectric continuum model was used to calculate the wavelength of the surface potential perturbation. The field emission (FE) measurements were carried out to determine the correlation between the FE and the shape of nanostructures. It was observed that tapered nanowires with an average diameter of 60 nm were excellent field emitters with low turn-on fields ranging from 5.5 to 11.67 V/μm and the field enhancement factors from 1940 to 890, respectively, at various cathode distances. By the two-region field emission model, the absolute amplification factor was calculated and found to be comparable to that of carbon nanotubes. The fitting of the experimental points with the Fowler−Nordheim equation concluded that the emission is from the tips of nanowires. The present study on FE showed that the tapered nanowires can be potential materials for the novel optoelectronic devices.

■

standing applications, such as field emitters,10 field effect transistors (FETs),11 sensors,12−14 photocatalysts,15 etc. The work function of ZnS (7.0 eV) is higher than that of some other popular FE materials, such as C nanotubes (5.0 eV), ZnO (5.3 eV), and CdS (4.2 eV); thus, ZnS is not among the best FE materials. However, ZnS has a tendency to grow in numerous morphologies with controlled shapes, and due to its wide band gap of 3.91 eV, it has high band bending in the large applied fields. These properties made it a potential material for the FE applications. Production of different types of ZnS nanostructures, namely, nanobelts,16 nanowires arrays,17 and nanorods,18 with less or more efficiency has already been reported. There are several processes to synthesize ZnS nanostructures, namely, thermal evaporation,19,20solvothermal technique,21 and chemical vapor deposition.22 Among those, the thermal evaporation is the simplest way to get large-density, high-quality, and single-crystalline 1-D nanostructures.23 It is not trivial to master the control on the morphology of the growing nanostructures because the growth process is very complex due to numerous unseen mechanisms taking place during growth. Using the thermal evaporation technique, the morphology of the ZnS nanostructures can be changed from

INTRODUCTION Research and development in the field of one-dimensional (1D) nanostructures, such as nanowires, nanobelts, and nanotubes, presents some of the most exciting breakthroughs in the modern era of science and technology.1−5In recent years, the research on field emission properties of 1D wide-band-gap semiconductors became crucial due to growing demands of micro/nanotechnologies The Fowler−Nordheim (FN) theory6 relates the field emission to the two important parameters, namely, the work function of an emitting material and the fieldenhancement factor.7 The work function is an intrinsic parameter that varies from 1 to 7 eV for most of the inorganic semiconductors, the lower being better for the higher emission efficiency.8 The field-enhancement factor is defined as the ratio of a local field to an applied field and is determined by the shape and morphology of the emitter. The work function can be modified by the shape of the material because confined electrons near the surface edge become concentrated and more localized.9 Morphological variations may have prominent effects on the field emission properties, such as turn-on field and field enhancement factor, etc. Thus, it is possible to achieve strong field emission from a given nanostructure material if it is long and sharp. As an important II−VI semiconductor, ZnS was one of the first semiconductors discovered and is probably one of the most important electronic and optoelectronic materials with out© 2012 American Chemical Society

Received: February 1, 2012 Revised: March 9, 2012 Published: March 15, 2012 8297

dx.doi.org/10.1021/jp3010635 | J. Phys. Chem. C 2012, 116, 8297−8304

The Journal of Physical Chemistry C

Article

Figure 1. (a) 3D AFM image of 1 nm Au-coated substrate annealed at 850 °C. (b, c) Illustrations of the possible mechanisms at the substrate surfaces and comparison of the surface diffusion processes.

Figure 2. Experimental setup for two growth experiments.

The synthesized nanostructures were characterized using AFM (Agilent’s PicoPlus), field emission SEM (Hitachi SU 8000), HRTEM (JEOL JEM 2100F) equipped with an X-ray energy-dispersive spectrometer (EDS), and XRD (RINT 2200HF) for surface topography, imaging, and phase identification, respectively. For the optical properties measurements at room temperature, a Raman spectrometer (Horiba Jobin-Yvon T6400 with 514 nm excitation wavelength) and a micro laser photoluminescence (PL) spectrometer (Horiba Jobin-Yvon S.A.S using He−Cd laser with an excitation wavelength of 325 nm) were used. The field emission properties were studied at room temperature in a high-vacuum chamber (4 × 10−6 Pa) using a 1 mm2 cross-sectional area copper anode. A dc voltage sweeping from 100 to 1100 V was applied to the samples.

nanodots to nanorods, to nanowires, to nanobelts, and to nanosheets by changing the growth temperature24 and carrier gas flow rate,13 but the growth of tapered nanostructures is quite difficult and complex. We present a novel way to synthesize 1D tapered nanostructures under an efficient control on the extent of their tapering by tuning the surface diffusion, which is the result of minimization of Gibbs free energy. In this paper, we report the efficient way to introduce controlled tapering in ZnS 1D nanostructures by using the vapor−liquid−solid (VLS) mechanism. Some calculations on flow dynamics allowed us to achieve a careful control on the velocity of the vapor species to make a uniform flow. Atomic force microscopy (AFM), scanning electron microscopy (SEM), and high-resolution transmission electron microscopy (HRTEM) were used to check the effect of growth conditions on the morphology and sizes of the grown nanostructures. Xray diffraction (XRD), Raman spectroscopy, and photoluminescence (PL) spectroscopy have been used to determine the effect of growth conditions on the structure and quality (study of surface and bulk defects) of nanostructures. The dielectric continuum (DC) model was used to correlate the theoretical and experimentally observed surface optical phonons. Finally, the field emission measurements of the tapered nanowires were undertaken and compared with other documented FE results for ZnS nanowires.

■

■

RESULTS AND DISCUSSION We were targeting the sharp-edged (tapered) 1-D nanostructures having controlled tapering. By using a vapor−liquid−solid (VLS) mode of growth, this objective was achieved by careful control on the density and sizes of the catalyst nanoparticles. These were controlled by varying the thickness of the catalyst (Au) thin film and growth temperature. Figure 1a shows the 3D AFM topographic images of catalyst droplets formed under annealing of the 1 nm Au-coated Si substrate at 850 °C. The surface coverage was around 65%. For such a low surface coverage deposition, a decent amount of vapors will flow over the substrate surface or side facets of the growing nanostructures. The VLS mechanism is purely a energy minimization process; the vapor species that come to the surface of the droplet minimize its Gibb’s free energy through adsorbing on the surface of the droplet, diffusing through the droplet, and precipitating at the liquid−solid interface. The tapered wires were grown under the high supersaturation conditions, as shown in Figure 1b; the vapor species that landed on the substrate surface had a high Gibb’s free energy, which results in a high rate of surface diffusion to minimize it. Another possible mechanism is that the small catalyst droplet has a high value of surface tension. It is difficult for the vapor species to diffuse through the catalyst droplet; instead, they surface diffuse over the surface of the droplet, as depicted in the Figure 1c.

EXPERIMENTAL SECTION

Si(100) substrates with a very thin (2−3 nm) native oxide layer were used for the growth of nanostructures. A 1 nm Au (catalyst) thin film was deposited by an e-beam evaporator at room temperature. The Au-coated substrates were then loaded in a central zone of a horizontal tube furnace along with 99.99% pure ZnS powder (1 g) in an alumina boat placed at the center of the tube. In the two growth experiments, ZnS powder was heated to 1120 and 1000 °C, and the substrates were kept in the downstream. N2 gas was used as the carrier gas to transport the evaporated flux from the ZnS boat to the Si wafers. The flow rate of the carrier gas was maintained at 50 sccm for each experiment. Separately, Au-coated substrates were heated to 850 °C, and these samples were characterized by AFM to determine the size and density of the Au droplets. 8298

dx.doi.org/10.1021/jp3010635 | J. Phys. Chem. C 2012, 116, 8297−8304

The Journal of Physical Chemistry C

Article

To observe the effect of gas flow dynamic conditions on the growth of nanostructures, specific experimental settings of growth collection substrates for the two experiments were developed, as shown in Figure 2. In the two experiments, the temperature is different, so the dynamic viscosity, density (gas, vapor species), and kinematic viscosity change, which thus will affect the Reynolds number. The characteristic of flow (either laminar flow or turbulent flow) can be determined by the Reynolds number and is given by25

Re =

QDρ μA

(1)

Q is the volumetric flow rate, D is the diameter of the tube, ρ is the density of the gas, μ is the dynamic viscosity, and A is the area of the cross section. The dynamic viscosity is a purely temperature-dependent quantity, and for the two different temperatures, 1120 and 1000 °C, its value is 5.14 × 10−5 and 4.85 × 10−5 kg/m·s, respectively.26 The density of the gas has a direct relation with temperature and pressure. The density of nitrogen gas at 0 °C and at 1 atm pressure is 1.25 kg/m3, and for 1120 and 1000 °C, ρ is calculated as 0.0242 and 0.0265 kg/ m,3 respectively. If we consider that nothing is blocking the gas in the tube, and if the diameter of the tube is 0.036 m and the area of the cross section, A = π(0.018)2 = 1.02 × 10−3 m2, the volumetric flow rate becomes 50 sccm = 50 cm3/60 s = 8.33 × 10−7 m3/s. The calculated Re number is then 1.4 × 10−2 and 1.62 × 10−2 for 1120 and 1000 °C, respectively. Since, the flowing gas vapors are the mixtures of nitrogen and ZnS vapor, the total maximum vapor density approximated in the tube is 1.85 kg/ m3.25 This means that the Re number was less than 1.07 and 1.12 for 1120 and 1000 °C, respectively. This number is much smaller than the threshold number, 2300, which distinguishes a laminar flow from a turbulent flow. Therefore, the flow is purely laminar, but on the way, we have placed alumina boats containing the substrates. The height of the boat is 0.6 cm, and it is located 1.6 cm away from the bottom of the tube. Such conditions create a blocking situation. It is known that, when a gas flowing in a pipe encounters the entrance divergence, abrupt enlargement, or the entrance blocking situation, a laminar flow characteristic could change to a turbulent flow locally due to the dominant gas−solid entrance wall boundary friction effect.27 The gas flow is blocked by the alumina boat and may change the flow from the laminar to the turbulent one. The typical characteristic of the turbulent flow is that its velocity profile is flat; that is, throughout the cross section of the tube, the velocity is the same so that, in the turbulent flow, the velocity of the vapor species is uniform.28 For the second sample, the blocking situation happens twice, which helps the vapor species to move at a more uniform velocity. Hence, we were able to make a uniform motion of the vapor species that strike on the surface of the deposition substrate. From the above discussion, it is concluded that the tapering can be achievable on the substrate under the low surface coverage, small density, and small-sized catalysts droplets while having a uniform velocity of a vapor flux. Figure 3a−d shows the SEM and TEM micrographs of ZnS nanostructures synthesized in the first experiment. Low- and high-magnification FESEM images show that tens of micrometers long, dense tapered nanobelts with an average tip diameter in the range of 150 ± 10 nm and an average width of ∼1 μm were obtained, as shown in Figure 3a,b. The growth of nanobelts in this experiment suggested that the growth

Figure 3. (a, b) Low- and high-magnification FESEM images of samples grown in the first experiment. (c, d) TEM and HRTEM images of a single tapered nanobelt; the inset in (d) shows a SAED pattern. (e) The energy-dispersive X-ray spectrum of a single nanobelt. (f) XRD pattern of tapered ZnS nanobelts.

temperature and supersaturation conditions are very high. Fang et al. have also reported that, in the growth of Aucatalyzed ZnS nanostructures, at high temperature and small distance between the source and the substrate, the morphology of a synthesized product is beltlike.24 The inset in Figure 3b shows high-magnification SEM images of the single nanobelt having a catalyst droplet at the tip end, indicating the pure VLS growth. Figure 3c shows the TEM image of a single tapered nanobelt, indicating that nanobelts have rectangular cross sections. Figure 3d shows the HRTEM image of an individual nanobelt revealing that the nanobelt is a single crystal with the [100] growth direction. A selected area electron diffraction (SAED) pattern further confirms that the growth direction is along the [100] orientation, as shown in the inset of Figure 3d. The energy-dispersive X-ray spectrum (EDS) is shown in Figure 3e. This confirmed that the nanobelt has a stoichiometric 1:1composition of Zn and S, (the Cu signal comes from the grid). Figure 3f shows an XRD pattern of the ZnS tapered nanobelts. All the diffraction peaks were identified to the hexagonal wurtzite 2H structure of ZnS with the lattice parameters a = 3.82 Å and c = 6.26 Å, in good agreement with the literature (JCPDF card number 05-0492). No other characteristic peaks of any other phases and materials were observed in the measured range. The sharpness of the peaks reflects the good quality of the sample. The XRD spectrum showed a strong peak of [100], which also confirmed the growth direction, in accordance with the HRTEM results. Figure 4a−d shows the SEM and TEM micrographs of ZnS nanostructures synthesized in the second experiment. Low- and high-magnification FESEM images showed that tens of micrometers long, dense tapered nanowires with an average diameter in the range of 60 ± 5 nm and the tip neck diameters 8299

dx.doi.org/10.1021/jp3010635 | J. Phys. Chem. C 2012, 116, 8297−8304

The Journal of Physical Chemistry C

Article

of vapor species, so the diameter was small and the surface diffusion was controllable. X-ray diffraction patterns of the grown nanobelts confirmed that the grown products have wurtzite-type crystal structures. A wurtzite ZnS has the space group C6V4 (C63mc) with two formula units per primitive cell, and all atoms occupy C3V sites. Group theory predicts the following phonon symmetries with momentum q = 0: χatom = 1A1 + 2B1 + 1E1 + 2E2, where A1 and E1 symmetry phonons are both Raman- and IR-active, two E2 pairs of modes are only Raman-active, while the two B1 modes are optically inactive.29 We have observed symmetry breaking at the surface of nanobelts under HRTEM imaging (Figure 3d). This type of symmetry breaking arises because of diameter modulation along the wire due to instability during the vapor−liquid−solid (VLS) growth.30 In the case of nanobelts, the deposition temperature was high and the crystal growth rate was directly related to the availability of chemical species in the droplet (substrate is near to source) and its diffusion rate through the droplet. Therefore, under such conditions, the diameter modulation occurs in the growing nanostructures, which is responsible for the symmetry breaking at the surface. This symmetry breaking phenomenon may activate the surface optical (SO) mode. Thus, Raman spectroscopy was employed for the evaluation of such a SO mode. A Raman spectrum obtained from the ZnS tapered nanobelts is shown in Figure 5a; this was taken at room temperature with an excitation wavelength of 514 nm. All the peaks are labeled in the spectrum along with their peak positions. The dominant peak is due to the longitudinal optical (LO) phonon of ZnS at

Figure 4. (a, b) Low- and high-magnification FESEM images of samples grown in the second experiment. (c, d) TEM and HRTEM images of a single tapered nanowire; the inset in (d) shows the SAED pattern.

of 40 ± 3 were obtained, as shown in Figure 4a,b. This shows that high supersaturation conditions were achieved, since the growth rate of nanowires depends on such conditions, which were controlled by the tunable gas flow rate. The inset in Figure 4b shows high-magnification SEM images of the nanowires. Considerations of the contact angle between the droplets and the nanowires can help to estimate the strength of the surface tension of the small droplets as their shape is retained. Figure 4c shows the TEM image of a single tapered nanowire with a diameter of 62 nm and the neck diameter of 41 nm. The HRTEM image displays the lattice fringes of the nanowire, and it is clear that it is a single crystal with the [100] growing direction (Figure 4d). A SAED pattern further confirms that the growth direction was along [100], as shown in the inset of Figure 4d. The synthesized nanostructures in two experiments have the [100] growth direction, which is in good agreement with the reported results by Fang et al.24 They showed that, in ZnS nanostructures grown using the VLS mechanism, the growth direction was [100], and if the growth mode is VS, then the growth direction will be [001]. In our case, the lateral growth direction is [001], and in wurtzite crystal structures, the [001] is the fast growing direction with low activation energy. In first growth experiment, the substrate temperature is high, and thus the formation of nanobelts is favorable due to activation of growth in the low-energy [001] plane. In the second growth experiment, due to low substrate temperature, lateral growth is forbidden and a wirelike morphology was obtained. The tip diameters of nanostructures varied in the two experiments, which was due to the change in temperature and variations in the flux velocity. At the high temperature, the size of the catalyst droplet increased due to Ostwald ripening. This resulted in the large tip diameters and low densities of the droplets. From the SEM images, it is clear that the density of nanobelts is low as compared with that of nanowires, because the growth temperature of nanowires was lower as compared with that of the nanobelts. At the high temperature and near to the source, the surface diffusion rate is high, which resulted in enhanced tapering and large-sized structures. For the second experiment, we have low temperature and a more uniform flux

Figure 5. (a) Micro-Raman spectrum of tapered ZnS nanobelts grown in the 1st experiment, (b) Gaussian peak fittings on the twin peak of TO phonons, and (c) LO and SO phonons. 8300

dx.doi.org/10.1021/jp3010635 | J. Phys. Chem. C 2012, 116, 8297−8304

The Journal of Physical Chemistry C

Article

347.7 cm−1, as allowed by the Raman selection rules for the phonon modes of a one-dimensional system. There is a twin peak for the transverse optical (TO) phonon. The Lorentzian fit shows that A1/E1 (TO) is at 267 cm−1 and E2 (TO) is at 280 cm−1, as shown in Figure 5b. ZnS in the wurtzite crystal structure exhibits polar behavior. In the high-aspect-ratio wurtzite 1D nanostructures, the long-range dipolar interaction can give rise to an additional splitting of the TO mode.31 This happens because the dipole sums, which determine TO mode frequencies, are sensitive to the 1D nanostructures' aspect ratio. The bulk values of LO and TO modes for ZnS (wurtzite-2H) are 350 and 276 cm−1, respectively.32 The Lyddane−Sachs− Teller (LST) relation was used for the calculation of the actual TO frequency and is given by33 ω2TO =

ω2LOε∞ ε0

(2)

As an isotropic approximation, from ref 32, ε0 = 8.29, ε∞ = 5.11 was taken and the LO phonon energy, as determined from the Raman spectrum, ωLO = 347.7 cm−1 was used, and thus the frequency of TO phonons was calculated by using the LST relation; it came out to be ωTO = 272.9 cm−1. As expected after the observation of symmetry breaking in the HRTEM images of nanobelts, we have observed a surface optical (SO) phonon peak at 336.17 cm−1, as shown in Figure 5c. It has been known that the presence of surface potential leads to symmetry breaking and is responsible for the observation of SO phonon modes. The average wavelength of the surface potential perturbations was determined by using the dielectric continuum model for the rectangular cross section of the nanowires.34 The symmetric (S) and asymmetric (AS) modes of SO phonons were used to establish the dispersion relation. Using this dispersion relation, the value of wave vector q was calculated. In Figure 6a, the curves show the dispersion of S and AS modes of SO phonon ωSO(q) as a function of qL/2 for ZnS nanobelts with square cross sections Lx = Ly = L; thus, q = √2qi (i = x,y) in air. The experimental SO frequency in air is ωSO(q) 336.17 cm−1.We can then estimate the important wave vector q of the surface perturbation responsible for activating the SO modes. We found that the experimental value (dotted line) of the SO phonons intersects the S mode at qL = 3.22. If we take the average wire square cross-sectional dimensions to be 50 × 50 nm2, we can then calculate the wavelength for the surface potential perturbation that breaks the symmetry and activates the SO mode of Raman scattering; that is, λ = 2π/q = 98 nm. Such a value is in good agreement with the experimentally observed one during TEM analysis (92 nm), as shown in Figure 6b. The average distance between the arrows is 92 nm in the TEM image and HRTEM image (Figure 6c), thus showing that the symmetry breaking is of the order of several atomic layers. The existence of the surface potential perturbation and symmetry breaking were further confirmed by the PL measurements, which demonstrate the emission due to surface defect states. Figure 7a shows the PL spectra of tapered nanostructures taken at room temperature using a He−Cd laser with an excitation energy of 325 nm. The PL spectrum of tapered nanowires shows a broad peak. The peak was fitted with two Gaussian functions peaked at 524 and 541 nm (Figure 7b). For the tapered ZnS nanobelts, the two peaks seen are the following: one at 412 nm and the other a very broad peak at

Figure 6. (a) SO phonon dispersion relation calculated for tapered nanobelts by using corresponding equations for symmetric modes (S) and asymmetric (AS). (b) TEM images of a ZnS nanobelt showing the surface modulation along the wire axis indicated by the arrows; an average distance between the arrows is ∼92 nm. (c) HRTEM image displaying that the surface roughness is about several atomic layers.

around 550 nm. The Gaussian multipeaks fit (Figure 7c) showed that this broad peak was a summation of two peaks positioned at 544 and 578 nm. It is well known that surface

Figure 7. (a) Micro-PL spectra of tapered ZnS nanobelts and tapered nanowires. (b, c) Corresponding Gaussian peak fittings of the spectra. 8301

dx.doi.org/10.1021/jp3010635 | J. Phys. Chem. C 2012, 116, 8297−8304

The Journal of Physical Chemistry C

Article

Figure 8. (a) J−E plot of tapered nanowires measured at a 100 μm gap between the anode and sample; the inset shows the corresponding F-N plot. (b) J−E plots and (c) corresponding F-N plots from tapered nanowires with different vacuum gaps. (d) Variation in turn-on field (at 10 μA/cm2), threshold field (at 100 μA/cm2), and the field-enhancement factor with a change in vacuum gap d.

defect states are responsible for the blue emission at 410 nm.35 There are several reports for the peak around 520 nm, which is explained due to the Au-catalyst impurity deep level emission.36−38 The green emission at 540 nm may originate from Zn vacancy states, element sulfur species on the surface, or interstitial states.39 The emission band at 570 nm may be attributed to some self-activated centers and vacancy states.40 For the nanowires, one peak is due to Au impurities and the other one is due to Zn vacancies; probably, Au replaced Zn due to high solubility of small-sized catalyst particles. Small catalyst droplets have high solubility as compared with the large droplets; that is why we have not observed Au peaks for the nanobelts.41 For the nanobelts, the three peaks are due to surface defects, Zn vacancy states, and self-activated centers, which is probably due to symmetry breaking at the surface and high temperature growth. From the morphological and optical studies, it is concluded that the growth conditions have a large impact on the size, morphology, and quality of nanostructures. High (low)temperature growth results in relatively low (high) density, and large (small)-sized nanostructures with surface defects (relatively smooth surface), respectively. Because of the smooth surface, high density of nanowires with tapered sharp tips, and the presence of Au traces in the nanowires, the FE measurements of nanowires were carried out at room temperature. Actually, in the FE measurements, the electrons always flow along the surface of a nanowire, and the surface charges play the key role. FE measurements of ZnS tapered nanowires showed that they were potentially good field emitters. Figure 8a shows the field emission measurements of ZnS tapered nanowires measured at a 100 μm gap between the anode and a sample in a vacuum chamber maintained at a pressure of 4 × 10−6 Pa. From the curve of the emission current density as a function of

the applied field (JE), it is concluded that the present structures had the best field emission properties with a low turn-on field (Eto, defined as E at which J becomes 10 μA/cm2) of 6.2 V/μm, and a threshold field (defined as E at which J becomes 0.1 mA/ cm2) of 8.2 V/μm. The FE current voltage characteristics were further analyzed by the Fowler−Nordheim (F-N) equation42 ⎛ Bϕ3/2 ⎞ Aβ2E2 ⎟ exp⎜⎜ − βE ⎟⎠ ϕ ⎝

(3)

⎛ J ⎞ Bϕ3/2 Aβ2 ln⎜ 2 ⎟=ln − ⎝E ⎠ βE ϕ

(4)

J=

or

where A and B are constants with values of 1.54 × 10−6 A eV V−2 and 6.83 × 103 V/μm eV−3/2, respectively, J is the current density, β is the field-enhancement factor, E is the applied field, and Φ is the work function (7.0 eV for ZnS) of the emitting materials. It is known that the field enhancement factor is related to the emitter geometry (such as aspect ratios), crystal structure, vacuum gaps, and the spatial distribution of emitting centers.43 The inset shows that F-N plot at a distance of 100 μm and the linear variation of ln(J/E2) with (1/E) implies that the electron emission from ZnS tapered nanowires follows the FN behavior. We were able to calculate the field enhancement factor from the slope of the F-N plot, which was 1562. Some deviation from the FN theory and current saturation was observed in the samples; this type of deviation has also been reported for the CNTs. This was probably due to the presence of catalyst particles at the tips of nanowires or the contact resistance between the nanowires and the substrate.44 8302

dx.doi.org/10.1021/jp3010635 | J. Phys. Chem. C 2012, 116, 8297−8304

The Journal of Physical Chemistry C

Article

basis of the two-region field emission (TRFE) model,44 we found that our experimental data was fitted to a straight line and could be approximated by (1/β) = (d2/d) + (1/β0), where d2 is the width of the field-enhancement region near the nanowires' surface and β0 is the absolute amplification factor, which is intrinsically determined by emitters and independent of d and the applied field. The values of d2 and β0 were determined by the linear fitting and were given by the slope and intercept, respectively. The value of d2 is determined as 55 nm, while β0 is 6992. The value of β0 revealed by the present tapered nanowires was comparable to that of the carbon nanotubes grown on a Si wafer (β0 = 7900),44 while it is much larger than that for CdS nanocones grown on a Si substrate (β0 = 4933),47 ZnO nanorods grown on a Si substrate (β0 = 3738),48 and an Al-doped ZnO emitter (β0 = 1845)49 However, d2 is smaller than the values reported in the literature, which is probably due to the strong screening effect caused by the higher density and catalyst particles at the ends of the nanowires. Table 1 shows the comparative key FE parameters

Figure 8b illustrates the FE curves measured at different vacuum gaps (d). When a vacuum gap increased from 60 to 160 μm, the turn-on field decreased from 11.67 to 5.5 V/μm, and the threshold field also decreased from 12.83 to 7.8 V/μm. Figure 8c shows the F-N plot at different distances. The calculated field enhancement factors from the slope of the linear part of the F-N plot were 891, 1301, 1562, 1680, 1709, and 1942 for d = 60, 80, 100, 120, 140, and 160 μm, respectively. Figure 8d demonstrates the variation of field enhancement factor, turn-on field, and threshold field as a function of the vacuum gap. After the curve fitting based on the trends of Eto and Eth, the following equation was developed ⎛ d⎞ Turn‐on field or Threshold field=A × exp⎜ − ⎟ + C ⎝ B⎠

(5)

where A, B, and C are constants having the following values: for Eto [A = 212, B = 16.90, and C = 5.67], whereas for Eth [A = 227.16, B = 15.80, and C = 7.6]. By putting the value of d, we can get the turn-on or threshold field at any point in the measurement range. The turn-on field (5.5 V/μm @ 10 μA/ cm2) and β = 1942 revealed by tapered nanowires is comparable with CNTs having catalyst particles on the tips with a turn-on field (4.9 V/μm @ 1 μA/cm2) and β = 1870.45 The FE properties are much better than the aligned ZnO nanowires with a turn-on field (6 V/μm @ 0.1 μA/cm2) and β = 847,46 and the quasi aligned CdS nanowires array with a turnon field (12.2 V/μm @ 10 μA/cm2) and β = 555.47 When we plot 1/β versus 1/d, as shown in Figure 9a, it becomes clear that 1/β follows the straight line for 1/d; that is, the larger the vacuum gap, the higher is the value of β. On the

Table 1. Comparison of Key FE Parameters for ZnS Nanowires and Nanorods Found in the Literature with the Present Results for Tapered Nanowires ZnS emitter

synthesis method

nanowires

thermal evaporation

nanorods

radio frequency magnetron sputtering vapor phase deposition thermal evaporation

nanowires tapered nanowires

turn-on field (V/μm) 6.5 @ 0.01 mA/cm2 2.9−6.3 @ 2.452 μA/cm2 11.7 @ 0.1 μA/cm2 5.5−11.6 @ 0.01 mA/cm2

β

ref

1529

17

420−105

18

522

50

1942−891

this work

of ZnS nanowires and nanorods found in the literature with our results for the present nanostructures. Stability of the field emitters is another important parameter related to potential applications. FE stability measurements were performed on the nanowires by keeping an electric field at 10 V/μm over a period of more than 8 h. As shown in Figure 9b, there were not any notable fluctuations during this period. Good emission stability demonstrates that the present structures can find potential applications in the cold-cathodebased electronics.

■

CONCLUSIONS High-quality ZnS tapered 1D nanostructures were successfully synthesized using vapor−liquid−solid (VLS) growth after having careful control on the size of the catalyst and growth temperature. Morphological, structural, and optical characterizations were undertaken to check the effect of growth conditions. Surface optical phonons were observed due to symmetry breaking at the surface. The dielectric continuum (DC) model was successfully used to calculate the wavelength of surface potential perturbation. Field emission measurements show that tapered nanowires are excellent field emitters with a turn-on field ranging from 5.5 to 11.67 V/μm and the field enhancement factor varying from 1940 to 890, respectively, at various cathode distances. The absolute amplification factor was calculated by using the two-region field emission (TRFE) model and was found to be comparable to carbon nanotubes. Fitting of experimental data to the Fowler−Nordheim equation

Figure 9. (a) Linear fit to the experimental data based on the TRFE model. (b) Field emission stability data collected over 8 h and acquired at the field of 10 V/μm. 8303

dx.doi.org/10.1021/jp3010635 | J. Phys. Chem. C 2012, 116, 8297−8304

The Journal of Physical Chemistry C

Article

(23) Zhai, T.; Fang, X.; Zeng, H.; Xu, X.; Bando, Y.; Golberg, D. Pure Appl. Chem. 2010, 82, 2027−2053. (24) Fang, X. S.; Ye, C. H.; Zhang, L. D.; Wang, Y. H.; Wu, Y. C. Adv. Funct. Mater. 2005, 15, 63−68. (25) Kim, D.; Shimpi, P.; Gao, P. X. Sci. Adv. Mater. 2010, 2, 421− 427. (26) http://ddbonline.ddbst.de/EE/ 1056%20VIS%20(Dynamic%20Viscosity).shtml. (27) McCabe, W. L.; Smith, J. C.; Harriott, P. Unit Operation of Chemical Engineering; McGraw-Hill, Inc.: Singapore, 1985. (28) Giling, Z. J. J. Phys. (Paris) 1982, 43, 235−246. (29) Arguello, C. A.; Rousseau, D. L.; Porto, S. P. S. Phys. Rev. 1969, 181, 1351−1363. (30) Givargizov, E. I. J. Cryst. Growth 1973, 20, 217−226. (31) Mahan, G. D.; Gupta, R.; Xiong, Q.; Adu, C. K.; Eklund, P. C. Phys. Rev. B 2003, 68, 073402−073405. (32) Ebisuzaki, Y.; Nicol, M. J. Phys. Chem. Solids 1971, 33, 763−766. (33) Yu, P. Y.; Cardona, M. Fundamentals of Semiconductors: Physics and Materials Properties, 2nd ed; Springer: Berlin, 1999. (34) Xiong, Q.; Wang, J.; Reese, O.; Lew, L. C.; Voon, Y.; Eklund, P. C. Nano Lett. 2004, 4, 1991−1996. (35) Senthikumaar, S.; Selvi, R. T. J. Appl. Sci. 2008, 8, 2306−2311. (36) Rosenberg, R. A.; Shenoy, G. K.; Heigl, F.; Lee, S. T.; Kim, P. S. G.; Zhou, X. T.; Sham, T. K. Appl. Phys. Lett. 2005, 87, 253105− 253107. (37) Wang, Y. W.; Zhang, L. D.; Liang, C. H.; Wang, G. Z.; Peng, X. S. Chem. Phys. Lett. 2002, 357, 314−318. (38) Hu, J.; Wang, G.; Guoa, C.; Li, D.; Zhang, L.; Zhao, J. J. Lumin. 2007, 122, 172−175. (39) Tiwary, C. S.; Kumbhakar, P.; Mondal, A. K.; Mitra, A. K. Phys. Status Solidi A 2010, 207, 1874−1879. (40) Lan, C.; Gong, J.; Jiang, Y.; Song, Y.; Yang, S. CrystEngComm. 2012, 14, 708−712. (41) Hafeez, M.; Manzoor, U.; Bhatti, A. S.; Kaynar, M. B.; Shah, S. I. J. Appl. Phys. 2012, 111, 024313−024319. (42) Li, L.; Fang, X. S.; Chew, H. G.; Zheng, F.; Liew, T. H.; Xu, X. J.; Zhang, Y. X.; Pan, S. S.; Li, G. H.; Zhang, L. D. Adv. Funct. Mater. 2008, 18, 1080−1088. (43) Gautam, U. K.; Fang, X. S.; Bando, Y.; Zhan, J. H.; Golberg, D. ACS Nano 2008, 2, 1015−1021. (44) Zhong, D. Y.; Zhang, G. Y.; Liu, S.; Sakurai, T.; Wang, E. G. Appl. Phys. Lett. 2002, 80, 506−508. (45) Jang, H. S.; Lee, H. R.; Kim, D. H. Thin Solid Films 2006, 500, 124−128. (46) Lee, C. J.; Lee, T. J.; Lyu, S. C.; Zhang, Y.; Ruh, H. Appl. Phys. Lett. 2002, 81, 3648−3650. (47) Zhai, T. Y.; Fang, X. S.; Bando, Y.; Liao, Q.; Xu, X.; Zeng, H.; Ma, Y.; Yao, J.; Golberg, D. ACS Nano 2009, 3, 949−959. (48) Li, C.; Fang, G. J.; Liu, N. S.; Li, J.; Liao, L.; Su, F. H.; Li, G. H.; Wu, X. G.; Zhao, X. Z. J. Phys. Chem. C 2007, 111, 12566−12571. (49) Xue, X. Y.; Li, L. M.; Yu, H. C.; Chen, Y. J.; Wang, E. G.; Wang, T. H. Appl. Phys. Lett. 2006, 89, 043118−043120. (50) Chang, Y. Q.; Wang, M. W.; Chen, X. H.; Ni, S. L.; Qiang, W. J. Solid State Commun. 2007, 142, 295−298.

revealed that the emission is from the nanowire tips. The present FE study revealed that the tapered nanowires could be potential materials for the novel optoelectronic devices.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected] (A.S.B.), zhai.tianyou@ gmail.com (T.Z.). Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS

■

REFERENCES

This work was supported by the World Premier International (WPI) Research Center on Material Nanoarchitechtonics (MANA), MEXT, Japan. M.H. is thankful to the IRSIP Program of Higher Education Commission, Pakistan (HEC), and the National Institute for Materials Science (NIMS) for financial support and an Internship award.

(1) Zhai, T. Y.; Li, L.; Ma, Y.; Liao, M.; Wang, X.; Fang, X. S.; Yao, J.; Bando, Y.; Golberg, D. Chem. Soc. Rev. 2011, 40, 2986−3004. (2) Fang, X. S.; Zhai, T. Y.; Gautam, U. K.; Li, L.; Wu, L.; Bando, Y.; Golberg, D. Prog. Mater. Sci. 2011, 56, 175−287. (3) Li, Z.; Cheg, L. N.; Sun, Q.; Zhu, Z. H.; Riley, M. J.; Ajiada, N.; Cheng, Z. X.; Wang, X. L.; Qiao, S. Z.; Smith, S. C.; Lu, G. Q. Angew. Chem., Int. Ed. 2010, 49, 2777−2781. (4) Li, Z.; Kurtulus, O.; Fu, N.; Myalitsin, A.; Wang, Z.; Kornowski, A.; Pietsch, U.; News, A. Adv. Funct. Mater. 2009, 19, 3650−3661. (5) Zhai, T. Y.; Fang, X. S.; Liao, M. Y.; Xu, X. J.; Li, L.; Liu, B. D.; Koide, Y.; Ma., Y.; Yao, J. N.; Bando, Y.; Golberg, D. ACS Nano 2010, 4, 1596−1602. (6) Junpeng, L.; Tao, Y.; Minrui, Z.; Haur, S. C. J. Phys. Chem. C 2011, 115, 8816−8824. (7) Ye, C. H.; Bando, Y.; Fang, X. S.; Shen, G. Z.; Golberg, D. J. Phys. Chem. C 2007, 111, 12673−12676. (8) Fang, X. S.; Bando, Y.; Gautam, U. K.; Ye, C. H.; Golberg, D. J. Mater. Chem. 2008, 18, 509−522. (9) Sun, C. Q. Prog. Solid State Chem. 2007, 35, 1−159. (10) Chen, Z. G.; Cheng, L.; Xu, H. Y.; Liu, J. Z.; Zou, J.; Sekiguchi, T.; Lu, G. Q. M.; Cheng, H. M. Adv. Mater. 2010, 22, 2376−2380. (11) He, J. H.; Zhang, Y. Y.; Liu, J.; Moore, D.; Bao, G.; Wang, Z. L. J. Phys. Chem. C 2007, 111, 12152−12156. (12) Fang, X. S.; Bando, Y.; Liao, M.; Zhai, T. Y.; Gautam, U. K.; Li, L.; Koide, Y.; Golberg, D. Adv. Funct. Mater. 2010, 20, 500−508. (13) Hafeez, M.; Manzoor, U.; Bhatti, A. S. J. Mater. Sci.: Mater. Electron. 2011, 22, 1772−1777. (14) Zhai, T.; Li, L.; Wang, X.; Fang, X.; Bando, Y.; Golberg, D. Adv. Funct. Mater. 2010, 20, 4233−4248. (15) Chen, D.; Huang, F.; Ren, G.; Li, D.; Zheng, M.; Wang, Y.; Lin, Z. Nanoscale 2010, 2, 2062−2064. (16) Fang, X. S.; Bando, Y.; Ye, C. H.; Golberg, D. Chem. Commun. 2007, 3048−3050. (17) Biswas, S.; Ghoshal, T.; Kar, S.; Chakrabarti, S.; Chaudhuri, S. Cryst. Growth Des. 2008, 8, 2171−2176. (18) Ghosh, P. K.; Maiti, U. N.; Jana, S.; Chattopadhyay, K. K. Appl. Surf. Sci. 2006, 253, 1544−1550. (19) Liang, C.; Shimizu, Y.; Sasaki, T.; Umehara, H.; Koshizaki, N. J. Phys. Chem. B 2004, 108, 9728−9733. (20) Zhai, T.; Gu, Z.; Fu, H.; Ma, Y.; Yao, J. Cryst. Growth Des. 2007, 7, 1388−1392. (21) Chai, L.; Du, J.; Xiong, S.; Li, H.; Zhu, Y.; Qian, Y. J. Phys. Chem. C 2007, 111, 12658−12662. (22) Lei, M.; Fu, X. L.; Li, P. G.; Tang, W. H. J. Alloys Compd. 2011, 509, 5769−5772. 8304

dx.doi.org/10.1021/jp3010635 | J. Phys. Chem. C 2012, 116, 8297−8304