This content has been downloaded from IOPscience. Please scroll down to see the full text.

Download details:

This content was downloaded by: fpattini

IP Address: 160.78.60.169

This content was downloaded on 14/05/2014 at 15:00

Please note that terms and conditions apply.

Solution-free and catalyst-free synthesis of ZnO-based nanostructured TCOs by PED and

vapor phase growth techniques

View the table of contents for this issue, or go to the journal homepage for more

2012 Nanotechnology 23 194008

(http://iopscience.iop.org/0957-4484/23/19/194008)

Home Search Collections Journals About Contact us My IOPscience

IOP PUBLISHING NANOTECHNOLOGY

Nanotechnology 23 (2012) 194008 (7pp) doi:10.1088/0957-4484/23/19/194008

Solution-free and catalyst-free synthesisof ZnO-based nanostructured TCOs byPED and vapor phase growth techniques

D Calestani, F Pattini, F Bissoli, E Gilioli, M Villani and A Zappettini

IMEM-CNR, Parco Area delle Scienze 37/A, PARMA 43124, Italy

E-mail: calle@imem.cnr.it

Received 30 December 2011, in final form 3 February 2012Published 27 April 2012Online at stacks.iop.org/Nano/23/194008

AbstractZinc oxide (ZnO) is one of the most promising materials for realizing three-dimensional (3D)nanostructured transparent conducting oxides (TCOs) on large scale, because it is cheap, it canbe modified with large concentrations of trivalent elements (such Al, Ga or In) and it ischaracterized by good electron mobility, wide bandgap and visible-range transparency. But,above all, it can be easily obtained in the form of different nanostructures with a large numberof growth techniques. A solution-free and catalyst-free approach has been explored here bythe vapor phase synthesis of vertically aligned ZnO nanorods on ZnO:Al (AZO) films grownby pulsed electron deposition (PED). The obtained nanostructured TCOs resulted to behomogeneous on large areas and easily patternable by means of mechanical masks. Themorphology, crystalline structure, electrical and optical properties of the obtained sampleshave been characterized in depth. The possible use of such a nanostructured TCO in excitonic(e.g. DSSC) or low-reflectivity traditional solar cells is discussed.

(Some figures may appear in colour only in the online journal)

1. Introduction

The continuous technological research for improving theefficiency of modules in traditional photovoltaics or forexploring and exploiting the possible alternatives in excitonicsolar cells cannot exclude the development of low-cost andbetter-performing transparent conducting oxides (TCOs).

Zinc oxide (ZnO) has attracted a lot of attention in thisfield not only because it is a cheap and transparent widebandgap semiconductor, with natural n-type conductivityand good electron mobility [1–3], but also because it canbe doped with a high concentration of trivalent atoms(e.g. Al, Ga or In) to increase the electrical conductivity,with minor effects on the crystal structure and transparencyof the material [1, 4–7]. But, above all, ZnO has also thepeculiar property of being easily obtained in the form ofnanostructures [8–13], allowing researchers to design morecomplex three-dimensional (3D) structures for TCOs withsmarter light exploitation (e.g. creating low-reflectivity layers)and more efficient short-range charge collection in excitonic

solar cells. Moreover, a TCO that is completely based on asingle material, such as ZnO, is also desirable to promotelong-range electron transfer inside a continuous mediumconductor, not affected by band alignment issues typical ofmulti-layered structures.

TCOs made of vertically aligned ZnO nanorods (NRs)grown over different oxide films (e.g. ITO, FTO, AZO,etc) are widely reported in the literature (e.g. see [14–19])and, in most cases, reproducible results are obtained onlyby means of wet chemical methods (hydrothermal growth,chemical bath deposition, electrochemical deposition, etc) ormore expensive vapor phase growth techniques with metalcatalysts (typically Au). Solution-based growth of ZnO-NRsis generally the most suitable technique for cheap andlarge-scale production, but it cannot always be matchedeasily with the deposition processes of other cell components.Moreover, in those cell structures where ZnO-NRs have anactive role in the light adsorption and exciton separation,the intrinsic doping from solution inclusions may degradethe device performance. In contrast, vapor phase growth is

10957-4484/12/194008+07$33.00 c© 2012 IOP Publishing Ltd Printed in the UK & the USA

Nanotechnology 23 (2012) 194008 D Calestani et al

generally less controllable and more expensive, especiallywhen metal–organic precursors or metal catalyst are used inorder to improve the synthesis reproducibility.

In the present work the authors have explored analternative solution-free and catalyst-free synthesis process,which combines two different techniques developed atIMEM-CNR, i.e. a pulsed electron deposition (PED) ofZnO:Al (AZO) films on commercial glass substrates anda self-catalyzed vapor phase growth of vertically alignedZnO-NRs at relatively low temperature.

Different deposition techniques are generally usedto obtain high quality ZnO-based TCO layers, such asRF magnetron sputtering [20], chemical vapor deposition(CVD) [21] and pulsed laser deposition (PLD). In particular,doped-ZnO thin films grown by PLD exhibit high opticaltransparency [22], high conductive property and excellentcrystalline quality [23]. However, due to the cost of thelaser source, PLD is not a reliable production technique forlarge-scale production of TCOs. PED, instead, is a morerecent and cheaper alternative to PLD for growing highquality pure and doped-ZnO thin films, and it shares some ofthe main advantages with PLD, like accurate stoichiometriccontrol, high deposition rate and large flexibility of usablematerials [24–27].

The AZO films obtained by PED are not simply goodTCO layers, but also have the right features, in terms ofsurface roughness and preferential grain orientation, to beused for the self-catalyzed nucleation and vapor phase growthof vertical ZnO-NRs along the same crystalline axis ([001]of hexagonal wurtzite structure). This NR growth process,although conducted at 480 ◦C, is limited to a few minutes, iscompatible with commercial glass substrates and makes useof a simple metallic Zn source.

The complete AZO-film/ZnO-NRs structure obtainedwith these techniques is extremely interesting for use asTCO. Its morphology, crystal structure, patterning, electricalconductivity and optical properties have been characterizedand in some case compared with those of an undopedZnO-film/ZnO-NR sample. The obtained results are herediscussed, as well as the possible use of such a TCO inexcitonic (e.g. DSSC) or low-reflectivity traditional solarcells.

2. Experimental details

2.1. Deposition of ZnO and AZO films by PED

The PED technique is based on rapid material ablation from atarget due to the collision of a highly energetic pulsed electronbeam. Targets of undoped ZnO were prepared starting fromcommercial powder (Sigma Aldrich, 4N), while in AZO thetarget ZnO was mixed with Al2O3 powder (2 wt%, SigmaAldrich, 5N). Free-standing 40 mm diameter pellets wereprepared with these powders by pressing at 280 bar for 5 min,after a fine grinding in a ball mill with iso-propanol for30 min. The obtained targets were then sintered at 1000 ◦C for12 h in order to increase their toughness while maintaining asufficient porosity and finally polished to obtain a smooth and

uniform surface. Soda-lime glass was used both as substratesand for mechanical masking of the desired portion of thesubstrates.

After high vacuum pumping, the deposition chamberpressure was set to 1.5–2.5×10−3 mbar in Ar to optimize thePED plume, while the substrate temperature was set to 200 ◦C.These conditions were chosen in order to deposit a film withthe best transparency and conductivity characteristics [28].A commercial PED source (supplied by Neocera Inc., USA)was used to ablate the target. The acceleration voltage of theincident pulsed electron beam was fixed to 14 kV for bothpure ZnO and AZO, with a pulse repetition rate of 10 Hz.

In these conditions the electron beam impacts on a4 mm2 target area with a pulse energy of about 12 J cm−2.The substrates were finally cooled to room temperaturewhile keeping the same base pressure value. No post-growthannealing treatments were performed.

Soda-lime glass was used here because it is a typicallow-cost substrate, with good transparency properties, oftenused for photovoltaic applications, and that can reach 480 ◦C(for NR growth) without softening. Any other glass substratewith these properties, such as, e.g., borosilicate glass, can bealternatively used.

2.2. Vapor phase growth of ZnO nanorods

ZnO-NRs were grown by the optimized vapor phase techniquedescribed in [29] directly on the AZO and ZnO filmsobtained by PED. Commercial Zn (powder, Sigma Aldrich,4.5N) was used as source material after a smooth etchingin diluted HCl. The source material and substrates wereplaced side-by-side in a tubular furnace, with the substratesin the downstream position and the source material insidea quasi-closed container (with only a small window for Znvapor production). The growth temperature was limited to480 ◦C in order to be compatible with the glass substrateswhile Ar was used as an inert transporting gas. O2 wasintroduced into the reactor for oxidation (1:20 O2 to Ar ratio,50 sccm) only when the maximum temperature was reached.The furnace was kept at 480 ◦C for only 5 min and then thesystem was cooled down to room temperature.

Also in this case it was possible to limit the growthof ZnO-NRs to specific areas on the substrate by means ofmechanical masks (glass or alumina).

2.3. Characterization

The morphology of the ZnO and AZO films was studiedby atomic force microscopy (AFM Digital InstrumentsNANOSCOPE IIIA), while that of the obtained nanostruc-tured samples was characterized by means of a scanningelectron microscope (Philips 515 SEM). The electricalproperties of the AZO films used were measured by afour-probe Van der Pauw setup. The crystal structure andorientation were investigated by x-ray diffraction (XRD)spectra and rocking-curves (Siemens D-500 diffractometerwith a Cu Kα source—λ 1.540 A), while the opticalproperties were examined by room temperature absorption

2

Nanotechnology 23 (2012) 194008 D Calestani et al

Figure 1. SEM images of the deposited AZO films: (a) film roughness is low and a very low density of ‘droplet’-shaped bumps is revealed;(b) low magnification image of a sample area across a deposition pattern. Both images are tilted by 45◦ in order to better highlight thesample roughness.

Figure 2. AFM measurement of a 5× 5 µm2 area of an AZO film deposited by PED.

(JASCO spectrometer in the range 300–1100 nm) andphotoluminescence (325 nm excitation by He–Cd laser)measurements.

3. Results and discussion

The morphology of the deposited AZO films was character-ized by both SEM and AFM images. SEM images were usedto evaluate the deposition homogeneity on the substrate andto have a first estimation of the film roughness (figure 1(a)).The depositions resulted to be uniform on the whole substrates(about 1 cm2) and, in those cases where mechanical maskingwas used, the patterning effectiveness was successfullyconfirmed (figure 1(b)). Only a few isolated ‘droplets’,intrinsic in this kind of deposition technique [30–32], werepresent on the film thanks to the optimization of the depositionconditions; their presence on the film did not affected thefollowing growth of NRs.

AFM analysis was limited to smaller areas, randomlychosen on the substrates, for a better quantification of filmgrain size and mean surface roughness. For the growth ofaligned nanorods, indeed, a sufficiently flat surface is requiredin order to obtain a homogeneous distribution of nuclei and,then, nanostructures. A representative image of the obtainedresults is reported in figure 2. An average roughness of 9 nm(RMS value) was typically found.

This value completely satisfies the roughness require-ments, since usually a maximum RMS threshold limit for the

growth of aligned NRs was experimentally observed at about50–60 nm. Below this value, the distribution of Zn clusters(that are at the base of the NR nucleation [29]) is evidently notaffected, while probably above this value cluster aggregationis promoted at surface points with the highest wettability andlarge crystals with random orientation can form below thenanostructures. The peaks and pits that are visible in the AFMmaps are small in size and height and they generally do notproduce visible effects on the following NR growth.

X-ray diffraction (XRD) was used to verify whether apreferential orientation was present in the film grains. Infigure 3(a) a typical XRD spectrum for the obtained AZOfilms is reported and it clearly shows that the film grains,although randomly arranged on the substrate planes, havea common orientation of the [001] axis (the polar onein wurtzite structure), which is orthogonal to the substratesurface. Preferential orientation in ZnO films prepared bydifferent techniques is not unusual (e.g. PLD, sputtering,etc) and the PED films were confirmed to have this feature.This behavior was ascribed to the formation and growth ofcolumnar crystals that coalesce to form a continuous film oforiented grains [28]. In figure 3(b) a rocking curve for the002 peak is reported. This kind of XRD analysis can providebetter structural information for the misalignment range inoriented grains and the result is an FWHM value of 5.2◦. Thisvalue, although not so low, is in agreement with deposition onan amorphous substrate (glass) that induces a non-crystallinegrowth of the first deposited AZO layers.

3

Nanotechnology 23 (2012) 194008 D Calestani et al

Figure 3. XRD analysis of the obtained samples. (a) XRD spectra of an obtained AZO film and a complete ZnO-NR/AZO-film sample;(b) rocking curve for the 002 peak.

Figure 4. Transmittance plot of an AZO film deposited by PEDafter subtraction of the glass substrate contribution. The inset showsan optical image of a sample with three different areas: clean glass(‘GL’), NR-free AZO film (‘AZO’) and nanostructuredZnO-NR/AZO film (‘NRs’).

Similar structural and morphological results wereobtained also for the reference undoped ZnO films.

As TCO layers, AZO films must fulfil the followingrequirements: high optical transparency (>80% in the visibleand NIR regions) to let the sunlight reach the absorber layerand low electrical resistivity (ρ < 10−3 � cm) to minimizethe solar cell series resistance. The best results for AZO filmsdeposited by PED, in terms of transmittance and resistivity,were obtained starting from a target doped with 2 wt% ofAl2O3 [28].

The optical transmittance of the AZO films (figure 4)was obtained from absorption measurements in the range300–1100 nm by subtracting the glass contribution. The filmsshowed high transparency in all the visible range.

Resistivity measurements were performed in theVan der Pauw configuration with In–Ga Ohmic contacts. TheAl concentration and deposition conditions were optimized inorder to reach resistivity values of 5× 10−4 � cm.

These AZO films were used as a functional seeding layerfor the growth of vertically aligned ZnO-NRs. In figure 5the results of vapor phase growth are shown. SEM imagesrevealed that the average thickness of the obtained NRs was inthe range 20–50 nm, while their density and length dependedon the growth parameters. The NR length can be increasedby a small increase in growth temperature and/or time; thetypical length with the parameters reported in section 2 is1.5 µm. The density, instead, can be modified by varying thenucleation conditions (inert gas flow and temperature, sourcematerial temperature and window size, etc), although an upperlimit is intrinsic in the mechanism described in [29], becausean excess in Zn cluster density on the seeding film producesa continuous Zn layer that inhibits the anisotropic formationof NRs. With the growth parameters reported in section 2 anaverage NR-to-NR distance of 100–150 nm was obtained.

SEM images revealed also that ZnO-NRs are generallyaligned orthogonally to the substrate plane. By analyzing theSEM images, it was possible to determine that the typicalangle formed by the nanorods with respect to the normaldirection to the substrate plane ranges is generally lower than5◦. This spread of angles for the nanowire growth directionis in agreement with the misalignment of the nanocrystalsforming the AZO films (as evaluated by x-ray analysis infigure 3(b)). Only NRs longer than 3 µm are generally bentas a result of the high length-to-thickness ratio (figure 5(d)).

XRD analysis confirmed that the ZnO-NRs grewcoherently with the bottom AZO film, maintaining thepreferential orientation of the [001] axis orthogonal to thesubstrate (figure 3(a)).

No appreciable difference was observed in the samplemorphology when ZnO-NRs were grown on a referenceundoped ZnO film.

Mechanical masking was effective also for the growthof ZnO-NRs (figure 6(a)), so even in this case it is possibleto easily create growth patterns. In figure 6(b) the effectof masking in different stages of the growth is shown,obtaining clean glass surfaces, NR-free AZO-film areas andnanostructured ZnO-NR/AZO-film areas. The difference inreflectivity between the area with ZnO-NRs and the one withAZO only is also visible in figure 6(b).

Other important requirements for the TCO are thechemical and thermal stability, in compatibility with the whole

4

Nanotechnology 23 (2012) 194008 D Calestani et al

Figure 5. SEM images of ZnO-NRs grown on AZO film deposited by PED. (a) Tiled view of typically grown aligned ZnO-NRs; (b) crosssection of the complete TCO structure with glass substrate, AZO film, intermediate wetting layer and ZnO-NRs; (c) homogeneousdistribution of NRs on a larger area (tilted view); (d) 5 µm long NRs that bend due to the excessive length-to-thickness ratio (tilted view);(e) top view of a sample with aligned NRs on AZO film.

Figure 6. Patterning of NR deposition on AZO film by means of mechanical masking. (a) SEM image of a masked AZO sample thatresulted in a selective ZnO-NR growth only on the left side; (b) optical image of a sample that underwent a sequential masking, showingclean glass, NR-free AZO film and nanostructured ZnO-NR/AZO film.

process of solar cell building. AZO-film conductivity wastested after the growth of NRs, because in principle anythermal treatment at high temperature may affect the film’sresistivity. After a thermal treatment at 480 ◦C (NR growthtemperature), a decrease in the AZO-film resistivity wasobserved, so the vapor phase growth process was optimizedin order to reduce as much as possible the time at hightemperature, and the process described in section 2 is thebest compromise for the growth setup used. After thisshort exposure at high temperature the AZO-film resistivitydecreased only to 1.0–1.2× 10−3 � cm, which is still a goodvalue for a TCO.

On the contrary, the NRs were expected to have a lowerconductivity, basically associated to residual n-type doping ofZnO (generally due to the low formation energies of pointdefects, e.g. Zn interstitials). But in the described growth

process Al may diffuse from the AZO film into the growingZnO-NRs. A bare estimation of the Al content in the NRs canbe done by measuring the bandgap shift [28, 33–35]. ZnO,indeed, has a room temperature bandgap of about 3.3 eV,while 2% of Al can raise it up to 3.7 eV.

In figure 7 the PL spectra of NRs grown on an undopedZnO film and on an AZO film are compared. No appreciableshift is observed in the near band edge (NBE) emission peak,hence only a negligible amount of Al may have drifted in thenanostructures during their fast growth.

On the other hand, in the PL plots a different behaviorin the relative ratio between the NBE emission and thewide emission band in the visible region, generally ascribedto surface defect states, is clearly visible. Since no otherdifference was observed in the growth of NRs on AZO andZnO films, this evidence suggests that, however, a small

5

Nanotechnology 23 (2012) 194008 D Calestani et al

Figure 7. PL spectra of two samples of ZnO-NRs grown on anAZO film and on an undoped ZnO film respectively (laser excitationat 325 nm). The maximum of the NBE emission peak is located at376 nm for both samples.

concentration of Al probably entered the NRs, affectingthe recombination processes. This Al doping creates donorlevels that further enhance the n-type conductivity of thenanostructures, but its concentration is clearly orders ofmagnitude lower than that of the AZO films.

During the first stage of the NR growth a thin ZnO layergenerally forms over the seeding film (figure 5(b)), allowingthe transition from the flat and continuous seeding film to theanisotropic and not continuous NR morphology [29]. The PLspectrum in figure 7 is the result of emissions from both theNRs and the wetting layer. The same considerations regardingAl diffusion can be extended to this layer: equivalentPL spectra are obtained when the vapor phase growth isinterrupted at the end of this first stage and the NR growthis inhibited.

This ZnO wetting layer and the ZnO-NRs create ann-type 3D ZnO structure over the AZO film that is, forexample, the ideal one for the realization of excitonic cellswhere ZnO is used as the anode (figure 8). In excitonic solarcells (DSSC, ETA, QDSSC, etc), indeed, the diffusion lengthof the charge carriers in light adsorbing materials is generallyvery short (sketch ‘(a)’ in figure 8), and so only thin layers ofthese materials can grant an efficient transfer to the transparentoxide. On the other hand the photon absorption probabilityin such thin layers is very low. So, generally, polycrystalline

porous films filled with the light absorbing material areused to simultaneously overcome these two issues, creatingthicker packed multiple layers of absorber and charge carriertransporting material (sketch ‘(b)’ in figure 8).

Unfortunately, even in this case, the film thickness isgenerally limited, because the charge carrier has to crossa large number of interfaces and grain boundaries beforereaching the TCO, with an increasing proportional probabilityof recombination. In contrast, vertical and single-crystal NRscan collect charge carriers at a longer distance from the TCO(sketch ‘(c)’ in figure 8), because in that case only a minimumnumber of interfaces is present on the carrier path, i.e. athicker cell can be built with the use of these nanostructures.In principle, if the distance between the NRs is not too high,as in the case of the obtained ZnO-NRs, a better exploitationof the empty volume can be alternatively achieved by fillingit with a traditional polycrystalline material of the properdimensions [36] (sketch (d) in figure 8) that, in this case, willalways be at a short distance from a fast transfer NR path.Moreover, the good alignment of the NRs should, in principle,facilitate the infiltration process.

In all these cases, the presence of the wetting layer atthe NRs’ base ensures that no direct contact between the lightabsorber or the hole-transporter and the TCO can occur.

In the described configuration, the electron mobilitywithin the grown NRs and the complete anode structure isvery important. A vapor phase process intrinsically produceslower impurities and defect concentrations than solutiongrowth methods and, in this sense, the described solution-freeand catalyst-free method is clearly favored.

Finally, the 3D structure created by the NRs improvesthe light transport and diffusion within the cell volume,increases the photon absorption probability and reducesthe light reflection. Low-reflectivity TCOs are of generalinterest also in traditional photovoltaic cells and the describednanostructured ZnO-based one may be implemented as asubstitute for flat ZnO layers to improve the cells’ efficiency.

Although the described results have been obtained onthe laboratory scale (few square centimeters per run), thescalability of this process is currently under investigation, asthe simple two-step process may be sequentially implementedin a single vacuum production line.

Figure 8. Sketch of different possible configurations of the elements in an excitonic solar cell with a transparent anode; black arrows areused to indicate the electron transfer from the light adsorbing material towards the TCO; see the text for a detailed description of thedifferent cases. (a) Thick absorber layer, (b) porous polycrystalline structure, (c) vertically aligned and single-crystal NRs, (d) mixedstructure with NRs and polycrystalline material.

6

Nanotechnology 23 (2012) 194008 D Calestani et al

4. Conclusions

An alternative solution-free and catalyst-free synthesismethod has been explored for the realization of 3Dnanostructured ZnO-based TCOs. This method combinesAZO films with high transparency and conductivity depositedby PED and the vapor phase growth of vertically alignedZnO-NRs on the top of them. Both synthesis steps arecompatible with the commercial glass substrates typicallyused for photovoltaic applications and make no use ofexpensive catalysts or metal–organic precursors (only cheapZnO, Al2O3 and Zn powders are used as the source materials).

The obtained samples are homogeneous on the growtharea (few square centimeters on the laboratory scale) and boththe film deposition and the NR growth can be easily patternedby means of mechanical masks.

The resulting nanostructured TCO is suitable for efficientexcitonic solar cell design and is a good alternative for cellproduction processes that are not compatible with chemicalbaths or high impurity levels. Moreover, the low reflectivityof the ZnO-NR/AZO-film structure is also promising forimplementation in traditional photovoltaic cells.

Acknowledgment

This work has been partially supported by the MIST E-RConsortium.

References

[1] Klingshrin C 2007 ChemPhysChem 8 782[2] Park W I and Yi G C 2004 Adv. Mater. 16 87[3] Ozgur U, Alivov Y I, Liu C, Teke A, Reshchikov M A,

Dogan S, Avrutin V, Cho S J and Morkoc H 2005 J. Appl.Phys. 98 041301

[4] Hochbaum A I and Yang P D 2010 Chem. Rev. 110 527[5] Agashe C, Kluth O, Schope G, Siekmann H, Hupkes J and

Rech B 2003 Thin Solid Films 442 167[6] Law M, Greene L E, Johnson J C, Saykally R and

Yang P D 2005 Nature Mater. 4 455[7] Gonzalez-Valls I and Lira-Cantu M 2009 Energy Environ. Sci.

2 19[8] Wang Z L 2006 Mater. Today 7 26[9] Lu J G, Chang P and Fan Z 2006 Mater. Sci. Eng. R 52 49

[10] Schmidt-Mende L and MacManus-Driscoll J L 2007 Mater.Today 10 40

[11] Zha M Z, Calestani D, Zappettini A, Mosca R, Mazzera M,Lazzarini L and Zanotti L 2008 Nanotechnology 19 325603

[12] Zanotti L, Calestani D, Villani M, Zha M Z, Zappettini A andPaorici C 2010 Cryst. Res. Technol. 45 667

[13] Lazzarini L, Salviati G, Fabbri F, Zha M Z, Calestani D,Zappettini A, Sekiguchi T and Dierre B 2009 ACS Nano3 3158

[14] Baxter J B, Walker A M, Van Ommering K andAydil E S 2006 Nanotechnology 17 S304

[15] Qiu J J, Li X M, Zhuge F W, Gan X Y, Gao X D, He W Z,Park S J, Kim H K and Hwang Y H 2010 Nanotechnology21 195602

[16] Yu E T and Van De Lagemaat J 2011 MRS Bull. 36 424[17] Lupan O, Guerin V M, Tiginyanu I M, Ursaki V V, Chow L,

Heinrich H and Puaporte T 2010 J. Photochem. Photobiol.A 211 65

[18] Yengantiwar A, Sharma R, Game O and Banpurkar A 2011Curr. Appl. Phys. 11 S113

[19] Cao B Q, Zuniga-Perez J, Czekalla C, Hilmer H, Lenzner J,Boukos N, Travlos A, Lorenz M and Grundmann M 2010J. Mater. Chem. 20 3848

[20] Jeong W J, Kim S K and Park G C 2006 Thin Solid Films506/507 180

[21] Purica M, Budianu E, Rusu E, Danila M and Gavrila R 2002Thin Solid Films 403/404 485

[22] Zhu B L, Sun X H, Zhao X Z, Su F H, Li G H, Wu X G, Wu J,Wu R and Liu J 2008 Vacuum 82 495

[23] Bellingeri E, Marre D, Pallecchi I, Pellegrino L, Canu G andSiri A S 2005 Thin Solid Films 486 186

[24] Strikovski M and Harshavardan K S 2003 Appl. Phys. Lett.82 853

[25] Christen H M, Lee D F, List F A, Cook S W, Leonard K J,Heatherly L, Martin P M, Paranthaman M, Goyal A andRouleau C M 2005 Supercond. Sci. Technol. 18 1168

[26] Rampino S, Bissoli F, Gilioli E and Pattini F 2011 Prog.Photovolt., Res. Appl. at press, doi:10.1002/pip.1234

[27] Zanettini S, Bissoli F, Nasi L, Ranzieri P and Gilioli E 2011Cryst. Res. Technol. 46 881

[28] Bissoli F, Pattini F, Gilioli E and Rampino S 2011 Thin SolidFilms submitted

[29] Calestani D, Zha M Z, Zanotti L, Villani M andZappettini A 2011 CrystEngCommun. 13 1707

[30] Chen L C 1994 Particulates Generated by Pulsed LaserAblation in Pulsed Laser Deposition of Thin Filmsed D B Chrisey and G K Hubler (New York: Wiley) p 190

[31] Craciun V and Craciun D 1999 Appl. Surf. Sci. 138/139 218[32] Krebs H U and Bremert O 1993 Appl. Phys. Lett. 62 2341[33] Kim K H, Park K C and Ma D Y 1997 J. Appl. Phys. 81 7764[34] Sernelius B E, Berggren K F, Jin Z C, Hamberg I and

Granqvist G 1988 Phys. Rev. B 37 10244[35] Shan F K and Yu Y S 2003 Thin Solid Films 435 174[36] Kevin M, Fou Y H, Wong A S W and Ho G W 2010

Nanotechnology 21 315602

7