Home
Search
Collections
Journals
About
Contact us
My IOPscience
Evaluation of the piezoelectric properties and voltage generation of flexible zinc oxide thin films
This content has been downloaded from IOPscience. Please scroll down to see the full text. 2015 Nanotechnology 26 215704 (http://iopscience.iop.org/0957-4484/26/21/215704) View the table of contents for this issue, or go to the journal homepage for more
Download details: IP Address: 130.133.8.114 This content was downloaded on 10/05/2015 at 13:06
Please note that terms and conditions apply.
Nanotechnology Nanotechnology 26 (2015) 215704 (9pp)
doi:10.1088/0957-4484/26/21/215704
Evaluation of the piezoelectric properties and voltage generation of flexible zinc oxide thin films M Laurenti1,2,5, S Stassi2,5, M Lorenzoni3,4, M Fontana1,2, G Canavese1,2, V Cauda1 and C F Pirri1,2 1
Center for Space Human Robotics, Istituto Italiano di Tecnologia, C.so Trento 21, 10129 Turin, Italy Department of Applied Science and Technology, Politecnico di Torino, C.so Duca degli Abruzzi 24, I-10129 Turin, Italy 3 Nanophysics, Fondazione Istituto Italiano di Tecnologia, Via Morego, 30, I-16163 Genoa, Italy 4 Barcelona Microelectronics Institute, IMB-CNM (CSIC), 08193 Bellaterra, Spain 2
E-mail: [email protected] Received 28 January 2015, revised 23 March 2015 Accepted for publication 2 April 2015 Published 6 May 2015 Abstract
Local piezoresponse and piezoelectric output voltage were evaluated on ZnO thin films deposited by radio-frequency magnetron sputtering on hard Si/Ti/Au and flexible Cu-coated polyimide substrates. Three different thicknesses of ZnO films were studied (285 nm, 710 nm, and 1380 nm), focusing on characteristics like crystallinity, grain size, surface roughness, and morphology. Independent of the nature of the metal layer and the substrate, our results show that thicker films presented a higher level of crystallinity and a preferential orientation along the c-axis direction, as well as a lower density of grain boundaries and larger crystal sizes. The improvement of the crystalline structure of the material directly enhances its piezoelectric properties, as confirmed by the local characterizations performed by piezoresponse force microscopy and by the evaluation of the output voltage generation under the application of a periodical mechanical deformation on the whole film. In particular, the highest value of the d33 coefficient obtained (8 pm V−1) and the highest generated output voltage (0.746 V) belong to the thickest films on hard and flexible substrates, respectively. These results envision the use of ZnO thin films—particularly on flexible substrates—as conformable, reliable, and efficient active materials for use in nanosensing, actuation, and piezoelectric nanogenerators. Keywords: Zinc oxide, piezoelectricity, sputtering, flexible (Some figures may appear in colour only in the online journal) 1. Introduction
[1, 6, 7] and energy-harvesting systems [2, 8, 9]. High piezoelectric coefficient d33 values ranging from 14.3 pmV−1 to 26.7 pmV−1 were found in the characterization of ZnO nanobelts [10], while the first attempts to convert mechanical energy into electrical energy were carried out by means of ZnO NWs [11]. In addition, flexible, high-output nanogenerators based on ZnO NWs were successfully demonstrated [12]. These results supported the application of ZnO nanostructures as nanosensors and nanogenerators. The possibility of regulating the charge-carrier transport properties in ZnObased nanomaterials was also reported on both hard and flexible devices based on one-dimenisonal (1D) ZnO
Zinc oxide (ZnO) is a well known wide-bandgap (∼3.37 eV) metal oxide semiconductor that can be easily prepared in the form of thin films and different micro- and nanostructures, such as micro- [1] and nanowires (NWs) [2–4] and nanorods (NRs) [5]. Because of its semiconducting nature and the presence of piezoelectricity, in recent years there has been increased interest in studying ZnO in hopes of exploiting its physical properties in the fabrication of sensing devices 5
Contributed equally to this work.
0957-4484/15/215704+09$33.00
1
© 2015 IOP Publishing Ltd Printed in the UK
M Laurenti et al
Nanotechnology 26 (2015) 215704
single-point PFM spectroscopy, reported here for the first time for a sputtered ZnO thin film grown on a conductive support. Further insight into the piezoelectric characterization was obtained through piezoelectric output voltage measurements applied to flexible sputtered ZnO thin films, with the further aim of using them as functional materials for sensing and actuating applications or as piezoelectric nanogenerators. In particular, ZnO thin films on flexible substrates are shown to be ideal candidates for the fabrication of conformable, reproducible, large-area distributed sensor arrays, thanks to the possibility of depositing piezoelectric thin films through a completely low-temperature and plasma-assisted deposition method that is fully compatible with semiconductor technology. Piezoelectric phenomena of the resulting flexible ZnO thin films can thus be successfully exploited to convert and detect mechanical stress into electrical signals. Piezoelectric ZnO thin films on flexible substrates are also interesting for energy recovery from voluntary and involuntary body movements, since they are easily conformable to curved surfaces and lead to remarkable energy generation.
nanostructures [2, 13]. Although 1D nanostructures are attractive for fundamental research studies, some limitations restrict their use in industrial applications. Indeed, the synthesis techniques for 1D nanostructures still do not guarantee sufficient uniformity of the as-grown materials, especially in terms of dimensions and morphology, with a strong lack of reproducibility and a worsening in the performance of the obtained devices. For this reason, the well-known thin film technology could be a valid alternative for the synthesis of piezoelectric materials, being fully compatible with the semiconductor technology and with the microfabrication processes. Piezoelectric ZnO thin films have been widely used to fabricate pressure sensors [14], actuators [15], ultrahigh-frequency surface acoustic wave devices [16], and energy-harvesting architectures [17]. Nevertheless, most of these applications (i.e., acoustic wave-based devices) rely on the transversal piezoelectric effect, which has been thoroughly investigated for ZnO-based applications [18]. On the contrary, the piezoelectric properties and the charge-generation phenomena related to the longitudinal piezoelectric effect in (002)-oriented ZnO thin films are still under investigation [19, 20]. Recently, the piezoelectric response of porous ZnO films was successfully demonstrated [8]. However, the presence of a polycrystalline structure in this case is expected to lower the piezoresponse, since highly (002)-oriented ZnO films, together with a compact nanostructure, should be pursued to maximize the piezoelectricity of the material [21]. Sputtered ZnO thin films commonly grow with a preferential (002) crystal orientation, and d33 values ranging from 2 up to 12.4 pmV−1 have already been reported for highly (002)oriented ZnO films grown by different techniques [22] and analyzed by piezoresponse force microscopy (PFM). However, to the best of our knowledge no results concerning the evaluation of the converse piezoelectric effect of ZnO nanostructures have been reported. PFM is a commonly used technique for the investigation of the polarity distribution, local piezoelectric properties, and estimation of the longitudinal piezoelectric coefficient (d33) [10, 23, 24]. A helpful approach to maximize the piezoelectric response of ZnO films is the use of thermal treatments, which are generally required to increase the resistivity of ZnO films and to promote their (002) crystal orientation [25]. However, room-temperature depositions are preferable in view of using polymeric substrates for the fabrication of flexible devices. In this work, we report on the deposition and the exhaustive characterization of ZnO thin films grown by the sputtering technique at room temperature on both hard Si/Ti/ Au substrates and flexible Cu-coated polyimide (PI) foils. The morphology and crystal structure of ZnO films were analyzed by means of field emission scanning electron microscopy (FESEM) and x-ray diffraction (XRD). The PFM was exploited to study the local piezoelectric response of the ZnO thin films grown on flat Si/Ti/Au conductive substrates, and to estimate the piezoelectric coefficient, d33. PFM also allowed us to evaluate the local piezoelectric behavior of the ZnO film when working in the actuating mode. In this case, the analysis of the local mechanical displacement under the application of an external electric field was investigated by
2. Experimental 2.1. Deposition of ZnO thin films
ZnO films were deposited by the radio-frequency (RF) magnetron sputtering technique. Both hard Si/Ti/Au wafers and commercial flexible PI/Cu foils were considered as conductive substrates. All the depositions were carried out at room temperature, starting from a ceramic ZnO target. ZnO depositions were performed in a mixed atmosphere of Ar (95%) and O2 (5%) at a pressure of 5 × 10−3 Torr. Suitable vacuum conditions, with base pressures ranging between 3.1 × 10−7 and 3.5 × 10−7 Torr, were achieved before starting the deposition processes. Different film thicknesses were obtained by using three deposition times: 1, 2, and 4 h. The average thicknesses of ZnO thin films are summarized in table 1. 2.2. Materials characterization
The crystal structure and orientation of the ZnO thin films were evaluated by using a Panalytical X’Pert x-ray diffractometer in the Bragg–Brentano configuration (characteristic wavelength, λ = 1.54 059 Å), while the morphology was investigated by FESEM measurements obtained with a Carl Zeiss Dual Beam Auriga. PFM characterization was carried out on ZnO thin films deposited on hard and flat Si/Ti/Au substrates using MFP-3D (Asylum Research, an Oxford Instruments company, USA). The experimental measurements were performed at the first-mode contact resonance of 300 kHz with a silicon Pt-coated cantilever (AC240TM Olympus ‘ElectriLever’) with a nominal stiffness of 2 N m−1 and a first-mode resonance frequency of 70 kHz. To get semiquantitative information, the PFM setup was calibrated on a standard sample, and the following measurements were performed under the same conditions, using the same tip. The 2
M Laurenti et al
Nanotechnology 26 (2015) 215704
orientation along the c-axis direction. Independently from the kind of substrate and metal, the intensity of the (002) peak increases with an increase in the film’s thickness. The (002) peak position for ZnO reported by JCPDSICDD (card n. 89-1397) and considered as a reference in this work is 2θ = 34.37°. Comparing this value with those reported in table 2, one can see that the peak position of the thickest samples, ZnO_4h (34.44°), is closer to the reference sample, while it slightly shifts towards higher 2θ angles for thinner films. The FWHM values estimated for each sample are summarized in table 2. A decrease in this value can be observed when the film thickness increases, with the minimum values reported in the case of the thickest samples, ZnO_4h. Since the deposition parameters—except for the deposition time—are the same for all three samples, the observed changes in the crystal properties of the analyzed samples can be related to the presence of different metal layers underneath the film (i.e., Au and Cu). Since the lattice constants of these two metals are different from each other, the lattice mismatch at the metal-oxide interface is also different and could affect the crystal properties of the overlying ZnO layer. From the comparison of the (002) peak positions and the FWHM values reported in table 2, some differences can be observed between samples ZnO_1h and ZnO_2h when the metal layer changes from Au to Cu, while a good agreement in the reported data is observed for samples ZnO_4h, independent of the metal used. The effect of the different metal layers is thus more pronounced for the thinner samples, ZnO_1h and ZnO_2h, but when the ZnO thickness increases up to 1380 nm, the influence of the underlying metal on the ZnO crystal properties is no longer visible. Despite the use of different interface metal layers, the better crystallinity of samples ZnO_4h can be highlighted, with a 2θ angle closer to the JCPDS-ICDD reference position, and a minimum in the FWHM value (close to 0.2°). Figure 2 shows FESEM images of ZnO thin films deposited on hard and flexible substrates, while FESEM images representative of the Au and Cu layer morphology are shown in figures 3(a) and (b), respectively. The average grain size increases with an increase in the ZnO film thickness. In the case of Si/Ti/Au substrates, a fine-grained, smooth surface with small, rounded grains (average grain size between 20 nm and 40 nm), together with the presence of some voids, is obtained for sample ZnO_1h (see figure 2(a)). When the film thickness is increased to 710 nm, the ZnO surface is characterized by irregularly sized, densely packed grains with no voids and an average dimension between 30 nm and 80 nm (see figure 2(b)). Figure 2(c) shows the surface morphology of sample ZnO_4h. Despite the fact that the grain size does not increase further (the average grain size is similar to the one of sample ZnO_2h), a change in the grain shape can be highlighted. In particular, one can see that the crystal grains grow as long hexagonal rods for sample ZnO_4h, while circular grains are still present in samples ZnO_1h and ZnO_2h. FESEM results of ZnO films deposited on PI/Cu foils are shown in figures 2(d)–(f). The surface morphology changes from the island shape with small crystal grains in samples ZnO_1h and ZnO_2h (see figures 2(d) and (e)) to the more
Table 1. Average thickness, d33 value, and output voltage of sputtered ZnO thin films.
Sample
Average thickness [nm]
d33a [pmV−1]
Output voltageb [V]
ZnO_1h ZnO_2h ZnO_4h
285 710 1380
5.0 5.3 8.0
0.058 0.361 0.746
a
Values refer to ZnO thin films deposited on hard Si/Ti/Au substrates. Values refer to ZnO thin films deposited on flexible PI/Cu foils, under a periodical mechanical stimulus of 30 N. b
periodically poled lithium niobate test sample consists of a 3 mm × 3 mm LiNbO3 transparent die that is 0.5 mm thick. The active area consists of an alternating pattern of oppositely poled stripe domains that are parallel to one axis of the die and cover the entire die surface. The pitch of the domains is 10 μm. By comparing the expected and measured vertical displacements on the reference test sample, the Q factor was calculated. Assuming only vertical polarization, the piezoelectric constant, d33, could be evaluated as: d 33 =
A VAC ⋅ Q
where A is the amplitude signal during scan, VAC is the alternate bias applied during scan, and Q is the quality factor estimated on the reference sample as Q ≈ 10. To minimize the error on the measurement, we adopted the Q from the reference sample measurement, as suggested by the PFM manufacturer (Asylum). Piezoelectric output voltage measurements were performed in the case of flexible ZnO films grown on PI/Cu substrates, supplying the dynamic compressive load to the ZnO film samples with a shaker controlled in a closed loop (TV51110, Tira Gmbh, Germany). The shaker was driven at different voltages and frequencies by means of a function generator and a controller (VR 9500, Vibration Research, USA) to produce a cyclic force stimulus with the desired magnitude and period. Force and acceleration were controlled with a load cell and an accelerometer screwed on the plate of the shaker. The samples were attached to the plate and brought into compression during each cycle by pressing on a fixed, rigid frame placed above the shaker. The piezoelectric voltage from the ZnO film was recorded with a data acquisition board in an open-circuit configuration [26].
3. Results and discussion The XRD pattern of ZnO_1h, ZnO_2h, and ZnO_4h films deposited on hard and flexible conductive substrates are shown in figures 1(a) and (b), respectively. Apart from the diffraction contributions of the Au and Cu metal layers, only reflections coming from the (002) crystal planes are detected in both cases, confirming that all the ZnO films exhibit the hexagonal wurtzite-like crystal structure with a preferential 3
M Laurenti et al
Nanotechnology 26 (2015) 215704
Figure 1. XRD spectra of ZnO_1h, ZnO_2h, and ZnO_4h samples deposited (a) on hard Si/Ti/Au substrates and (b) on flexible PI/Cu foils.
shape quite similar to the domains of the Cu surface (figure 3(b)). This last aspect is very evident in the case of the thinnest sample, ZnO_1h (figures 2(d) and (e)), and becomes less pronounced for the thickest sample, ZnO_4h. From XRD and FESEM characterizations, the high quality of hard- and flexible-sample ZnO_4h can be described as having highly (002)-oriented crystallites and the lowest FWHM, together with large and regular grains characterized by a low grainboundary density. PFM topographic images show an increase in the surface roughness with an increase in the thickness of ZnO films on hard substrates (figure 4). The root mean square values of roughness for samples ZnO_1h, ZnO_2h, and ZnO_4h are 1.9 nm, 2.5 nm, and 2.9 nm, respectively. From figures 2 and 4, one can also see that a high grain-boundary density is present in the thinnest film, whereas the grain-boundary density is lower for the other samples, ZnO_2h and ZnO_4h, which are characterized by larger grains. This characteristic is found to be independent of the particular metal underneath the ZnO layer, since it is observable for samples grown on both hard and flexible substrates. Thus, we conclude that the grainboundary density just depends on the thickness of the ZnO layer (i.e., the sputtering deposition time). It is reported that grain boundaries act as trapping sites for charge carriers. This aspect—together with the presence of a (002) preferential orientation—is believed to strongly influence the final piezoelectric response of the ZnO film [29]. The d33 is characteristic of each piezoelectric material and can be considered as representative of the quality of the analyzed sample. In the present work, PFM measurements were performed to evaluate the local piezoelectric properties and the d33 of the sputtered ZnO thin films deposited on hard, conductive Si/Ti/Au substrates. ZnO thin films grown on flexible PI foil were not investigated by the PFM technique since the higher level of the substrate roughness screened the functional response, distorting the piezoelectric analysis. For the PFM measurements carried out in this work, only the vertical displacement was considered, using the so-called
Table 2. XRD peak positions and full width half maximum (FWHM) values for ZnO thin films deposited on hard Si/Ti/Au and flexible PI/ Cu foils.
Substrate
Sample
(002) peak position
FWHM
Si/Ti/Au
ZnO_1h ZnO_2h ZnO_4h ZnO_1h ZnO_2h ZnO_4h
34.86° 34.72° 34.44° 34.62° 34.54° 34.43°
0.34° 0.30° 0.19° 0.59° 0.39° 0.23°
PI/Cu
compact and pyramid-like grains in sample ZnO_4h (see figure 2(f)). In general, one can note that the different ZnO morphologies observed in figure 2 are strongly related (i) to the increase of ZnO film thickness [27] and (ii) to the morphology of the underlying metal substrate reported in figure 3. Concerning the film thickness, for thinner ZnO films (Samples ZnO_1h), the surface is mainly formed by small grains. When the ZnO deposition goes on, it results in thicker films, and grains further grow in lateral size and height (samples ZnO_2h and ZnO_4h). These aspects are in accordance with the two-regimes growth model already proposed by Vasco et al [28]. With regard to the influence of the underlying substrate on ZnO morphology, FESEM analysis pointed out that the Au layer is characterized by a smooth, flat surface with small, round grains (figure 3(a)). On the contrary, the Cu layer shows irregular, bigger, and nonuniform grains (figure 3(b)). As noted above, the strong differences in the surface morphology of the two metal layers directly affect the morphology of the overlying ZnO thin film. Indeed, by comparing figures 2 and 3, it is evident that ZnO samples grown on hard Si/Ti/Au substrates present regular, round grains, reflecting the morphology of the underlying Au layer (see the inset of figure 3(a)). In the case of ZnO films grown on PI/Cu foils (shown in figures 2(d)–(f)), the presence of irregular crystal grains packed into bigger islands presents a 4
M Laurenti et al
Nanotechnology 26 (2015) 215704
Figure 2. FESEM images of: (a), (d) ZnO_1h, (b), (e) ZnO_2h, and (c), (f) ZnO_4h samples, deposited on hard and flexible substrates.
Figure 3. FESEM images representing the surface morphology of (a) Au and (b) Cu metal layers.
single-frequency PFM. To slightly enhance the piezoresponse while avoiding excessive topographic cross-talk, the working frequency was set below the contact resonance (off peak). In this condition, the amplitude signal can be related to the longitudinal piezoelectric coefficient, d33, by estimating the quality factor Q, as previously described. The PFM characterization highlighted the better piezoelectric properties of the ZnO_4h sample (i.e., the thickest film prepared with the longest sputtering time), with respect to the other samples. This sample presents a more uniform response over the whole scanned region and a higher value of the d33 constant when compared to the ZnO_1h and ZnO_2h samples. Indeed, as highlighted by the XRD characterization, the thickest film has a high level of crystallinity, similar to the bulk material for what concerns the piezoelectric properties [10]. The d33 values estimated for the ZnO thin films on hard substrates are reported in table 1 and are comparable with those already reported in the literature for sputtered ZnO films. Figure 4 shows the topography, the amplitude of the out-of-plane displacement, and the phase obtained with PFM
on the three samples. The bright area in the topography maps represents the top of the ZnO grains (or cluster of grains). These zones, compared with the amplitude image, result in the regions with the highest piezoresponse. In contrast, the regions around the top of the grains are darker in the amplitude map, since the piezoelectric properties are reduced or almost null in the proximity of the grain boundaries. The PFM maps show that all the grains present the same orientation of the polarization because no phase shift is observable, most likely due to the controlled growth direction during deposition on the (002) crystal orientation. The average phase value on the grains is close to 0°, indicating that the polarization vector is directed upward, parallel with the direction of the applied field between the tip and the substrate. The amplitude and phase images related to the thinnest samples (ZnO_1h) register a high cross-talk with the topography map because of a strong electrostatic coupling between the metallic tip and the gold layer underneath. Despite this problem, from the measurements it was possible to evaluate the piezoelectric d33 constant and the surface roughness as the root mean square. 5
M Laurenti et al
Nanotechnology 26 (2015) 215704
Figure 4. Vertical PFM scans of the (a) ZnO_1h, (b) ZnO_2h, and (c) ZnO_4h samples. The three columns show, from left to right, the
topography, amplitude, and phase maps. The maps correspond to a scanned area of 1 μm × 1 μm.
PFM spectroscopy was performed on the center of a ZnO grain of sample ZnO_4h, and it was allowed to analyze the local mechanical deformation of the piezoelectric material directly grown on a conductive substrate under the application of an external electric field. This aspect is of particular importance for nanoactuating systems, which are based on the aforementioned physical principle. The amplitude signal reported in figure 5 shows the deformation of the ZnO crystalline cell under the effect of the applied electric field. From the slope of the amplitude curve, it is also possible to evaluate the d33 coefficient from the grain (8.7 pm V−1) that is in line with the average piezoelectric properties computed from the amplitude image (figure 4(c)) and reported in table 1. When the applied voltage has positive values, the ZnO grain locally expands, while when the voltage values become negative, the material contracts, and thus the phase has a switch of around 180°, as shown in the graph in figure 5. Theoretically, the phase inversion should happen at zero voltage value, and the amplitude response should be symmetric around this value. However, these behaviors are normally modified by surface charge defects that generate a local electric field in contrast or
Figure 5. Single-point spectroscopy performed on sample ZnO_4h.
Phase (upper) and amplitude (lower) signals recorded by applying a voltage sweep (5 cycles +/− 10 V) at 0.5 Hz. No sign of hysteresis appears in either amplitude or phase signal, confirming the absence of any ferroelectric phenomena in the analyzed ZnO region. 6
M Laurenti et al
Nanotechnology 26 (2015) 215704
Figure 6. (a) Piezoelectric peak-to-peak voltage generation upon mechanical stimulation of the ZnO samples. (b) Piezoelectric voltage (up) as a function of time generated by the ZnO_4h sample upon the application of a force signal (down). (c) Image and (d) scheme of the experimental setup for the piezoelectric voltage generation measurements.
and coupled to another PI/Cu foil working as the top electrode. The high quality of the ZnO films deposited on the flexible substrates is confirmed by the FESEM and XRD measurements discussed previously, suggesting the reason for a higher piezoelectric voltage generation for the thicker sample, as reported in figure 6(a). The generated peak-to-peak voltage with a periodic mechanical stimulus of 30 N was 0.058 V for the ZnO_1h sample, 0.361 V for the ZnO_2h sample, and 0.746 V for the ZnO_4h sample. Flexible ZnObased nanogenerators have been widely investigated in the literature, especially by using 1D nanostructures. A few works also discussed the study of thin film-based nanogenerating systems. However, there were some critical points linked to the rise of Schottky barriers at the metal/oxide interface, as well as to the screening effect of the piezopotential, which strongly limited the efficiency of the investigated systems. Recently, Yoon et al showed that screening effects and Schottky barriers in ZnO thin films can be reduced by p-type doping. In this case, a maximum voltage peak of 0.3 V has been reported [31]. An all-solution-processed, flexible ZnO thin film nanogenerator was reported by Kim et al [32]. In that case, the authors used a simple aqueous ZnO ink as the thin film, resulting in a maximum peak voltage of 0.6 V. Nevertheless, a p-type polymer blend and a hole transport layer were also added to the investigated system to reduce the aforementioned limiting effects. In the present case, the maximum voltage peak of 0.746 V has been
in phase with the tip-applied signal [30]. This phenomenon is observable in the ZnO_4h spectroscopy graphs, where the inversion point happens at around 1.5 V. All the results concerning the PFM characterization show the marked piezoelectric behavior of the thickest sample, ZnO_4h. Measurements are performed on ZnO samples grown on conductive substrates, which are mandatory in view of the fabrication of different classes of devices. The piezoelectric response of the materials analyzed can thus be practically exploited, and the most promising sample, ZnO_4h, in particular could be readily integrated into the fabrication process of sensors, actuators, or energy-harvesting systems. In direct-mode piezoelectric characterization, a periodical mechanical deformation is applied to the piezoelectric thin film samples generating a synchronous voltage signal that can be used for both sensing and energy-harvesting applications. This kind of analysis was adopted to investigate ZnO thin films grown on flexible foils. ZnO samples grown on hard Si/ Ti/Au substrates were not considered because they are fragile and break easily, invalidating the measurement. Piezoelectric output voltage measurements were performed by supplying the dynamic compressive load to the ZnO film samples with a shaker controlled in a closed loop. An image and a scheme of the experimental setup used during these measurements are reported in figures 6(c) and (d). For these characterizations, the ZnO samples prepared on flexible PI/Cu foils with an area of 10 mm × 10 mm were used 7
M Laurenti et al
Nanotechnology 26 (2015) 215704
output voltage measurements were carried out on flexible ZnO thin films deposited on PI/Cu foils to show the potential applications of such films for sensing and energy-harvesting systems. An increment of the piezoelectric voltage generation was underlined as the ZnO thickness increased, reaching a maximum generated peak-to-peak voltage of 0.746 V in the thickest ZnO film. This aspect further supports PFM results, which pointed out the best piezoresponse of our thickest ZnO film. All the results concerning the functional characterization of the piezoelectric properties envision the use of ZnO thin films on both hard and flexible conductive substrates in sensor applications and as conformable, reliable, and efficient piezoelectric nanogenerators.
achieved in the case of undoped ZnO, without using an additional interlayer. It is believed that such a generated output voltage directly follows the high-quality characteristics of the thickest sample, regarding the crystal structure, morphology, and the resulting piezoelectric properties. Comparing the generated voltage at 10 N, 20 N, and 30 N, the signal is not linear as expected, but tends to saturate for higher forces because of the shape of the mechanical applied stimulus. Indeed, even if the actuation shaker signal had a sinusoidal shape, the force stimulus recorded by the load cell has the trend shown in figure 6(b) because of a recoil effect during the compression of the sample on the fixed frame that increases mutually with the force amplitude. When the sample is compressed against the fixed bar, the crystalline cell of the ZnO is distorted and the center of gravity of the negative charges will no longer coincide with the positive charges, producing an electric dipole. This deformation will globally induce the generation of electric potential between the two faces of the piezoelectric thin film, moving the electrons inside the external circuit to compensate for the piezoelectric potential. The variation of the force signal due to the recoil effect and elastic properties of the material will produce fast oscillation of the voltage signal, both positive and negative, up to the decompression step (the flat part in the force signal curve). Indeed, when the stress is released, the electric dipole in the ZnO cell disappears and the electrons flow back through the external circuit, producing a negative electrical potential slower peak. All the results concerning the piezoelectric output voltage estimation show that flexible devices based on the thickest sample, ZnO_4h, can be successfully fabricated following a complete room-temperature deposition process.
References [1] Cauda V, Motto P, Perrone D, Piccinini G and Demarchi D 2014 pH-triggered conduction of amine-functionalized single ZnO wire integrated on a customized nanogap electronic platform Nanoscale Res. Lett. 9 1–10 [2] Rivera V F, Auras F, Motto P, Stassi S, Canavese G, Celasco E, Bein T, Onida B and Cauda V 2013 Lengthdependent charge generation from vertical arrays of highaspect-ratio ZnO nanowires Chem. Eur. J. 19 14665–74 [3] Laurenti M, Cauda V, Gazia R, Fontana M, Rivera V F, Bianco S and Canavese G 2013 Wettability control on ZnO nanowires driven by seed layer properties Eur. J. Inorg. Chem. 14 2520–7 [4] Ottone C, Bejtka K, Chiodoni A, Farías V, Canavese G, Roppolo I, Stassi S and Cauda V 2014 Comprehensive study of the templating effect on the ZnO nanostructure formation within porous hard membranes New J. Chem. 38 2058–65 [5] Wang Z L 2004 Zinc oxide nanostructures: growth, properties and applications J. Phys.: Condens. Matter 16 R829–58 [6] Alenezi M R, Henley S J and Silva S R P 2014 On-chip fabrication of high performance nanostructured ZnO UV detectors Sci. Rep. 5 8516 [7] Das S N, Kar J P, Choi J H, Lee T, Moon K J and Myoung J M 2010 Fabrication and characterization of ZnO single nanowire-based hydrogen sensor J. Phys. Chem. C 114 1689–93 [8] Gazia R, Motto P, Stassi S, Sacco A, Virga A, Lamberti A and Canavese G 2013 Photodetection and piezoelectric response from hard and flexible sponge-like ZnO-based structures Nano Energy 2 1294–302 [9] Stassi S, Cauda V, Ottone C, Chiodoni A, Pirri C F and Canavese G 2015 Flexible piezoelectric energy nanogenerator based on ZnO nanotubes hosted in a polycarbonate membrane Nano Energy 13 474–81 [10] Zhao M H, Wang Z L and Mao S X 2004 Piezoelectric characterization individual zinc oxide nanobelt probed by piezoresponse force microscope Nano Lett. 4 587–90 [11] Wang Z L and Song J 2006 Piezoelectric nanogenerators based on Zinc oxide nanowire arrays Science 312 242–6 [12] Zhu G, Yang R, Wang S and Wang Z L 2010 Flexible highoutput nanogenerator based on lateral ZnO nanowire array Nano Lett. 10 3151–5 [13] Xue F, Zhang L, Tang W, Zhang C, Du W and Wang Z L 2014 Piezotronic effect on ZnO nanowire film based temperature sensor ACS Appl. Mater. Interfaces 6 5955–61 [14] Kuoni A, Holzherr R, Boillat M and De Rooij N F 2003 Polyimide membrane with ZnO piezoelectric thin film pressure transducers as a differential pressure liquid flow sensor J. Micromech. Microeng. 13 S103–7
4. Conclusions In this work, piezoelectric ZnO thin films were grown by RF magnetron sputtering on both hard and flexible conductive substrates following a complete room-temperature deposition process. A comprehensive characterization of their properties was presented. FESEM and XRD results showed that high caxis-oriented ZnO films were obtained in both cases, and confirmed that a high-quality crystal structure with few grain boundaries can be obtained by increasing the ZnO thickness up to 1380 nm. In addition, it was observed that the underlying metal substrate has s strong influence on the morphology of the ZnO thin film surface. PFM characterization was performed on ZnO thin films deposited on hard, conductive Si/Ti/Au substrates, and d33 values ranging from 5.0 pm V−1 to 8.0 pm V−1 were estimated for the thinnest and thickest films, respectively, demonstrating that the superior degree of crystallinity in the thickest film leads to an increase in the piezoresponse. Moreover, PFM single-point spectroscopy was carried out for the first time on such films to analyze the mechanical expansion and compression of a crystal grain under the application of an external electric field. Results pointed out that the developed material can be successfully integrated into nanoactuation systems, where the mechanical expansion of the material can be exploited. Piezoelectric 8
M Laurenti et al
Nanotechnology 26 (2015) 215704
[24] Cauda V, Torre B, Falqui A, Canavese G, Stassi S, Bein T and Pizzi M 2012 Confinement in oriented mesopores induces piezoelectric behavior of polymeric nanowires Chem. Mater. 24 4215–21 [25] Hsu Y H, Lin J and Tang W C 2008 RF sputtered piezoelectric zinc oxide thin film for transducer applications J. Mater. Sci., Mater. Electron. 19 653–61 [26] Wang Z L 2009 ZnO nanowire and nanobelt platform for nanotechnology Mater. Sci. Eng. R 64 33–71 [27] Rosa A M, Da Silva E P, Amorim E, Chaves M, Catto A C, Lisboa-Filho P N and Bortoleto J R R 2012 Growth evolution of ZnO thin films deposited by RF magnetron sputtering J. Phys.: Conf. Ser. 370 012020 [28] Vasco E, Rubio-Zuazo J, Vázquez L, Prieto C and Zaldo C 2001 Submicron structure and acoustic properties of ZnO films deposited on (100) InP by pulsed laser deposition J. Vac. Sci. Technol. B 19 224–9 [29] Gardeniers J G E, Rittersma Z M and Burger G J 1998 Preferred orientation and piezoelectricity in sputtered ZnO films J. Appl. Phys. 83 7844–54 [30] Morozovska A N, Svechnikov S V, Eliseev E A, Rodriguez B J, Jesse S and Kalinin S V 2008 Local polarization switching in the presence of surface-charged defects: microscopic mechanisms and piezoresponse force spectroscopy observations Phys. Rev. B 78 054101 [31] Lee E, Park J, Yim M, Jeong S and Yoon G 2014 Highefficiency micro-energy generation based on free-carriermodulated ZnO:N piezoelectric thin films Appl. Phys. Lett. 104 213908 [32] Chung S Y, Kim S, Lee J H, Kim K, Kim S W, Kang C Y, Yoon S J and Kim Y S 2012 All-solution-processed flexible thin film piezoelectric nanogenerator Adv. Mater. 24 6022–7
[15] Shibata T, Unno K, Makino E, Ito Y and Shimada S 2002 Characterization of sputtered ZnO thin film as sensor and actuator for diamond AFM probe Sensors Actuators A 102 106–13 [16] Molarius J, Kaitila J, Pensala T and Ylilammi M 2003 Piezoelectric ZnO films by r.f. sputtering J. Mater. Sci., Mater. Electron. 14 431–5 [17] Choi D, Lee K Y, H Lee K, Kim E S, Kim T S, Lee S Y, Kim S-W, Choi J-Y and Kim J M 2010 Piezoelectric touchsensitive flexible hybrid energy harvesting nanoarchitectures Nanotechnology 21 405503 [18] Magnusson E B, Williams B H, Manenti R, Nam M S, Nersisyan A, Peterer M J, Ardavan A and Leek P J 2015 Surface acoustic wave devices on bulk ZnO crystals at low temperature Appl. Phys. Lett. 106 063509 [19] Wen X, Wu W, Ding Y and Wang Z L 2013 Piezotronic effect in flexible thin-film based devices Adv. Mater. 25 3371–9 [20] Acharya S, Chouthe S, Sturm C, Graener H, Schmidt-Grund R, Grundmann M and Seifert G 2010 Charge carrier dynamics of ZnO and ZnO-BaTi03 thin films J. Phys.: Conf. Ser. 210 012048 [21] Gazia R, Canavese G, Chiodoni A, Lamberti A, Stassi S, Sacco A, Bianco S, Virga A, Tresso E and Pirri C F 2014 Novel spongelike nanostructured ZnO films: Properties and applications J. Alloys Compd. 586 S331–5 [22] Bdikin I K, Gracio J, Ayouchi R, Schwarz R and Kholkin A L 2010 Local piezoelectric properties of ZnO thin films prepared by RF-plasma-assisted pulsed-laser deposition method Nanotechnology 21 235703 [23] Bdikin I K, Shvartsman V V and Kholkin A L 2003 Nanoscale domains and local piezoelectric hysteresis in Pb(Zn1/3Nb2/3) O3-4.5%PbTiO3 single crystals Appl. Phys. Lett. 83 4232–4
9
Search
Collections
Journals
About
Contact us
My IOPscience
Evaluation of the piezoelectric properties and voltage generation of flexible zinc oxide thin films
This content has been downloaded from IOPscience. Please scroll down to see the full text. 2015 Nanotechnology 26 215704 (http://iopscience.iop.org/0957-4484/26/21/215704) View the table of contents for this issue, or go to the journal homepage for more
Download details: IP Address: 130.133.8.114 This content was downloaded on 10/05/2015 at 13:06
Please note that terms and conditions apply.
Nanotechnology Nanotechnology 26 (2015) 215704 (9pp)
doi:10.1088/0957-4484/26/21/215704
Evaluation of the piezoelectric properties and voltage generation of flexible zinc oxide thin films M Laurenti1,2,5, S Stassi2,5, M Lorenzoni3,4, M Fontana1,2, G Canavese1,2, V Cauda1 and C F Pirri1,2 1
Center for Space Human Robotics, Istituto Italiano di Tecnologia, C.so Trento 21, 10129 Turin, Italy Department of Applied Science and Technology, Politecnico di Torino, C.so Duca degli Abruzzi 24, I-10129 Turin, Italy 3 Nanophysics, Fondazione Istituto Italiano di Tecnologia, Via Morego, 30, I-16163 Genoa, Italy 4 Barcelona Microelectronics Institute, IMB-CNM (CSIC), 08193 Bellaterra, Spain 2
E-mail: [email protected] Received 28 January 2015, revised 23 March 2015 Accepted for publication 2 April 2015 Published 6 May 2015 Abstract
Local piezoresponse and piezoelectric output voltage were evaluated on ZnO thin films deposited by radio-frequency magnetron sputtering on hard Si/Ti/Au and flexible Cu-coated polyimide substrates. Three different thicknesses of ZnO films were studied (285 nm, 710 nm, and 1380 nm), focusing on characteristics like crystallinity, grain size, surface roughness, and morphology. Independent of the nature of the metal layer and the substrate, our results show that thicker films presented a higher level of crystallinity and a preferential orientation along the c-axis direction, as well as a lower density of grain boundaries and larger crystal sizes. The improvement of the crystalline structure of the material directly enhances its piezoelectric properties, as confirmed by the local characterizations performed by piezoresponse force microscopy and by the evaluation of the output voltage generation under the application of a periodical mechanical deformation on the whole film. In particular, the highest value of the d33 coefficient obtained (8 pm V−1) and the highest generated output voltage (0.746 V) belong to the thickest films on hard and flexible substrates, respectively. These results envision the use of ZnO thin films—particularly on flexible substrates—as conformable, reliable, and efficient active materials for use in nanosensing, actuation, and piezoelectric nanogenerators. Keywords: Zinc oxide, piezoelectricity, sputtering, flexible (Some figures may appear in colour only in the online journal) 1. Introduction
[1, 6, 7] and energy-harvesting systems [2, 8, 9]. High piezoelectric coefficient d33 values ranging from 14.3 pmV−1 to 26.7 pmV−1 were found in the characterization of ZnO nanobelts [10], while the first attempts to convert mechanical energy into electrical energy were carried out by means of ZnO NWs [11]. In addition, flexible, high-output nanogenerators based on ZnO NWs were successfully demonstrated [12]. These results supported the application of ZnO nanostructures as nanosensors and nanogenerators. The possibility of regulating the charge-carrier transport properties in ZnObased nanomaterials was also reported on both hard and flexible devices based on one-dimenisonal (1D) ZnO
Zinc oxide (ZnO) is a well known wide-bandgap (∼3.37 eV) metal oxide semiconductor that can be easily prepared in the form of thin films and different micro- and nanostructures, such as micro- [1] and nanowires (NWs) [2–4] and nanorods (NRs) [5]. Because of its semiconducting nature and the presence of piezoelectricity, in recent years there has been increased interest in studying ZnO in hopes of exploiting its physical properties in the fabrication of sensing devices 5
Contributed equally to this work.
0957-4484/15/215704+09$33.00
1
© 2015 IOP Publishing Ltd Printed in the UK
M Laurenti et al
Nanotechnology 26 (2015) 215704
single-point PFM spectroscopy, reported here for the first time for a sputtered ZnO thin film grown on a conductive support. Further insight into the piezoelectric characterization was obtained through piezoelectric output voltage measurements applied to flexible sputtered ZnO thin films, with the further aim of using them as functional materials for sensing and actuating applications or as piezoelectric nanogenerators. In particular, ZnO thin films on flexible substrates are shown to be ideal candidates for the fabrication of conformable, reproducible, large-area distributed sensor arrays, thanks to the possibility of depositing piezoelectric thin films through a completely low-temperature and plasma-assisted deposition method that is fully compatible with semiconductor technology. Piezoelectric phenomena of the resulting flexible ZnO thin films can thus be successfully exploited to convert and detect mechanical stress into electrical signals. Piezoelectric ZnO thin films on flexible substrates are also interesting for energy recovery from voluntary and involuntary body movements, since they are easily conformable to curved surfaces and lead to remarkable energy generation.
nanostructures [2, 13]. Although 1D nanostructures are attractive for fundamental research studies, some limitations restrict their use in industrial applications. Indeed, the synthesis techniques for 1D nanostructures still do not guarantee sufficient uniformity of the as-grown materials, especially in terms of dimensions and morphology, with a strong lack of reproducibility and a worsening in the performance of the obtained devices. For this reason, the well-known thin film technology could be a valid alternative for the synthesis of piezoelectric materials, being fully compatible with the semiconductor technology and with the microfabrication processes. Piezoelectric ZnO thin films have been widely used to fabricate pressure sensors [14], actuators [15], ultrahigh-frequency surface acoustic wave devices [16], and energy-harvesting architectures [17]. Nevertheless, most of these applications (i.e., acoustic wave-based devices) rely on the transversal piezoelectric effect, which has been thoroughly investigated for ZnO-based applications [18]. On the contrary, the piezoelectric properties and the charge-generation phenomena related to the longitudinal piezoelectric effect in (002)-oriented ZnO thin films are still under investigation [19, 20]. Recently, the piezoelectric response of porous ZnO films was successfully demonstrated [8]. However, the presence of a polycrystalline structure in this case is expected to lower the piezoresponse, since highly (002)-oriented ZnO films, together with a compact nanostructure, should be pursued to maximize the piezoelectricity of the material [21]. Sputtered ZnO thin films commonly grow with a preferential (002) crystal orientation, and d33 values ranging from 2 up to 12.4 pmV−1 have already been reported for highly (002)oriented ZnO films grown by different techniques [22] and analyzed by piezoresponse force microscopy (PFM). However, to the best of our knowledge no results concerning the evaluation of the converse piezoelectric effect of ZnO nanostructures have been reported. PFM is a commonly used technique for the investigation of the polarity distribution, local piezoelectric properties, and estimation of the longitudinal piezoelectric coefficient (d33) [10, 23, 24]. A helpful approach to maximize the piezoelectric response of ZnO films is the use of thermal treatments, which are generally required to increase the resistivity of ZnO films and to promote their (002) crystal orientation [25]. However, room-temperature depositions are preferable in view of using polymeric substrates for the fabrication of flexible devices. In this work, we report on the deposition and the exhaustive characterization of ZnO thin films grown by the sputtering technique at room temperature on both hard Si/Ti/ Au substrates and flexible Cu-coated polyimide (PI) foils. The morphology and crystal structure of ZnO films were analyzed by means of field emission scanning electron microscopy (FESEM) and x-ray diffraction (XRD). The PFM was exploited to study the local piezoelectric response of the ZnO thin films grown on flat Si/Ti/Au conductive substrates, and to estimate the piezoelectric coefficient, d33. PFM also allowed us to evaluate the local piezoelectric behavior of the ZnO film when working in the actuating mode. In this case, the analysis of the local mechanical displacement under the application of an external electric field was investigated by
2. Experimental 2.1. Deposition of ZnO thin films
ZnO films were deposited by the radio-frequency (RF) magnetron sputtering technique. Both hard Si/Ti/Au wafers and commercial flexible PI/Cu foils were considered as conductive substrates. All the depositions were carried out at room temperature, starting from a ceramic ZnO target. ZnO depositions were performed in a mixed atmosphere of Ar (95%) and O2 (5%) at a pressure of 5 × 10−3 Torr. Suitable vacuum conditions, with base pressures ranging between 3.1 × 10−7 and 3.5 × 10−7 Torr, were achieved before starting the deposition processes. Different film thicknesses were obtained by using three deposition times: 1, 2, and 4 h. The average thicknesses of ZnO thin films are summarized in table 1. 2.2. Materials characterization
The crystal structure and orientation of the ZnO thin films were evaluated by using a Panalytical X’Pert x-ray diffractometer in the Bragg–Brentano configuration (characteristic wavelength, λ = 1.54 059 Å), while the morphology was investigated by FESEM measurements obtained with a Carl Zeiss Dual Beam Auriga. PFM characterization was carried out on ZnO thin films deposited on hard and flat Si/Ti/Au substrates using MFP-3D (Asylum Research, an Oxford Instruments company, USA). The experimental measurements were performed at the first-mode contact resonance of 300 kHz with a silicon Pt-coated cantilever (AC240TM Olympus ‘ElectriLever’) with a nominal stiffness of 2 N m−1 and a first-mode resonance frequency of 70 kHz. To get semiquantitative information, the PFM setup was calibrated on a standard sample, and the following measurements were performed under the same conditions, using the same tip. The 2
M Laurenti et al
Nanotechnology 26 (2015) 215704
orientation along the c-axis direction. Independently from the kind of substrate and metal, the intensity of the (002) peak increases with an increase in the film’s thickness. The (002) peak position for ZnO reported by JCPDSICDD (card n. 89-1397) and considered as a reference in this work is 2θ = 34.37°. Comparing this value with those reported in table 2, one can see that the peak position of the thickest samples, ZnO_4h (34.44°), is closer to the reference sample, while it slightly shifts towards higher 2θ angles for thinner films. The FWHM values estimated for each sample are summarized in table 2. A decrease in this value can be observed when the film thickness increases, with the minimum values reported in the case of the thickest samples, ZnO_4h. Since the deposition parameters—except for the deposition time—are the same for all three samples, the observed changes in the crystal properties of the analyzed samples can be related to the presence of different metal layers underneath the film (i.e., Au and Cu). Since the lattice constants of these two metals are different from each other, the lattice mismatch at the metal-oxide interface is also different and could affect the crystal properties of the overlying ZnO layer. From the comparison of the (002) peak positions and the FWHM values reported in table 2, some differences can be observed between samples ZnO_1h and ZnO_2h when the metal layer changes from Au to Cu, while a good agreement in the reported data is observed for samples ZnO_4h, independent of the metal used. The effect of the different metal layers is thus more pronounced for the thinner samples, ZnO_1h and ZnO_2h, but when the ZnO thickness increases up to 1380 nm, the influence of the underlying metal on the ZnO crystal properties is no longer visible. Despite the use of different interface metal layers, the better crystallinity of samples ZnO_4h can be highlighted, with a 2θ angle closer to the JCPDS-ICDD reference position, and a minimum in the FWHM value (close to 0.2°). Figure 2 shows FESEM images of ZnO thin films deposited on hard and flexible substrates, while FESEM images representative of the Au and Cu layer morphology are shown in figures 3(a) and (b), respectively. The average grain size increases with an increase in the ZnO film thickness. In the case of Si/Ti/Au substrates, a fine-grained, smooth surface with small, rounded grains (average grain size between 20 nm and 40 nm), together with the presence of some voids, is obtained for sample ZnO_1h (see figure 2(a)). When the film thickness is increased to 710 nm, the ZnO surface is characterized by irregularly sized, densely packed grains with no voids and an average dimension between 30 nm and 80 nm (see figure 2(b)). Figure 2(c) shows the surface morphology of sample ZnO_4h. Despite the fact that the grain size does not increase further (the average grain size is similar to the one of sample ZnO_2h), a change in the grain shape can be highlighted. In particular, one can see that the crystal grains grow as long hexagonal rods for sample ZnO_4h, while circular grains are still present in samples ZnO_1h and ZnO_2h. FESEM results of ZnO films deposited on PI/Cu foils are shown in figures 2(d)–(f). The surface morphology changes from the island shape with small crystal grains in samples ZnO_1h and ZnO_2h (see figures 2(d) and (e)) to the more
Table 1. Average thickness, d33 value, and output voltage of sputtered ZnO thin films.
Sample
Average thickness [nm]
d33a [pmV−1]
Output voltageb [V]
ZnO_1h ZnO_2h ZnO_4h
285 710 1380
5.0 5.3 8.0
0.058 0.361 0.746
a
Values refer to ZnO thin films deposited on hard Si/Ti/Au substrates. Values refer to ZnO thin films deposited on flexible PI/Cu foils, under a periodical mechanical stimulus of 30 N. b
periodically poled lithium niobate test sample consists of a 3 mm × 3 mm LiNbO3 transparent die that is 0.5 mm thick. The active area consists of an alternating pattern of oppositely poled stripe domains that are parallel to one axis of the die and cover the entire die surface. The pitch of the domains is 10 μm. By comparing the expected and measured vertical displacements on the reference test sample, the Q factor was calculated. Assuming only vertical polarization, the piezoelectric constant, d33, could be evaluated as: d 33 =
A VAC ⋅ Q
where A is the amplitude signal during scan, VAC is the alternate bias applied during scan, and Q is the quality factor estimated on the reference sample as Q ≈ 10. To minimize the error on the measurement, we adopted the Q from the reference sample measurement, as suggested by the PFM manufacturer (Asylum). Piezoelectric output voltage measurements were performed in the case of flexible ZnO films grown on PI/Cu substrates, supplying the dynamic compressive load to the ZnO film samples with a shaker controlled in a closed loop (TV51110, Tira Gmbh, Germany). The shaker was driven at different voltages and frequencies by means of a function generator and a controller (VR 9500, Vibration Research, USA) to produce a cyclic force stimulus with the desired magnitude and period. Force and acceleration were controlled with a load cell and an accelerometer screwed on the plate of the shaker. The samples were attached to the plate and brought into compression during each cycle by pressing on a fixed, rigid frame placed above the shaker. The piezoelectric voltage from the ZnO film was recorded with a data acquisition board in an open-circuit configuration [26].
3. Results and discussion The XRD pattern of ZnO_1h, ZnO_2h, and ZnO_4h films deposited on hard and flexible conductive substrates are shown in figures 1(a) and (b), respectively. Apart from the diffraction contributions of the Au and Cu metal layers, only reflections coming from the (002) crystal planes are detected in both cases, confirming that all the ZnO films exhibit the hexagonal wurtzite-like crystal structure with a preferential 3
M Laurenti et al
Nanotechnology 26 (2015) 215704
Figure 1. XRD spectra of ZnO_1h, ZnO_2h, and ZnO_4h samples deposited (a) on hard Si/Ti/Au substrates and (b) on flexible PI/Cu foils.
shape quite similar to the domains of the Cu surface (figure 3(b)). This last aspect is very evident in the case of the thinnest sample, ZnO_1h (figures 2(d) and (e)), and becomes less pronounced for the thickest sample, ZnO_4h. From XRD and FESEM characterizations, the high quality of hard- and flexible-sample ZnO_4h can be described as having highly (002)-oriented crystallites and the lowest FWHM, together with large and regular grains characterized by a low grainboundary density. PFM topographic images show an increase in the surface roughness with an increase in the thickness of ZnO films on hard substrates (figure 4). The root mean square values of roughness for samples ZnO_1h, ZnO_2h, and ZnO_4h are 1.9 nm, 2.5 nm, and 2.9 nm, respectively. From figures 2 and 4, one can also see that a high grain-boundary density is present in the thinnest film, whereas the grain-boundary density is lower for the other samples, ZnO_2h and ZnO_4h, which are characterized by larger grains. This characteristic is found to be independent of the particular metal underneath the ZnO layer, since it is observable for samples grown on both hard and flexible substrates. Thus, we conclude that the grainboundary density just depends on the thickness of the ZnO layer (i.e., the sputtering deposition time). It is reported that grain boundaries act as trapping sites for charge carriers. This aspect—together with the presence of a (002) preferential orientation—is believed to strongly influence the final piezoelectric response of the ZnO film [29]. The d33 is characteristic of each piezoelectric material and can be considered as representative of the quality of the analyzed sample. In the present work, PFM measurements were performed to evaluate the local piezoelectric properties and the d33 of the sputtered ZnO thin films deposited on hard, conductive Si/Ti/Au substrates. ZnO thin films grown on flexible PI foil were not investigated by the PFM technique since the higher level of the substrate roughness screened the functional response, distorting the piezoelectric analysis. For the PFM measurements carried out in this work, only the vertical displacement was considered, using the so-called
Table 2. XRD peak positions and full width half maximum (FWHM) values for ZnO thin films deposited on hard Si/Ti/Au and flexible PI/ Cu foils.
Substrate
Sample
(002) peak position
FWHM
Si/Ti/Au
ZnO_1h ZnO_2h ZnO_4h ZnO_1h ZnO_2h ZnO_4h
34.86° 34.72° 34.44° 34.62° 34.54° 34.43°
0.34° 0.30° 0.19° 0.59° 0.39° 0.23°
PI/Cu
compact and pyramid-like grains in sample ZnO_4h (see figure 2(f)). In general, one can note that the different ZnO morphologies observed in figure 2 are strongly related (i) to the increase of ZnO film thickness [27] and (ii) to the morphology of the underlying metal substrate reported in figure 3. Concerning the film thickness, for thinner ZnO films (Samples ZnO_1h), the surface is mainly formed by small grains. When the ZnO deposition goes on, it results in thicker films, and grains further grow in lateral size and height (samples ZnO_2h and ZnO_4h). These aspects are in accordance with the two-regimes growth model already proposed by Vasco et al [28]. With regard to the influence of the underlying substrate on ZnO morphology, FESEM analysis pointed out that the Au layer is characterized by a smooth, flat surface with small, round grains (figure 3(a)). On the contrary, the Cu layer shows irregular, bigger, and nonuniform grains (figure 3(b)). As noted above, the strong differences in the surface morphology of the two metal layers directly affect the morphology of the overlying ZnO thin film. Indeed, by comparing figures 2 and 3, it is evident that ZnO samples grown on hard Si/Ti/Au substrates present regular, round grains, reflecting the morphology of the underlying Au layer (see the inset of figure 3(a)). In the case of ZnO films grown on PI/Cu foils (shown in figures 2(d)–(f)), the presence of irregular crystal grains packed into bigger islands presents a 4
M Laurenti et al
Nanotechnology 26 (2015) 215704
Figure 2. FESEM images of: (a), (d) ZnO_1h, (b), (e) ZnO_2h, and (c), (f) ZnO_4h samples, deposited on hard and flexible substrates.
Figure 3. FESEM images representing the surface morphology of (a) Au and (b) Cu metal layers.
single-frequency PFM. To slightly enhance the piezoresponse while avoiding excessive topographic cross-talk, the working frequency was set below the contact resonance (off peak). In this condition, the amplitude signal can be related to the longitudinal piezoelectric coefficient, d33, by estimating the quality factor Q, as previously described. The PFM characterization highlighted the better piezoelectric properties of the ZnO_4h sample (i.e., the thickest film prepared with the longest sputtering time), with respect to the other samples. This sample presents a more uniform response over the whole scanned region and a higher value of the d33 constant when compared to the ZnO_1h and ZnO_2h samples. Indeed, as highlighted by the XRD characterization, the thickest film has a high level of crystallinity, similar to the bulk material for what concerns the piezoelectric properties [10]. The d33 values estimated for the ZnO thin films on hard substrates are reported in table 1 and are comparable with those already reported in the literature for sputtered ZnO films. Figure 4 shows the topography, the amplitude of the out-of-plane displacement, and the phase obtained with PFM
on the three samples. The bright area in the topography maps represents the top of the ZnO grains (or cluster of grains). These zones, compared with the amplitude image, result in the regions with the highest piezoresponse. In contrast, the regions around the top of the grains are darker in the amplitude map, since the piezoelectric properties are reduced or almost null in the proximity of the grain boundaries. The PFM maps show that all the grains present the same orientation of the polarization because no phase shift is observable, most likely due to the controlled growth direction during deposition on the (002) crystal orientation. The average phase value on the grains is close to 0°, indicating that the polarization vector is directed upward, parallel with the direction of the applied field between the tip and the substrate. The amplitude and phase images related to the thinnest samples (ZnO_1h) register a high cross-talk with the topography map because of a strong electrostatic coupling between the metallic tip and the gold layer underneath. Despite this problem, from the measurements it was possible to evaluate the piezoelectric d33 constant and the surface roughness as the root mean square. 5
M Laurenti et al
Nanotechnology 26 (2015) 215704
Figure 4. Vertical PFM scans of the (a) ZnO_1h, (b) ZnO_2h, and (c) ZnO_4h samples. The three columns show, from left to right, the
topography, amplitude, and phase maps. The maps correspond to a scanned area of 1 μm × 1 μm.
PFM spectroscopy was performed on the center of a ZnO grain of sample ZnO_4h, and it was allowed to analyze the local mechanical deformation of the piezoelectric material directly grown on a conductive substrate under the application of an external electric field. This aspect is of particular importance for nanoactuating systems, which are based on the aforementioned physical principle. The amplitude signal reported in figure 5 shows the deformation of the ZnO crystalline cell under the effect of the applied electric field. From the slope of the amplitude curve, it is also possible to evaluate the d33 coefficient from the grain (8.7 pm V−1) that is in line with the average piezoelectric properties computed from the amplitude image (figure 4(c)) and reported in table 1. When the applied voltage has positive values, the ZnO grain locally expands, while when the voltage values become negative, the material contracts, and thus the phase has a switch of around 180°, as shown in the graph in figure 5. Theoretically, the phase inversion should happen at zero voltage value, and the amplitude response should be symmetric around this value. However, these behaviors are normally modified by surface charge defects that generate a local electric field in contrast or
Figure 5. Single-point spectroscopy performed on sample ZnO_4h.
Phase (upper) and amplitude (lower) signals recorded by applying a voltage sweep (5 cycles +/− 10 V) at 0.5 Hz. No sign of hysteresis appears in either amplitude or phase signal, confirming the absence of any ferroelectric phenomena in the analyzed ZnO region. 6
M Laurenti et al
Nanotechnology 26 (2015) 215704
Figure 6. (a) Piezoelectric peak-to-peak voltage generation upon mechanical stimulation of the ZnO samples. (b) Piezoelectric voltage (up) as a function of time generated by the ZnO_4h sample upon the application of a force signal (down). (c) Image and (d) scheme of the experimental setup for the piezoelectric voltage generation measurements.
and coupled to another PI/Cu foil working as the top electrode. The high quality of the ZnO films deposited on the flexible substrates is confirmed by the FESEM and XRD measurements discussed previously, suggesting the reason for a higher piezoelectric voltage generation for the thicker sample, as reported in figure 6(a). The generated peak-to-peak voltage with a periodic mechanical stimulus of 30 N was 0.058 V for the ZnO_1h sample, 0.361 V for the ZnO_2h sample, and 0.746 V for the ZnO_4h sample. Flexible ZnObased nanogenerators have been widely investigated in the literature, especially by using 1D nanostructures. A few works also discussed the study of thin film-based nanogenerating systems. However, there were some critical points linked to the rise of Schottky barriers at the metal/oxide interface, as well as to the screening effect of the piezopotential, which strongly limited the efficiency of the investigated systems. Recently, Yoon et al showed that screening effects and Schottky barriers in ZnO thin films can be reduced by p-type doping. In this case, a maximum voltage peak of 0.3 V has been reported [31]. An all-solution-processed, flexible ZnO thin film nanogenerator was reported by Kim et al [32]. In that case, the authors used a simple aqueous ZnO ink as the thin film, resulting in a maximum peak voltage of 0.6 V. Nevertheless, a p-type polymer blend and a hole transport layer were also added to the investigated system to reduce the aforementioned limiting effects. In the present case, the maximum voltage peak of 0.746 V has been
in phase with the tip-applied signal [30]. This phenomenon is observable in the ZnO_4h spectroscopy graphs, where the inversion point happens at around 1.5 V. All the results concerning the PFM characterization show the marked piezoelectric behavior of the thickest sample, ZnO_4h. Measurements are performed on ZnO samples grown on conductive substrates, which are mandatory in view of the fabrication of different classes of devices. The piezoelectric response of the materials analyzed can thus be practically exploited, and the most promising sample, ZnO_4h, in particular could be readily integrated into the fabrication process of sensors, actuators, or energy-harvesting systems. In direct-mode piezoelectric characterization, a periodical mechanical deformation is applied to the piezoelectric thin film samples generating a synchronous voltage signal that can be used for both sensing and energy-harvesting applications. This kind of analysis was adopted to investigate ZnO thin films grown on flexible foils. ZnO samples grown on hard Si/ Ti/Au substrates were not considered because they are fragile and break easily, invalidating the measurement. Piezoelectric output voltage measurements were performed by supplying the dynamic compressive load to the ZnO film samples with a shaker controlled in a closed loop. An image and a scheme of the experimental setup used during these measurements are reported in figures 6(c) and (d). For these characterizations, the ZnO samples prepared on flexible PI/Cu foils with an area of 10 mm × 10 mm were used 7
M Laurenti et al
Nanotechnology 26 (2015) 215704
output voltage measurements were carried out on flexible ZnO thin films deposited on PI/Cu foils to show the potential applications of such films for sensing and energy-harvesting systems. An increment of the piezoelectric voltage generation was underlined as the ZnO thickness increased, reaching a maximum generated peak-to-peak voltage of 0.746 V in the thickest ZnO film. This aspect further supports PFM results, which pointed out the best piezoresponse of our thickest ZnO film. All the results concerning the functional characterization of the piezoelectric properties envision the use of ZnO thin films on both hard and flexible conductive substrates in sensor applications and as conformable, reliable, and efficient piezoelectric nanogenerators.
achieved in the case of undoped ZnO, without using an additional interlayer. It is believed that such a generated output voltage directly follows the high-quality characteristics of the thickest sample, regarding the crystal structure, morphology, and the resulting piezoelectric properties. Comparing the generated voltage at 10 N, 20 N, and 30 N, the signal is not linear as expected, but tends to saturate for higher forces because of the shape of the mechanical applied stimulus. Indeed, even if the actuation shaker signal had a sinusoidal shape, the force stimulus recorded by the load cell has the trend shown in figure 6(b) because of a recoil effect during the compression of the sample on the fixed frame that increases mutually with the force amplitude. When the sample is compressed against the fixed bar, the crystalline cell of the ZnO is distorted and the center of gravity of the negative charges will no longer coincide with the positive charges, producing an electric dipole. This deformation will globally induce the generation of electric potential between the two faces of the piezoelectric thin film, moving the electrons inside the external circuit to compensate for the piezoelectric potential. The variation of the force signal due to the recoil effect and elastic properties of the material will produce fast oscillation of the voltage signal, both positive and negative, up to the decompression step (the flat part in the force signal curve). Indeed, when the stress is released, the electric dipole in the ZnO cell disappears and the electrons flow back through the external circuit, producing a negative electrical potential slower peak. All the results concerning the piezoelectric output voltage estimation show that flexible devices based on the thickest sample, ZnO_4h, can be successfully fabricated following a complete room-temperature deposition process.
References [1] Cauda V, Motto P, Perrone D, Piccinini G and Demarchi D 2014 pH-triggered conduction of amine-functionalized single ZnO wire integrated on a customized nanogap electronic platform Nanoscale Res. Lett. 9 1–10 [2] Rivera V F, Auras F, Motto P, Stassi S, Canavese G, Celasco E, Bein T, Onida B and Cauda V 2013 Lengthdependent charge generation from vertical arrays of highaspect-ratio ZnO nanowires Chem. Eur. J. 19 14665–74 [3] Laurenti M, Cauda V, Gazia R, Fontana M, Rivera V F, Bianco S and Canavese G 2013 Wettability control on ZnO nanowires driven by seed layer properties Eur. J. Inorg. Chem. 14 2520–7 [4] Ottone C, Bejtka K, Chiodoni A, Farías V, Canavese G, Roppolo I, Stassi S and Cauda V 2014 Comprehensive study of the templating effect on the ZnO nanostructure formation within porous hard membranes New J. Chem. 38 2058–65 [5] Wang Z L 2004 Zinc oxide nanostructures: growth, properties and applications J. Phys.: Condens. Matter 16 R829–58 [6] Alenezi M R, Henley S J and Silva S R P 2014 On-chip fabrication of high performance nanostructured ZnO UV detectors Sci. Rep. 5 8516 [7] Das S N, Kar J P, Choi J H, Lee T, Moon K J and Myoung J M 2010 Fabrication and characterization of ZnO single nanowire-based hydrogen sensor J. Phys. Chem. C 114 1689–93 [8] Gazia R, Motto P, Stassi S, Sacco A, Virga A, Lamberti A and Canavese G 2013 Photodetection and piezoelectric response from hard and flexible sponge-like ZnO-based structures Nano Energy 2 1294–302 [9] Stassi S, Cauda V, Ottone C, Chiodoni A, Pirri C F and Canavese G 2015 Flexible piezoelectric energy nanogenerator based on ZnO nanotubes hosted in a polycarbonate membrane Nano Energy 13 474–81 [10] Zhao M H, Wang Z L and Mao S X 2004 Piezoelectric characterization individual zinc oxide nanobelt probed by piezoresponse force microscope Nano Lett. 4 587–90 [11] Wang Z L and Song J 2006 Piezoelectric nanogenerators based on Zinc oxide nanowire arrays Science 312 242–6 [12] Zhu G, Yang R, Wang S and Wang Z L 2010 Flexible highoutput nanogenerator based on lateral ZnO nanowire array Nano Lett. 10 3151–5 [13] Xue F, Zhang L, Tang W, Zhang C, Du W and Wang Z L 2014 Piezotronic effect on ZnO nanowire film based temperature sensor ACS Appl. Mater. Interfaces 6 5955–61 [14] Kuoni A, Holzherr R, Boillat M and De Rooij N F 2003 Polyimide membrane with ZnO piezoelectric thin film pressure transducers as a differential pressure liquid flow sensor J. Micromech. Microeng. 13 S103–7
4. Conclusions In this work, piezoelectric ZnO thin films were grown by RF magnetron sputtering on both hard and flexible conductive substrates following a complete room-temperature deposition process. A comprehensive characterization of their properties was presented. FESEM and XRD results showed that high caxis-oriented ZnO films were obtained in both cases, and confirmed that a high-quality crystal structure with few grain boundaries can be obtained by increasing the ZnO thickness up to 1380 nm. In addition, it was observed that the underlying metal substrate has s strong influence on the morphology of the ZnO thin film surface. PFM characterization was performed on ZnO thin films deposited on hard, conductive Si/Ti/Au substrates, and d33 values ranging from 5.0 pm V−1 to 8.0 pm V−1 were estimated for the thinnest and thickest films, respectively, demonstrating that the superior degree of crystallinity in the thickest film leads to an increase in the piezoresponse. Moreover, PFM single-point spectroscopy was carried out for the first time on such films to analyze the mechanical expansion and compression of a crystal grain under the application of an external electric field. Results pointed out that the developed material can be successfully integrated into nanoactuation systems, where the mechanical expansion of the material can be exploited. Piezoelectric 8
M Laurenti et al
Nanotechnology 26 (2015) 215704
[24] Cauda V, Torre B, Falqui A, Canavese G, Stassi S, Bein T and Pizzi M 2012 Confinement in oriented mesopores induces piezoelectric behavior of polymeric nanowires Chem. Mater. 24 4215–21 [25] Hsu Y H, Lin J and Tang W C 2008 RF sputtered piezoelectric zinc oxide thin film for transducer applications J. Mater. Sci., Mater. Electron. 19 653–61 [26] Wang Z L 2009 ZnO nanowire and nanobelt platform for nanotechnology Mater. Sci. Eng. R 64 33–71 [27] Rosa A M, Da Silva E P, Amorim E, Chaves M, Catto A C, Lisboa-Filho P N and Bortoleto J R R 2012 Growth evolution of ZnO thin films deposited by RF magnetron sputtering J. Phys.: Conf. Ser. 370 012020 [28] Vasco E, Rubio-Zuazo J, Vázquez L, Prieto C and Zaldo C 2001 Submicron structure and acoustic properties of ZnO films deposited on (100) InP by pulsed laser deposition J. Vac. Sci. Technol. B 19 224–9 [29] Gardeniers J G E, Rittersma Z M and Burger G J 1998 Preferred orientation and piezoelectricity in sputtered ZnO films J. Appl. Phys. 83 7844–54 [30] Morozovska A N, Svechnikov S V, Eliseev E A, Rodriguez B J, Jesse S and Kalinin S V 2008 Local polarization switching in the presence of surface-charged defects: microscopic mechanisms and piezoresponse force spectroscopy observations Phys. Rev. B 78 054101 [31] Lee E, Park J, Yim M, Jeong S and Yoon G 2014 Highefficiency micro-energy generation based on free-carriermodulated ZnO:N piezoelectric thin films Appl. Phys. Lett. 104 213908 [32] Chung S Y, Kim S, Lee J H, Kim K, Kim S W, Kang C Y, Yoon S J and Kim Y S 2012 All-solution-processed flexible thin film piezoelectric nanogenerator Adv. Mater. 24 6022–7
[15] Shibata T, Unno K, Makino E, Ito Y and Shimada S 2002 Characterization of sputtered ZnO thin film as sensor and actuator for diamond AFM probe Sensors Actuators A 102 106–13 [16] Molarius J, Kaitila J, Pensala T and Ylilammi M 2003 Piezoelectric ZnO films by r.f. sputtering J. Mater. Sci., Mater. Electron. 14 431–5 [17] Choi D, Lee K Y, H Lee K, Kim E S, Kim T S, Lee S Y, Kim S-W, Choi J-Y and Kim J M 2010 Piezoelectric touchsensitive flexible hybrid energy harvesting nanoarchitectures Nanotechnology 21 405503 [18] Magnusson E B, Williams B H, Manenti R, Nam M S, Nersisyan A, Peterer M J, Ardavan A and Leek P J 2015 Surface acoustic wave devices on bulk ZnO crystals at low temperature Appl. Phys. Lett. 106 063509 [19] Wen X, Wu W, Ding Y and Wang Z L 2013 Piezotronic effect in flexible thin-film based devices Adv. Mater. 25 3371–9 [20] Acharya S, Chouthe S, Sturm C, Graener H, Schmidt-Grund R, Grundmann M and Seifert G 2010 Charge carrier dynamics of ZnO and ZnO-BaTi03 thin films J. Phys.: Conf. Ser. 210 012048 [21] Gazia R, Canavese G, Chiodoni A, Lamberti A, Stassi S, Sacco A, Bianco S, Virga A, Tresso E and Pirri C F 2014 Novel spongelike nanostructured ZnO films: Properties and applications J. Alloys Compd. 586 S331–5 [22] Bdikin I K, Gracio J, Ayouchi R, Schwarz R and Kholkin A L 2010 Local piezoelectric properties of ZnO thin films prepared by RF-plasma-assisted pulsed-laser deposition method Nanotechnology 21 235703 [23] Bdikin I K, Shvartsman V V and Kholkin A L 2003 Nanoscale domains and local piezoelectric hysteresis in Pb(Zn1/3Nb2/3) O3-4.5%PbTiO3 single crystals Appl. Phys. Lett. 83 4232–4
9
