A Facile Integration of Zero- (I–III–VI Quantum Dots) and One

Letter pubs.acs.org/NanoLett

A Facile Integration of Zero- (I−III−VI Quantum Dots) and One(Single SnO2 Nanowire) Dimensional Nanomaterials: Fabrication of a Nanocomposite Photodetector with Ultrahigh Gain and Wide Spectral Response Meng-Lin Lu,† Chih-Wei Lai,‡ Hsing-Ju Pan,‡ Chung-Tse Chen,† Pi-Tai Chou,*,‡ and Yang-Fang Chen*,† †

Department of Physics, National Taiwan University, Taipei 106, Taiwan, Republic of China Department of Chemistry, National Taiwan University, Taipei 106, Taiwan, Republic of China

‡

S Supporting Information *

ABSTRACT: Via the integration of nanocomposites comprising I−III−VI semiconductor quantum dots (QDs) decorated onto a single SnO2 nanowire (NW), we successfully fabricate an ultrahigh-sensitivity and wide spectral-response photodetector. Under the illumination of He−Cd laser (325 nm) with the photon energy larger than the band gap of SnO2 nanowire, remarkably, an ultrahigh photocurrent gain up to 2.5 × 105 has been achieved, and an enhancement factor can reach up to 700% (cf. bare SnO2 NW) as light illumination onto the wire with an excitation intensity of 15 W/m2. Also, a high gain value up to 1.3 × 105 is attained with the excited photon energy (488 nm) smaller than the band gap of SnO2 nanowire. Several key factors contribute to ultrahigh photocurrent gain and wide spectral response. First, the decorated quantum dot processes an inherent nature of a large absorption coefficient above its band gap. Furthermore, the single SnO2 nanowire provides an excellent conduction path for the photogenerated carriers as well as bears a large surface-to-volume ratio so that the coupling strength with quantum dots can be greatly enhanced. Most importantly, the spatial separation of photogenerated electrons and holes can be easily achieved due to the charge transfer arising from a type II band alignment between QDs and SnO2 NW. This work thus demonstrates a new approach in which by selectively decorating suitable QDs the photocurrent gain of SnO2 NWs can be greatly enhanced and extended to a wide spectral range of photoresponse previously inaccessible, providing a very useful guideline to create cheap, nontoxic, and highly efficient photodetectors. KEYWORDS: Tin oxide nanowires, wide spectral response and high gain, nanocomposite photodetector, I−III−VI quantum dots

H

known to be their high toxicity despite numerous advantageous features. To circumvent this disadvantage, a number of synthetic methodologies have been focusing on the fabrication of cadmium-free QDs, among which representatives are those ternary nontoxic I−III−VI QDs, such as CuInS2 (CIS) and CuInSe2 (CISe).11,12 These QDs with a direct band gap and a large absorption coefficient have frequently been used in photovoltaic devices.13 Also, certain CIS (core only) and CIS/ ZnS (core/shell) QDs in red and near-infrared emission region (650−830 nm) have been exploited for bioimaging both in vitro and in vivo.14,15 Different from the binary semiconductor QDs, for which the associated band gaps is tuned by the quantum size effect,16,17 the energy gap of ternary CuInS2 or quaternary ZnS−CuInS2 (ZCIS) QDs can be conveniently finetuned by invoking variation of the composition in multicomponents.18,19 This composition control provides a more

ybrid nanocomposites consisting of multicomponent nanomaterials have attracted significant attention due to their niches in multifunctionality and versatility, which cannot be found in one single component material. Especially, because nanomaterials possess the inherent nature of a large surface to volume ratio, the coupling strength among the constituents in the nanocomposite can be greatly enhanced, which enables us to generate many novel properties.1−4 Here, we provide a seminal attempt with the integration of zero- and one-dimensional nanomaterials, forming a nanocomposite photodetector with ultrahigh gain and wide spectral response. In this approach the colloidal luminescent zero-dimensional semiconductor nanoparticles (NPs), also known as quantum dots (QDs), have attracted our attention due to their unique optical properties such as a tunable wavelength, good photostability, and a high extinction coefficient.5,6 Among various types of semiconductor QDs, cadmium- and lead-based QDs have been widely applied in a variety of aspects, such as biolabeling, lasers, solar cells, and light-emitting diodes7−10 by taking these superiorities. However, a crucial limitation is © 2013 American Chemical Society

Received: November 8, 2012 Revised: March 18, 2013 Published: April 10, 2013 1920

dx.doi.org/10.1021/nl3041367 | Nano Lett. 2013, 13, 1920−1927

Nano Letters

Letter

As for the preparation of ZCIS QDs, a mixture of CuI (0.03 mmol), In(Ac)3 (0.3 mmol), and zinc diethyldithiocarbamate (0.03 mmol) were loaded into a three-necked flask containing ODE (4 mL), oleyamine (0.4 mmol), oleic acid (0.4 mmol), and dichlorodiphenyltrichloroethane (DDT, 1 mmol). Under vacuum with magnetic stirring for 1 h, the flask was refilled with nitrogen, and then the temperature was raised to 180 °C. After about 30 min, the resulting mixture was cooled to room temperature, and the purification of the yielding ZCIS QDs follows the same procedure as that of CIS QDs. As for the formation of ZCIS/ZnS core/shell NCs, the surface passivation of ZCIS core QDs was practiced via the cation exchange process.18 The cation precursor, zinc stearate (0.05 mmol), dissolved in ODE (0.5 mL), was added into the crude solution containing ZCIS prepared above at 180 °C. The mixture was then elevated up to 210 °C for a half hour, allowing the passivation of ZnS shell. After completion of surface passivation, the core/shell QDs were then obtained and purified via routine process elaborated above. The size of the as-prepared CIS, ZCIS, and ZCIS/ZnS QDs was measured with a transmission electron microscope (TEM, Philips/FEI Tecnai 20 G2 S-Twin, 200 kV). Typically, the sample was dropped on a carbon-membrane-coated nickel grid. The characterization of the as-prepared CIS, ZCIS, and ZCIS/ ZnS QDs was made by powder X-ray diffraction (XRD, model PANalytical X9 Pert PRO). Further compositional verification of each nanoparticle was provided by energy-dispersive X-ray spectroscopy (EDX), which was equipped on TEM (Philips/ FEI Tecnai 20 G2 S-Twin). Ultraviolet−visible steady-state absorption and emission spectra were recorded with a Hitachi (U-3310) spectrophotometer and an Edinburgh (FS920) fluorimeter, respectively. X-ray photoelectron spectroscopy (XPS) measurement was performed with a PHI 5000 VersaProbe scanning microprobe (ULVAC-PHI, Japan), which is a surface-sensitive technique to analyze the surface chemical compositions of core and core/shell QDs. The comparison of high-resolution scans of Cu 2p and Zn 2p photoelectron of CIS, ZCIS, and ZCIS/ZnS QDs was also made, which will be elaborated in later sections. Time-resolved photoluminescence (TRPL) measurement was carried out using the technique of time-correlated single-photon counting (TSPC), in which a pulsed diode laser (374 nm) with a repetition rate of 5 MHz served as the excitation source. The collected luminescence was directly projected into a Triax 320 monochromator and detected with a high-speed PMT. The photoluminescence signal is fed into a Time Harp counting card triggered with a signal from the diode laser. The growth of SnO2 NWs was based on vapor−liquid−solid process (VLS). In the synthesis process, an Au layer of 10 nm was first deposited on M-plane (100) sapphire (Al2O3) to serve as the catalyst. Then, Sn powder was placed on a ceramic boat and put into a furnace with Argon flow as gas carriers with a flow rate of 200 sccm. The temperature was increased from room temperature to 1000 °C at a rate of 100 °C/min. After the growth of 30 min at 1000 °C, the SnO2 NWs were obtained. The SnO2 NWs devices were made by the shadow-mask method,22 and Ni (15 nm)/Au (150 nm) electrodes were deposited on SnO2 NW to achieve Ohmic contacts. A treatment of thermal annealing at 500 °C for 60 s after contact deposition was necessary to minimize the contact resistance. Subsequently, the Cd-free I− III−VI QDs were then deposited by drop casting on the single SnO2 NW device. In order to improve the interface contact between SnO2 NW and QDs, after the drop casting, the

feasible way to obtain a specific range of emission wavelength without strictly controlling sizes and reaction times. In yet another approach, metal−oxide semiconductors based on one-dimensional (1D) nanostructures have been extensively investigated, among which 1D SnO2 nanowires (NWs) possessing a wide direct band gap and many unique optoelectronic properties have received intensive attention.20,21 For example, SnO2 NWs have a band gap of 3.6 eV and high quantum efficiency in emission in ultraviolet (UV) region, providing potential applications in many practical devices such as photodetectors and solar cells.22,23 SnO2 NWs have also been used in gas sensors, including CO, NO2, and so forth.24,25 Owing to oxygen defects on the SnO2 NWs surface, the upward band-bending forms a low-conductivity depletion layer. Upon photoexcitation to generate electron−hole pairs, holes drift to the surface through the build-up electric field, leaving unpaired electrons inside; thereupon the recombination probability of electrons and holes reduces, and the lifetime increases.26,27 Because NWs have a high surface-to-volume ratio, the surface of NWs influences the conductivity drastically. Recently, composite materials have been developed via combining catalytically active metals and/or p-type semiconductors on semiconductor NWs in an aim to improve the device sensitivity.28−32 In addition, by introducing the local piezopotential, it was able to tune Schottky barrier height at the metal contact and enhance the NW photodetector by up to 530%.33 In this contribution, new composite devices based on the functionalization of a single SnO2 NW surface by a series of Cdfree I−III−VI semiconductor QDs, including CIS, ZCIS, and ZCIS/ZnS, are fabricated and investigated. It is found that the photodetectors based on these new composites exhibit an ultrahigh gain and wide spectrum response covering from ultraviolet to visible radiation. With the decoration of CIS QDs on a single SnO2 NW, the photocurrent gain can be enhanced by up to 2.5 × 105 under the illumination of He−Cd laser (325 nm), and the gain value can be up to 1.3 × 105 with the excited photon energy (488 nm) that is much smaller than the band gap of the SnO2 nanowire. The results are remarkable and can be well-rationalized in term of the roles played by several important factors simultaneously, including the large absorption coefficient of QDs above band gap, the excellent conduction channel along the axis of NW, the strong coupling between QDs and NW due to a large surface to volume ratio, and the charge transfer arising from the designated type II band alignment between QDs and NW. This seminal study therefore may pave a new avenue for the generation of highly efficient, panchromatic photodetectors. Experimental Section. The preparation of the CIS core QDs for one bath experiment was carried out according to previous reports with a slight modification.14,42 The CIS QDs were made in a mixture of 0.03 mmol of CuI, 0.3 mmol of indium acetate, 1.2 mmol of 1-dodecanthiol, 0.4 mmol of oleic acid, 0.4 mmol of oleyl amine, and 4 mL of 1-octadecene in a three-necked flask and then was purged by N2 gas and heated at 100 °C for 1 h to remove any moisture and oxygen. Subsequently, the flask was refilled with nitrogen and the solution heated to 180 °C. After ∼20 min, heating was stopped immediately, and the reaction solution was cooled down to room temperature. The resulting CIS QDs were precipitated by adding excess amount of methanol and centrifuged at 9000 rpm for 5 min. This as-purified CIS will be used for characterization. 1921

dx.doi.org/10.1021/nl3041367 | Nano Lett. 2013, 13, 1920−1927

Nano Letters

Letter

samples were baked at 100 °C for two minutes to evaporate the deposited chemical solvent. This is a crucial step to obtain high photocurrent sensitivity. For the photocurrent measurement, we used a He−Cd laser working at 325 nm and an argon laser working at 488 nm as the excitation light source. A measurement system (Keithley 236) was utilized to supply the dc voltage (1 V) and to record the photocurrent. Results and Discussion. The TEM images of the asprepared CIS, ZCIS, and ZCIS/ZnS QDs, shown in Figure 1a, b, and c, respectively, revealed that the particle sizes were within 2.6−3.0 nm. Their uniformity was evidenced by the standard deviation of, e.g., σ < 15% in size distribution. The

enlarged TEM images clearly show lattice fringes, indicating that the studied QDs are single crystal (see Supporting Information 1). The XRD characteristic peaks of CIS and ZCIS nanocrystals were well-resolved for CuInS2 (JCPDS 38-0777) and Cu0.4In0.4Zn0.2S (JCPDS 47-1370) patterns as shown in Figure 1d. In comparison to the ZCIS core nanocrystals, the shift of characteristic peaks of ZCIS/ZnS to higher angles is attributed to the ZnS shell coating onto the core, providing an evidence to support for the successful core/shell formation.18 The EDX data shown in Figure 1e (CIS) and f (ZCIS) unveiled the elemental characteristic peaks of Cu, In, S, and Zn. In further quantifying analysis, the alloying composition of CIS, ZCIS, and ZCIS/ZnS are Cu:In:S = 19:37:44, Zn:Cu:In:S = 21:15:29:35, and Zn:Cu:In:S = 20:15:26:39, respectively. The combination of TEM, XRD, and EDX results therefore indicate that the preparation of CIS and ZCIS QDs and the passivation of ZnS shell onto the ZCIS QDs (less In atom composition) are successful. Figure 2a shows the FE-SEM image of the morphology of SnO2 NWs taken from a JSM-6500F FE-SEM system. The

Figure 2. (a) Field emission scanning electron microscope (FE-SEM) image of the as-grown SnO2 nanowires. (Inset: The X-ray diffraction (XRD) pattern of the as-grown SnO2 nanowires.) (b) Scanning electron microscope (SEM) image of the single SnO2 nanowire device. (c) I−V characteristics of the pristine SnO2 NW device. (d) Transmission electron microscope (TEM) image of QDs decorated single SnO2 nanowire.

XRD spectrum of SnO2 NWs, shown in inset of Figure 2a, can be employed to identify the structure of the as-synthesized product. Most of the peaks can be perfectly indexed according to the tetragonal rutile structure of SnO2. The typical SEM image of a single SnO2 NW device is illustrated in Figure 2b, in which, the channel distance of the channel between two terminals is estimated to be 10 μm, and the diameter of the NW is about 100 nm. As shown in Figure 2c, the I−V curves of the NWs reveal a well-behaved Ohmic characteristic. The TEM image of CIS QDs decorated on SnO2 NW is shown in Figure 2d, which displays a uniformly random distribution along the length of the NW. The estimated coverage of QDs on the whole surface of the SnO 2 NW is about 60%. The corresponding selected-area electron diffraction pattern (SAED) reveals that the SnO2 nanowire is structurally uniform and single crystalline with tetragonal rutile structure (see Supporting Information 2).

Figure 1. (a, b, c) Transmission electron microscope (TEM) images of CIS, ZCIS and ZCIS/ZnS core/shell QDs, and the average diameters are 2.7 ± 0.4 nm, 2.6 ± 0.2, and 3.0 ± 0.3 nm, respectively. CIS, ZCIS, and ZCIS/ZnS QDs with similar absorption peak at 460 nm were selected for TEM measurement. The particles are pointed out using a white dashed line. (d) X-ray diffraction (XRD) pattern of CIS, ZCIS and ZCIS/ZnS core/shell QDs. (e, f, g) Energy-dispersive X-ray spectroscopy (EDX) spectra of CIS, ZCIS, and ZCIS/ZnS core/ shell QDs, respectively. 1922

dx.doi.org/10.1021/nl3041367 | Nano Lett. 2013, 13, 1920−1927

Nano Letters

Letter

Figure 3. (a) Photoresponse of the pristine and other three NW devices deposited by CIS, ZCIS, and ZCIS/ZnS core/shell QDs NW devices with a bias of 1 V and under the illumination of He−Cd laser working at 325 nm with an excitation intensity of 15 W/m2. (b) The relation between the illumination power and the absolute photocurrent of the pristine and other three NW devices deposited by CIS, ZCIS, and ZCIS/ZnS core/shell QDs NW devices with a bias of 1 V and under the illumination of He−Cd laser working at 325 nm. (c) Photoresponse of the three NW devices deposited by CIS, ZCIS, and ZCIS/ZnS core/shell QDs with a bias of 1 V and under the illumination of an argon laser working at 488 nm with an excitation intensity of 15W/m2. (d) The relation between the illumination power and the absolute photocurrent of the three NW devices deposited by CIS, ZCIS, and ZCIS/ZnS core/shell QDs with a bias of 1 V and under the illumination of argon laser working at 488 nm.

SnO2 NW under the illumination of He−Cd laser as shown in Figure 3a. To have a more detailed examination, using the ZCIS/ZnS core/shell QDs decorated sample as a basis for comparison, the enhancement of the photocurrent for CIS QDs deposited on SnO2 NW is more than 400%, while that of the ZCIS QDs decorated sample can also be up to 200%. Figure 3d displays the corresponding photocurrents as a function of excitation power. The photocurrent gain (Γ) of the photoresponse, which determines the efficiency of electrons induced by photon and collected during the photocurrent measurement, can be obtained by22

Traditionally, upon illumination with photon energy larger than the energy band gap of SnO2 (3.6 eV), the conductivity of SnO2 NWs is expected to increase greatly due to the photogeneration of electron−hole pairs. Indeed, as shown in Figure 3a, the photoresponse of the pristine NW device under a bias of 1 V, and the illumination of He−Cd laser working at 325 nm with an excitation intensity of 15 W/m2 is very pronounced. The corresponding light power illuminated onto the NW device is estimated to be 15 pW. Quite interestingly, the photocurrents of the composite SnO2 NW decorated by I− III−VI QDs show a significant improvement, for which the enhancement can reach up to 700%, 400%, and 200% for CIS, ZCIS, and ZCIS/ZnS core/shell QDs, respectively. Figure 3b shows the dependence of the photocurrent on the excitation power, indicating that the photocurrent indeed increases with the excitation power. Once the photon energy is smaller than the band gap, in theory, electron−hole pairs should not be generated, resulting in negligible photocurrent. Figure 3c shows the photocurrent of the four samples for the pristine SnO2 NW and SnO2 NW decorated with CIS, ZCIS, and ZCIS/ZnS core/shell QDs under a bias of 1 V and the illumination of a cw argon laser working at 488 nm with an excitation intensity of 15 W/m2. As expected, for the pristine SnO2 NW, the photocurrent is hardly detectable. However, quite surprisingly, even though the photon energy of 488 nm electromagnetic wave is smaller than the band gap of SnO2 NW, the measured photocurrent of these composite devices is as pronounced as that of the pristine

Γ=

Δi/q 1 × P /hν η

(1)

where Δi is the current difference between photocurrent and dark current, q is the electron charge, and hν is the photon energy of incident light. P is the power of photon that the nanowire absorbs, i.e., P = I × l × d where I is the excitation intensity illuminating the pattern, and l and d are the length and the width of the nanowire, respectively. In eq 1, the efficiency (η) of the photon absorption to generate electron−hole pairs is dominated by the quantum efficiency ηq = 1 − e−αd, where α is the absorption coefficient. In this approach, η is set to be unity for simplicity as long as the tendency rather than absolute value is emphasized throughout this work. Under the illumination of He−Cd laser (325 nm) with an excitation intensity of 1 W/m2, the photocurrent gain for the 1923

dx.doi.org/10.1021/nl3041367 | Nano Lett. 2013, 13, 1920−1927

Nano Letters

Letter

pristine single SnO2 NW is about 4 × 104. Quite interestingly, the gain values can be up to 2.5 × 105, 1.3 × 105, and 5.8 × 104 for the single SnO2 NW devices decorated by CIS, ZCIS, and ZCIS/ZnS core/shell QDs, respectively. Meanwhile, under the illumination of argon laser (488 nm) with an excitation intensity of 1 W/m2, the gains for the ZCIS/ZnS core/shell QDs and ZCIS QDs decorated SnO2 NW can reach up to 1.9 × 104 and 7 × 104, respectively. More pronounced result is observed for the CIS QDs decorated NW, for which the photocurrent gain is measured to be as high as 1.3 × 105. Notably, under similar experimental conditions, the photocurrent gain of the CIS-NW composite device in the visible range is larger than that of the pristine NW device in the UV range. The performance of the fabricated devices is quite reproducible under the similar experimental conditions (see Supporting Information 3). Let us now compare our present result with some works published previously.32,33 To enhance the photoresponsivity, Yang et al. introduced a local strain to manipulate the Schottky barrier height at the metal electrodes.33 Compared with our current investigation, except a smaller enhancement factor, the requirement of an external mechanical strain in their study is less convenient for practical application. Besides, the response spectrum is still limited by the band gap absorption of the NW. For the comparison with our preliminary work,32 our current result has a much higher photocurrent gain, and the decorated QDs are cheaper and more environmentally friendly. Therefore, it is of prime importance to gain insight into the underlying mechanism that causes the gigantic photocurrent enhancement after the deposition of ternary nontoxic I−III−VI QDs. It is well-known that the ternary nontoxic I−III−VI QDs have a direct band gap and a large absorption coefficient, which have been used in optoelectronic devices previously.13 This inherent nature of the optical property held by the QDs definitely provides an important contribution to the above observed phenomena. Since SnO2 is a wide band gap material of 3.6 eV, the high gain of QDs/SnO2 NW composite devices under 488 nm illumination is therefore more plausibly attributed to the deposited Cd-free I−III−VI QDs. Next, the channel along the axis of NW can serve as an excellent conduction path for the transport of photogenerated electron− hole pairs.22,23 In addition, due to a large surface-to-volume ratio, the coupling strength between QDs and NW can be greatly enhanced, which makes the effect of the decoration of QDs more pronounced. Most importantly, let us examine the band alignment between the QDs and NWs studied here. Comparing the energy levels between the titled QDs and SnO2 NW (see Figure 4),32 the studied composites possess a type II band alignment.34,35 Upon visible-light excitation, the photogenerated electrons in QDs drift into the conduction channel of SnO2 NW and contribute to the detected photocurrent, the consequence of which reduces the rate of exciton recombination. The spatial separation of electrons and holes further induces changes of the built-in surface electric field. According to the result of simulation reported by Garrido et al., the photocurrent induced by the modulation of surface charge regions (SCRs) gives a current gain Γ that follows an inverse power law with respect to the excitation intensity, i.e., Γ ∝ I−κ where κ in the exponent term is between 0.5 and 0.9.36 Figure 5a and b shows the gain logarithmic plot versus intensity of pristine NW and QDs-NW composite devices under 325 and 488 nm illumination, respectively. Accordingly, κ is deduced to

Figure 4. Proposed operation mechanism of QDs and SnO2 NWs composite device under the argon laser (488 nm) illumination.

be in between 0.6 and 0.8 for both cases, which is in good agreement with the aforementioned theoretical prediction of surface band-bending effect. The above proposal, i.e., the photoinduced electron transfer under the type II band alignment, seems to promptly explain our observation. However, upon taking a closer inspection, further investigation is necessary to gain more insight into the difference of the performance among various QD−SnO2 NW composites, in which the experiment has been carried out under the condition that these Cd-free I−III−VI QDs were strategically synthesized so that they possess similar band gap, with the lowest lying absorption peak wavelength (shoulder) around ∼470−490 nm (see Supporting Information S4). Although the peak wavelengths of ZCIS and ZCIS/ZnS QDs are noted to be slightly deceased upon incorporating Zn (c.f. CIS),18 such a slim difference in band gap should not be fully responsible to the substantial difference in photocurrent enhancement. To render a more in-depth explanation, let us consider the recombination process of charge carriers in QDs. On the one hand, the quantum yield (QY) of CIS QDs can be enhanced by substituting Cu with Zn atoms, forming ZCIS, due to the fact that the diffused Zn species might preferentially replace Cu sites since the ionic diameter of Zn2+ is close to that of Cu+. Such a Zn substitution of Cu was known to be energetically favorable to prevent the formation of antisite defects.18 On the other hand, QY of ZCIS/ZnS core/shell QDs is larger than that of ZCIS QDs based on the superiority offered by core/shell configuration: (i) Lattice mismatch between CIS and ZnS is relatively low (2−3%), resulting in the formation of a smooth, low-defect interface. (ii) The confinement of charge carriers in the core region due to the fact that the band gap of ZnS (Eg = 3.6 eV) is much larger than that of CIS (Eg = 1.5 eV) with a type I band alignment.18,37 In order to support the above delineation we have performed time-resolved PL (TRPL) measurement of the titled QDs. Upon the generation of electron−hole pairs by the incident light, the observed lifetime of exciton is governed by both radiative and nonradiative deactivation processes. The increase of nonradiative pathways may imply more surface vacancies to trap electrons, resulting in lower efficiency of the radiative recombination and consequently shortening of the measured lifetime. In a type II QDs-SnO2 NW band alignment, upon visible light excitation of QDs, the trapped electrons will flow across the interface into the conduction band of SnO2 NW, increasing the probability of charge transfer and hence the 1924

dx.doi.org/10.1021/nl3041367 | Nano Lett. 2013, 13, 1920−1927

Nano Letters

Letter

Figure 5. (a) The gain logarithmic plot versus intensity of the pristine NW and other NW devices deposited by CIS, ZCIS, and ZCIS/ZnS core/ shell QDs under the illumination of He−Cd laser (325 nm). (b) The gain logarithmic plot versus intensity of the three NW devices deposited by CIS, ZCIS, and ZCIS/ZnS core/shell QDs under the illumination of argon laser (488 nm).

obviously enhanced. Based on the experimental data, the radiative decay rate constant of the as-prepared QDs was deduced to be in the similar magnitude of 1.1 × 106 to 2.0 × 106 s−1, and the nonradiative decay rate constant was 22.9 × 106 s−1 for CIS, 9.5 × 106 s−1 for ZCIS, and 5.6 × 106 s−1 for ZCIS/ZnS core/shell QDs. The calculated result indicates that the amount of electrons trapped at the nonradiative defects is largest for CIS QDs. Obviously, CIS QDs possess the shortest observed lifetime as well as lowest QY and hence the greatest population for surface trapped electrons (nonradiative pathway) among three titled QDs studied. Accordingly, the CIS QDs decorated SnO2 NW sample is expected to exhibit the highest photocurrent gain, consistent with the photocurrent measurement shown in Figures 3 and 5. Figure 6 also shows the PL decay curves of CIS, ZCIS, and ZCIS/ZnS core/shell QDs coated on the SnO2 NWs. Upon a close examination, their deduced lifetime (long component) is shorter than that without depositing on NWs, which are 11 ns for CIS, 39 ns for ZCIS, and 108 ns for ZCIS/ZnS core/shell QDs. This observation indicates that the photocurrent enhancement is affiliated with the charge transfer through surface defects. More insight into the trapping states can be provided via the XPS measurement. Given a similar size, Zn2+ may simultaneously replace the surface’s Cu+ ions during the growth process of ZCIS core QDs or passivation treatment of ZCIS/ ZnS core−shell QDs.18 As shown in Figure 7a, the spin−orbit coupled Cu 2p3/2 and 2p1/2 peaks of the CIS QDs exhibit a higher signal intensity than those of ZCIS QDs. Moreover,

amplification of photocurrent gain as shown in Figure 4. Along this line, the PL decay dynamics for CIS, ZCIS, and ZCIS/ZnS core/shell QDs are measured and shown in Figure 6, and the

Figure 6. Photoluminescence (PL) decay curves of CIS, ZCIS, and ZCIS/ZnS core/shell QDs (solid); the PL decay curves of CIS, ZCIS, and ZCIS/ZnS core/shell QDs on the SnO2 nanowires (hollow).

corresponding PL decay times of CIS, ZCIS, and ZCIS/ZnS core/shell QDs in the solid state are fitted to be 41, 88, and 130 ns, respectively, for the long decay component. Concurrently, QYs for CIS, ZCIS core, and ZCIS/ZnS core/shell QDs were measured to be 4.5%, 13.5%, and 26%, respectively. Compared to the QY of intact core QDs, that of ZnS-passivated QDs was

Figure 7. (a, b) X-ray photoelectron spectroscopy (XPS) spectra of CIS, ZCIS, and ZCIS/ZnS core/shell quantum dots. 1925

dx.doi.org/10.1021/nl3041367 | Nano Lett. 2013, 13, 1920−1927

Nano Letters

Letter

for the creation of cheap, nontoxic, and panchromatic photodetectors with high efficiency.

upon ZnS shell passivation on ZCIS QDs, Cu 2p peaks of core−shell QDs are further suppressed, whereas the Zn 2p peaks in Figure 7b unveil stronger signal intensity than that of core-only ZCIS QDs, in a qualitative manner, explaining that the ZCIS/ZnS QDs consist of a Cu-deficient nanostructure. Evidently, the aforementioned data draw a conclusion that the population of Cu atom on QDs’ surface is in the order of CIS > ZCIS > ZCIS/ZnS, implying that there are more Cu-based defects on the surface of CIS QDs.18 It is well-known that one of the major chalcopyrite semiconductors crystal defects is related to the Cu deficiency, which provides facile nonradiative relaxation pathways and consequently results in the decrease of emission intensity.37,38 With the introduction of Zn atoms using the deposition process, the existing Cu-deficient defects, such as Cu vacancies and, in part, antisite defects can be efficiently eliminated,18 the result of which reduces the probability of charge trapping and the number of nonradiative deactivation pathways. Thus, it became convincing that, under the light illumination, the increasing number of photoexcited electrons can be accumulated on the surface trapped sites for CIS QDs, followed by their transferring to the conduction band of SnO2 NWs, leading to the significant enhancement of photocurrent. The lowest photocurrent being enhanced in ZCIS/ZnS core/sell QDs may be due additionally to the type I band gap arrangement, from which the charge transfer, in theory, may have to take place from a tunneling mechanism, so that the delocalization of carrier wave functions can be extended into the outer semiconductor (SnO2 NW) cladding.39 In order to provide the additional evidence to support the mechanism for the transfer of photogenerated carriers from QDs to NW, we have performed the photoluminescence (PL) measurement (see Supporting Information S5). The PL intensity of ZCIS/ ZnS core/shell QDs deposited on SnO2 NWs is quenched. This phenomenon is consistent with the expectation that the photogenerated electrons can transfer from the conduction band (CB) or from radiative defect levels of QDs into the conduction band of SnO2 NW, which enables to reduce the radiative recombination probability in QDs.40,41 Conclusion. In summary, based on a novel nanocomposite comprising I−III−VI QDs and SnO2 NW with a type II band alignment, the sensitivity and spectral response of the asprepared photodectors have been remarkably improved. The gain value can reach up to 2.5 × 105 as CIS QDs are deposited on the SnO2 NW under the illumination of He−Cd laser, which represents a large enhancement factor of 700% compared with the bare SnO2 NW. Even though the incident photon energy is smaller than the band gap of SnO2 NW, the gain value can also be up to 1.3 × 105 of the illumination of 488 nm laser, hence greatly extending the detectivity toward the visible range. Several key factors play important roles simultaneously to achieve the obtained results, including the large absorption coefficient of QDs, the excellent conduction path along the axis of NW, the strong coupling between QDs and NW due to a large surface-to-volume ratio, and the charge transfer arising from the type II band alignment between QDs and NW. The underlying mechanism incorporating photoinduced charge transfer has been supported by PL, TRPL, and XPS measurements. Knowing that the high gain of photocurrent conventionally only exists under the illumination with photon energy larger than the band gap of the active e.g. SnO2 NW, this new approach therefore shows a brightening prospect

■

ASSOCIATED CONTENT

* Supporting Information S

Enlarged TEM image of QDs, structure information for tin oxide NWs, statistics of the device performance of the CIS QDs decorated SnO2 NW photodetector, absorption spectra of ternary I−III−VI QDs, and comparison of PL spectra of ZCIS/ ZnS core−shell QDs with and without SnO2 NW. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

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

M.-L.L. and C.-W.L. contributed equally to this work. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work is supported by grants from National Science Council and Ministry of Education of the Republic of China. We like to thank Professor Liang for support concerning the TEM image.

■

REFERENCES

(1) Novak, B. M. Adv. Mater. 1993, 5, 422−433. (2) Lee, J. M.; Choung, J. W.; Yi, J.; Lee, D.; Samal, M.; Yi, D. K.; Lee, C.-H.; Yi, G.-C.; Paik, U.; Rogers, J. A.; Park, W. I. Nano Lett. 2010, 10, 2783−2788. (3) Zhu, R.; Chung, C. H.; Cha, K. C.; Yang, W. B.; Zheng, Y. B.; Zhou, H. P.; Song, T. B.; Chen, C. C.; Weiss, P. S.; Li, G.; Yang, Y. ACS Nano 2011, 5, 9877−9882. (4) Lu, L. M.; Weng, T. M.; Chen, J. Y.; Chen, Y. F. NPG Asia Mater. 2012, 4, e26. (5) Weller, H. Adv. Mater. 1993, 5, 88−95. (6) Murray, C. B.; Norris, D. J.; Bawendi, M. G. J. Am. Chem. Soc. 1993, 115, 8706−8715. (7) Bruchez, M.; Moronne, M.; Gin, P.; Weiss, S.; Alivisatos, A. P. Science 1998, 281, 2013−2016. (8) Klimov, V. I.; Mikhailovsky, A. A.; Xu, S.; Malko, A.; Hollingsworth, J. A.; Leatherdale, C. A.; Eisler, H. J.; Bawendi, M. G. Science 2000, 290, 314−317. (9) Colvin, V. L.; Schlamp, M. C.; Alivisatos, A. P. Nature 1994, 370, 354−357. (10) Michalet, X.; Pinaud, F. F.; Bentolila, L. A.; Tsay, J. M.; Doose, S.; Li, J. J.; Sundaresan, G.; Wu, A. M.; Gambhir, S. S.; Weiss, S. Science 2005, 307, 538−544. (11) Nakamura, H.; Kato, W.; Uehara, M.; Nose, K.; Omata, T.; Otsuka-Yao-Matsuo, S.; Miyazaki, M.; Maeda, H. Chem. Mater. 2006, 18, 3330−3335. (12) Malik, M. A.; O’Brien, P.; Revaprasadu, N. Adv. Mater. 1999, 11, 1441−1444. (13) Panthani, M. G.; Akhavan, V.; Goodfellow, B.; Schmidtke, J. P.; Dunn, L.; Dodabalapur, A.; Barbara, P. F.; Korgel, B. A. J. Am. Chem. Soc. 2008, 130, 16770−16777. (14) Li, L.; Daou, T. J.; Texier, I.; Tran, T. K. C.; Nguyen, Q. L.; Reiss, P. Chem. Mater. 2009, 21, 2422−2429. (15) Pons, T.; Pic, E.; Lequeux, N.; Cassette, E.; Bezdetnaya, L.; Guillemin, F.; Marchal, F.; Dubertret, B. ACS Nano 2010, 4, 2531− 2538. (16) Alivisatos, A. P. Science 1996, 271, 933−937. (17) Lai, C. W.; Wang, Y. H.; Chen, Y. C.; Hsieh, C. C.; Uttam, B. P.; Hsiao, J. K.; Hsu, C. C.; Chou, P. T. J. Mater. Chem. 2009, 19, 8314− 8319. 1926

dx.doi.org/10.1021/nl3041367 | Nano Lett. 2013, 13, 1920−1927

Nano Letters

Letter

(18) Zhang, W. J.; Zhong, X. H. Inorg. Chem. 2011, 50, 4065−4072. (19) Feng, J.; Sun, M.; Yang, F.; Yang, X. Chem. Commun. 2011, 47, 6422−6424. (20) Lee, S. H.; Jo, G.; Park, W.; Lee, S.; Kim, Y. S.; Cho, B. K.; Lee, T.; Kim, W. B. ACS Nano 2010, 4, 1829−1836. (21) Lu, M. L.; Lin, H. Y.; Chen, T. T.; Chen, Y. F. Appl. Phys. Lett. 2011, 99, 0911106. (22) Lin, C. H.; Chen, R. S.; Chen, T. T.; Chen, H. Y.; Chen, Y. F.; Chen, K. H.; Chen, L. C. Appl. Phys. Lett. 2008, 93, 112115. (23) Snaith, H. J.; Ducati, C. Nano Lett. 2010, 10, 1259−1265. (24) Kolmakov, A.; Zhang, Y. X.; Cheng, G. S.; Moskovits, M. Adv. Mater. 2003, 15 (12), 997−1000. (25) Choi, Y. J.; Hwang, I. S.; Park, J. G.; Choi, K. J.; Park, J. H.; Lee, J. H. Nanotechnology 2008, 19, 095508. (26) Soci, C.; Zhang, A.; Xiang, B.; Dayeh, S. A.; Aplin, D. P. R.; Park, J.; Bao, X. Y.; Lo, Y. H.; Wang, D. Nano Lett. 2007, 7, 1003− 1009. (27) Prades, J. D.; Hernandez-Ramirez, F.; Jimenez-Diaz, R.; Manzanares, M.; Andreu, T.; Cirera, A.; Romano-Rodriguez, A.; Morante, J. R. Nanotechnology 2008, 19 (46), 465501. (28) Kolmakov, A.; Klenov, D. O.; Lilach, Y.; Stemmer, S.; Moskovits, M. Nano Lett. 2005, 5, 667−673. (29) Chen, X. H.; Moskovits, M. Nano Lett. 2007, 7, 807−812. (30) Lin, C. H.; Chen, T. T.; Chen, Y. F. Opt. Express 2008, 16, 16916−16922. (31) Lu, M. L.; Lin, T. Y.; Weng, T. M.; Chen, Y. F. Opt. Express 2011, 19, 16266−16272. (32) Lu, M. L.; Lin, C. H.; Chen, Y. F. Appl. Phys. Lett. 2011, 99, 081109. (33) Yang, Q.; Guo, X.; Wang, W.; Zhang, Y.; Xu, S.; Lien, D. H.; Wang, Z. L. ACS Nano 2010, 4, 6285−6291. (34) Chou, P. T.; Chen, C. Y.; Cheng, C. T.; Pu, S. C.; Wu, K. C.; Cheng, Y. M.; Lai, C. W.; Chou, Y. H.; Chiu, H. T. ChemPhysChem 2006, 7, 222−228. (35) Zhong, H.; Lo, S. S.; Mirkovic, T.; Li, Y.; Ding, Y.; Li, Y.; Scholes, G. D. ACS Nano 2010, 4, 5253−5262. (36) Garrido, J. A.; Monroy, E.; Izpura, I.; Munoz, E. Semicond. Sci. Technol. 1998, 13, 563−568. (37) Nam, D. E.; Song, W. S.; Yang, H. J. Mater. Chem. 2011, 21, 18220−18226. (38) Uehara, M.; Watanabe, K.; Tajiri, Y.; Nakamura, H.; Maeda, H. J. Chem. Phys. 2008, 129, 134709. (39) Kuno, M.; Fromm, D. P.; Johnson, S. T.; Gallagher, A.; Nesbitt, D. J. Phys. Rev. B 2003, 67, 125304. (40) Wu, H. C.; Huang, Y. C.; Ding, I. K.; Chen, C. C.; Yang, Y. H.; Tsai, C. C.; Chen, C. D.; Chen, Y. T. Adv. Funct. Mater. 2011, 21, 474−479. (41) Yang, Y.; Zhong, H. Z.; Bai, Z. L.; Zou, B. S.; Li, Y. F.; Scholes, G. D. J. Phys. Chem. C 2012, 116, 7280−7286. (42) Zhong, H. Z.; Zhou, Y.; Ye, M. F.; He, Y. J.; Ye, J. P.; He, C.; Yang, C. H.; Li, Y. F. Chem. Mater. 2008, 20, 6434−6443.

1927

dx.doi.org/10.1021/nl3041367 | Nano Lett. 2013, 13, 1920−1927