Epitaxial growth of ZnO films

Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138 www.elsevier.com/locate/pcrysgrow

Epitaxial growth of ZnO films R. Triboulet a,*, Jacques Perrie`re b a

CNRS, Laboratoire de Physique des Solides et de Cristallogene`se, 1 Place Aristide Briand, F 92195 Meudon Cedex, France b Groupe de Physique des Solides, Universite´ Paris VI, Campus Boucicaut, 140 rue de Lourmel, 75015 Paris, France

Abstract After summing up the main physical properties of ZnO and its subsequent applications the aim of this article is to review the growth of ZnO epitaxial films by PLD, MBE, MOCVD and sputtering under their various aspects, substrates, precursors, reaction chemistry, assessment of the layers etc. ., keeping constantly in mind some key issues for the device applications of ZnO in optoelectronics, surface acoustic filters and spintronics, amongst which the growth of high quality epitaxial layers of both n- or p-type conductivity, the possibility of dissolving transition elements in the layers, the growth of ZnO related alloys and heterostructures are of major significance. Ó 2005 Elsevier Ltd. All rights reserved. PACS: 81.05.Dz; 81.15.Kk Keywords: A3 molecular beam epitaxy; A3 physical vapor deposition processes; A3 metalorganic chemical vapor deposition; A3 laser epitaxy; B2 ZnO; B2 semiconducting IIeVI materials

* Corresponding author. Tel.: C33 1 4507 5088; fax: C33 1 4507 5841. E-mail address: [email protected] (R. Triboulet). 0960-8974/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.pcrysgrow.2005.01.003

66

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

1. Introduction ZnO can be considered as an ‘old’ semiconductor. It has been compelling research attention for a long time because of its applications in many scientific and industrial areas such as piezoelectric transducers, optical waveguides, acoustooptic media, conductive gas sensors, transparent conductive electrodes, varistors [1]. It has now received increasing attention and is recognized as a promising candidate for applications related to its optoelectronic possibilities in the UV range. Furthermore, its piezoelectric properties could allow the development of SAW filters to be integrated in future analog circuits for portable electronics for which there is a strong need. These potential applications have boosted research related to the growth of high quality ZnO thin films by a lot of different techniques, which include: -

chemical spray pyrolysis or pyrosol process [2], screen printing [3], solegel technique [4], thermal oxydation of ZnS [5] or Zn [6], post-thermal annealing of zinc-implanted silica [7], sputtering [8], reactive deposition or CVD [9], close spaced vapor transport (CSVT) [10], radical beam getter epitaxy [11], pulsed-laser deposition (PLD) with different variants like electron cyclotron resonance-assisted PLD [12], femtosecond and nanosecond PLD [13], UVassisted [14] or radical assisted PLD [15], - molecular beam epitaxy with also variants like laser MBE or PLD [16], plasmaassisted [17] or electron cyclotron resonance-assisted [18] MBE, gas source or radical source MBE [19] and metalorganic MBE [20], - metalorganic chemical vapor deposition (MOCVD). Amongst all these techniques, PLD, MBE, MOCVD and sputtering are the main ones used for producing epitaxial films and structures for making photonic devices. The aim of this article is, after summing up the main physical properties of ZnO and its subsequent applications, to review the growth of ZnO epitaxial films by PLD, MBE, MOCVD and sputtering under their various aspects, substrates, precursors, reaction chemistry, assessment of the layers etc. . We will keep constantly in mind some key issues for device application of ZnO in optoelectronics, surface acoustic filters and spintronics, amongst which the growth of high quality epitaxial layers of both n- or p-type conductivity, the possibility of transition element incorporation in layers, the growth of ZnO related alloys and heterostructures are primary considerations.

2. ZnO physical properties Amongst the most significant characteristics of ZnO, let us cite [1] the large band gap of 3.37 eV at RT; a unique combination of piezoelectric (e33 Z 1.2 C/m2, among

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

67

the highest values of all semiconductors), conducting, thermal (thermal conductivity of 0.54 W cm
1 K
1, to compare with 0.5, for example, for GaAs) and optical properties; the largest exciton binding energy of all classical IIeVI and IIIeV semiconductors, 60 meV, allowing excitonic stimulated emission up to 550 K, a characteristic already demonstrated. Unlike GaN, high quality substrates are now produced by a variety of techniques, meaning that the ZnO homoepitaxy is now possible. The ternary system CdOeZnOeMgO covers a larger band gap range than nitrides with a smaller variation of the lattice parameter [3.24e3.26 A˚ from ZnCdO (2.8 eV) to ZnMgO (4 eV)]. ZnO has a large shear modulus, a parameter which has been identified to be a key material signature expressing the stability of the crystal. The shear modulus of ZnO has been found to be w45.5 GPa. By comparison, the shear modulus of ZnSe has been estimated to be 18.35, 32.60 for GaAs, and 51.37 for Si. The p-type conductivity of ZnO remains a problematical issue but the possibility of its control has now been reported by several authors. Furthermore, it has been suggested that the incorporation of transition elements like Co or Mn in ZnO could lead to a ferromagnetic phase transition close to room temperature, induced by a gas of electrons and holes. This should pave the way to the development of devices based on the control of the spin state (spintronics) such as quantum computers. Finally, ZnO has the same crystallographic structure as GaN, with a lattice mismatch ! 1.8%, allowing it to be used as an alternative substrate for GaN. Thus ZnO is rich in device opportunities that require the growth of good epitaxial layers.

3. Zinc oxide MOCVD growth Table 1 is a chronological survey of more than 80 papers of the growth modes, goals, growth conditions and characteristics of ZnO layers grown by MOCVD. Two distinct periods can be clearly distinguished in the MOCVD growth of ZnO depending on the applications aimed at. During the first period, roughly from 1964 to 1999, the films were mainly dedicated to such applications as solar cell transparent electrodes, piezoelectric devices or SAW filters. After 1998, given the hope of p-type doping, the main application aimed at was photonic devices. During the first period, the ‘‘epitaxial’’ quality of the films was not as essential as it became after 1998. The evolution of the number of communications versus time (Fig. 1) shows clearly an increasing interest from 1998. Premature reaction between the Zn metalorganic compounds and the oxidants, leading to unwanted deposits upstream from the susceptor, has been the main problem needing to be solved to achieve successful MOCVD growth of ZnO. To solve this key issue, less reactive Zn metalorganic compounds have been used, mainly during the first period, in combination with various oxidants, but also some adducts. Thus, different growth modes such as low pressure MOCVD and photo-enhanced or laser-induced MOCVD are required to increase the growth rate often severely lowered by these less reactive precursors. Separate inlets to inject the metalorganic compound (Fig. 2a and b) and the oxidant have then been used to get rid of the

Table 1 Growth modes, goals, growth conditions and quality and properties of ZnO layers grown by MOCVD Growth mode

Goal

Growth conditions Precursors

LP vertical reactor LP vertical reactor Horiz. PE- and UVE-SS-MOCVD Atm. P rotary reactor Atm. P horiz. reactor

Atm. P vert. and horiz. geometries

PE-MOCVD low P (1.2 mmHg) Horiz. reactor separate inlets Atm. P planetary gear system Laser-induced MOCVD (2 Torr)

Zn propionate, O2

w400

Zn propionate, Zn acetylacetonate, O2 Acoustic transformers

Acoustic, acoustooptic, piezoelectric, solar cell coating Acoustoelectricoptic Electro-optic Photoconductive SAW transducers, acoustooptic Acoustic, acoustooptic, electro-optic Acoustic, acoustooptic, piezoelectric Acoustoelectric, acoustooptic, TCO

Atm. P

Atm. P horiz. reactor

Layer quality and properties Carrier Growth gas T (  C)

Imaging devices

400e680

Substrate

Orientation

Glass

Unoriented films

Glass, metal, mica, Ge, Si, NaCl Si, Ge, GaAs, Al2O3

Refs.

Resistivity (U cm)

n or p (cm
3)

m (cm2/ Vs)

XRD FWHM (arcsec) or PL or transmission

Growth rate (mm/h) [75]

5

8

0.065 at 670  C

10 e10

[76]

Zn propionate, or Zn acetylacetonate

Ar or N2

380e420

DEZ, O2

N2

275

Si

Oriented textured films on mica at 630e670  C Oriented structure for films up to 2 mm thick Vitreous films

DEZ, O2

Ar

200e500

Corning glass 7059

c-axis orientation at 400  C

[24]

400e730

ð01
12Þ sapphire

Epi. films at 400 and Low resistivity 500  C respectively with H2O and with H2O and N2O N2O

[25]

Epitaxial ð01
12Þ layer with 80 W RF power c-axis oriented polycrystalline films

DEZ H2O, CO2/ N2O/ DEZ, CO2

H2 N2 Ar

200

ð01
12Þ sapphire

DEZ, O2

Ar

280e350

Pyrex

DMZ e O2

N2

200e500

SiO2, (100)-Si, -GaAs

DMZ, NO2 or N2O He

20e220

Si

DMZ C4H4O, C4H8O C5H10O DEZn, tBuOH

H2

350e400

He

300

(100) GaP/ Al2O3/ Glass/ Glass

c-axis above 325  C

Polycrystalline films Single Polycrystalline films c-axis oriented polycrystalline films

106e1010, dep. on the growth temp.

Better results with RF discharge

[78]

[23]

10
2
50

0.06e52 n depending on T and O2/DMZ 103e10
1 dep. on DMZ/NO2

!2880 arcsec

[26]

90%

[27]

[78]

Best films with NO2 and irradiation at 193 nm Total absence of pre-reaction Pyrolysis mechanisms

[79]

[80]

[28]

Atm. P horiz. reactor

Optoelectronic devices

DMZ, THF

H2

375e425

(0001) Sapphire

(0002) Epilayers

10
3e1

LP photoMOCVD

Acoustooptic, piezoelectric, solar cell coating

DEZ, O2 or NO2

Ar

150

Corning glass 7059

c-axis oriented polycrystalline films

Decrease of r with UV irradiation

Atm. P vert. reactor

SAW devices

DMZeTHF, H2O

N2, He

350e500

c-axis oriented films

0.1e50

Horiz. LP-SSMOCVD PE-MOCVD, horiz. & vert. reactors

TCO

Zn acetate

(100) Si, glass, (100) Si/SiO2, ð1
102Þ Al2O3 InP, Si, GaAs or soda glass Borosilicate glass

Laser-induced Acoustooptic, MOCVD sep. inlets piezoelectric, solar cell coating Atm. P TCO, varistors, gas sensors, SAW devices LP horiz. reactor Solar cell coating Atm. P horiz. reactor Solar cell transparent electrodes SS-CVD Atm. P Piezoelectric devices Photo-MOCVD Solar cell (253.7 nm) electrodes

350e420

PL domin. by bound exciton emission Improvement in c-axis orientation with UV irradiation

[81,82]

[ with UV light

[29]

[91]

0.03e1

[85]

N2

50e400

DEZ NO2

Ar

300e500

Zn acetylacetonate O2 C H2O

Ar

500e600

Glass

c-axis oriented polycrystalline films

w2.2e2.4 at 550  C

[88]

DEZeH2O

Ar

100e300

Textured films

367e444

10 with B doping 3 ! 10
4

[31]

DEZ, TEA, ethanol He

Corning glass 7059 Glass

Zn acetate

180e500

DEZeH2O, B2H6 dopant

N2 or air Ar

125e140

c-axis oriented polycrystalline films above 200  C
ð0112Þ sapphire/ ð01
12Þ films (0001) Sapphire/ (0001) Films

4

DMZ CO2 or O2

Borosilicate or soda glass plates Corning glass 7059

Disc like structure, diam.100e1000 nm, height 30e60 nm c-axis oriented polycrystalline films Uniform films over 10 ! 10 cm2

Zn(C5H7O2)2, H2O N2 or air, Al(C5H7O2)2

350 (H2O), Corning glass 550 (air) 7059

Textured milky layers

PECVD, 0.8 Torr, vert.

DEZ, N2O

Ar

100e300

Glass, p-Si (100)

c-axis oriented (RF power 200 W)

Zn(C5H7O2)2, H2O, Ar H2O2 or air, Al(C5H7O2)2

350e550

Corning glass 7059

Textured milky layers

Atm. and LP horiz. reactor

w50

Polycrystalline films

Atm. P separate Solar cell inlets, horiz. reactor transparent electrodes SAW device subs., window mater. Opto devices, window coating

1017 e 1020

[86]

0.1e2

n 5e8 ! 4e10 1018

n 2e 10e35 8 ! 1020

[30]

85%

[32]

[86] Improved elec. prop. with UV irradiation, r w 1.3 ! 10
3 3.4 ! 3.4 ! 10
3 (550  C, air) 1020 (350  C, H2O)

[33]

81%

[95]

[34]

4e6 ! 10
3 (550  C, H2O, H2O2)

3.6 ! 1020 Al-doped (350  C, LP)

[96]

(continued on next page)

Table 1 (continued ) Growth mode

Goal

Growth conditions

Layer quality and properties

Precursors

Carrier Growth gas T (  C)

LP-MOCVD horiz. Opto and reactor acousto devices, Solar cells PECVD Piezo-devices

Methylzinc, isopropoxide, or tert-butoxide

N2

240e400 Glass

Zn(AA)2, O2

Ar

200e250 Sapphire, SiO2/Si

LP-MOCVD

Photochromic device

Zn acetate dihydrate C H2O C O2

LP-MOCVD vert high speed rotation reactor SS-CVD

TCO

DEZ, O2, TEGa

ALD, photoMOCVD

SAW devices, TCO, Zn acetate optical modulators TCO DEZ, H2O, B2H6

Substrate

Orientation

c-axis oriented crystallites ð01
12Þ films

350

ð01
12Þ sapphire

Ar

400

Corning 7059 glass Grain size 40 nm

H2O

400

(100) Si

Ar

150

c-axis oriented polycrystalline films Corning glass 7059 Textured films

Corning glass 7059

TCO

DEZ, H2O, B2H6

150

LP-MOCVD rotating disc, vert. reactor, sep. injectors

SAW filters

DEZ, O2

350e450 R-Al2O3

epi. ð11
20Þ ZnO films

Atm. P, horiz. Reactor LP, rotating disc vert. reactor, sep. injectors PEMOMBE

Solar cell coating, optoelectronic UV modulators, SAW devices, UV lasers, GaN buffer layers Nonlinear optics, UV lasers

DEZ, tBuOH, NBuCl, TEGa DEZ, O2

H2

300e400 Glass

N2 C Ar

400e450 R-Al2O3

c-axis oriented polycrystalline films Smooth epi. ð11
20Þ films

550e600 (001) GaAs

Refs. n or p (cm
3)

m (cm2/ Vs)

XRD FWHM (arcsec) or PL or transmission

Growth rate (mm/h) 0.2e4.4

ALD, photoMOCVD

DEZ, O2

Resistivity (U cm)

ZB films with a ZnS buffer layer

108

O90%

0.274e2.07 after annealing in O2 atm.

XRD FWHM 2.8e6.1  , no excitonic emission in the near UV 85%

2.6 ! 10
4

[99]

[92] 25e125 nm/h

[87]

[35]

[100] 2 ! 10
4 on ALD-ZnO/photoMOCVD-ZnO (two step process) Instability after air exposure for MOCVD and photo-MOCVD, stability for ALD

[36]

[37]

Improved crystallinity after annealing at 850  C in O2 C N2 ambience 3 ! 10
4(doped) 60

[38]

[39] 6 meV PL FWHM peak at 3.363 eV (11 K)

[21,40]

Main PL peak at 3.27 eV (120 meV FWHM)

[41]

LP rotating disc V reactor LP remote plasma enhanced CVD Atm. P

UV detectors

DEZ, O2, NH3

350e600 R-Al2O3

Smooth ð11
20Þ films

Light-emitting devices

DEZ, CO2

400

c- and R-Al2O3

Smooth c- and R-epi. films

Optical devices

O2 and Zn acetylacetonate

N2

475

ð01
12Þ Al2O3

LP (5 Torr), Horiz. type

Photonic devices

DEZ, O2

Ar

250e550 (0001) Al2O3

c-axis oriented films

LP (5 Torr), Horiz. type

Photonic devices

DEZ, O2

Ar

400e500 SiO2/Si (100)

c-axis oriented polycrystalline films

Atm. P

Optical devices

DEZ, NO2

500e800 a-, c-Al2O3, (001)GaAs or GaP

LP, sep. injectors

UV photodetectors

DEZ, O2, NH3

380e420 R-Al2O3

epi. ð11
20Þ ZnO films

PECVD

Optical devices

DEZ, O2

600

c-oriented films, better quality N-doped

Atm. P

Photonic devices

DEZ, NO2

Atm. P, sep. injectors, H reactor LP, sep. injectors, H flow reactor LP, sep. injectors, horiz. flow reactor

Photonic and acoustic devices

DMZ-TEN, tBuOH

H2

270e450 c-Al2O3

c-oriented films in a clean reactor

Photonic devices

DEZ, i-PrOH

H2

380

Layer mosaicity dep. on VI/II ratio

Photonic devices

DEZn, i-PrOH

H2

N2 C Ar

(0006) Al2O3

c-Al2O3, bulk ZnO, GaN templates on c-Al2O3

Sharp 380 nm peak, broad 620 nm band, broad 3.8 and 3.3 eV peaks under O-rich XRD FWHM 0.29  , strong CL peak at 379 nm Strong UV emission (growth at 500  C), strong stimulated emission @ RT XRD FWHM 2  , dominant PL emission at 3.360 eV (15 K), sharp (6 meV) stimulated emission at 3.181 eV at RT 4 meV FWHM of D0X with a 500  C grown ZnO buffer on Al2O3 (15 K) Schottky diodes with 1.5 A/W at 5 V bias, 1 nA leakage current at 5 V High resistive N-doped films

1e2

[42]

12.7e 1.3 nm/min.

[93,94]

[89]

[43]

1e1.5

w1017

100

3e4 ! 1017

50

XRD FWHM 353 arcsec, 3 meV FWHM of D0X (15 K) XRD FWHM 400 arcsec, ! 3 meV FWHM of D0X (2 K) XRD FWHM 850 arcsec, ! 4 meV FWHM of D0X XRD FWHM 100 arcsec and 5 meV FWHM of D0X for homoepitaxial ZnO

[44]

[48]

[46]

1

1 ! 1018

MBE-grown ZnO on a-plane Al2O3

c-Al2O3

High r at 390  C (NH3)

[47]

[48]

[97]

[49]

0.95

[98]

(continued on next page)

Table 1 (continued ) Growth mode

Atm. P, sep. inlets

Goal

Photonic devices

LP, sep. Photonic devices injectors, horiz. flow reactor Atm. P, sep. inlets, Photonic and horiz. flow acoustic devices reactor

Growth conditions

Layer quality and properties

Precursors

Carrier Growth gas T (  C)

Substrate

Orientation

Zn(AA)2, O2

N2

475

ð01
12Þ Al2O3

ð11
20Þ ZnO epilayers

DEZ, CO2, NO dopant

H2

360e390 (0002) Al2O3

c-axis oriented films

DEZ, tBuOH

H2

270e450 c-Al2O3

c-axis oriented films

375e400 c-Al2O3

c-axis oriented films

c-axis oriented films

MOMBE

Photonic devices

DEZ, H2O

LP, (200 Torr)

Photonic devices

DEZ, N2O

N2

800

LP

Photonic devices

DEZ, NO2/

N2

DEZ, N2O/

N2

500  C e 6 Torr 800  C ZnO/ZnO/Si/ e 200 Torr 600 (0006) Al2O3

ZnO/sapphire (540 arcsec), and ZnO/Si ZnO/Si

PEMOCVD

Photonic devices

DEZ, O2

N2

PEMOCVD

Photonic devices

DEZ, O2

Ar, N2

Atm. P, SSMOCVD

360 (111) Si, (100) InP, Photonic devices, Zn acetylacetonate, Ar and O2 glass SAW devices, gas Zn(C5H7O2)2 sensors, piezo transducers Photonic devices DEZ, N2O N2 400e900 c-Al2O3, (0001) ZnO

LP, (200 Torr), H flow reactor

600

c-Plane sapphire

Resistivity (U cm)

Refs. n or p (cm
3)

m (cm2/ Vs)

c-axis oriented films

4 ! 1017 24

c-axis oriented films

1.8 ! 1015 (700  C annealing)

Nanorods (10e50 nm) at 500  C, epilayers at 800  C

Growth rate (mm/h)

XRD FWHM 0.35  for 0.30 mm thickness, DAP bands C D0X (4.5 meV FWHM) 17 nm FWHM for D0X at 4 K (doped sample) XRD FWHM 0.3  (growth at 420  C), 3 meV FWHM of the 3.359 eV peak XRD FWHM 0.17  (growth at 370  C), 7 meV FWHM of the 3.36 eV BE line (8 K) XRD FWHM 353 arcsec for ZnO/ZnO/ Al2O3, 3 meV for D0X 3 meV FWHM for BX, appearance of FX (9 K)

c-axis oriented films

(112) ZnO on (111) Si, c-axis ZnO on (100) InP

XRD FWHM (arcsec) or PL or transmission

3 ! 10
4

XRD FWHM 0.56  , 16 meV FWHM of the 3.377 eV BE line (10 K) XRD FWHM 0.16  21.58 (700  C) (004), transmission 96%

[90]

[22]

[50]

[51]

[52]

[53]

1

[54]

[55]

Transmittance O 80%

[102]

0.5 meV FWHM of the 3.357 eV BE line (9 K) of homoepitaxial layers

[56]

LP, H flow reactor

Photonic devices

DEZ, i-PrOH or tBuOH

H2

320e580 c-plane sapphire

PECVD

Photonic devices

DEZ, CO2

H2

230

LP, H flow reactor

Photonic devices

DMZn, acetone i-PrOH or N2O

H2 w450 (prop.)

GaN/Si (111) templates

c-axis oriented films at 19 W RF power c-axis oriented films for i-PrOH or NO2

PEMOCVD

Photonic devices

DEZ, O2

Ar

c-Plane sapphire

c-axis oriented films

PEMOCVD

Photonic devices

DEZ, O2, plasma NH3

SS-CVD LP-MOCVD

Photonic devices

Zinc acetate dihydrate DEZ, NO

600

(001) Si

500e600

p-Type after doping with NH3, (1016 cm
3) but ‘unstable’ c-axis oriented films

O2

200e275 (100) Si, c-Al2O3

N2 or Ar

200e500 Corning glass 1737

Ar

250e720 (001) GaAs

(002) ZnO films

?

c-Al2O3

Same c-orientation when flowing DEZn before introducing O2 Strong c-oriented growth

Vert. LP-MOCVD Optical devices

DEZ, O2

LP-MOCVD

GaN substrates

DEZ, O2

LP-rotating disk vert. react., sep. injectors LP-MOCVD, sep. injectors

Photonic devices

DEZ, O2

Ar

600

c-Al2O3

Optical devices

DEZ, O2

Ar

600

c-Al2O3

LP-rotating disk vert. react., sep. injectors LP, sep. inject. reactor

Optical devices

DEZ, O2

Ar

600

(001) Si

Photonic devices

DEZ, O2, NH3

610

R-Al2O3

PECVD LP vert. react. PECVD

Photonic devices Photonic devices Photonic devices

DEZ, CO2 DEZ, O2 DEZ, CO2

N2 and Ar H2 Ar H2

230 (001) Si 100e200 (001) Si 120 (001) Si

OOptical quality with tBuOH, 1.1 meV FWHM of D0X line at 2.1 K XRD FWHM 0.21  , intense FE emission XRD FWHM 590 arcsec (800  C grown films at 300 mbar) Better quality after 700  C annealing, 87% transmission XRD FWHM 0.26  at 500  C, 0.18  after annealing at 700  C 90% Transmittance 20

P 1.0 ! 1015e 1.0 ! 1018 n5! 1018e 5! 1019

41 at 650  C

[85]

[59]

[109]

[101] [61]

Better films at 650  C

[105]

O0.5

Increase from 2.4 to 1100 after annealing under O2

[57]

[58]

0.1

n Y with [ P (O2)

Quality improvement after thermal annealing under vacuum Crystallinity degradation when increasing P (O2)

c-axis oriented films

[62]

[63]

[64]

[65]

102

c-axis oriented films c-axis oriented films c-axis oriented films

2.6 at 400 mbar

p w 1016 3.6

Films of better crystallinity when N-doped XRD FWHM 0.21  XRD FWHM 0.44  Film quality improved by post-annealing up to 900  C

[66]

[67] [68] [69]

(continued on next page)

Table 1 (continued ) Growth mode

Goal

Growth conditions

Layer quality and properties

Precursors

Carrier gas

Growth T (  C)

Substrate

Orientation

LP, sep. injectors, horiz. flow reactor LP, vertical reactor

Photonic devices

DEZ, O2

H2

340e460

(111) Si

c-axis oriented films

Photonic devices

DEZ, O2

Ar

450

GaN (0001)/ 6H-SiC (0001)

c-axis oriented films

LP-MOCVD

Photonic devices

Zn acetylacetonate, O2

Ar

520

ZnO buffer/Si (111)

c-axis oriented films

Resistivity (U cm)

Refs. n or p (cm
3)

m (cm2/ Vs)

XRD FWHM (arcsec) or PL or transmission XRD FWHM 0.19  for a growth temp. of 400  C Growth in the SK mode on O-terminated GaN (0001) Strong PL peak at 378 nm

Growth rate (mm/h) [70]

[71]

[72]

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

75

Papers reporting on the MOCVD growth of ZnO Number of papers

35 30 25 20 15 10 5 0 1960

1970

1980

1990

2000

2010

Years Fig. 1. Number of papers related to the MOCVD growth of ZnO versus time.

problem of pre-reaction. They appeared from 1981 [27] and have been then generalized during the second period. Various carrier gases, different geometries, horizontal or vertical reactors, high speed rotation reactors have been used as well. 3.1. Precursors The main criteria, which have guided the choice of precursors, have been their cost, safety, purity, availability, growth rate and the key problem of premature reactions. DEZn, (C2H5)2Zn, has been and remains the most used Zn precursor [21e71]. Because of premature homogeneous reactions with some oxidants in the gas phase occuring upstream from the reactor which severely limit the process, several other less reactive metalorganic compounds have been used mainly during the first period, before 1999: zinc propionate (Zn(C3H5O2)2) [75e77], dimethyl zinc (Zn(CH3)2) [78e 80,82e84], zinc acetate (Zn(CH3COO)2) [85e87], zinc acetylacetonate (Zn(AA)2) Zn(C5H7O2)2 [72,73,76,77,88e90] considered as low cost and easy to handle and some adducts like DMZeTHF or DEZeTHF [91].

Fig. 2. Separate inlet reactors according to: (a) Gorla et al. [21] and (b) Ye et al. [22].

76

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

As oxidants, in addition to the very reactive O2 [21,23,24,27,29,35,38,40,42e 44,46,47,54,55,59,60,62e66,68,70e72,75,76,78,83,87e90,92], alternative less reactive compounds such as CO2 [22,25,26,58,67,69,83,93,94], H2O [31,33,36,37,51,87,88,91, 95,96], H2O2 [96], N2O [25,34,52,53,56,79,84], NO2 [29,30,45,48,53,79], NO [61], the heterocyclic furan [80] (C4H4O), tetrahydrofuran [80,81] (C4H8O), tetrahydrophyran [80] (C5H10O) and alcohols, mainly tBuOH [28,39,50,57,97], and i-PrOH [49,57,84, 98] have been used. The advantages of tBuOH over i-PrOH have been demonstrated [57]. The low deposition rates obtained with such less reactive precursors and oxidants as CO2, N2O, NO2 or H2O are often enhanced either by an excitation using a plasma [34,47,54,55,58e60,83,92e94] or UV light [26,29,30,33,36,37,77,79]. Alkylzinc alkoxides like methylzinc isoprotoxide MeZn(OPri) [85], and methylzinc tert-butoxide MeZn(OBut) [99], and Zn acetate [85,86,100,101] and Zn acethylacetonate [102] have been shown to be single source precursors for the growth of ZnO films, i.e. allowing the growth of ZnO without any added oxygen source and which then avoids premature reactions. Note that the designation ‘‘single source precursor’’ is sometimes incorrectly used when oxygen is the carrier gas. At the present time, DEZn as a Zn precursor and O2, CO2, N2O or tBuOH as oxygen sources are mostly used with a separate injector geometry. To dope the layers, B2H6 [33,36,37], Al(C5H7O2)2 [95,96] and TEGa [35,39] have been used as a source of B, Al and Ga, respectively, and NO [22], NH3 [42,46,60] and diallylamine [103] as an N source. The decomposition mechanisms and kinetics associated with the main precursors, DEZn and DMZn, have not been intensively studied so far. It has been shown [104] from a mass spectrum analysis of tertiarybutanol and DEZn introduced in the reactor separately or together under H2 that the most abundant products of the separate pyrolysis were ethane and ethylene for DEZn while in the case of tertiarybutanol, water and isobutene were detected. The formation of isobutene was attributed to the recombination of methyl and propyl radicals. Another possibility is homolytic fission of tertiarybutanol. In addition to these products, ethylene was obtained when DEZn and tertiarybutanol were interacted in the reactor. The pyrolysis of these sources separately or in a mixture occurs by homolytic fission and radical recombination. tBuOH appears to be thermally more stable than DEZn, with complete decomposition at 430  C and 360  C, respectively. The percentage of decomposition of tert-butanol and tert-butanol/ZnO as a function of temperature is displayed in Fig. 3. Apart from these experimental studies, the following reactions have been proposed in the case of DEZn, which is a very reactive electron deficient compound with two vacant orbitals, with oxygen [24]: ðC2 H5 Þ2 ZnC7O2 ZZnOC4CO2 C5H2 O with H2O [25]: ðC2 H5 Þ2 ZnC2H2 OZZnðOHÞ2 C2C2 H6 ZnðOHÞ2 ZZnOCH2 O

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

77

Fig. 3. Decomposition % of tert-butanol and tert-butanol/ZnO versus temperature (from [104]).

with CO2 [25]: ðC2 H5 Þ2 ZnC2CO2 ZðC2 H5 COOÞ2 Zn ðC2 H5 COOÞ2 ZnZZnOCgaseous products and with alcohols [28]: ðC2 H5 Þ2 ZnCðReOHÞZZn alcoholates ðReOZnOeRÞ or adducts ðC2 H5 Þ2 ZnORH ðalkyl Zn alkoxidesÞ The same type of reactions has been proposed in the case of DMZn with tBuOH [39]: ðC2 H5 Þ2 ZnCðCH3 Þ3 COHZðCH3 Þ3 COZnCH3 CCH4 At higher temperature or a large partial pressure of tBuOH, the alkoxides react in a second step to form a bisalkoxide: ÿ ðCH3 Þ3 COZnCH3 CðCH3 Þ3 COHZ ðCH3 Þ3 COÞ2 ZnCCH4 The following reaction has been proposed in the case of DMZn with CO2 [86]: ðC2 H5 Þ2 ZnC2CO2 ZðCH3 COOÞ2 Zn Zinc acetate can then react with water according to the following reaction to form ZnO [86]: ðCH3 COOÞ2 ZnCH2 OZZnOC2CH3 COOH The different reaction mechanisms, kinetically limited, mass-transport limited and desorption limited growth, have been studied in the case of the DEZn and O2 sources [105] and of the DEZn and tButanol [50] and or isopropanol sources [57]. In Figs. 4 and 5 are displayed the corresponding Arrhenius plots of the growth rate versus

78

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

Fig. 4. Arrhenius plots of the growth rate versus reciprocal temperature using DEZn and O2 sources (from [105]).

reciprocal temperature. In agreement with such results, the growth temperatures with such sources lie generally in the 350e600  C temperature range. It has been shown [105] that the surface morphology, structural, optical properties of ZnO films grown using DEZn and O2 were improved with increasing growth temperature up to 650  C, whilst the properties of the films grown at higher temperature (720  C) were degraded due to the decomposition of the films. This increase in the growth temperature in order to improve the properties of the layers appears to be a recent trend. 3.2. Substrates During the first period mentioned above, the main substrates used have been Corning 7059 glass and more rarely sapphire with different orientations, ð01
12Þ or (0001).

Fig. 5. Arrhenius plots of the growth rate versus reciprocal temperature using DEZn and t-butanol sources (from [50]).

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

79

After 1998, in order to get epitaxial layers, such substrates as a- [45,48], c- [22,43,45,47,49e51,54e57,59,62e64,74,93,94,97,98], and r-sapphire [21,40,42,46, 66,89,90,93,94], (001)-GaAs [41,45,74,78], -GaP [45,80], and -Si [34,58,65,67e 69,78,91), (111)-Si [70,72], GaN/c-Al2O3 [98], GaN/(111)Si [84], GaN/AlN/ 6HSiC(0001) [71] and finally more recently ZnO [56,98] or ZnO/a-Al2O3 [48], ZnO/sapphire [52] and ZnO/Si [52,53] have been used. Note that c-plane sapphire remains the most used substrate. The epitaxial relationships of ZnO on sapphire have been identified for both R- [38] and c-sapphire [106]:  ZnO=ð01
12ÞAl2 O3 0 ð11
20Þ ZnO k ð01
12Þ Al2 O3 0 ½0001 ZnO k ½0
111 Al2 O3  ZnO/(0001) Al2O3 0 (0001) ZnOk(0001) Al2O3 with the following in-plane orientation relationships:  ½10
10 sapphire k ½11
20 ZnO
 ½1120 sapphire k ½01
10 ZnO The in-plane orientation of the films influences their quality. Two important issues in epitaxial growth in the [0001] direction of wurtzite semiconductors such as ZnO is the control of the polarity of the epitaxial film, which can influence significantly the crystalline, optical and electrical properties, and the elimination of rotation domains frequently observed in the growth of ZnO films on c-Al2O3. The conditions for obtaining no-twist ZnO films have been determined experimentally by Zhang et al. [62]: flowing DEZn before introducing O2 with TZn % 260  C and 0.1 % v % 1.5 mmoles/ min. The in-plane orientation, however, was demonstrated to have no relation with the surface polarity and the films grown by MOCVD all had Cc polarity [62]. It has been shown that the morphology of the layers was very sensitive to the presence of deposits over the walls [97] and was better on R- than on c-sapphire. Significant differences have been found between the morphologies of ZnO layers grown on sapphire depending on the substrate orientation (a-, c-, r-, and m-planes) [107] SFM images of the surface morphology of ZnO layers after 25 s of deposition on the four substrate orientations are shown in Fig. 6. The films grown on a- (Fig. 6a), c- (Fig. 6b) and r-planes (Fig. 6c) display periodic strips formed by small ZnO grains over the whole surface, while the films deposited on m-planes cover uniformly the surface. These strips have been found to decorate initial substrate steps, copying the sapphire substrate morphology. In order to integrate future ZnO devices with Si IC technology, the growth of ZnO films on Si substrates has attracted a great deal of attention. Highly c-axis oriented films have been obtained on Si (100) [34,58,65,67e69,78,91] and Si (111) [70,72] at temperatures as low as 100  C [68], while the best structural quality is generally reached at higher temperatures (w400  C [70]). The improvement of the film quality by annealing has been demonstrated [69] as well as the beneficial deposition of ZnO buffer layers on Si substrates [72].

80

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

Fig. 6. SFM images of the surface morphology of ZnO films after 25 s deposition on a- (a), c- (b), r- (c) and m- (d) sapphire planes (from [107]).

3.3. Layer characteristics During the first period, c-axis oriented polycrystalline films have been essentially reported when grown on glass. Some epitaxial films have been reported when the growth occurred on c- and r-sapphire templates. Transmission, reaching 90%, and electrical properties have been among the main characteristics reported for these layers. Resistivities ranging from 10
2 to 108 U cm have been measured on as-grown layers, depending on the growth conditions (VI/II ratio, temperature, nature of precursors) and on eventual post-growth annealing (under O2 pressure), and resistivities ranging from 2 ! 10
4 to 6 ! 10
3 U cm for films doped with B, Al or Ga. Electron concentrations in the 1017e3.6 ! 1020 cm
3 have been reported with mobilities ranging from 4 to 60 cm2/Vs. From 1999, the FWHM of band edge emissions, generally bound exciton lines (D0X ), have been reported; 1.1 meV at 2.1 K for heteroepitaxial layers on sapphire [57] and 0.5 meV at 9 K for homoepitaxial films [56] are among the best values reported so far. Rocking-curve FWHMs of 255 and 100 arcsec have been measured, respectively, on heteroepitaxial layers grown on c-Al2O3 and homoepitaxial layers [99]. Electron concentrations ranging from 1.8 ! 1015 to 1018 cm
3 with mobilities ranging from 20 to 100 cm2/Vs have been reported. Rocking-curve FWHM and electron concentration can be considered as more significant characteristics of the layers than the D0X FWHM which depends on the

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

81

experimental conditions and on the area and place of the spot measured. But the rocking-curve FWHM highly depends as well on the experimental conditions which are rarely given. 3.4. Doping It has been shown that hydrogen behaves essentially as a shallow donor in ZnO [60] and hence has to be carefully considered in view of p-type doping. The inevitable presence of hydrogen in the MOCVD growth of ZnO, either coming from the carrier gas or from the organic radicals of Zn precursors, could make the MOCVD technique unfavorable for p-type doping. On the other hand, hydrogen could have the beneficial effect of increasing the acceptor solubility [108] and suppressing compensation by native defects. It has been shown [109] that hydrogen had a highly unusual behaviour in ZnO. While hydrogen behaves as an amphoteric impurity in most semiconductors, meaning that it acts as a compensating center, it behaves exclusively in ZnO as a shallow donor (HC only) which activates the oxygen vacancies, passivates the Zn vacancies and increases the solubility of acceptors. In that sense hydrogen could be the donor impurity used in the framework of a codoping approach [110,111] to get p-type layers. Attempts to achieve p-type doping in ZnO MOCVD layers have not been so far very successful. Using N as a dopant from NH3 activated by a plasma, Wang et al. [66] have obtained ZnO layers with a hole concentration w1016 cm
3, but this p-type conductivity turned out to be unstable as a function of time. The instability of the electrical properties of the layers as a function of time, regardless of the deposition method, has been reported for a long time [27] and shown to be closely related to the change in surface conductance due to oxygen chemisorption. Using NO as an oxidant and DEZn as a Zn precursor, Li et al. [61] obtained p-type polycrystalline films with p ranging from 1.0 ! 1015 to 1.0 ! 1018 cm
3 and m w 0.1 cm2/Vs but showing also an unstable state. Room temperature Hall measurements yielding hole concentrations as high as mid 1018 cm
3 using N have been said to be obtained at CERMET [112]. Conversion to p-type was not achieved after Ga and N ion implantation followed by thermal post-annealing, although postimplant annealing, up to 800  C restored the optical and structural quality of the samples to a high degree [113]. A two step strategy for achieving p-type doping has been developed by Rommelue`re et al. [103]. Experimental conditions leading to ‘‘intrinsic’’ ZnO have been first determined. They have been shown to depend, amongst others, on the growth temperature and the VI/II ratio (Fig. 7). Two different donor doping mechanisms have been found to occur on both sides of some kind of ‘‘doping window’’. In a second step, a hole concentration reaching 6.5 ! 1017 cm
3 at RT associated with a hole mobility as high as w16 cm2/Vs has been obtained by post-growth annealing such layers under high pressure conditions stemming from the decomposition of NH4NO3. The decomposition of this compound at high temperature can establish both a nitrogen pressure, for nitrogen doping, and an

82

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

Fig. 7. Electron concentration versus growth temperature for various VI/II ratios.

oxygen pressure, to restore the stoichiometry and to make the layer less oxygen thirsty and then more stable. In addition, the annealing process activates the acceptors, according to a mechanism which has not been yet been identified. 3.5. ZnO related materials The development of ZnO-based semiconductor devices requires band gap engineering. Ternary alloys such as MgZnO and CdZnO can tailor the band gap from w2.8 eV to 4 eV, depending on the Cd or Mg content and enable the preparation of low dimensional heterostructures suitable for making light emitters as well as ultraviolet photodetectors. Zn1
xMgxO thin films up to x Z 0.49 at 500e650  C were epitaxially grown on cAl2O3 substrates using DEZn, bis-cyclopentadienyl-Mg (cp2Mg) and oxygen as the reactants with argon as a carrier gas by Park et al. [44]. No significant phase separation was observed as determined by X-ray diffraction measurements. Due to the Mg incorporation, the near-band-edge emission was blue-shifted from 3.364 eV for x Z 0 to 4.05 eV for x Z 0.49 eV at 15 K. These alloys were grown as well with Mg concentrations up to 9% by Waag et al. [114] using the same Zn and Mg precursors and alcohols like i-PrOH or tBuOH for low temperature growth (300
500  C) and N2O for high temperature growth (600e950  C). Due to the incorporation of Mg, the c-lattice constant became smaller, whereas a became larger. This unusual behaviour was attributed to the increasing ‘‘cubic influence’’ with increasing Mg concentration since the binary MgO has a rocksalt lattice. Furthermore, such ZnMgO layers were incorporated as barriers in ZnMgO/ZnO quantum well structures. More recently, Zn1
xMgxO (0 % x % 0.33) thin films were deposited on r-sapphire at w400  C at typical growth rates w0.6e0.7 mm/h using MCp2Mg and DEZn as a precursor transported with ultrahigh purity argon in an axisymetric rotating disk vertical flow reactor [115]. A thin ZnO buffer layer of the order of 50 A˚ was found to be critical for the growth of epitaxial films with a wurtzite-type structure. The epitaxial relationship between the wurtzite-type

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

83

Zn1
xMgxO films and r-sapphire was determined to be ð11
20Þ Zn1
x Mgx O k ð01
12Þ Al2 O3 and ½0001 Zn1
x Mgx O k ½0
111 Al2 O3 . ZnCdO epitaxial layers with a Cd concentration up to 2% were grown by the same group on ZnO/sapphire substrates [116]. A narrowing of the fundamental band gap up to 300 meV was observed, while introducing a lattice mismatch of only 0.5% with respect to binary ZnO. Nevertheless, distinct Cd concentrations were distinguishable within the layers. 3.6. Conclusions Very promising results have already been obtained in the MOCVD growth of ZnO films. As usual, the best layers have been obtained by homoepitaxy, but the quality of the layers remains limited by the quality of the substrates. The availability of the substrates and the difficult technological issue of the preparation of their surface remain also limiting as factors. Concerning the key issue of p-type doping, the detrimental compensation of acceptors as a result of the inevitable presence of residual hydrogen donors could be balanced by an increase of the solubility of acceptors. Preliminary results demonstrate the possibility of p-type doping by nitrogen of MOCVD grown ZnO films, either during the growth process or by post-growth annealing. As far as the incorporation of transition elements in the ZnO lattice for spintronic applications the physical and chemical mechanisms of ZnO MOCVD growth close to thermodynamic equilibrium make this technique less favorable than MBE or laser ablation, which are much further from thermodynamic equilibrium than is MOCVD. 4. Zinc oxide epitaxial thin film growth by PLD 4.1. Introductory remarks Pulsed-laser deposition (PLD) of thin films can be considered as a simple deposition process which uses pulsed laser radiation to vaporise by photon absorption the surface of the material (target) to be deposited as a thin film on a surface [117e120]. A schematic pulsed-laser deposition system for the growth of thin films is shown in Fig. 8. Intense laser pulses of nanosecond duration range are focused in a vacuum chamber onto a target surface where they are absorbed. Above a threshold power density depending upon the target material (generally around 50 MW/cm2), significant material removal from the target occurs in the form of an ejected luminous plume (see Fig. 8) whose species are collected on a substrate which can be heated to ensure the growth of crystalline material. An empirical description of PLD involves successive steps. First the laserematter interaction leads to the melting of the target surface and vaporisation in the shape of a plume of the thin upper layer of the molten surface. The plume propagates then in a direction normal to the target (Fig. 8), with a possible interaction with an ambient gas. Finally the film forms at the surface of the substrate. Each step plays a role in the composition, crystalline quality and surface morphology of the deposited material.

84

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

Fig. 8. Schematic description of a PLD experimental set-up.

The success of high Tc superconducting thin film formation by PLD has led to the intensive development of this technique since 1987 [119]. Nowadays, PLD has become one of the most widely used techniques in research laboratories to grow high quality thin films. The method itself and its advantages and specificities have been previously described in some review papers [118] in which the laserematter interactions, plasma formation and expansion (see Fig. 8), and effects of the various parameters (laser fluence and wavelength, pulse duration) are analyzed and discussed. Let us just recall here one of the most important points: PLD allows the growth of thin films in reactive oxygen pressure (from 10
6 to 1 mbar) with the same composition as that of the target (congruent laser ablation). This makes PLD particularly well suited for growing thin films of oxides, even those with complex compositions and having a wide range of physical properties: superconductors, ferroelectrics, magnetic or optical materials. [121]. Epitaxial films of oxides have been grown on well suited substrates and multilayers or supperlattices have also been obtained by sequential PLD from various targets [122]. 4.1.1. Droplets The main drawback of the pulsed-laser deposition which limits its industrial applications is related to the ejection of liquid droplets by the target during the laser irradiations. This phenomenon leads to PLD materials consisting of rather smooth

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

85

films with macroscopic particulates (in the micrometer range) superimposed on it [118]. A careful selection of target quality, laser wavelength and intensity, and homogeneisation of the laser power density across laser spot are known to slightly reduce this effect [123]. Besides the optimisation of ablation conditions, various solutions have been proposed to avoid the deposition of these droplets or to reduce their density and size: a mechanical velocity filter [124], off-axis deposition [125], the use of a second laser in order to eliminate the droplets in the case of crossed laser induced plasma plumes [126] or preheating the target up to the molten state just before ablation [127]. Various phenomena have been considered to explain the presence of these droplets at the surface of the PLD films. The phase explosions, splashing effect or even exfoliation have given rise to contradictory interpretations [123,128e130]. The origin of these particulates is thus still a matter of discussion. It is now generally admitted that the presence of droplets at the film surface is the consequence of thermal effects taking place in the target during the laserematter interaction, liquid material being ejected from the molten zone of the irradiated target. In that sense, ZnO can be considered a priori as a very good material for the formation by PLD of droplet free films, according to the fact that it melts and sublimes at 2248 K without phase transformation from the molten liquid to the evaporative plume. The absence of the molten layer should thus avoid the ejection of droplets. However, it has been experimentally shown [131] that even though ZnO sublimes at its melting point, its molten state exists for a finite duration and could be the source of the ejection of liquid droplets which are observed on ZnO films grown by PLD using 532 or 1064 nm laser wavelengths [132]. In the case of wide band gap insulators, it has been shown [133] that the surface morphology of the PLD films greatly depends on the target optical properties, the higher the optical absorption coefficient a, the lower is the droplet density on the surface of the deposited films. In fact, laser irradiations of a low a value target lead to a large volume heating with the formation of a thick molten layer and thus to an important ejection of liquid droplets. It follows that a high value of the optical absorption of the target material at the laser wavelength is necessary for the formation of droplets free films. In the case of ZnO this condition is fulfilled for energies of photons higher than the band gap (3.3 eV), i.e. for laser wavelength lower than 380 nm. This is the reason why it has been found that ZnO films deposited with the use of ArF (193 nm), KrF (248 nm) excimer lasers or frequency quadrupled (266 nm) Nd:YAG lasers exhibited a far superior morphology than films grown by PLD using the second harmonic (532 nm) or the fundamental frequency (1064 nm) of Nd:YAG lasers [134,135]. Except in the particular case of ultrashort (femtosecond) laser pulses, PLD of ZnO films is nowadays essentially carried out with ArF or KrF excimer lasers, or frequency quadrupled Nd:YAG lasers. 4.1.2. ZnO growth of thin films The main processing parameters of PLD such as substrate temperature, background gas pressure, target to substrate distance, pulse repetition rate and laser fluence play a more or less important role on the growth and properties of ZnO

86

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

films. For example, crystalline ZnO films can be obtained by PLD at temperatures as low as 200  C [135], the crystalline quality of the films being improved with increasing substrate temperatures [136]. This low temperature for crystalline thin film formation is one of the specificities of the PLD technique. It can be related to the kinetic energy of the species (10e100 eV) emitted by the target (largely higher than the corresponding kinetic energies observed in the other methods of vapor growth), enabling the growth of crystalline films via an enhanced surface mobility of these species [120]. This low growth temperature of crystalline films could lead to interesting applications in various domains (displays, solar cells or sensors), when soft substrates like plastics which cannot withstand high temperatures are used. The high kinetic energy for the emitted species is also at the origin of a low surface roughness of the films deposited at low temperature. Indeed simulations have shown that the roughness for energetic particle deposition (like PLD) is smaller than for thermal deposition, the difference decreasing with increasing temperatures [137]. An oxygen ambient gas is generally needed in order to form stoichiometric and crystalline oxide films by PLD. In the laser ablation process, under low oxygen pressure (from vacuum up to 0.5 mbar), atoms and ions are emitted by the target during laser irradiations and then condense on the substrate to form the films. At higher pressures (1e20 mbar), the emitted species undergo a great number of collisions with gas molecules and condensation occurs in the gas phase leading to the formation of nanosize particles which precludes the growth of smooth and dense films at the surface of the substrate. The mechanisms of oxygen incorporation in oxide films grown by PLD have been studied by isotopic tracing techniques [138]. It appeared that only one part of the oxygen present in the film is directly incorporated from the target. The exact fraction of the oxygen coming from the target depends upon the material, while the remaining part comes from the ambient gas. Therefore, ZnO films grown by PLD under vacuum (10
6 mbar) contain oxygen vacancies which have strong effects on their electrical properties and may also play an important role on the crystalline quality of the films [139]. Actually, oxygen deficient zinc oxide films show low crystallite size and high mosaicity. This lack of oxygen can be related to the high volatility of oxygen with respect to zinc, and to the fact that a Zn excess is observed in the near surface region of ZnO target after laser irradiation [140]. This Zn excess in the ZnO target being due to the Zn atoms re-condensation on the surface after repeated laser irradiation, while oxygen is lost by vacuum pumping, or/and to the surface reduction of the ZnO target under the high temperature conditions (during laser ablation) under very low oxygen pressure. In order to avoid the formation of off stoichiometric and poorly crystallized zinc oxide films, PLD has to be carried out under an ambient oxygen pressure. The published works on the determination of the optimised value of oxygen pressure for ZnO thin film growth by PLD show differences in these optimised oxygen pressures. The reason why is clearly due to the differences in experimental systems used for the growth and in particular to differences in the target to substrate distance. Indeed, it has been established that a relationship exists between the oxygen pressure P and the target to substrate distance d, during PLD of oxide. In order to grow the best quality films it is found that P d g Z constant [141], where g depends upon the oxide compound.

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

87

This relationship can be simply explained by assuming that an optimum exists for the kinetic energy of the species reaching the surface of the growing film. Too low an energy does not allow the surface diffusion of these species and thus precludes the formation of high quality crystalline films. On the other hand, too high an energy leads to surface sputtering and the creation of defects in the growing film. The P d g Z constant relationship corresponds to the fact that an increase in pressure will induce collisions of the emitted species with the gas molecules and thus to a decrease of their kinetic energy. To maintain the optimum kinetic energy, it is therefore necessary to decrease the target to substrate distance. The last experimental parameter, the laser energy fluence, plays a noticeable role on the growth of ZnO films [142]. At low laser fluence, the growth of ZnO occurs via a 3D-island mode with a low deposition rate. This can be related to the low kinetic energy of the species in these experimental growth conditions which decreases their surface mobility and accordingly leads to an island growth mode [137]. On the other hand, too high a laser fluence causes a degradation of crystalline, electrical and optical quality of ZnO films through the bombardment of the growing film by energetic species. An optimal laser fluence window, from 1.2 to 2.5 J/cm2, can be defined to obtain high quality ZnO films. Taking into consideration all these points, the classical experimental conditions for the growth of high quality ZnO epitaxial films by pulsed-laser deposition are the following: use of an excimer (ArF or KrF) or frequency quadrupled laser, with a fluence around 2 J/cm2, a substrate temperature in the 600e800  C range under an oxygen pressure lying between 10
6 and 10
1 mbar, for a target to substrate distance in the 5e10 cm range. 4.2. ZnO epitaxial growth 4.2.1. Sapphire substrates Although numerous applications of ZnO thin films only need polycrystalline materials, the development of optoelectronics applications based on ZnO involves epitaxial thin films. Due mainly to the fact that sapphire substrates can be obtained in large area wafer at low cost, a large part of the studies devoted to the growth of epitaxial ZnO thin films by PLD used c-cut Al2O3 substrates. Epitaxial ZnO films have been obtained at low temperatures (300  C), but the crystalline quality of the films (assessed from the FWHM of X-ray diffraction peaks, rocking curves and f scan peaks), increases with increasing temperatures up to 750e800  C. The in-plane alignment of the ZnO films was reported [143] to be a function of the substrate temperature, two orientations being present at low temperatures, while a single one is observed with increasing temperatures. Fig. 9aec shows the pole figures of the ð10
10Þ planes of the ZnO films grown at various temperatures on c-cut sapphire substrates, while for comparison purposes the pole figure of ð10
10Þ planes of the sapphire is shown in Fig. 9d [144]. For substrate temperatures lower than 500  C, the in-plane epitaxial relationship deduced from Fig. 9a is the following: ZnO ½10
10 k Al2 O3 ½10
10, i.e. the direct superposition of hexagons of the basal plane of ZnO and sapphire, the

88

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

Fig. 9. Pole figure for ð10
10Þ ZnO grown on sapphire at various T (from [144]).

corresponding lattice mismatch for this in-plane orientation being 31.8%. With increasing temperatures, a second in-plane orientation appears in Fig. 9b, i.e. ZnO ½10
10 k Al2 O3 ½11
20, which presents a reduced value for the lattice mismatch (18.3%) and corresponds to a 30  rotation of the hexagon of ZnO with respect to the substrate. Two types of crystallites are thus present in the ZnO films, but at substrates temperatures higher than 700e750  C, a perfect alignment with the ZnO ½10
10 k Al2 O3 ½11
20 orientation is achieved [143,144]. This last orientation corresponds to the epitaxial relationship classically observed in the growth of ZnO on a sapphire substrate, whatever the growth method. In this in-plane orientation, the hexagonal close-packed oxygen sublattices in both substrate and film have an identical in-plane orientation, although the lattice mismatch is as large as 18% [145]. By reducing the deposition rate, this epitaxial relationship can be observed in the low temperature domain where the other in-plane orientation is the dominant one (Fig. 9a). This means that the ZnO ½10
10 k Al2 O3 ½11
20 orientation is thermodynamically more stable. It was concluded [145] that at high substrate temperature and low deposition rate, a thermodynamically favorable orientation is chosen, whereas the orientation choice is based on the local interface energy when the growth is governed by kinetics, i.e. at low substrate temperature and high deposition rate. This change in the in-plane epitaxial relationships with temperature was accompanied by a correlated improvement in the crystalline quality of the ZnO films, as shown by the decrease of the FWHM of the rocking curves for the (000l) family planes with increasing temperatures [143]. Rocking-curve measurements characterize the angular distribution of the c-axis of the ZnO crystallites. To gain information on the intrinsic quality of the crystallites (presence of point and

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

89

extended defects), Rutherford backscattering spectroscopy (RBS) in channelling geometry can be used. Fig. 10 therefore represents the random and aligned spectra for an epitaxial ZnO film grown at 750  C [143]. The large decrease in the surface backscattering yield for the aligned spectrum is due to the channelling of the incident ions along the c-axis and it demonstrates the high crystalline quality of the film. The cmin value (ratio of the yield in channelling and random geometries) deduced from Fig. 10 (2%) is similar to that measured on bulk single crystal. Moreover, an increase in the backscattering yield is observed towards the filmesubstrate interface. It characterizes the presence of defects at this interface which act as de-channelling centers for the incident ions. These results show that the growth by PLD of high quality epitaxial ZnO layers on sapphire requires processing temperatures higher than 750  C. To reduce this growth temperature without decreasing the quality of the ZnO films an interesting approach has been proposed: the ultraviolet-assisted PLD technique (UVPLD), where an UV source provides in situ irradiation with UV photons during the PLD process [146]. The use of this UVPLD resulted in the growth of epitaxial ZnO layers on sapphire substrates that exhibited similar quality to the ZnO layers deposited at substrate temperatures higher by at least 200  C [146,147]. As a matter of fact, Fig. 11 shows that ZnO films grown at 550  C by UVPLD exhibited a narrower FWHM rocking curve (0.17  ) than that measured on ZnO films grown by PLD at the same temperature (0.57  ). Moreover, the cmin value measured by RBS in channelling geometry (2%) was equal to that reported in Fig. 10 (for ZnO film grown at 750  C), indicating a similar crystalline quality. The origin of this effect seems to be the fact that UV photons can efficiently dissociate molecular oxygen at the surface of the growing film and enhance the surface mobility of the

Fig. 10. RBS spectra in random and channelling geometry for a ZnO film on sapphire (from [143]).

90

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

Fig. 11. XRD for UVPLD ZnO and PLD ZnO (from [146]).

oxygen atoms which can travel larger distances till they find a low energy lattice position. This can lead to the better crystallinity measured from films deposited at lower temperature under otherwise identical conditions, as has been observed in the case of UVPLD of other oxide compounds (Y2O3, BaxSr1
xTiO3). The defects induced in epitaxial ZnO films which are related to the large lattice mismatch between ZnO and sapphire have been the subjects of numerous works. In addition to the RBS analyses (Fig. 10), transmission electron microscopy (TEM) experiments have been used to study the ZnO/sapphire interface and the nature and concentration of defects present at this interface [148]. This interface was found atomically flat with most of the defects being dislocations and planar defects (stacking faults). Their high density at the interface decreases rapidly with the

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

91

distance from the interface, in agreement with the results reported in Fig. 10. The misfit dislocations resulting from the epitaxial growth were revealed by high resolution TEM, when every 7th ð2
1
10Þ plane of sapphire terminates at this interface [148]. A correlation between these defects localized near the interface and the optical properties of the film was established by the study of the in-depth luminescence in ZnO [149]. The green emission intensity increases towards the filmesubstrate interface, due to the poor interfacial crystal quality. More generally, a lot of work has been carried out on the optical and electrical properties of ZnO and on their dependence on the structural quality of the films [150e159]. The main conclusion drawn from these works was the fact that a strong dependence exists between the crystalline state of the ZnO films and their physical properties, i.e. mosaic spread, structural defects, doping or impurity in the films broaden or shift the exciton resonance and increase the broadband deep level emission which are observed in photoluminescence experiments. In this frame, the problem of strain and stress in ZnO films has been considered. Whatever the deposition method, stresses are induced in thin films deforming the unit cell and introducing unwanted anisotropy. This can have large effects on the physical properties of the material. For example, it has been shown for GaN films epitaxially grown on sapphire [160] that the energy gap, effective masses and exciton resonance energy are strongly dependent upon the strain of wurtzitic GaN films due to the residual stresses developed during the growth process. The residual stresses in ZnO epitaxial films grown on sapphire substrates at 550  C have been measured [13] and compressive stresses equal to
550 MPa were obtained. These residual stresses can be considered as the combination of: (i) stresses accumulated during the growth process due to the lattice mismatch between film and substrate and to specific phenomena (like bombardment of the growing film by energetic species), and (ii) stresses due to the thermal behaviour of the system. This last thermal component is the consequence of the mismatch in thermal expansion between the film and the substrate related to the difference in the thermal expansion coefficient of sapphire (7.5 ! 10
6/K) and ZnO (2.9 ! 10
6/K). According to these values, a compressive thermal stress can be deduced. Moreover, the epitaxial growth of ZnO on Al2O3 with the 18% lattice mismatch leads also to a compressive stress induced in the ZnO layer [13]. Such residual compressive stresses inducing a strain field in the films can modify the band structure and could have thus important effects on the optical properties of the ZnO epitaxial films. The studies on this point gave rather contradictory results since it was first reported that the exciton resonance energy showed a very weak dependence as a function of the strain of the ZnO epitaxial layer [161]. In contrast, further studies [156] indicated that the excitonic transition energies increased with increasing strain and eventually the energies were resolved into two bands at a strain of 1.63%. These results show that the effects of the residual stresses on the physical properties of ZnO epitaxial films are still a matter of discussion which need further work, based first on the precise measurements of stresses in such films and then on their correlation with the electrical and optical properties of the films. Once ZnO epitaxial films have been grown on sapphire substrates, the improvement of crystalline quality and physical properties of these films has given

92

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

rise to numerous works. At first, the effects of high temperature postdeposition annealing under oxygen atmosphere have been studied [162,163]. By using the excitonic structure to investigate and quantify the effects of oxygen annealing at 800  C, it was found that the excitonic features, luminescence properties and resistivity were improved [162]. These improvements were believed to be due to the reduction of excess zinc, the removal of point defects and better crystallinity. However, this conclusion appears to contradict further studies which showed that annealing at 1000  C increased the film mosaı¨ city and FWHM of the reflection peaks (i.e. decrease of the out of plane crystallite size and increase of the strain in crystallites), while the lateral grain size was found to increase [163]. At the same time, the electron mobility increased with the lateral grain size reaching 100 cm2/Vs, certainly via a decrease of charge scattering at the grain boundaries. This value of 100 cm2/Vs is still smaller than that of bulk single crystal (230 cm2/Vs) and indicates that further improvements of grain size, mosaic spread and impurity concentration have to be achieved. Although high temperature postdeposition annealing treatment could lead to partial improvement of the film quality, it does not seem to be the solution to the problems of defects induced in ZnO films during the PLD process. The oxygen pressure during the growth plays an important role on the epitaxy, surface morphology and optoelectronic properties [164]. Schematically, ZnO films grown at low oxygen pressure present a high structural quality with a surface morphology dominated by a typical honeycomb-like structure. In contrast, ZnO films grown at higher oxygen pressure (10
2e10
1 mbar) contain additional structural defects (dislocations, interstitials.), but present a smooth surface resulting from a change in growth mode. These points led to a two step process, with a low oxygen pressure (10
4e10
3 mbar) in the initial stages of growth to control the epitaxy and the quality of the filmesubstrate interface, while a higher oxygen pressure is used in further growth to achieve the formation of smooth films [164]. This approach has been further developed by the introduction of a buffer layer of ZnO, deposited at a relatively low temperature (around 500  C), followed by the growth at the usual substrate temperature (700e750  C) [165]. By optimisation of the thickness and growth temperature of the homo-buffer layer, a decrease in the mosaic spread is obtained, as well as a smoother surface. This can be directly observed in Fig. 12 which shows AFM images of ZnO films grown without and with a ZnO buffer layer. Without the buffer layer, the honeycomb-like structure is observed (Fig. 12a), while increasing the thickness of the buffer layer (Fig. 12b and c) induces an increase of the 2D growth leading to a smooth surface. A further increase in buffer layer thickness leads to a 3D dominant growth with island formation (Fig. 12d). AFM images, X-ray diffraction and RHEED studies led to the optimisation of the growth temperature of the homobuffer layer, in order to improve the structural quality of the ZnO films. The changes associated with this homo-buffer layer were believed to be due to the relaxation of the strain due to the lattice mismatch between ZnO and sapphire. Finally following these works, a multi-step PLD process has been proposed [166]. First a low temperature and low pressure thin (30 nm) ZnO relaxation layer is grown. Then higher temperatures and pressures are used to grow thicker ZnO epitaxial layers.

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

93

Fig. 12. AFM images of ZnO films grown with and without homo-buffer layer (from [165]).

Such an optimised method was found to give ZnO films with narrow X-ray diffraction peaks and FWHM rocking curves, and atomically flat surfaces. Moreover, high electron mobility (up to 155 cm2/Vs) in a narrow carrier concentration range (2e5 ! 1016 cm
3) was observed in these films, as well as narrow photoluminescence line widths [166]. 4.2.2. Other substrates The presence of structural defects due to the large lattice mismatch in ZnO epitaxial films grown on c-cut sapphire substrates has deleterious effects on the optical and electrical properties of the films. To overcome this problem, the use of lattice-matched substrates is necessary and among various possibilities the case of ScMgAlO4 has been considered for PLD of ZnO films [167]. This hexagonal oxide compound can be described as a natural superlattice composed of alternate stacking of wurtzite (Mg,Al)Ox and rocksalt ScOy layers, with an in-plane lattice constant of 0.3246 nm. This value leads to a lattice mismatch with ZnO as low as 0.09%, which has thus led to the study of the epitaxial growth of ZnO on such ScMgAlO4 single crystal substrates. Whatever the growth temperature (in the 350e1000  C range under 10
5 mbar), (000l) oriented ZnO films were grown on (000l) ScMgAlO4 substrates and showed [167] the following in-plane epitaxial relationship: ZnO ½11
20 k ScMgAlO4 ½11
20, that we can call a perfect ‘‘hexagon on hexagon growth’’, by analogy with the well known ‘‘cube-on-cube growth’’ in cubic systems. Moreover, this epitaxial relationship was present without traces of any other in-plane orientation domains similar to

94

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

those observed in ZnO films grown on c-cut sapphire substrates at relatively low temperature. The surface morphology of ZnO films was greatly improved by the use of ScMgAlO4 substrates by comparison with sapphire. For the same growth conditions [168], ZnO films formed on cleaved ScMgAlO4 substrates showed very smooth surfaces consisting of flat terraces with 0.26 nm step heights corresponding to the charge neutral unit of ZnO, while the films grown on sapphire substrates showed rough surface with about a 20 nm roughness [167]. The beneficial effect of the use of such ScMgAlO4 substrates on the crystalline quality can be clearly observed in Fig. 13, which represents the rocking-curve measurements for the (0002) and ð10
11Þ reflection peaks recorded on ZnO films grown on ScMgAlO4 and sapphire substrates. Fig. 13a characterizes the mosaicity of the films (angular distribution of the c-axis), and large differences are observed in the FWHM of the rocking curves, respectively, 39 and 378 arcsec for ZnO films grown on ScMgAlO4 and sapphire. The same behaviour, i.e. broader angular distribution, is observed through the rocking curve of the ð10
11Þ ZnO reflection peak in Fig. 13b for the in-plane distribution of the crystallites. Thus, using ScMgAlO4 substrates greatly improved the quality of ZnO epitaxial films in terms of surface morphology and crystallinity. In addition, the physical properties of the ZnO epitaxial films were also ameliorated. For instance, the electronic properties of such ZnO films showed both high electron mobility (w100 cm2/Vs) and low residual carrier concentration

Fig. 13. Rocking curves for ZnO grown on ScMgAlO4 and sapphire (from [168]).

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

95

(w1015 cm
3) when compared to the films grown on sapphire under similar conditions [167]. Moreover, by Al doping (0.5%) it was possible to obtain high crystalline quality films showing high conductivity (O103 S cm
1), which cannot be grown directly on sapphire substrates. The improvement of ZnO film quality by the use of ScMgAlO4 substrates was also confirmed by optical characterizations [169e 171]. The optical absorption spectra for these films showed a clear split by 7e 8 meV for A and B excitons which has not been observed for films grown on sapphire, and by photoluminescence measurements, the line width of the excitonic emission line was found less than 0.8 meV, which was claimed to be the smallest value recorded [171]. To further improve the crystalline quality of epitaxial ZnO thin films grown on ScMgAlO4 single crystal substrates, an ex-situ high temperature (1000  C) annealed ZnO buffer layer has been checked [172]. The surface of the annealed buffer layer was atomically smooth, enabling persistent layer-by-layer growth of ZnO film as demonstrated by RHEED intensity oscillation. The high crystalline quality (reduction of structural defects) of these ZnO films was deduced from the optical properties (photoluminescence and optical reflection spectroscopies) which were found similar to those observed on the best bulk ZnO single crystals. Hexagonal LiNbO3 (000l) substrates have also been considered for epitaxial ZnO film growth by PLD. The smaller lattice misfit (8.5%) between the ½10
10 ZnO direction and LiNbO3 along the ½11
20 direction, as compared with that in the case of c-cut sapphire substrates favored the epitaxial growth of ZnO films which has been observed on this substrate [173]. Although a complete characterization of the optical properties of the epitaxial ZnO layers has not been reported, the overall quality of these films, which can be deduced from the published results, is inferior to that obtained with the ScMgAlO4 substrates. This is due to the larger lattice mismatch of LiNbO3 substrates with ZnO. In these two preceding cases, the single crystal substrates possess the same hexagonal structure as ZnO. Other substrates having specific properties (conductivity, dielectric.) could be favorable for potential applications of ZnO thin films, and have thus been suggested. In this frame, an interesting approach has been proposed, based on some cubic crystals, according to the fact that the (111) plane of cubic materials presents a hexagonal symmetry identical to that of the basal plane of ZnO. A rather detailed study has been reported in the particular case of (LaAlO3)0.3(Sr0.5Ti0.5O3)0.7 (LSAT) single crystal substrate [174]. High quality (000l) ZnO films were formed by PLD on LSAT (111) substrates at substrate temperature higher than 550  C and the epitaxial growth can be deduced from Fig. 14 which shows the RHEED patterns recorded for such films. In Fig. 14a-1 and a-2, which correspond to film orientations differing from 30  from each other, fine symmetrical streaky line patterns are characteristic of the epitaxial growth of (000l) ZnO films having surface flatness on the nanometer scale. From the lattice spacing, the RHEED patterns were identified as diffraction by the ½10
10 and ½11
20 ZnO planes [174]. The following in-plane epitaxial relationship was deduced: ZnO ½01
10 k LSAT ½110, corresponding to a rotation of the ZnO hexagon relative to the substrate hexagon by 30  . This rotation is due to the large lattice

96

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

Fig. 14. RHEED patterns of ZnO on LSAT substrates (from [174]).

mismatch (more than 40%) between the ZnO (000l) and the LSAT (111) planes in a direct superposition of the respective hexagons, which is reduced to 2.9% with this 30  rotation [174]. For comparison purposes, Fig. 14c-1 and c-2 corresponding to a ZnO film grown under the same experimental conditions on a c-cut sapphire substrate gave streaky line RHEED patterns, indicating the formation of epitaxial films with the classical in-plane orientation. The approach based on (111) oriented single crystal cubic substrate for epitaxial growth of ZnO films has been also used with yttria stabilized zirconia (YSZ) single crystal [175]. (000l) ZnO films grown at a 800  C substrate temperature presented an atomically flat surface with the following in-plane orientation: ZnO ½11
20 k YSZ ½110. The influence of film thickness was demonstrated by a correlated increase of crystalline grain size and Hall mobility in the epitaxial ZnO layers [175]. Let us mention that this (111) substrate has been further used for the growth of various multilayers (ITO, ZnO.) in the field of pen junction applications. In order to be able to combine the advantage of the various functional properties of ZnO with advanced Si electronics on the same substrate, (111) oriented silicon has thus been checked as a substrate for ZnO epitaxial growth. The pulsed-laser deposited ZnO films on bare (111) Si only show a textured (000l) growth without any

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

97

in-plane orientation with the substrate axes. This result has been thought to be due to the formation of an amorphous silica layer on Si during the first stage of the ZnO film deposition process. The use of a buffer layer was needed to grow epitaxial ZnO films [176,177]. AlN with its hexagonal structure was considered as a good candidate, its in-plane parameter (0.311 nm) leading to a 4.2% lattice mismatch with ZnO. Despite this difference, AlN epitaxially grown on (111) Si was used as a buffer layer for ZnO epitaxial growth. (000l) ZnO films were formed on (000l) AlN films on (111) Si [178]. This work was extended to the case of a GaN buffer layer, which has lower lattice mismatch with ZnO (2.2%) and very similar results were obtained. g-Al2O3 has been also used as a buffer layer on Si, according to the fact that this material can be epitaxially grown on (111) Si by MOCVD. Highly oriented (000l) ZnO films were grown by PLD on such buffer layers, with a rather smooth surface, i.e. the rms value being of the order of 2 nm [177]. The in-plane relationships deduced from RHEED analyses were the following: ZnO ½10
10 k g
Al2 O3 ½112 k Si ½110. Promising optical properties were found in these ZnO films grown on g-Al2O3 buffer layers, as indicated by the photoluminescence spectrum which does not show any broad emission tail corresponding to the deep level emission. The film quality was also deduced from photoluminescence studies with the fine structures at 3.22 and 3.29 eV in the emission spectra [177]. It can be noticed that similar films grown by RF sputtering do not show equivalent structural and optical properties, a possible explanation of this fact can be the difference in the kinetic energy of the species reaching the substrate during the growth by PLD and RF sputtering. Despite the difference in symmetry, various works have also been carried out on the growth of ZnO thin films on (100) oriented cubic substrates. For example, (100) oriented (LaAlO3)0.3(Sr0.5Ti0.5O3)0.7 (LSAT) substrates have been used and the Xray diffraction analysis reveals that the films are mainly (110) oriented ZnO films with small proportions of (000l) oriented component crystals [174]. This (110) texture comes from the fact that the ZnO (110) plane presents a rectangular lattice (0.26 ! 0.28 nm) whose dimensions can be reasonably matched with the cubic lattice parameter of LSAT (0.77 nm). The RHEED images of ZnO films grown on LSAT (1000) substrates are shown in Fig. 14b. Two streaky patterns with different spacing and intensities can be observed and are identified as diffraction from (0002) and ð10
10Þ planes. These patterns confirm the formation of films consisting of two types of component crystallites with different orientation in the plane. These different textures are also observed in the case of ZnO growth on other cubic substrates, YSZ (100) or SrTiO3 (100). In this last case, the La doping of the strontium titanate leads to unexpected results. Indeed, very surprisingly, different results were obtained by photoluminescence on SrTiO3 and La-doped (0.5%) SrTiO3. When grown on Ladoped SrTiO3 single crystal substrates, the deep level emission of ZnO disappeared almost completely [180], the origin of this effect being not yet explained. (100) oriented GaAs and InP single crystal substrates have also been considered. Although some interesting results have been reported on the growth modes of ZnO [181], the films grown on these substrates always showed the classical (000l) texture without any in-plane orientations with the substrates axes in the case of GaAs

98

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

[182,183] or InP [184]. The use of an adequate buffer layer should be tried in order to grow epitaxial ZnO thin films on such promising substrates for applications. Finally the growth of ZnO thin films has been studied on NaCl cubic single crystals [185] and interesting results were obtained. Indeed, ZnO thin films with the wurtzitic structure were grown on this (001) oriented cubic substrate with two distinct textures as revealed by electron diffraction in a TEM. Selected area diffraction patterns indicate the presence of ð11
20Þ-oriented grains with their c-axis in the plane of the film. The c-axis spacing (0.52 nm) being close to a spacing of the NaCl structure (0.55 nm), it was suggested that these ð11
20Þ-oriented grains grow with the [0002] direction aligned with the [100] direction of the NaCl single crystal substrate [185]. The selected diffraction patterns also clearly identify the presence of a ½10
11 zone axis, meaning that the c-axis was inclined to the surface normal (½10
11 axis). It was not possible to precisely define the epitaxial relationship with the substrate due to the fact that the ½10
11 zone axis of ZnO does not correspond to a plane of atoms in real space. These results give rise to various questions on the origin of this specific last texture and more generally to the epitaxial growth of films on substrates having structures presenting distinct symmetries like hexagonal films on cubic substrates. 4.3. ZnO related compounds One of the main advantages of the PLD concerns the ability to grow thin films with a wide range of composition. In fact with the use of a multitarget system, it is possible to obtain alloyed films with a complex composition, even if the bulk material cannot be synthesized. A further improvement in PLD is the combinatorial approach [186], which strongly reduces the time spent in exploring new compositions or continuously spread compositions, via the deposition of a large number of thin films on substrates. The PLD combinatorial deposition system is generally based on a conventional multitarget carousel and by the use of masks which can be rapidly switched under computer control. It is thus possible to vary the film composition on each substrate [187,188]. A practical combinatorial library can be obtained in a single deposition run with such an approach. Since ZnO alloyed with various atoms can lead to new or enhanced functional properties with interesting potential applications, all the possibilities of PLD have been largely used to grow epitaxial thin films of ZnO related compounds. Zinc in the wurtzite structure can be easily substituted by divalent cations with similar radius (Mg2C, Co2C or Mn2C for example), which can noticeably modify the physical properties of ZnO. The solubility of these elements in ZnO is generally rather low (only a few percent). However, it is worth noticing that the solubility limits in epitaxial thin films can differ significantly from those of corresponding bulk materials due to the phase stabilizing effect of the substrate [189]. This epitaxial stabilisation of oxide phases has thus been advantageously used to form ZnO alloyed films having interesting properties. As a matter of fact, intensive research on combinatorial thin film preparation of transition metal doped ZnO has been carried out to obtain ferromagnetic thin films

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

99

above room temperature, in order to use such films as spin injectors in the devices envisaged in the new type of electronics, spintronics [190]. Indeed, the prediction that some wide band gap semiconductors (GaN, ZnO) could be ferromagnetic at or above room temperature [191], has led to the PLD formation of Mn-, Ni-, V-, Co-, Cr-doped ZnO films [179,192e194]. The solubility limits of these transition metals in ZnO have been examined via X-ray diffraction and optical transmission analyses, and Fig. 15 summarizes these results [186]. Fig. 15 indicates that the solubility limit varies quite a lot with the species, but Mn and Co are the most soluble dopants. Two factors play a role in the solubility limit, the valence state and the ionic radii of the substituted species. The high Mn or Co concentrations preserving the wurtzite structure of ZnO are mainly due to the fact that the ionic radii of Mn2C (0.066 nm) and Co2C (0.058 nm) in fourfold coordination are the closest values to that of Zn2C (0.06 nm). It was therefore possible to grow single phase epitaxial ZnO doped films on sapphire substrates with a high crystalline quality, but the measurements of the magnetic properties of these films gave somewhat contradictory results. For example, in Mn doped ZnO films, no signature of ferromagnetic ordering could be detected down to liquid He [179]. In the case of epitaxial Co-doped ZnO films, ferromagnetic features with a Curie temperature higher than room temperature were sometimes observed with a poor reproducibility [195], while in another report [186], ferromagnetism was not observed in the films despite their crystalline quality. A possible explanation of the behaviour of Co-doped ZnO films could be the formation of magnetic Co clusters in the films during the growth [196,197]. Recently, Co-doped ZnO thin films grown by PLD on c-cut sapphire substrates were found ferromagnetic at room temperature [198,199]. These films were obtained either by reactive PLD of Zn and Co metallic target under oxygen pressure [198] or by direct

Fig. 15. Solubility limits of 3D metals in ZnO thin films (from [186]).

100

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

PLD of a compound target [199]. In the two cases, epitaxial Co-doped ZnO films were grown with the classical 30  rotation of the in-plane cell axes with respect to those of the substrate and these films exhibited ferromagnetism with a Curie temperature close to room temperature. The ferromagnetism in these films was due to the intrinsic properties of the ZnxCo1exO material, and did not result from segregation effects or Co clusters formation in the deposited films. Indeed, the linear dependence of the out of plane parameter as a function of Co content [198], or the precise imaging of the films by Auger electron spectroscopy [199], can exclude the presence of Co clusters in the Co-doped ZnO films. Another interesting result of the alloying of ZnO with various species is the ability to control its energy band gap, for instance with Mg or Cd. In the case of Mg, this can be clearly seen in Fig. 16, which represents the variation of the band gap as a function of the Mg concentration in MgxZn1
xO films [200], the nature of the crystalline phases observed in the films also being indicated. Fig. 16 shows that Mg concentrations up to 34% preserve the hexagonal structure, while the band gap varies linearly with Mg content between 3.3 and 4.2 eV. Beyond 34% a phase separation into MgO and ZnO takes place, while for Mg concentrations higher than 62%, the MgxZn1
xO films crystallize in the cubic structure with a parameter close to that of MgO. At the same time, the band gap varies between 5.6 and 6.8 eV. This behaviour is due to the fact that the ionic radii of Mg2C (0.057 nm) and Zn2C (0.06 nm) are sufficiently close to avoiding large deformation of the wurtzite structure in the low Mg concentration domain, or of the MgO cubic structure for low Zn concentrations. Regarding the low Mg concentration domain, the theoretical solubility limit is 0.04, while epitaxial MgxZn1
xO films are obtained with x values up to 34% [201e 203]. The MgxZn1
xO should thus be considered as a metastable phase and accordingly its stability has been studied under various temperatures up to 1000  C

Fig. 16. Band gap as a function of Mg concentration in MgxZn1
xO films (from [200]).

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

101

[204]. MgxZn1
xO (with x Z 0.23) films did not show any segregation effects at temperatures up to 850  C and the evolution of the energy band gap at higher temperatures led to the conclusion that an apparent solubility was x Z 0.15. MgO segregation and interdiffusion takes place at temperatures higher than 850  C, hence high quality epitaxial MgxZn1
xO could easily be grown around 700  C on sapphire substrates [201,202]. As a result, the out of plane and in-plane mosaic spread were found similar to those observed for epitaxial ZnO films on sapphire and the in-plane crystal orientation of MgxZn1
xO on c-cut sapphire substrates corresponded to the classical 30  rotation of the film axes with respect to those of the substrate. MgxZn1
xO films with a cubic structure having high Mg concentrations were epitaxially grown by PLD on various substrates: c-cut sapphire, MgO (100) and Si (100) substrates [205e207]. Due to the in-plane symmetry, (111) oriented MgxZn1
xO films were grown on sapphire with a low mosaic spread indicated by a 0.38  of the FWHM of rocking curves. The high crystalline quality of the films was further confirmed by RBS in channelling geometry [205,206]. (00l) Oriented films were formed on MgO substrates with a lattice constant close to that of MgO but increasing slightly with the fraction of MgO. Moreover, due to the nearly perfect lattice match, defects at the film interface were not revealed by cross-section TEM experiments or RBS in channelling geometry analysis [205]. In the case of Si (100) substrates, a buffer layer was used to overcome the large lattice mismatch between film and substrate. Since the lattice constant of cubic MgxZn1
xO films is very close to that of TiN it is possible to grow epitaxial TiN thin films on Si (100), epitaxial cubic MgxZn1
xO films were grown on TiN with a cube-on-cube alignment [205]. A SrTiO3 buffer layer has also been considered as a buffer layer on Si substrates and cubic MgxZn1
xO films were grown with a moderate mosaic spread (0.23  ) and a high crystalline quality with a very low cmin value (4%) in channelling experiments. The MgxZn1
xO and SrTiO3 films follow a cube-on-cube alignment with a 45  rotation of the unit cell with respect to the in-plane c-axes [200]. UV photodetectors based on metalesemiconductoremetal structure have been demonstrated using MgxZn1
xO films [200,207,208]. The photodetectors based on hexagonal low Mg concentration epitaxial MgxZn1
xO films gave a peak response centered at 308 nm, in good agreement with the energy band gap of the films [207]. Photodetectors with a different wavelength domain have been obtained with thin MgxZn1
xO films in the cubic structure, i.e. the high Mg concentration phase. By varying the Mg concentration, it could be possible to detect in a wide range of UV spectrum, from 157 to 230 nm [200]. C-axis oriented alloyed CdxZn1
xO films were grown by PLD on (000l) sapphire [209] or (000l) ScMgAlO4 [210] substrates. Although the normal thermodynamic solubility limit of CdO in ZnO is around 2%, CdxZn1
xO films with a wurtzite phase were grown by PLD with Cd concentrations up to about 7e8%. At higher Cd concentrations, segregation in the films seems to occur. This can be deduced from Fig. 17 which shows the optical transmission spectra of CdxZn1
xO films at various x values [209]. The absorption edge of these films is shifted to the red with increasing Cd concentrations up to x around 8%. Above this value, the transmission spectra

102

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

Fig. 17. Transmission spectra of CdxZn1
xO films as a function of Cd concentration (from [209]).

showed an unusual behaviour with a large broadening of the transition which was believed to be due to the segregation of CdO phases in the films. The inset in Fig. 17 shows the variation of the band gap as a function of Cd content, with a monotonic decrease from 3.3 to 2.9 eV with Cd concentration increasing up to 8%. The crystalline quality of the CdxZn1
xO films was found to degrade with Cd concentration, the FWHM of the diffraction peaks and rocking curves being broadened with an increase in Cd content. This has been attributed to the difference in ionic radii of Cd2C (0.074 nm) and Zn2C (0.06 nm) which could lead to an increase of the cell volume when Cd is substituted for Zn in the wurtzite structure [210]. Doping ZnO with group III elements (Al, Ga, and In) or Si and H decreases its resistivity to low values (10
4 Ucm), while keeping its high optical transparency in the visible wavelength domain. However, the main potential applications of ZnO films as transparent conducting electrodes (flat panel displays, solar cells) do not need the use of epitaxial films. As a result, a few papers only report on the epitaxial growth of doped ZnO films. The correlation between structural and physical properties of Al-doped zinc oxide films (AZO) epitaxially grown on (000l) sapphire substrate has been studied [211]. Epitaxial AZO films were grown in the 400e700  C temperature range, with the classical epitaxial relationship on c-cut sapphire substrate: AZO ½10
10 k Al2 O3 ½11
20. However, the crystalline quality of the films appears lower than that of undoped material. Indeed, FWHM of the rocking curves lies between 3.4 and 0.3  in the 250e680  C temperature domain, while the c-axis lattice parameter was found slightly larger than that of undoped ZnO. These observations can be explained by the substitutional incorporation of Al3C ions into Zn2C sites which can induce defects creation and lattice deformations. The lowest resistivity in these films (2.2 ! 10
4 U cm) was obtained for a substrate temperature (250  C), far from the optimised conditions for high crystalline quality AZO film growth.

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

103

Another example of epitaxial thin films of ZnO related compounds is furnished by the transparent conductive oxides exhibiting the empirical formula In2O3(ZnO)m (m being an integer). This compound can be considered as a superlattice structure of wurtzite ZnO with the alternate stacking of InO2 and InO(ZnO)5 layers along the c-axis direction, in which a high electron mobility is expected. The growth of single crystalline thin films of this material using a vapor phase method is a complex task because of the difficulty of composition control due to the large vapor pressure of Zn and In at high growth temperatures. In order to avoid composition deviations in the films which can lead to spurious phase formation, epitaxial thin films of this compound have been grown by a method derived from PLD [212]. First, a ZnO epitaxial layer was formed by PLD on a (111) YSZ single crystal substrate [175]. Then films of In2O3(ZnO)4 were deposited by PLD followed by annealing the films at 1450  C in air. By solid state diffusion InO(ZnO)5 epitaxial films were formed, preserving the initial epitaxial relationship of the ZnO layer on the YSZ substrate, i.e. out of plane and in-plane orientations of the film with the substrate were InO(ZnO)5 (000l)kYSZ (111) and InOðZnOÞ5 ½11
20 k YSZ ½110. A particular aspect of this growth method is the fact that by varying the film thickness ratio of the initial epitaxial ZnO to the In2O3(ZnO)4 layers, it was possible to grow a compound with a specific m value having a general formula In2O3(ZnO)m [212]. In this solid-phase epitaxial method of growth, the epitaxial ZnO thin layer plays an essential role in determining the crystallographic orientation and allowing the growth of single crystalline films with controlled chemical compositions. This solid-phase epitaxy technique has also been used to form epitaxial ZnRh2O4 thin films [213]. This compound has the normal spinel-type structure and exhibits p-type conduction. In that case, the ZnO epitaxial layer grown on YSZ (or on epitaxial ITO on YSZ), was covered by a ZnRh2O4 film deposited at room temperature by PLD. Then, the solid-phase epitaxy led to the following epitaxial relationships: ZnRh2 O4 ½110 k ZnO ½11
20 k YSZ ½110 (or ZnRh2 O4 ½110 k ZnO ½11
20 k ITO ½110 k YSZ ½110). With this growth method, pen heterojunction diodes were obtained with an abrupt interface exhibiting the expected rectifying IeV characteristics [213]. 4.4. Superlattices and heterostructures Such systems are thin film structures presenting a periodicity in one dimension in composition, structure or in both composition and structure, and this leads to enhanced or new physical properties. Superlattices cannot be found in the nature, except some very specific materials [212], and the formation of these layered artificial materials is mainly the result of the development of thin film growth methods. In particular the advantages of PLD (multitargets system and wide range of oxygen pressure) have been used to the formation of superlattices and heterostructures. Moreover, the concepts and methods of combinatorial chemistry have been extended to the case of PLD [188]. Owing to their interest in the field of optoelectronics, ZnO/MgxZn1
xO superlattices have been grown on sapphire substrates by PLD [214,215], the value of x (0.2) leading to a 0.13% in-plane lattice mismatch between ZnO and

104

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

Fig. 18. Schema of the ZnO/ZnxMg1
xO superlattices, with RHEED and AFM (from [214]).

Zn0.2Mg0.8O corresponding to a 0.5 eV difference in the energy band gap of the layers. As shown in Fig. 18 a thick (100 nm) ZnO buffer layer grown at low temperature (550  C) was used as a buffer layer, to obtain a high quality surface morphology for the superlattices. Indeed, Fig. 18 shows fine streaks in the RHEED patterns which are indicative of an atomically flat surface, while the AFM images show hexagonally shaped grains for both the ZnO buffer layer and the superlattice. Optical studies of such superlattices indicated a clear shift of the photoluminescence peak to higher energies as the thickness of the ZnO layers decreases [214]. This quantum confinement effect was further checked on superlattices with ZnO layer thicknesses less than 5 nm, with a blue shift of the absorption edge and photoluminescence peak [215]. To improve the crystalline quality of the superlattices, single crystalline ScMgAlO4 substrates have been used, leading to a negligibly small lattice mismatch (0.08%) at the superlatticeesubstrate interface. As a result, superlattices with high crystalline quality and thickness homogeneity, atomically flat surfaces and interfaces were grown [216]. These morphological and structural improvements were accompanied by improved optical properties [216e218], as demonstrated by an efficient stimulated emission up to 377 K, related to the combined effects of excitons and quantum confinment. The threshold of the stimulated emission (11 kW/cm2) was far better than that observed in ZnO epitaxial films and by changing the thickness of the barrier layer, it was possible to tune the emission energy between 3.2 and 3.4 eV. This demonstration of stimulated emission of excitonic origin at room temperature opens the way to the achievement of a low threshold violet diode laser composed of ZnO-based multiquantum wells structures [217].

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

105

A first example of a heterostructure based on a ZnO epitaxial film has been demonstrated by its use as a buffer layer on a sapphire substrate used to grow high quality GaN thin films. Indeed, ZnO and GaN have the same wurtzite structure with a low lattice mismatch (2%) and thus epitaxial ZnO films grown on (000l) Al2O3 substrates have been used as buffer layer for the growth of GaN layers [219e221]. The positive effect of the ZnO buffer layer on the growth of GaN thin films can be observed in Fig. 19 through the comparison of the X-ray diffraction diagrams of GaN films grown with and without the ZnO layer [219]. In Fig. 19, the GaN film grown directly on the basal plane of sapphire has a mixed crystal orientation (polycrystalline structure). When the growth occurs with a thin epitaxial ZnO layer, the GaN film grows with a single c-axis orientation and low mosaic spread. Under optimised growth conditions, the following epitaxial relationship is observed:GaN ½10
10 k ZnO ½10
10 k Al2 O3 ½11
20. The surface morphology of the GaN films was also greatly improved by comparison with the growth on bare sapphire substrates, whilst optical transmission and photoluminescence experiments confirm the high quality of the GaN on the ZnO heterostructure. It is worth noting that ZnO on GaN heterostructures showing good luminescence properties have been grown on sapphire substrate [222]. This use of GaN as a material for a buffer layer on sapphire is rather surprising, in view of the large lattice mismatch between these compounds, but it was found that 0.1e1.5 mm thick

Fig. 19. X-ray diagram for GaN with and without ZnO buffer layer (from [219]).

106

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

GaN film grown on (000l) Al2O3 substrate led to the growth of high quality epitaxial ZnO films [222] with a perfect in-plane alignment of the ZnO and GaN axes. Moreover, the low lattice mismatch avoided the presence of a high density of defects at the ZnO/GaN interface in contrast to the case of ZnO/sapphire interface. These points lead to the conclusion that ZnO on GaN and GaN on ZnO heterostructures with sharp interfaces provide new opportunities for the fabrication of hybrid ZnO/ GaN optoelectronic devices on sapphire [222]. Other ZnO-based heterostructures have been studied in view of the successful development of pen heterojunctions [223e225]. Such devices are formed on an YSZ single crystal oriented on the (111) thus allowing the epitaxial growth of an ITO layer acting as electrode. Then n-type ZnO and p-type SrCu2O2 films are successively grown on the ITO template layer [223,224]. The optimised PLD growth of these multilayered films leads to an atomically flat surface and interface with in-plane epitaxial relationships which are summarized in Fig. 20. This figure shows that (000l) ZnO films are grown with an in-plane alignment characterized by ZnO ½11
20 k ITO ½110 k YSZ ½110, whatever the substrate temperature. Then, the growth of the SrCu2O2 layer depends upon temperature, since at 350  C (112) textured SrCu2O2 films are formed with the following in-plane orientation: SrCu2 O2 ½110 k ZnO ½11
20, but at elevated temperatures (600  C)

Fig. 20. Cross-sectional HR TEM ZnO/SrCu2O2 (from [224]).

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

107

a different behaviour is observed. In fact, (100) textured SrCu2O2 films are formed with the SrCu2 O2 ½011 k ZnO ½11
20 epitaxial relationship [224]. An interpretation of these particular epitaxial relationships was based on the large lattice mismatch in this system and on the ionicity of these oxides. Another approach for the formation of pen heterojunctions was the use of p-type NiO films grown on an ITO layer, an n-type ZnO film being further epitaxially grown on top of the NiO film [225]. Finally, pen junctions based on p-type manganite (La0.7Sr0.3MnO3) and n-type ZnO layers were also grown by PLD on c-cut sapphire substrates [226]. ZnO thin films were first epitaxially grown on the sapphire substrate. ZnO was chosen as the counter electrode, because it exhibits n-type behaviour, and by changing the oxygen stoichiometry (through the oxygen pressure during the growth), it is possible to control the carrier concentration. The La0.7Sr0.3MnO3 film further grown on ZnO showed two distinct textures (100) and (110), with random in-plane orientations, certainly due to the difference in symmetry in the growth plane between La0.7Sr0.3MnO3 and ZnO. Despite these structural imperfections, these junctions were claimed [226] to provide an opportunity to integrate various magnetic and magnetoresistive properties of manganites with the nonlinear and optoelectronic applications of ZnO. 4.5. PLD using ultrashort laser pulses The most important recent development in the field of PLD is certainly the use of ultrashort (in the femtosecond range) laser pulses to grow thin films. Indeed, the availability of reliable and powerful femtosecond (fs) lasers has led to the development of a wide variety of works in the field of laserematter interaction and in particular on laser ablation using ultrashort laser pulses. The interest of such fs lasers for PLD was related to the fact that there were thought to be a solution to the problem of droplets. Very schematically, it was assumed that when the laser pulse temporal length becomes shorter than the time needed to couple the electronic energy to the lattice (a few picoseconds), thermal effects could not play a significant role in the target during laser irradiation [227] and thus the ejection of liquid droplets from the molten target should be precluded. Moreover, the multiphoton absorption, which is very likely to occur at high laser intensities (O1013 W/cm2) currently reached with fs lasers, should also overcome the problem of materials having too low an optical absorption coefficient, i.e. laser ablation of highly insulating material was thus thought to be possible with the use of ultrashort laser pulses. Based on these assumptions, fs PLD has been used to grow thin films of various materials [228e230]. More specifically, oxide thin films were grown in this way, SnO2 [231], BaTiO3 [232] and ZnO [233,234]. The surface morphology of such films grown by fs PLD was found highly dependent upon the material and experimental conditions (laser fluence and focalisation). Schematically two extreme cases can be encountered, with (i) films appearing as formed by the random stacking of clusters or aggregates (in the 10e100 nm range) leading to highly perturbed surface morphology and (ii) smooth and dense films free from any particulates or clusters. The first case is well represented by the fs PLD of Ti [235] or BaTiO3 [232], while

108

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

SnO2 or ZnO corresponds to the second case under well defined experimental growth conditions. This significant difference is not yet clearly explained, but is partly related to the effects of the laser power density, since it is possible to change for a given material, the surface morphology of the deposited film by changing the fluence or focalisation of the laser. Femtosecond PLD of ZnO in the 500e700  C range under an oxygen pressure in the 10
6e10
3 mbar domain leads to the growth of smooth, dense and stoichiometric films with the hexagonal structure [233,234]. It is worth noting that the growth of epitaxial ZnO thin films by nanosecond PLD occurs at substrate temperatures around 700  C, under a wide range of oxygen pressure (10
6e10
1 mbar). With fs PLD, it is not possible to grow epitaxial ZnO films at a pressure higher than 10
3 mbar. In fact at increasing pressures, a phenomenon of gas phase condensation occurs, leading to the formation of nano (and micro) particulates which condense on the surface of the substrate. As a result, ZnO films grown under these conditions, present a high surface roughness due to the random stacking of the nano and micro particulates. (000l) ZnO films have been grown by fs PLD on various substrates (sapphire, (100) Si or amorphous SiO2), and an initial difference between femtosecond and nanosecond PLD has been found in the rocking-curve measurements which indicate a larger FWHM (around 1.5  ), for films grown by the former method. This broader mosaic spread was observed independently of the nature of the substrate (amorphous or crystalline), while it is known that the use of single crystalline substrates decreases the mosaic spread of the films. The broad angular distribution of the ZnO c-axis is therefore an intrinsic characteristic of the growth by PLD using ultrashort laser pulses. Despite this large out of plane angular distribution, epitaxial relationships were observed for ZnO films grown on c-cut sapphire. As a matter of fact, Fig. 21a which represents the f scan of the ð10
11Þ ZnO reflection peak shows the presence of peaks separated by 60  and demonstrates the sixfold symmetry with a rotation of 30  of the ZnO axis with respect to the sapphire substrate, i.e. the ½10
10 and ½11
20 directions being aligned for the ZnO film and the Al2O3 substrate. For comparison purposes, Fig. 21b represents the same f scan recorded on a ZnO film grown on a sapphire substrate by PLD using nanosecond pulses. The in-plane mosaic spread of the epitaxial orientations can be deduced from the FWHM of the peaks in the f scans (the inserts in the Fig. 21 show the details of one of the peaks in each case). A larger FWHM value is observed for the ZnO film grown by fs PLD, indicating a worse crystalline quality for these films in comparison with ZnO grown by nanosecond PLD. Another difference between fs and ns PLD ZnO films was deduced from the width of the reflection peaks in the qe2q Xeray diagrams. By using the Scherrer formula, the size of the diffracting crystallites can be estimated and values in the 10e15 nm range were typically deduced for ZnO epitaxial films grown by fs PLD, while values around 50 nm (equivalent to film thickness) were measured for ZnO films grown by ns PLD. This means that the ZnO fs PLD films can be described as formed by the stacking of ZnO nanocrystallites, while the ns PLD leads to classical columnar growth. Thus the structural quality of the ZnO films grown using ultrashort laser pulses is very different from that observed in the films grown using nanosecond lasers.

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

109

Fig. 21. f Scans of the ð10
11Þ ZnO reflection peak for fs (a) and ns (b) PLD (from [13]).

Additional information was obtained by RBS analyses in the channelling geometry which revealed the presence of defects in ZnO nanocrystallites. As shown in Fig. 22, the comparison of the RBS spectra in random and channelling geometry indicates a rather poor crystalline quality with cmin values around 80% [234]. This cmin value demonstrates that the defect density in such ZnO films is much higher compared to that obtained in ZnO films grown by ns PLD which have a low cmin value of 2% (see Fig. 10). cmin values in the 60e80% range were obtained whatever the thickness of the ZnO films. This indicates that the presence of a high density of defects at the ZnOesapphire interface (due to the large lattice mismatch) cannot be accounted for by such high cmin values. Specific phenomena occurring during the

110

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

Fig. 22. RBS in random and channelling geometry for ZnO fs PLD (from [234]).

growth have thus to be considered in order to explain the presence of defects throughout the whole thickness of these epitaxial ZnO films grown by fs PLD. Finally, the residual stresses were measured in ZnO films grown using these fs PLD lasers; smaller value (
330 MPa) compressive stresses were obtained compared to those using nanosecond (
550 MPa) lasers [13]. The origin of this difference in crystalline quality of ZnO films grown by fs and ns PLD needs to be researched as one of the most important differences between the fs and ns PLD processes, i.e. the large difference in the kinetic energy of the species reaching the surface of the growing ZnO films. Indeed, it has been shown that very energetic species (O1 KeV) are emitted by the target during femtosecond laser irradiations [13,236], energy at least 10 times higher than that of the species emitted during nanosecond laser ablation. This high energy bombardment during film growth could lead to microscopic effects such as defect creation and sputtering as well as macroscopic effects like densification of the films and the preferred orientation of the crystallites [237]. Moreover such an ion assisted growth can modify the microstructure of films (size and orientation of the crystallites) often resulting in smaller film stresses [238]. Thus the existence of these high kinetic energy species during fs PLD can explain the particular growth of ZnO nanocrystalline films [13]. The formation of ZnO films on cubic MgO (100) single crystal substrates by fs PLD has been also studied, and more unexpected results were obtained [13]. As shown by Fig. 23a, the pole figure recorded for the (0002) reflection of the ZnO wurtzite structure shows the presence of four poles which are located at a declination angle j equal to about 4.2  and with an azimuthal rotation f of 45  with respect to the [100] direction of the MgO substrate. The classical (000l) texture, which is generally obtained for the growth of ZnO on most of the substrates (amorphous or single crystalline) used, is therefore not observed in the case of (100) MgO substrates,

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

111

Fig. 23. (0002) and ð10
11Þ pole figure for ZnO grown on MgO (from [235]).

the c-axis being inclined by 4.2  with respect to the normal to the substrate. Moreover, epitaxial relationships exist with the MgO substrate due to the fourfold symmetry of the (0002) poles [13]. The existence of preferred in-plane orientations for the ZnO films was confirmed by the pole figure of the ð10
11Þ reflection of the ZnO, which is shown in Fig. 23b. Twelve poles can be observed with a rather large extension in the (f,j) plane, and each of them corresponds in fact to a double pole for the ð10
11Þ family, as it was revealed by a careful analysis of the pole figure. However, such analyses did not permit precise identification of the texture of the ZnO film on MgO. Indeed, in order to explain the 4.2  inclination of the c-axis with respect to the substrate normal,

112

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

a (h,k,j,l ) texture has to be assumed with low values for h and k and a rather large one for l. It was possible to show, through simulations of the stereographic projection of the (0002) and ð10
11Þ families of the ZnO, that a texture ð1; 0;
1; 25Þ with an in-plane alignment with [110] MgO could agree with the results of Fig. 23 [235]. However, other (h,k,j,l ) orientations can give a similar agreement, and it follows that more precise works are needed to precisely define the texture of fs PLD grown ZnO films. These results, and those reported for the growth of ZnO on NaCl [185] or AlN on MgO [239], give rise to some questions on the nature of the very first plane which is formed on the substrate, when hexagonal films are deposited on cubic substrates which leads to very specific and unexpected texture growth. Whatever the substrates used (c-cut sapphire, cubic MgO or even amorphous SiO2) nanocrystalline ZnO thin films are thus formed by fs PLD. Significantly more research will be required to understand the mechanisms of nanocrystalline epitaxial film formation by fs PLD in order to control the size of the nanocrystallites and the density of structural defects and the mosaicity of the films. However, it is worth noting that various potential applications exist for such ZnO nanocrystalline thin films. For example, ultraviolet laser emission has been observed from nanocrystalline thin films at room temperature [240]. These films grown by PLD which comprise closely packed and hexagonally shaped nanocrystals were found to give rise to a sharp free exciton emission at room temperature. When the nanocrystal size was adjusted to be optimum (50 nm), stimulated emission was observed at very low pumping intensity, i.e. above a threshold intensity as small as 24 kW/cm2 using 355 nm, 15 ps pumping laser pulses [240]. In the same way, it was demonstrated that ultraviolet laser emission could occur in a disordered and random medium like polycrystalline ZnO thin films [241,242]. Such effects are due the strong optical scattering on ZnO crystallites leading to the formation of a closed-loop path. These loops serving as ring cavities for light [242]. Moreover, the nonlinear optical properties of nanocrystalline ZnO films could be promising according to the measurements reported on the third-order optical nonlinearity (c(3)) in ZnO microcrystallite thin films [243] and the enhancement of the c(3) in semiconducting nanocrystallites [244] via quantum confinement effects. Recently second harmonic generation in nanocrystalline oxide thin films has been reported [245] whilst the second order susceptibilities observed in very thin (50 nm) epitaxial ZnO films are enhanced and can be larger than that of single crystal ZnO [246]. It would therefore be interesting to research the second harmonic generation in nanocrystalline ZnO films. Finally, ZnO nanocrystalline films formed by fs PLD could also be checked for sensor applications based on surface property measurements, as it has been studied in the case of SnO2 nanocrystalline thin films [247]. 4.6. p-Type doping The PLD synthesis of p-type ZnO films has been considered, and various ways have been followed to reach this aim. First, nitrogen has been tried as a dopant, but N incorporation in ZnO, even helped using nitrogen atomic species resulting from a plasma, did not lead to the formation of p-type ZnO [190]. Also, co-doping has been used to promote the formation of p-type films with the simultaneous incorporation of

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

113

donor (Ga) and acceptor (N). As a result the formation of p-type ZnO films by PLD has been reported [248], with a low resistivity (2 U cm and a carrier density around 4 ! 1019 cm
3). However, such results were found very difficult to reproduce [190]. As a matter of fact, further studies [249,250], systematically exploring the effects of N and Ga co-doping in a wide range of concentrations did not show any sign of p-type conductivity in such Co-doped ZnO films [250]. To illustrate the complexity of the situation, it has been reported [251] that p-type conductivity in ZnO thin films has been obtained by the co-doping (Ga and N) method, but the N doping was effective only when N2O gas was passed through a plasma source and not with the use of N2 gas. Finally it can be concluded that the growth by PLD of p-type ZnO films via Ga and N co-doping is far from being established. Phosphorus has been envisaged as a potential acceptor dopant, and its behaviour in PLD grown ZnO epitaxial films has been studied [252]. The effects of growth conditions and postdeposition annealing treatments were correlated using Hall voltage measurements. The conclusion drawn from this work was that only n-type conductivity was unambiguously determined in the phosphorus-doped ZnO films. Another approach for obtaining p-type ZnO films has been the use of arsenic dopants. A first report [253] mentioned the formation of p-type ZnO films grown by PLD on (001) GaAs substrates. The explanation for this effect was related to the fact that for substrate temperatures higher than 450  C, arsenic coming from the substrate diffuses easily into the growing ZnO films. Thus the As atomic concentration can reach the 1017e1021 atom/cm2 range [253], leading to a Hall mobility in the 0.1e50 cm2/Vs range. Further works were carried out in order to check these results and to understand precisely the effects of As doping in ZnO using a new growth process [254]. Arsenic doped ZnO films were grown on sapphire, ZnO or SiC substrates by combining ZnO film growth by pulsed-laser deposition of a pure ZnO target and a molecular beam of As for doping with an effusion cell. Using this hybrid beam deposition process, arsenic atoms are directly incorporated into the ZnO lattice during growth so that the substrate should not have any influence on the As doping phenomenon, in contrast to previous experiments. The effects of As doping on the electrical and optical properties of the films were determined and indicated that the As doped films show good p-type conductivity with hole carrier concentrations up to the mid-1017 cm
3 range at room temperature with a hole mobility around 35 cm2/Vs [254]. However, these results were obtained only on the O-face ZnO and Si-face SiC substrates. Therefore, the effect of the polarity of the substrate on the electrical behaviour of the As doped ZnO films has to be further studied to fully understand the origin of the p-type conductivity in these films. Despite this point, under certain conditions, this work shows that As doped ZnO films with good electrical and optical properties make them excellent candidates for p-type layers in ZnO-based devices [254].

5. MBE growth of ZnO Molecular beam epitaxial growth of ZnO has been developed more recently than MOCVD and PLD. The first MBE epitaxial growth of ZnO was reported in 1996

114

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

[255]. From this date, there has been an increasing interest in the MBE growth of ZnO using different techniques such as mainly plasma-assisted or radical sourceassisted MBE, electron cyclotron resonance-assisted MBE (ECR-assisted MBE), metalorganic MBE and laser MBE. This last technique, also called pulsed-laser deposition, is treated in part D. As in the case of MOCVD growth of ZnO, the interest for the MBE epitaxial growth of ZnO has been boosted from 1999 by the prospects of ZnO short wavelength photonic devices, as illustrated in Fig. 24. Amongst the main issues for the growth of ZnO thin films by MBE is the choice of suitable substrates, which as usual, is crucial for the successful growth of high quality layers, as is the necessity of generating active oxygen atoms. 5.1. Substrates

number of papers

In spite of a very large mismatch, sapphire, according to its (0001) c- [255e272], ð11
20Þ a- [274e286] and ð01
12Þ r-planes [258], has been so far the substrate preferred for the growth of ZnO films by MBE. The epitaxial relationship between ZnO films and c-plane sapphire has been found to be (0001) ZnOk(0001) Al2O3 with in-plane orientation relationships of ½
1100 ZnO k ½
2110 Al2 O3 [250], indicating a 30  rotation of ZnO relative to sapphire in the c-plane, and ð11
20Þ ZnO k ð11
20Þ Al2 O3 [274]. This 30  rotation results in a reduction in in-plane lattice mismatch (da/a) from 0.32 for the case where the a-axes are coincident to w0.19 for the case where they are offset by 30  [259], but the two types of in-plane rotation give rise to the presence of domains. In addition to the 30  -rotated domain, two other kinds of rotation domains have been observed by Wang et al. [287]. The XRD peaks of a dominant domain are observed at the same f position as those of the f scan for the (113) plane of Al2O3 with a standard epitaxial relationship with sapphire as ½10
10 ZnO k ½11
20 Al2 O3 , which is the same as that of GaN on sapphire. The third one is the 21.8  -rotated domain with the relationship ½11
20 ZnO k ½53
80 Al2 O3 . In order to surmount the very large lattice mismatch of about 18% and crystallographic difference between c-sapphire and ZnO and to eliminate the rotation domains, different buffer structures have been proposed. 20 18 16 14 12 10 8 6 4 2 0 1995

1996

1997

1998

1999

years

2000

2001

2002

2003

Fig. 24. Number of papers related to the MBE growth of ZnO versus time.

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

115

A several monolayer thick MgO layer has been developed [266]. The thin MgO buffer layer has been shown to facilitate the initial nucleation and to promote the lateral growth of ZnO leading to a great improvement of the ZnO film. As a result (3 ! 3) surface reconstruction of ZnO is observed and reflection high energy electron diffraction intensity oscillations have been recorded. FWHMs of 13 and 108 arcsec of the (0002) and ð10
15Þ U-rocking-curves, respectively, have been measured and are to be compared to 774 and 1640 arcsec without a buffer. Nitridation of the c-plane sapphire surface was used by Wang et al. [287] to eliminate the rotation domains and improve the quality of the ZnO films grown by RF-plasma-assisted MBE. It was found that a very thin hexagonal nitrogen polar AlN layer was formed by nitridation and this effectively served as a template for the following ZnO film growth, resulting in the elimination of the rotation domains. As a result of this nitridation, the quality of the films was markedly improved, with the FWHMs of (002) and (102) U-scans decreasing from 912 to 95 and 2870 to 445 arcsec, respectively. The same group [288] proposed further Ga decoration by using a few monolayers gallium in addition to sapphire nitridation to completely suppress the sub-domains. In a different approach, the same group [289] proposed the use of a low temperature (LT) GaN thin layer and a LT-ZnO layer as double buffer layers to improve the quality of ZnO films deposited on c-sapphire by RF-assisted plasma MBE. The FWHM values of (002) symmetric and (102) asymmetric U-scans were 90 and 430 arcsec, respectively. Following another approach, Sakurai et al. [263] have shown that twin crystal patterns and surface faceting observed with exactly c-plane oriented sapphire substrates were suppressed if the offset angles were enlarged from near-zero to 2.87  . In the growth on a-plane sapphire, high-sensitivity pole figures have indicated that the ZnO films were uniquely (0001) oriented with no trace of secondary orientation; it was also effective in the elimination of 30  rotation domains which usually appear in the case of growth on c-sapphire [275,276,279]. The orientation relationship between ZnO films and a-sapphire has been found to be ð0001Þ ZnO k ð11
20Þ sapphire and ð2
1
10Þ ZnO k ð0001Þ sapphire [277]. Other substrates have been chosen depending on their physical properties and availability using lattice accommodation and electrical conductivity criteria for vertical device structures such as laser diodes. 6H-SiC [255] has only a small lattice mismatch with ZnO (less than 5%) and can be highly conducting. Nakamura et al. [290] achieved the growth of ZnO films on 36  -rotated Y-cut ð01
12Þ LiTaO3, widely used for SAW devices, to obtain a higher electromechanical coupling factor of the shear horizontal-type SAW due to the piezoelectric contribution of ZnO. The epitaxial relationship in the film plane has been proved to be ð11
20Þ ZnO k ð01
12Þ LiTaO3 and ½0001 ZnO k ½0
111 LiTaO3 , which reveals that the c-axis of ZnO lies in the plane of the film and perpendicular to the X-axis of LiTaO3. Considering the future integration of monolithic devices composed of ZnO related light-emitting devices, Kumano et al. [291] demonstrated the growth of ZnO films on (001) GaAs substrates using ZnS buffer layers by metalorganic molecular beam

116

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

epitaxy using DEZn as a Zn precursor with an electron-resonance oxygen plasma. In addition, the authors stress the fact that the hypothetical cubic ZnO on (001) GaAs will be attractive because the [001] direction is free from a piezoelectric field, making it possible to explore the strong excitonic contributions to the optical features expected in heterostructures by reducing the strain-induced electric fields and the related Stark effects on excitons. Hexagonal ZnO layers have been grown on GaAs (111) substrates without a buffer layer [292]. (111) Si substrates have been used to integrate ZnO SAW devices into a Si integrated circuit [293] and to benefit from the high Si electrical conductivity in view of injection light-emitting devices [294]. Epitaxial GaN grown on sapphire, with a lattice mismatch to ZnO of 1.8%, 10 times lower than the one of ZnO on sapphire, and a smaller thermal mismatch than that of ZnO/Al2O3, has attracted great attention [295e301]. In addition to these advantages, {0001} surfaces of ZnO and GaN are polar surfaces, and polarity has strong effects on structural, optical and electrical properties. The ZnO/(111) CaF2 heterostructure has a misfit of w15.8%, smaller than the ZnO/Al2O3 one [302]. The tensile strain in the ZnO layer coming from this misfit is expected to be compensated by the compressive strain coming from the difference in thermal expansion coefficients. This is confirmed by the free exciton energy position which is in good agreement with the reported values in bulk ZnO, indicating the growth of strain-free films. Finally, Li-diffused bulk semi-insulating (0001) ZnO substrates have been used in the framework of a study of a p-type doping of layers using nitrogen atoms produced from a flux of N2 gas by an RF plasma source [303]. The effect of the O/Zn ratio on the homoepitaxial Znepolar ZnO films grown by plasma-assisted MBE has been investigated by Kato et al. [305]. The minimum line widths of (0002) and ð10
10Þ Urocking-curve were 42 and 46 arcsec, respectively. Furthermore, the FEA and FEB emissions and the n Z 2 state of FEA were clearly observed in PL spectra at 4.2 K, indicating a quality as high as that of bulk ZnO. 5.2. Growth modes In order to produce high quality ZnO layers by MBE, a source of oxygen atoms must be generated. The high molecular bond strength of diatomic oxygen limits the thermal dissociation of molecules into atoms and prevents the use of simple thermal gas cracking. Among the alternative means of breaking the molecules, radio frequency plasma or radical sources has been widely preferred [17,255,256,258e267, 269e271,274e276,278e280,282,285e288,291e299], but electron cyclotron resonance sources [257,268,273,277,288] and a simple ozone source [290] have been used as well. Fig. 25a and b shows typical plasma-assisted and ECR-assisted MBE systems. Typical experimental conditions are displayed in Table 2 for plasma-assisted MBE and Table 3 for ECR-assisted MBE. Some attempts to grow ZnO layers by metalorganic MBE (MOMBE) have been reported as well:

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

117

- Kumano et al. [291] have grown 0.9 mm thick ZnO layers at 550  C with an oxygen ECR plasma used as an oxygen source (flow rate 110 mmol/min) and DEZn as a Zn precursor. Prior to the growth of the ZnO films, approximately 0.3 mm-thick ZnS thin layers were grown at 400  C on the GaAs substrates as a buffer layer. The precursors used for the growth of ZnS were DEZn and DtBS. - Ashrafi et al. [281,283,284] have grown ZnO layers on a-face sapphire substrates at 400  C using either deionized H2O or an O2 ECR-excited plasma as an oxygen source and DEZn as a Zn precursor. Their layers were nitrogen-doped using monomethyl hydrazine as an N precursor [281]. - Shirakata et al. [272] carried out the growth of ZnO layers by MOMBE on the cplane of sapphire substrates at substrate temperatures ranging from 375 to 400  C using DEZn and H2O as source precursors. After chemical plus thermal cleaning at high temperature of the sapphire substrates prior to growth, their surfaces were usually exposed to a gentle oxygen [256] or Ar [258] plasma leading to a smooth and clean surface suitable for epitaxial growth. In order to improve surface morphology and crystal quality of ZnO epilayers, a buffer layer has been reported to be grown initially on the sapphire, Si, GaN or CaF2 surface by many authors, either MgO [261,266] or a thin ZnO layer grown at low temperature (w300e350  C) [267,276,279,286,289,291,295,297,298]. In some cases, either the initial buffer layers [297] or the final layers were postgrowth annealed at higher temperature under oxygen or nitrogen ambient to improve their surface smoothness and their structural, optical and electrical properties [262,292]. 5.3. Characteristics of the layers The structural, optical and electrical characterization of the MBE grown layers has been achieved more extensively and thoroughly than in the case of MOCVD.

Fig. 25. Plasma-assisted (a) and ECR-assisted (b) MBE systems (from [258] and [257], respectively).

118

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

Table 2 Experimental conditions of the growth of ZnO layers by plasma-assisted MBE Tsubstrate (  C)

400e550 500e550 300e400 375e500 350 350e650 350 (buffer), 550 (layer) 780 600 350 (buffer), 400 (layer) 300 (buffer), 700 (layer) 375e500 350 300 400 700 750 350 (buffer), 600 (layer) 500e550 350 (buffer), 650 (layer)

Zn flux (Torr or nm/s)

Oxygen plasma Pressure (Torr) or flow rate

Power (W)

5 ! 10
7e1 ! 10
5

5 ! 10
7e1 ! 10
4 1 ! 10
4 10
2

150e400 120 0e110

6 ! 10
7e1 ! 10
6 7 ! 10
7e4 ! 10
6

0.2e1 sccm 0.3 sccm

400 90e450 300

2 ! 10
7 8 ! 10
7 3 ! 10
7

0.3 sccm 0.3 sccm 0.3 sccm

300 450

5.5 ! 10
5
1 ! 10
4, 3.5 sccm 3 ! 10
7

0.3 sccm 3.5 sccm

0.15e0.2 nm/s 0.15 nm/s 5 ! 10
7e3 ! 10
6

3.5 sccm 1.5 sccm 0.3 sccm

0.15 nm/s

3.0 sccm

400 20 / 50 400

Growth rate (mm/h)

Refs.

1.6 0.1 1.7 0.3e1 0.6 0.2 0.25e0.50

[255] [256] [258] [259] [274] [260] [286]

0.15 0.3

[262] [263] [289]

0.4e0.75

[291]

0.4 0.3 0.1e1.5

[275] [264] [292] [265] [293] [294] [276]

0.5 0.4e0.6

[269] [297]

300 350 300

Nevertheless, despite the very large lattice mismatch between ZnO and sapphire, surprisingly narrow rocking-curve widths have been reported in the first studies related to the MBE growth of ZnO layers on sapphire [259]. In these studies, only the symmetric X-ray diffraction was taken into account. It has been clearly demonstrated since then that such symmetric X-ray diffraction results were not reliable and that asymmetric reflections such as the ð
1015Þ and ð
1012Þ ones were Table 3 Experimental conditions of the growth of ZnO layers by ECR-assisted MBE Tsubstrate (  C)

Zn flux

350 250e550 275 200e425

0.7e2 ! 1014 at/cm2 s

Oxygen ECR

Growth rate (mm/h)

Refs.

[287] [277] [268] [273]

Pressure (Torr) or flow rate

Power (W)

4.5 ! 10
5or 1.2 ! 10
4 10
3e10
4 2 ! 10
4 10
3e10
4

100

0.04e0.13

100

0.02e0.03

119

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

necessary to estimate the mosaicity around the [0001] axis [269,301]. In particular, it has been shown that although the two-dimensional layers exhibit a lower roughness, the three-dimensional ones present a better structural quality with a larger lateral coherence and a smaller in-plane mosaic spread [269]. Some structural and electrical properties of ZnO layers reported so far are displayed in Table 4. 5.4. p-Type doping As in the case of MOCVD, attempts to achieve p-type doping in ZnO MBE layers have not been so far quite successful. Ogata et al. [262] have found that thermal annealing in an O2 atmosphere at a low temperature seemed to be a suitable choice from the standpoint of decreasing the concentration of residual donors. Iwata et al. [264] have grown nitrogen-doped ZnO layers on sapphire substrates. An N-doped ZnO layer fabricated using an N2/O2 flow ratio of 10% was found to have a chemical nitrogen concentration of 1 ! 1019 cm
3. However, conversion from n- to p-type did not occur whilst large nitrogen incorporations were observed to induce extended defects. Table 4 Some structural and electrical properties of ZnO layers grown by MBE XRD-FWHM (arcsec)

XRD plane

216 2088 576 12

(0002) (0002) (110) (0002)

c-Al2O3 c-Al2O3 r-Al2O3 c-Al2O3

2772 13, 84

ð11
20Þ (0002), ð10
15Þ

ð01
12Þ LiTaO3 MgO/c-Al2O3

13, 108 2D layers 30 120 2900 3D layers 250 230 500 277

(0002), ð10
15Þ

Substrate

Si subst. epi. GaN MgO/c-Al2O3 a-sapphire

(0002) ð
1015Þ ð
1012Þ

c-Al2O3

(0002) (
1015) (
1012) (0002)

c-Al2O3

Electrical properties n (cm
3)

m (cm2/Vs)

9 ! 1018

260

!6 ! 1016

120

2.1 ! 1016 7 ! 1016

98 120

1.87 ! 1018 1 ! 1017 7.6 ! 1016

100 120

Refs.

[255] [256] [257] [258] [259] [274] [287] [261] [286] [288] [289] [291] [266] [276]

[269]

GaN a-sapphire

1017

90

[297] [283]

120

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

Nakahara et al. [280] found from Hall measurements an n-type conductivity in Ga and N co-doped ZnO layers grown by radical source MBE. They showed [17] that Zn-rich conditions were indispensable for nitrogen doping and that a high Ga concentration, necessary to enhance nitrogen incorporation, led to formation of the additional phase ZnGa2O4 in the films. Ashrafi et al. [281] were successful in producing reproducible p-type conductivity from nitrogen doping using H2O vapor assisted metalorganic MBE. As-grown p-type ZnO:N layers showed low net acceptor concentrations (NAeND) of w1014 cm
3, but thermal annealing of the N-doped samples as well as the optimisation of growth parameters increased the NAeND up to w5 ! 1016 cm
3. In N-doped ZnO layers grown by MBE on a Li-diffused bulk semi-insulating ZnO substrate, Look et al. [303] measured a hole concentration of 9 ! 1016 cm
3 with a hole mobility of 2 cm2/Vs. But such results were not reproducible. 5.5. ZnO related compounds Zn1
xMgxO layers have been grown by molecular beam epitaxy at 500  C on a-sapphire using Knudsen cells for Zn and Mg and RF oxygen plasma for a biosensing application [307]. Phase separation has been observed for x O 0.22. ZnO and Zn0.94Mg0.06O layers showed electron mobilities as high as 150 cm2/Vs and 100 cm2/Vs, respectively, indicating their high electrical quality. As an effect of a ZnO buffer layer, the complete series of Zn1
xMgxO (0 % x % 1) layers has been successfully grown by MBE on a-plane sapphire without severe phase mixing [308]. The structure of layers with x ! 0.5 were found to be wurtzite (0001) with the band gap continously changing from 3.3 to 4.5 eV, while those with x O 0.5 were rocksalt (001) rather than (111). Very thin Zn1
xMgxO layers (!25 nm) with x O 0.5, even MgO, were found to have the wurtzite (0001) structure on wurtzite ZnO buffer layers, opening the way to the possible formation of wurtzite ZnO/MgO superlattices which can be recognized as quasi-alloys. A blue luminescence of strong intensity with a large Stokes shift, whose mechanism remains unclear, was observed from ZnCdO films grown by RF-assisted MBE [309]. ZnO and CdO were found by the same group to be segregated from each other [310]. The hexagonal grains in the films consisted of pure ZnO, while Cd was concentrated at the grain boundaries. The current estimation of the authors is that the incorporation of CdO induces the formation of unknown deep levels in neighboring ZnO crystals, which emit bright luminescence despite their low density states.

6. RF magnetron sputtering In pulsed-laser deposition (PLD) of thin films, pulsed laser radiation is used to vaporize by photon absorption the surface of the material (target) to be deposited as a thin film on a surface. As a result of the sputtering process gas ions from the plasma are accelerated towards a target material. Material is sputtered from the

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

121

target and afterwards deposited on a substrate in the vicinity. The process is carried out in a closed chamber which is pumped down to a vacuum base pressure before deposition starts (Fig. 26). In RF-sputtering (radio frequency) an ac-voltage is applied to the target. In one phase, ions are accelerated towards the target surface and sputter material. In the other phase, charge neutrality is achieved. As a result sputtering of non-conducting material is possible. As in Fig. 26, a permanent magnet structure is located behind a target serving as a deposition source. Plasma confinement on the target surface is achieved by locating this permanent magnet structure behind the target surface. The resulting magnetic field forms a closed-loop annular path acting as an electron trap that reshapes the trajectories of the secondary electrons ejected from target into a cycloidal path, greatly increasing the probability of ionization of the sputtering gas within the confinement zone. Inert gases, specifically argon, are usually employed as the sputtering gas because they tend not to react with the target material or combine with any process gases and because they produce higher sputtering and deposition rates due to their high molecular weight. Positively charged argon ions from the plasma are accelerated toward the negatively biased target (cathode), resulting in material being sputtered from the target surface. In D.C. magnetron sputtering, a negative potential up to some hundred Volts is applied to the target. As a result, the Ar-ions are accelerated towards the target and set material free; on the other hand, they produce secondary electrons which cause a further ionization of the gas. The sputtering method has the advantages of high deposition rate (w1 mm) and low substrate temperature (200e300  C). The properties of the deposited thin films differ according to sputtering conditions, pressure, temperature, components of the target and atmosphere, distance of substrates and films growth ratios. Thin ZnO films grown by sputtering were initially developed for using either as a transparent and conductive coating (TCO) material, because of its low cost and high chemical stability, or as acoustoelectric and acoustooptic devices such as surface acoustic wave (SAW) filters and AO Bragg deflectors because ZnO thin films with c-axis orientation exhibit strong piezoelectric and piezooptic effects. As an example, TCO ZnO films, doped with group III impurities, reached resistivities as low as 10
4 Ucm [311e315]. ZnO pressed powder targets were used and the films were

Fig. 26. Typical RF sputtering system.

122

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

deposited on Corning 7059 glass substrates. Borosilicate substrates were used as well to get c-axis oriented films which were applied in acousto devices [316e319]. As a result amongst the numerous papers on ZnO thin film growth by sputtering, mainly during the early stages of the sputtering growth only a few were dedicated to the epitaxial growth of ZnO films. 6.1. Unconventional sputtering methods From the early stage of ZnO sputtering deposition, more sophisticated techniques than the classical DC or RF-magnetron sputtering have been developed. In conventional sputtering methods, it is difficult to optimise the deposition conditions, such as input power, substrate temperature, etc., to increase the deposition rate because the discharge of the sputtering system causes instability [319]. This difficulty may relate to a generation of both a plasma and a dependent sputtering phenomena occurring in the same region. In order to overcome this difficulty, the electron cyclotron resonance (ECR) sputtering method has been proposed. The ECR plasma is generated by microwaves with a magnetic field set-up by the ECR condition. In such a way, the phenomena can be easily controlled independently of the generation of plasma. A schematic diagram of an ECR sputtering apparatus is shown in Fig. 27. It is considered that a ZnO film deposited by the ECR sputtering process is of higher quality than the one deposited by conventional sputtering [320]. ð11
20Þ-plane epitaxial ZnO films have been deposited on the r-plane sapphire substrates at 250e 550  C by ECR sputtering [320]. Such films showed excellent piezoelectric properties and also showed room temperature free exciton emission on r-sapphire at 2 K and on c-sapphire.

Fig. 27. Schematic diagram of an ECR sputtering apparatus (from [319]).

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

123

Ion-beam sputtering is becoming as well an alternative to the conventional plasma configuration because it permits independent control of beam energy, direction and current density. Because of a very low deposition pressure (w10
2 Pa), thermalization of sputtered materials can be avoided and energy deposition of depositing particles is higher than in plasma sputtering. As a result, an enhancement of the material properties is observed together with an increase in film adhesion and surface coverage. According to this approach dual-ion-beam sputtering has been used for the deposition of ZnO films with the substrate at room temperature [321]. Polycrystalline ZnO films were grown on Corning 7059 glass plates with resistivities ranging from 10
3 to 105 U cm. A schematic diagram of the system configuration is shown in Fig. 28. A helicon-wave excited plasma (HWP) has been proposed to overcome the essential problem of surface damage caused by high energy sputtering particles, which are accelerated by a field between the target and the substrate [322]. HWP has higher plasma density, which is uniform in a large volume, and has lower ion energy than the usual capacitively coupled plasmas. Predominantly [0001]-oriented, transparent, low resistivity ZnO films with a FWHM of the (0002) X-ray diffraction peak reaching about 0.32  were grown by HWP on soda-lime glass substrates at temperatures ranging from 100 to 400  C [322]. The same group reported later the successful growth of completely a-axis-locked c(0001)-oriented ZnO epilayers on sapphire (0001) substrates having ultrasmooth surfaces with atomic steps [323]. A schematic diagram of the HWP apparatus is shown in Fig. 29. In the case of the on-axis sputtering configuration, in which the target and substrate surfaces face each other, films can present morphological differences, under some experimental conditions, as result of a competition between the surface diffusion/redistribution and the arriving rate of adatoms [324]. ZnO films grown by

Fig. 28. Schematic diagram of a dual-beam sputtering apparatus: (1) rotatable target holder, (2) target, (3) shutter, (4) temperature-controlled substrate holder, (5) deposition rate monitor, (6) sputtering ionbeam source, (7) assistance ion-beam source, (8) plasma bridge neutralizer, (9) flow control unit, and (11) turbomolecular pump (from [321]).

124

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

Fig. 29. Schematic diagram of a HWP sputtering apparatus (from [322]).

the off-axis deposition, in which the target surface and substrate surface are orthogonal to each other, at 660  C and 120 mTorr, are rather flat and show a very narrow X-ray rocking curve [325]. According to this off-axis sputtering deposition technique, schematized in Fig. 30, epitaxial and textured ZnO films have been deposited, respectively, on (0001) sapphire and glass substrates, with a FWHM of q rocking curves for epitaxial films less than 0.5  [326]. In order to control the ionic flux and the electron bombardment on the substrate, and hence in order to modify the microstructure of the films and to vary their defect content, ZnO thin films were prepared using an unbalanced magnetron puttering technique [327,328]. Unbalanced conditions, resulting in intense bombardment

Fig. 30. Schematic graph of an off-axis sputtering system (from [326]).

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

125

effects on the substrate, were obtained by progressively lowering the strength of the central magnet in the magnetron electrode to control the ionic flux and the electron bombardment on the substrate, as illustrated in Fig. 31. A faster photoresponse speed was observed in ZnO films prepared under unbalanced conditions. 6.2. Substrates In addition to the Corning glass substrates generally used for the growth of TCO ZnO films, various substrates have been used in a second stage mainly for the growth of epitaxial ZnO films for various specific applications. Utilizing diamond, which has the highest Young modulus of all materials, the highest SAW phase velocity can be achieved. The epitaxial growth of high quality (0001) ZnO films on a diamond (111) substrate has been successfully achieved by Hachigo et al. [329] at a temperature of 260  C and a deposition rate of 0.5e2.1 mm/h by the RF magnetron sputtering method. Epitaxial relations were found by RHEED to be (0001) ZnOk(111) diamond with a direction of ½11
20 ZnO k ½
101 diamond. The same group reported more recently a X-ray rocking-curve FWHM of 972 arcsec measured on the ZnO (0002) peak for epitaxial ZnO films deposited on diamond (111) planes at 260  C [330]. LiNbO3 has large Pockels coefficients as well as a large nonlinear susceptibility making it suitable for many applications in optical waveguide devices such as electro-optical switches and modulators, mode converters, nonlinear wavelength converters and waveguide lasers [331]. With the aim of optoelectronics integrated circuits (OEIC) for purposes such as light sources in conjunction with high speed

Fig. 31. Plasma confinement and target erosion patterns with (a) balanced magnetron, (b) slightly unbalanced magnetron, and (c) highly unbalanced magnetron electrode configurations (from [328]).

126

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

optical modulators, ZnO films were epitaxially grown on LiNbO3 (0001) substrates by RF magnetron sputtering [331]. A rocking-curve FWHM of the ZnO (0002) reflection of 0.46  has been measured for films grown at 550  C. The epitaxial relationship between the ZnO film and the LiNbO3 substrate was determined to be ½11
20 ZnO k ½10
10 LiNbO3 . The lattice mismatch for this orientation is about 9% compared to 18% for ZnO on sapphire (0001). The temperature necessary for epitaxy was found to be lower than that required for films grown on sapphire. Due mainly to the fact that silicon substrates can be obtained as large area wafers at low cost and that Si is not only of interest for the integration of optoelectronic devices but also easier to cleave in comparison to sapphire, several studies devoted to the growth of epitaxial ZnO thin films by sputtering using Si substrates. Si (001) substrates were used not only for structural characterization of ZnO films deposited specifically for piezoelectric applications [332] but also in order to study the main defects responsible for the visible emission of the films [333,334]. Highly c-axis oriented ZnO films were deposited by RF magnetron sputtering [335]. The films were found to be of better crystal quality when the RF power was increased and when the gas flux ratio of O2 to Ar was appropriately adjusted. Epitaxial ZnO films have been grown as well on Si (111) substrates by RF magnetron sputtering using a GaN buffer layer deposited by the same technique [336]. The films show crack-free morphology. To further improve the crystalline quality of epitaxial ZnO thin films grown on Si substrates, postdeposition annealing at 400  C under vacuum of ZnO films deposited by RF magnetron sputtering on unheated Si (100) substrates has been studied [337]. Such treatment has been shown to make the films with a higher resistivity, stronger c-axis (002) orientation, denser structure and a smoother surface with relieved stress. A well-oriented columnar structure and strong exciton emission have been found in ZnO films deposited by RF magnetron sputtering on GaAs (001) substrates and subsequently annealed at 550  C [338]. To apply SAW devices to optical fiber communications, optical waves with a wavelength 1.3e1.6 mm are needed. Preparing ZnO thin films on InGaAs/InP substrates becomes then an important issue. With this aim, ZnO thin films with a sharp X-ray diffraction peak with a 0.183  FWHM have been deposited at room temperature with 200 W RF power on (100)-oriented InP substrates by RF-magnetron sputtering by Chang et al. [339]. The resistivity of the deposited films reached 1010 Ucm and their refractive index was close to that of bulk ZnO. ZnO/InP based SAW devices were then successfully fabricated [340]. As in other techniques of thin film growth, sapphire substrates have been used in spite of their large lattice mismatch with ZnO. ZnO thin films epitaxially grown on sapphire (0001) substrates by RF magnetron sputtering at 600  C showed a FWHM of X-ray diffraction q-rocking-curve of 0.13  , indicating a highly c-axis oriented columnar structure [341,342]. Classical ZnO sputtering has been achieved on (0001) sapphire substrates [343]. A (0002) rocking-curve FWHM as low as 250 arcsec has been measured, but typical values lie more generally around 400 arcsec. The roughness is smaller than 1 nm (the best is 0.1 nm). The resistivity ranges from 3 to 105 U cm depending on the growth and coo1ing conditions. Under low pressure (3e 150 mTorr) sputtering, a rocking-curve FWHM w1080 arcsec [344] has been measured on sapphire substrates.

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

127

6.3. p-Type doping Several attempts of p-type doping of ZnO films grown by RF sputtering have been reported. As-grown n-type ZnO films deposited by RF-sputtering on sapphire substrates using P2O5 as a phosphorus dopant were converted to p-type by a thermal annealing process at a temperature above 800  C under an N2 ambient [345]. The layers showed a hole concentration of 1.0 ! 1017e1.7 ! 1019 cm
3, a mobility of 0.53e3.51 cm2/Vs and a low resistivity of 0.59e4.4 Ucm. Furthermore, the phosphorus-doped ZnO thin films showed a strong photoluminescence peak at 3.35 eV at 10 K, which is closely related to neutral acceptor bound excitons of the p-type ZnO. The thermal activation process was said to be very reproducible and effective in producing phosphorus-doped p-type ZnO thin films, and the produced p-type ZnO was very stable. A hole density of 9 ! 1016 cm
3 with a resistivity of 11.77 U cm have been measured in Ga C N co-doped ZnO films deposited at 250  C on glass substrates by conventional RF sputtering [346]. The type of conduction of the co-doped c-oriented films was said to be controllable by suppressing the oxygen vacancies by adjusting the oxygen partial pressure ratio into the sputtering chamber. p-Type N-doped ZnO films with highly c-axis orientation were grown by magnetron sputtering on silicon and sapphire substrates with NH3 as a nitrogen dopant source [347,348]. The hole carrier concentration of the p-type films grown, respectively, at 500  C and 450  C reached 3.2 ! 1017 cm
3 with a resistivity of 35 Ucm [347] and 8.02 ! 1018 cm
3 with a Hall mobility of 0.802 cm2/Vs [348]. ZnObased pen junctions were reported to be fabricated successfully on p-Si (100) substrates. The dependence of the film properties as a function of the ammonia concentration, for films prepared on sapphire (0001) substrates, showed that N-doped p-type ZnO films with c-axis orientation were achieved at ammonia concentrations of 25%, 50% and 75% [349]. At 0% ammonia concentration, intrinsic ZnO films with c-axis orientation were obtained, while at 100% ammonia concentration, the layers were zinc polycrystalline films. The same group reported on the growth of p-type ZnO thin films prepared by oxidation of Zn3N2 thin films deposited by DC magnetron sputtering [350]. Using an oxidation temperature between 350  C and 500  C, p-type ZnO films were obtained with a hole concentration as high as 5.78 ! 1017 cm
3 at 500  C, but with an oxidation temperature at 550  C, an n-type film was obtained. Note that the doped layers deposited by sputtering were polycrystalline and that parasitic electrical effects coming from the grain boundaries can be suspected.

7. Conclusions In this review the epitaxial growth of ZnO films have been reviewed by the most popular techniques; they include MOCVD, PLD, MBE and sputtering which have received spectacular regeneration following the first p-type conductivity announcements and the subsequent vision of photonic devices. The key issues for device

128

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

Table 5 Survey of the electrical characteristics of p-type doped ZnO thin films grown by MOCVD, PLD, MBE and sputtering Growth mode

p (cm
3)

Dopant

MOCVD MOCVD MOCVD C annealing PLD PLD MOMBE C annealing MBE sputtering C annealing Sputtering Sputtering sputtering

N from NH3 N from NO N from NH4NO3 Ga C N co-doping As N N P from P2O5 Ga C N co-doping N from NH3 Oxidation of Zn3N2

16

10 1.0 ! 1015e1.0 ! 1018 6.5 ! 1017 4 ! 1019 mid-1017 5 ! 1016 9 ! 1016 1.0 ! 1017e1.7 ! 1019 9 ! 1016 8.02 ! 1018 5.78 ! 1017

m (cm2/Vs)

Refs.

3.6 0.1 16 w0.08 35

[66] [61] [103] [190] [254] [281] [303] [345] [346] [348] [350]

2 0.53e3.51 5.9 0.802

application of ZnO in optoelectronics, surface acoustic filters and spintronics, include the growth of high quality epitaxial layers of both n- or p-type conductivity, the possibility of dissolving transition elements in the layers, the growth of ZnO related alloys and heterostructures all of which have been featured in this review. The significant aspects of each technique, such as precursors and decomposition mechanisms in MOCVD, various growth modes, substrates, p-type doping, ZnO related compounds and heterostructures by any technique have been specifically addressed. The results of p-type doping, whatever the epitaxial technique used, suffer still from some lack of reproducibility, instability and very low Hall mobilities of the holes. Parasitic electrical effects at the grain boundaries could furthermore be suspected in the case of the polycrystalline sputtered layers. As in the case of GaN, post-growth annealing could be very promising in respect of doping, activation of acceptors, restoration of the stoichiometry and passivation of the layer. The next table (i.e. Table 5) summarizes the results of p-type doping reported sofar. As for the structural characteristics of the films, it turns out that they seem to have reached some optimum values whatever the technique used, as illustrated in the following Table 6 displaying the rocking-curve FWHM in arcsec of ZnO layers grown by MOCVD, MBE, PLD and sputtering on various substrates. Table 6 Optimum rocking-curve FWHM measured on ZnO layers grown on various substrates by MOCVD, MBE, PLD and sputtering. Substrates c-Al2O3

MOCVD

MBE

PLD

Sputtering

255 [85]

30 [0002], 120 ½
1015, 2890 ½
1012 [269] 10e20 [0002], 80e90 ½1
105 [270]

240 [351]

250 [343]

MgO/c-Al2O3 GaN ZnO

230 [306] 40 [100]

42 [0002], 46 ½10
10 [305]

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

129

Rocking-curve FWHM of about 250 arcsec has been reported for the epitaxial growth on c-sapphire substrates whatever the technique used. The structural properties of homoepitaxial ZnO films have been shown to reproduce those of the ZnO substrates used. There is little doubt that homoepitaxial layers will soon benefit from the quality of the ZnO substrates produced, such as those hydrothermally grown in the Fukuda’s group [352] which have been shown to have an X-ray rocking-curve FWHM for the (002) reflection of 8 arcsec compared to about 40 arcsec for the CVT-grown substrates.

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28]

R. Triboulet, Proc. SPIE 4412 (2000) 1. R. Ayouchi, F. Martin, D. Leinen, J.R. Ramos-Barrado, J. Cryst. Growth 247 (2003) 497. B. Ismail, M. Abaab, B. Rezig, Thin Solid Films 383 (2003) 92. W. Tang, D.C. Cameron, Thin Solid Films 238 (1994) 83. Y.-Z. Yoo, Y. Osaka, T. Fukumura, Zhengwu Jin, M. Kawasaki, H. Koinuma, T. Chikyow, P. Ahmet, A. Setoguchi, S.F. Chichibu, Appl. Phys. Lett. 78 (2001) 616. S.J. Chen, Y.C. Liu, J.G. Ma, D.X. Zhao, Z.Z. Zhi, Y.M. Lu, J.Y. Zhang, D.Z. Shen, X.W. Fan, J. Cryst. Growth 240 (2002) 467. Y.X. Liu, Y.C. Liu, D.Z. Shen, G.Z. Zhong, X.W. Fan, X.G. Kong, R. Mu, D.O. Henderson, J. Cryst. Growth 240 (2002) 152. H. Yamamoto, N. Saiga, K. Nishimori, Appl. Surf. Sci. 169e70 (2001) 517. N. Takahashi, K. Kaiya, K. Omichi, T. Nakamura, S. Okamoto, H. Yamamoto, J. Cryst. Growth 209 (2000) 822. R. Triboulet, J.M. N’Tep, M. Barbe´, P. Lemasson, I. Mora-Sero, V. Munoz, J. Cryst. Growth 198/ 199 (1999) 968. T.V. Butkhuzi, A.V. Bureyev, A.N. Georbobiani, N.P. Kekelidze, T.G. Khulordava, J. Cryst. Growth 117 (1992) 366. X.L. Guo, H. Tabata, T. Kawai, J. Cryst. Growth 223 (2001) 135. J. Perrie`re, E. Millon, W. Seiler, C. Boulmer-Leborgne, V. Craciun, O. Albert, J.C. Loulergue, J. Etchepare, J. Appl. Phys. 91 (2002) 690. V. Craciun, R.K. Singh, J. Perrie`re, J. Spear, D. Craciun, J. Electrochem. Soc. 147 (2000) 1077. K. Matsubara, P. Fons, K. Iwata, A. Yamada, S. Niki, Thin Solid Films 422 (2002) 176. I. Ohkubo, Y. Matsumoto, A. Ohtomo, T. Ohnishi, A. Tsukazaki, M. Lippmaa, H. Koinuma, M. Kawasaki, Appl. Surf. Sci. 159e160 (2000) 514. K. Nakahara, H. Takasu, P. Fons, A. Yamada, K. Iwata, K. Matsubara, R. Hunger, S. Niki, J. Cryst. Growth 237e239 (2002) 503. K. Nakamura, T. Shoji, H.-B. Kang, Jpn. J. Appl. Phys. 39 (2000). K. Iwata, P. Fons, A. Yamada, K. Matsubara, S. Niki, J. Cryst. Growth 209 (2000) 526. A.B.M. Ashrafi, I. Suemune, H. Kumano, Jpn. J. Appl. Phys. 41 (2002) 2851. C.R. Gorla, N.W. Emanetoglu, S. Liang, W.E. Mayo, Y. Lu, J. Appl. Phys. 85 (1999) 2595. Jiangdong Ye, Shulin Gu, Shunmin Zhu, Tong Chen, Liqun Hu, Feng Qin, Rong Zhang, Yi Shi, Youdou Zheng, J. Cryst. Growth 243 (2002) 151. W. Kern, R.C. Heim, J. Electrochem. Soc. 117 (1970) 562. S.Z. Ghandi, R.J. Field, Appl. Phys. Lett. 37 (1980) 449. C.K. Lau, S.K. Tiku, K.M. Lakin, J. Electrochem. Soc. 127 (1980) 1843. T. Shiozaki, T. Yamamoto, M. Yagi, A. Kawabata, Appl. Phys. Lett. 39 (1981) 399. A.P. Roth, D.F. Williams, J. Appl. Phys. 52 (1981) 6685. S. Oda, H. Tokunaga, N. Kitajima, J.-I. Hanna, I. Shimuzu, H. Kokado, Jpn. J. Appl. Phys. 24 (1985) 1607.

130 [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65]

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138 M. Shimuzu, H. Kamei, M. Tanizawa, T. Shiosaki, A. Kawabata, J. Cryst. Growth 89 (1988) 365. M. Shimuzu, T. Takayama, Y. Tanaka, T. Shiosaki, A. Kawabata, J. Cryst. Growth 101 (1990) 171. W.W. Wenas, A. Yamada, M. Konagai, K. Takahashi, Jpn. J. Appl. Phys. 30 (1991) L 441. J. Hu, R.G. Gordon, J. Appl. Phys. 71 (1992) 880. M. Yoshino, W.W. Wenas, A. Yamada, M. Konagai, K. Takahashi, Jpn. J. Appl. Phys. 32 (1993) 726. Y.J. Kim, H.J. Kim, Mater. Lett. 21 (1994) 351. Y. Li, G.S. Tompa, S. Liang, C. Gorla, Y. Lu, J. Doyle, J. Vac. Sci. Technol. A 15 (1997) 1063. B. Sang, A. Yamada, M. Konagai, Jpn. J. Appl. Phys. 37 (1998) L1125. B. Sang, A. Yamada, M. Konagai, Jpn. J. Appl. Phys. 37 (1998) L206. S. Liang, C.R. Gorla, N. Emanetoglu, Y. Liu, W.E. Mayo, Y. Lu, J. Electron Mater. 27 (1998) L72. B. Hahn, G. Heindel, E. Pschorr-Schoberer, W. Gebhardt, Semicond. Sci. Technol. 13 (1998) 788. M. Wraback, H. Shen, S. Liang, C.R. Gorla, Y. Lu, Appl. Phys. Lett. 74 (1999) 507. A.B.M.A. Ashrafi, A. Ueta, A. Avramescu, H. Kumano, Y.-W. Ok, T.-Y. Seong, Appl. Phys. Lett. 76 (2000) 550. Y. Liu, C.R. Gorla, S. Liang, N. Emanetoglu, Y. Lu, S. Shen, M. Wraback, J. Electron. Mater. 29 (2000) 69. W.I. Park, G.-C. Yi, J. Electron. Mater. 30 (2001) L32. W.I. Park, G.-C. Yi, H.M. Jang, Appl. Phys. Lett. 79 (2001) 2022. K. Ogata, K. Maejima, Sz. Fujita, Sg. Fujita, J. Electron. Mater. 30 (2001) 659. S. Liang, H. Sheng, Y. Liu, Z. Huo, Y. Lu, H. Shen, J. Cryst. Growth 225 (2001) 110. Xinqiang Wang, Shuren Yang, Jinzhong Wang, Mingtao Li, Xiuying Jiang, Guotong Du, Xiang Liu, R.P.H. Chang, J. Cryst. Growth 226 (2001) 123. K.-I. Ogata, T. Kawanishi, K. Maejima, K. Sakurai, Sz. Fujita, Sg. Fujita, Jpn. J. Appl. Phys. 40 (2001) L657. Th. Gruber, C. Kirchner, K. Thonke, R. Sauer, A. Waag, Phys. Stat. Sol. (A) 192 (2002) 166. V. Sallet, C. Thiandoume, J.F. Rommeluere, A. Lusson, A. Rivie`re, J.P. Rivie`re, O. Gorochov, R. Triboulet, Mat. Lett. 53 (2002) 126. Sho Shirakata, Keisuke Saeki, Tomoaki Terasako, J. Cryst. Growth 237e239 (2002) 528. K.-I. Ogata, T. Kawanishi, K. Maejima, K. Sakurai, Sz. Fujita, Sg. Fujita, J. Cryst. Growth 237e239 (2002) 553. K. Ogata, S.-W. Kim, Sz. Fujita, Sg. Fujita, J. Cryst. Growth 240 (2002) 112. X. Wang, S. Yang, J. Wang, X. Jiang, G. Du, X. Liu, R.P.H. Chang, Optic. Quant. Electron. 34 (2002) 883. J.-Z. Wang, G.-T. Du, X.-Q. Wang, Y.-C. Chang, W. Yan, S.-R. Yang, Y. Ma, H.-S. Wang, D.-S. Gao, X. Liu, H. Cao, J.-Y. Xu, R.P.H. Chang, Chin. Phys. Lett. 19 (2002) 581. K.-I. Ogata, K. Maejima, Sz. Fujita, Sg. Fujita, J. Cryst. Growth 248 (2003) 25. C. Kirchner, Th. Gruber, F. Reuss, K. Thonke, A. Waag, Ch. Giessen, M. Heuken, J. Cryst. Growth 248 (2003) 20. B.S. Li, Y.C. Liu, D.Z. Shen, J.Y. Zhang, Y.M. Lu, X.W. Fan, J. Cryst. Growth 248 (2003) 179. G.-T. Du, J.-Z. Wang, X.-Q. Wang, X. Jiang, S.-R. Yang, Y. Ma, W. Yan, D.-S. Gao, X. Liu, H. Cao, J.-Y. Xu, R.P.H. Chang, Vacuum 69 (2003) 473. C.G. Van de Walle, Phys. Stat. Sol. (B) 229 (2002) 221. X. Li, Y. Yan, T.A. Gessert, C. DeHart, C.L. Perkins, D. Young, T.J. Coutts, Electrochem. SolidState Lett. 6 (2003) C56. B.-P. Zhang, L.-H. Manh, K. Wakatsuki, K. Tamura, T. Ohnishi, M. Lippmaa, N. Usami, M. Kawasaki, H. Koinuma, Y. Segawa, Jpn. J. Appl. Phys. 42 (2003) L264. X. Yang, G. Du, X. Wang, J. Wang, B. Liu, Y. Zhang, D. Liu, D. Liu, H.C. Ong, S. Yang, J. Cryst. Growth 252 (2003) 275. Y. Ma, G.T. Du, T.P. Yang, D.L. Qiu, X. Zhang, H.J. Yang, Y.T. Zhang, B.J. Zhao, X.T. Yang, D.L. Liu, J. Cryst. Growth 255 (2003) 303. Y. Zhang, G. Du, X. Wang, W. Li, X. Yang, Y. Ma, B. Zhao, H. Yang, D. Liu, S. Yang, J. Cryst. Growth 252 (2003) 180.

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

131

[66] J. Wang, G. Du, B. Zhao, X. Yang, Y. Zhang, Y. Ma, D. Liu, Y. Chang, H. Wang, H. Yang, S. Yang, J. Cryst. Growth 255 (2003) 293. [67] B.S. Li, Y.C. Liu, D.Z. Shen, J.Y. Zhang, Y.M. Lu, X.W. Fan, J. Cryst. Growth 249 (2003) 179. [68] H.W. Kim, K.S. Kim, C. Lee, J. Mater. Sci. Lett. 22 (2003) 1117. [69] Z.Z. Zhi, Y.C. Liu, X.T. Zhang, Y.M. Lu, D.Z. Shen, X.W. Fan, J. Phys. D Appl. Phys. 36 (2003) 719. [70] J. Ye, S. Gu, S. Zhu, T. Chen, W. Liu, F. Qin, L. Hu, R. Zhang, Y. Shi, Y. Zheng, J. Vac. Sci. Technol. A 21 (2003) 979. [71] T.P. Smith, W.J. Mecouch, P.Q. Miraglia, A.M. Roskowski, P.J. Hartlieb, R.F. Davis, J. Cryst. Growth 257 (2003) 255. [72] K. Haga, T. Suzuki, Y. Kashiwaba, H. Watanabe, B.P. Zhang, Y. Segawa, Thin Solid Films 433 (2003) 131. [73] Y. Kashiwaba, K. Sugawara, K. Haga, H. Watanabe, B.P. Zhang, Y. Segawa, Thin Solid Films 411 (2002) 87. [74] K.-H. Bang, D.-K. Hwang, M.-C. Jeong, K.-S. Sohn, J.-M. Myoung, Solid State Comm. 126 (2003) 623. [75] Ya.S. Savitskaya, L.A. Ryabova, Zh. Prikl. Khim. 37 (1964) 796. [76] L.A. Ryabova, Ya.S. Savitskaya, R.N. Sheftal, Izv. Akad. Nauk SSSR Neorg. Mater. 4 (1968) 602. [77] V.F. Korzo, P.S. Kireev, G.A. Lyashchenko, Neorg. Mater. 5 (1969) 367. [78] F.T.J. Smith, Appl. Phys. Lett. 43 (1983) 1108. [79] R. Solanki, G.J. Collins, Appl. Phys. Lett. 42 (1983) 662. [80] P.J. Wright, R.J.M. Griffiths, B. Cokayne, J. Cryst. Growth 66 (1984) 26. [81] P. Souletie, S. Bethke, B.W. Wessels, H. Pan, J. Cryst. Growth 86 (1988) 248. [82] S. Bethke, H. Pan, B.W. Wessels, Appl. Phys. Lett. 52 (1988) 138. [83] T. Maruyama, A. Nakai, Jpn. J. Appl. Phys. 28 (1989) L 346. [84] N. Oleynik, M. Adam, A. Krtschil, J. Bla¨sing, A. Dadgar, F. Bertram, D. Forster, A. Diez, A. Greiling, M. Seip, J. Christen, A. Krost, J. Cryst. Growth 248 (2003) 14. [85] O.F.Z. Khan, P. O’Brien, Thin Solid Films 173 (1989) 95. [86] T. Maruyama, J. Shionoya, J. Mater. Sci. Lett. 11 (1992) 170. [87] K. Kobayashi, T. Matsubara, S. Matsushima, S. Shirakata, S. Isomura, G. Okada, J. Mater. Sci. Lett. 15 (1996) 457. [88] M.F. Ogawa, Y. Natsume, T. Hirayama, H. Sakata, J. Mater. Sci. Lett. 9 (1990) 1351. [89] Y. Kashiwaba, F. Katahira, K. Haga, T. Sekiguchi, H. Watanabe, J. Cryst. Growth 221 (2000) 431. [90] Y. Kashiwaba, K. Haga, H. Watanabe, B.P. Zhang, Y. Segawa, K. Wakatsuki, Phys. Stat. Sol. (B) 229 (2002) 921. [91] Th. Kaufmann, G. Fuchs, M. Webert, Cryst. Res. Technol. 23 (1988) 635. [92] T.V. Tabenskaya, V.P. Ovsyannikov, E.A. Mazurenko, Journal de Physique IV, colloque C5, suppl. du Journal de Physique II 5 (1995) C5e711. [93] K. Haga, M. Kamidaira, Y. Kashiwaba, T. Sekiguchi, H. Watanabe, J. Cryst. Growth 214/215 (2000) 77. [94] T. Sekiguchi, K. Haga, K. Inaba, J. Cryst. Growth 214/215 (2000) 68. [95] T. Minami, H. Sato, H. Sonohara, S. Takata, T. Miyata, I. Fukuda, Thin Solid Films 253 (1994) 14. [96] H. Sato, T. Minami, T. Miyata, S. Takata, M. Ishii, Thin Solid Films 246 (1994) 65. [97] V. Sallet, J.F. Rommeluere, A. Lusson, A. Rivie`re, S. Fusil, O. Gorochov, R. Triboulet, Phys. Stat. Sol. (B) 229 (2002) 903. [98] Th. Gruber, C. Kirchner, A. Waag, Phys. Stat. Sol. (B) 229 (2002) 841. [99] J. Auld, D.J. Houlton, A.C. Jones, S.A. Rushworth, M.A. Azad, P. O’Brien, G.W. Critchlow, J. Mater. Chem. 4 (1994) 1249. [100] M.H. Koch, A.J. Hartmann, R.N. Lamb, M. Neuber, M. Grunze, J. Phys. Chem. B 101 (1997) 8231. [101] J. Lu, Z. Ye, J. Huang, L. Wang, B. Zhao, Appl. Surf. Sci. 207 (2003) 295. [102] M. Purica, E. Badianu, E. Rusu, M. Danila, R. Gavrila, Thin Solid Films 403e404 (2002) 485. [103] J.F. Rommelue`re, L. Svob, F. Jomard, J. Mimila-Arroyo, A. Lusson, V. Sallet, Appl. Phys. Lett., 83 (2003) 287.

132 [104] [105] [106] [107] [108] [109] [110] [111] [112] [113] [114] [115] [116] [117] [118] [119] [120] [121] [122] [123] [124] [125] [126]

[127] [128] [129] [130] [131] [132] [133] [134] [135] [136] [137] [138] [139] [140] [141]

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138 C. Thiandoume, V. Sallet, R. Triboulet, O. Gorochov, in press. K.-H. Bang, D.-K. Hwang, S.-W. Lim, J.-M. Myoung, J. Cryst. Growth 250 (2003) 437. F. Vigue´, Doctorate thesis, July 2001. C. Munuera, J. Zuniga-Pe´res, F. Soria, J.-F. Rommelue`re, V. Sallet, R. Triboulet, V. MunozSanjose´, C. Ocal, Journal of Crystal Growth 264 (2004) 70. A. Karchner, U. Haboeck, Martin Strassburg, Mathias Strassburg, G. Kaczmarczyk, A. Hofmann, C. Thomsen, A. Zeuner, H.R. Alves, D.M. Hofmann, B.K. Meyer, Appl. Phys. Lett. 80 (2002) 1909. C.G. Van de Walle, Phys. Stat. Sol. (B) 229 (2002) 221. T. Yamamoto, H. Katayama-Yoshida, Jpn. J. Appl. Phys. 38 (1999) L166. Y. Marfaing, Phys. Stat. Sol. (B) 229 (2002) 229. J. Nause, Private communication. F. Reuss, C. Kirchner, Th. Gruber, R. Kling, S. Maschek, W. Limmer, A. Waag, P. Zieman, Journal of Applied Physics 95 (2004) 3385. A. Waag, C. Kirchner, T. Gruber, F. Reuss, R. Kling, Phys. Stat. Sol. (C), in press. S. Muthukumar, Y. Chen, J. Zhong, F. Cosandey, Y. Lu, T. Siegrist, J. Cryst. Growth 261 (2004) 316. T. Gruber, C. Kirchner, R. Kling, F. Reuss, A. Waag, F. Bertram, D. Forster, J. Christen, M. Schreck, Appl. Phys. Lett. 83 (2003) 3290. K.L. Saenger, Proc. Adv. Mater. 2 (1993) 1. C. Belouet, Appl. Surf. Sci. 96e98 (1996) 630. D.B. Chrisey, G.K. Hubler (Eds.), Pulsed Laser Deposition of Thin Films, Wiley, New York, 1994. P.W. Willmott, J.R. Huber, Rev. Mod. Phys. 72 (2000) 315. J.C. Miller (Ed.), Laser Ablation and Desorption, Academic Press, San Diego, 1998. O. Nakagawara, T. Shimuda, T. Makino, S. Arai, H. Tabata, T. Kawai, Appl. Phys. Lett. 77 (2000) 3257. V. Craciun, N. Bassim, R.K. Singh, D. Craciun, J. Hermann, C. Boulmer-Leborgne, Appl. Surf. Sci. 186 (2002) 288. W.P. Barr, J. Phys. E 2 (1969) 112. S.G. Lee, D.S. Hwang, Y.K. Park, J.C. Park, Appl. Phys. Lett. 65 (1994) 764. A.A. Gorbunov, W. Pompe, A. Sewing, S.V. Gaponov, A.D. Akhsakhalyan, I.A. Zabrodin, I.A. Kas’kov, E.B. Klyenkov, A.P. Morozov, N.N. Salaschenko, R. Dietsch, H. Mai, S. Vo¨llmar, Appl. Surf. Sci. 96e98 (1996) 649. S. Witanachchi, K. Ahmed, P. Sakthivel, P. Mukherjee, Appl. Phys. Lett. 66 (1995) 1469. V. Craciun, D. Craciun, M.C. Bunescu, C. Boulmer-Leborgne, J. Hermann, Phys. Rev. B 58 (1998) 6787. A. Miotello, R. Kelly, Appl. Phys. A 69 (1999) 67. R. Kelly, A. Miotello, Phys. Rev. E 60 (1999) 2616. P. Mukherjee, S. Chen, J.B. Cuff, P. Sakthivel, S. Witanachchi, J. Appl. Phys. 91 (2002) 1828. N.J. Ianno, L. McConville, N. Shaikh, S. Pittal, P.G. Snyder, Thin Solid Films 220 (1992) 92. O. Guillot-Noe¨l, R. Gomez-san Roman, J. Perrie`re, J. Hermann, V. Craciun, C. Boulmer-Leborgne, P. Barboux, J. Appl. Phys. 80 (1996) 1803. V. Craciun, S. Amirhaghi, D. Craciun, J. Elders, J.G.E. Gardeniers, I.W. Boyd, Appl. Surf. Sci. 86 (1995) 99. V. Craciun, J. Elders, J.G.E. Gardeniers, I.W. Boyd, Appl. Phys. Lett. 65 (1994) 2963. S.L. King, J.G.E. Gardeniers, I.W. Boyd, Appl. Surf. Sci. 96e98 (1996) 811. S.G. Mayr, M. Moske, K. Samwer, M.E. Taylor, H.A. Atwater, Appl. Phys. Lett. 75 (1999) 4091. R. Gomez-San Roman, R. Perez Casero, C. Mare´chal, J.P. Enard, J. Perrie`re, J. Appl. Phys. 80 (1996) 1787. S. Im, B.J. Jin, S. Yi, J. Appl. Phys. 87 (2002) 4558. F. Claeyssens, A. Cheesman, S.J. Henley, M.N.R. Ashfold, J. Appl. Phys. 92 (2002) 6886. H.S. Kwok, H.S. Kim, D.H. Kim, W.P. Shen, X.W. Sun, R.F. Xiao, Appl. Surf. Sci. 109e110 (1997) 595.

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

133

[142] R.D. Vispute, S. Choopun, R. Enck, A. Patel, V. Talyansky, R.P. Sharma, T. Venkatesan, W.L. Sarney, L. Salamanca-Riba, S.N. Andronescu, A.A. Illiadis, K.A. Jones, J. Electron. Mater. 28 (1999) 275. [143] R.D. Vispute, V. Talyansky, Z. Trajanovic, S. Choopun, M. Downes, R.P. Sharma, T. Venkatesan, M.C. Woods, R.T. Laureau, K.A. Jones, A.A. Iliadis, Appl. Phys. Lett. 70 (1997) 2735. [144] I. Ohkubo, Y. Matsumoto, A. Ohtomo, T. Ohnishi, A. Tsukazaki, M. Lippmaa, H. Koinuma, M. Kawasaki, Appl. Surf. Sci. 159e160 (2000) 514. [145] I. Ohkubo, A. Ohtomo, T. Ohnishi, Y. Matsumoto, H. Koinuma, M. Kawasaki, Surf. Sci. 443 (1999) L1043. [146] V. Craciun, J. Perrie`re, N. Bassim, R.K. Singh, D. Craciun, J. Spear, Appl. Phys. A 69 (1999) 531. [147] V. Craciun, R.K. Singh, J. Perrie`re, J. Spear, D. Craciun, J. Electrochem. Soc. 147 (2000) 1077. [148] J. Narayan, K. Dovidenko, A.K. Sharma, S. Oktyabrsky, J. Appl. Phys. 84 (1998) 2597. [149] H.C. Ong, A.S.K. Li, G.T. Du, Appl. Phys. Lett. 78 (2001) 2667. [150] P.L. Washington, H.C. Ong, J.Y. Dai, R.P.H. Chang, Appl. Phys. Lett. 72 (1998) 3261. [151] X.W. Sun, H.S. Kwok, J. Appl. Phys. 98 (1999) 408. [152] S.H. Bae, S.Y. Lee, B.J. Jin, S. Im, Appl. Surf. Sci. 154e155 (2000) 458. [153] S.H. Bae, S.Y. Lee, B.J. Jin, S. Im, Appl. Surf. Sci. 169e170 (2001) 525. [154] B.J. Jin, S.H. Bae, S.Y. Lee, S. Im, Mater. Sci. Eng. B 71 (2000) 301. [155] B.J. Jin, H.S. Woo, S. Im, S.H. Bae, S.Y. Lee, Appl. Surf. Sci. 169e170 (2001) 521. [156] H.C. Ong, A.X.E. Zhu, G.T. Du, Appl. Phys. Lett. 80 (2002) 941. [157] E.U. Shim, H.S. Kang, J.S. Kang, J.H. Kim, S.Y. Lee, Appl. Surf. Sci. 186 (2002) 474. [158] N. Ashkenov, B.N. Mbenkum, C. Bundesmann, V. Riede, M. Lorenz, D. Spemann, E.M. Kaidashev, A. Kasic, M. Schubert, M. Grundmann, G. Wagner, H. Neumann, V. Darakchieva, H. Arwin, B. Monemar, J. Appl. Phys. 93 (2003) 126. [159] T. Koida, S.F. Chichibu, A. Uedeno, A. Tsukazaki, M. Kawasaki, T. Sota, Y. Segawa, H. Koinuma, Appl. Phys. Lett. 82 (2003) 532. [160] A. Shikanai, T. Azuhata, T. Sota, S. Chichibu, A. Kuramata, K. Horino, S. Nakamura, J. Appl. Phys. 81 (1997) 417. [161] T. Makino, T. Yasuda, Y. Segawa, A. Ohtomo, K. Tamura, M. Kawasaki, H. Koinuma, Appl. Phys. Lett. 79 (2001) 1282. [162] J.F. Muth, R.M. Kolbas, A.K. Sharma, S. Oktyabrsky, J. Narayan, J. Appl. Phys. 85 (1999) 7884. [163] A. Ohtomo, H. Kimura, K. Saito, T. Makino, Y. Segawa, H. Koinuma, M. Kawasaki, J. Cryst. Growth 214e215 (2000) 284. [164] S. Choopun, R.D. Vispute, W. Noch, A. Balsamo, R.P. Sharma, T. Venkatesan, A. Lliadis, D.C. Lock, Appl. Phys. Lett. 75 (1999) 3947. [165] T. Nakamura, Y. Yamada, T. Kusumori, H. Minoura, H. Muto, Thin Solid Films 411 (2002) 60. [166] E.M. Kaidashev, M. Lorenz, H. Von Weackstern, A. Rahn, H.C. Semmelhack, K.H. Han, G. Benndorf, C. Bundesmann, H. Hochmuth, M. Grundmann, Appl. Phys. Lett. 82 (2003) 3901. [167] A. Ohtomo, K. Tamura, K. Saikusa, T. Takahashi, T. Makino, Y. Segawa, H. Koinuma, M. Kawasaki, Appl. Phys. Lett. 75 (1999) 2635. [168] K. Tamura, A. Ohtomo, K. Sakusa, Y. Osaka, T. Makino, Y. Segawa, M. Sumya, S. Fuke, H. Koinuma, M. Kawasaki, J. Cryst. Growth 214e215 (2000) 59. [169] T. Makino, C.H. Chia, N.T. Tuan, Y. Segawa, M. Kawasaki, A. Ohtomo, K. Tamura, H. Koinuma, Appl. Phys. Lett. 76 (2000) 3549. [170] T. Makino, G. Isoya, Y. Segawa, C.H. Chia, T. Yusada, M. Kawasaki, A. Ohtomo, K. Tamura, H. Koinuma, J. Cryst. Growth 214e215 (2000) 289. [171] T. Makino, K. Tamura, C.H. Chia, Y. Segawa, M. Kawasaki, A. Ohtomo, H. Koinuma, J. Appl. Phys. 92 (2002) 7157. [172] A. Tzukazaki, A. Ohtomo, S. Yoshida, M. Kawasaki, C.H. Chia, T. Makino, Y. Segawa, T. Koida, S.F. Chichibu, H. Koinuma, Appl. Phys. Lett. 83 (2003) 2784. [173] J. Yin, Z.G. Liu, H. Liu, X.S. Wang, T. Zhu, J.M. Liu, J. Cryst. Growth 220 (2000) 281. [174] T. Nakamura, H. Minoura, H. Muto, Thin Solid Films 405 (2002) 109. [175] H. Ohta, H. Tanji, M. Orita, H. Hosono, H. Kawazoe, Mater. Res. Soc. Symp. Proc. 570 (1999) 309. [176] A. Tiwari, M. Park, C. Jin, H. Wang, D. Kumar, J. Narayan, J. Mater. Res. 17 (2002) 2480.

134 [177] [178] [179] [180] [181] [182] [183] [184] [185] [186] [187] [188] [189] [190] [191] [192] [193] [194] [195] [196] [197] [198] [199] [200] [201] [202] [203] [204] [205] [206] [207] [208] [209] [210] [211]

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138 M. Kumar, R.M. Mehra, A. Wakahara, M. Ishida, A. Yoshida, J. Appl. Phys. 93 (2003) 3837. C. Jin, A. Tiwari, A. Kvit, J. Narayan, Mater. Res. Soc. Proc. 722 (2002) 305. A. Tiwari, C. Jin, A. Kvit, D. Kumar, J.F. Muth, J. Narayan, Solid State Comm. 121 (2002) 371. M. Sugiura, Y. Nakashima, T. Nakasaka, T. Kobayashi, Appl. Surf. Sci. 197e198 (2002) 472. E. Vasco, C. Zaldo, L. Vasquez, J. Phys. Condens. Matter 13 (2001) L663. Y.R. Ryu, S. Zhu, D.C. Look, J.M. Wrobel, H.M. Jeong, H.W. White, J. Cryst. Growth 216 (2000) 326. Y.R. Ryu, S. Zhu, J.D. Budai, H.R. Chandrasekhar, P.F. Micelli, H.W. White, J. Appl. Phys. 88 (2000) 201. E. Vasco, J. Rubio-Zuazo, L. Vasquez, C. Prieto, C. Zaldo, J. Vac. Sci. Technol. B 19 (2001) 224. S.J. Henley, M.N.R. Ashfold, D. Cherns, Thin Solid Films 422 (2002) 69. Z. Jin, M. Murakami, T. Fukumura, Y. Matsumoto, A. Ohtomo, M. Kawasaki, H. Koinuma, J. Cryst. Growth 214e215 (2000) 55. H. Chang, C. Gao, I. Takeushi, Y. Yoo, J. Wang, P.G. Schultz, X.D. Xiang, R.P. Sharma, M. Downes, T. Venkatesan, Appl. Phys. Lett. 72 (1998) 2185. T. Ohnishi, D. Komiyama, T. Koida, S. Ohashi, S. Stauter, H. Koinuma, A. Ohtomo, M. Lippmaa, N. Nakagawa, M. Kawasaki, T. Kikuchi, K. Omote, Appl. Phys. Lett. 79 (2001) 536. A.R. Kaul, O.Y. Gorbenko, I.E. Graboy, S.V. Samoilenkov. in: M. Guilloux-Viry, A. Perrin (Eds.), Crystal Growth in Thin Solid Films, Research Signpost, Kerala, India, 2003, p. 265. S.J. Pearton, C.B. Abernathy, M.E. Overberg, G.T. Thaler, D.P. Norton, N. Theodoropulou, A.F. Hebrard, Y.D. Park, J. Kim, L.A. Boatner, J. Appl. Phys. 93 (2003) 1. T. Dietl, H. Ohno, F. Matsukara, J. Cibert, D. Ferrand, Science 287 (2000) 1019. T. Fukumura, Z. Jin, A. Ohtomo, H. Koinuma, M. Kawasaki, Appl. Phys. Lett. 75 (1999) 2366. T. Wakano, N. Fujimora, Y. Morigana, N. Abe, A. Ashida, T. Ito, Physica E 10 (2001) 260. H. Saeki, H. Tabata, T. Kawai, Solid State Comm. 120 (2001) 439. K. Ueda, H. Tabata, T. Kawai, Appl. Phys. Lett. 79 (2001) 988. H. Tabata, M. Saeki, S.L. Guo, J.H. Choi, T. Kawai, Physica B 308e310 (2001) 993. J.H. Kim, H. Kim, D. Kim, Y.E. Ihm, W.K. Choo, Physica B 327 (2003) 304. W. Prellier, A. Fouchet, B. Mercey, Ch. Simon, B. Raveau, Appl. Phys. Lett. 82 (2003) 3490. K. Rode, A. Anane, R. Mattana, J.P. Contour, O. Durand, R. LeBourgeois, J. Appl. Phys. 93 (2003) 7676. W. Yang, S.S. Hullavarad, B. Nagaraj, I. Takeuchi, R.P. Sharma, T. Venkatesan, R.D. Vispute, H. Shen, Appl. Phys. Lett. 82 (2003) 3424. A. Ohtomo, M. Kawasaki, T. Koida, K. Masubuchi, H. Koinuma, Y. Sakurai, Y. Yoshida, T. Yasuda, Y. Segawa, Appl. Phys. Lett. 72 (1998) 2466. A.K. Sharma, J. Narayan, J.F. Muth, C.W. Teng, C. Jin, A. Kvit, R.M. Kolbias, O.W. Holland, Appl. Phys. Lett. 75 (1999) 3327. C.W. Teng, J.F. Muth, U. Ozgur, M.J. Bergmann, H.O. Everitt, A.K. Sharma, C. Jin, J. Narayan, Appl. Phys. Lett. 76 (2000) 979. A. Ohtomo, R. Shiroki, I. Ohkubo, H. Koinuma, M. Kawasaki, Appl. Phys. Lett. 75 (1999) 4088. J. Narayan, A.K. Sharma, A. Kvit, C. Jin, J.F. Muth, O.W. Holland, Solid State Comm. 121 (2002) 9. S. Choopun, R.D. Vispute, W. Yang, R.P. Sharma, T. Venkatesan, H. Shen, Appl. Phys. Lett. 80 (2002) 1529. W. Yang, R.D. Vispute, S. Choopun, R.P. Sharma, T. Venkatesan, H. Shen, Appl. Phys. Lett. 78 (2001) 2787. I. Takeushi, W. Yang, K.S. Chang, M.A. Aronova, T. Venkatesan, R.D. Vispute, L.A. Bendersky, J. Appl. Phys. 94 (2003) 7336. P. Misra, P.K. Sahoo, P. Tripathi, V.N. Kulkarni, R.V. Nandekar, L.M. Kukreja, Appl. Phys. A 78 (2003) 37. T. Makino, Y. Segawa, M. Kawasaki, A. Ohtomo, R. Shiroki, K. Tamura, T. Yasuda, H. Koinuma, Appl. Phys. Lett. 78 (2001) 1237. H. Kim, J.S. Horwitz, S.B. Qadri, D.B. Chrisey, Thin Solid Films 420e421 (2002) 107.

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

135

[212] K. Nomura, H. Ohta, K. Ueda, M. Orita, M. Hirano, H. Hosono, Thin Solid Films 411 (2002) 147. [213] H. Ohta, H. Mizogushi, M. Hirano, S. Narushima, T. Kamiya, H. Hosono, Appl. Phys. Lett. 82 (2003) 823. [214] A. Ohtomo, M. Kawasaki, Y. Sakurai, I. Ohkubo, R. Shinoki, Y. Yoshida, T. Yasuda, Y. Segawa, H. Koinuma, Mater. Sci. Eng. B 56 (1998) 263. [215] A. Ohtomo, M. Kawasaki, I. Ohkubo, H. Koinuma, T. Yasuda, Y. Segawa, Appl. Phys. Lett. 75 (1999) 980. [216] A. Ohtomo, K. Tamura, M. Kawasaki, T. Makino, Y. Segawa, Z.K. Tang, G.K.L. Wong, Y. Matsumoto, H. Koinuma, Appl. Phys. Lett. 77 (2000) 2204. [217] H.D. Sun, T. Makino, N.T. Tuan, Y. Segawa, Z.K. Tang, G.K.L. Wong, M. Kawasaki, A. Ohtomo, K. Tamura, H. Koinuma, Appl. Phys. Lett. 77 (2000) 4250. [218] C.H. Chia, T. Makino, K. Tamura, Y. Segawa, M. Kawasaki, A. Ohtomo, H. Koinuma, Appl. Phys. Lett. 82 (2003) 1848. [219] X.W. Sun, R.F. Xiao, H.S. Kwok, J. Appl. Phys. 84 (1998) 5776. [220] R.P. Wang, H. Muto, Y. Yamada, T. Kusumori, Thin Solid Films 411 (2002) 69. [221] T. Ueda, T.F. Huang, S. Spruytte, H. Lee, M. Yuri, K. Itoh, T. Baba, J.S. Harris, J. Cryst. Growth 187 (1998) 340. [222] R.D. Vispute, V. Talyansky, S. Choopun, R.P. Sharma, T. Venkatesan, M. He, X. Tang, J.B. Halpern, M.G. Spencer, Y.X. Li, L.G. Salamanca-Riba, A.A. Lliadis, K.A. Jones, Appl. Phys. Lett. 73 (1998) 348. [223] H. Ohta, K. Kawamura, M. Orita, M. Hirano, N. Sarukara, H. Hosono, Appl. Phys. Lett. 77 (2000) 475. [224] H. Hosono, H. Ohta, K. Hayashi, M. Orita, M. Hirano, J. Cryst. Growth 237e239 (2002) 496. [225] H. Ohta, M. Hirano, K. Nakahara, H. Maruta, T. Tanebe, M. Kamiya, T. Kamiya, H. Hosono, Appl. Phys. Lett. 83 (2003) 1029. [226] A. Tiwari, C. Jin, D. Kumar, J. Narayan, Appl. Phys. Lett. 83 (2003) 1773. [227] M.D. Shirk, P.A. Molian, J. Laser Appl. 10 (1998) 18. [228] F. Qian, V. Craciun, R.K. Singh, D.S. Dutta, P.P. Pronko, J. Appl. Phys. 86 (1999) 2281. [229] Z. Zhang, P.A. VanRompay, J.A. Nees, R. Clarke, X.Q. Pan, P.P. Pronko, Appl. Surf. Sci. 154e155 (2000) 424. [230] E. Gyorgy, C. Ristocu, I.N. Mihailescu, A. Klini, N. Vainos, C. Fotakis, C. Ghica, G. Schmerber, J. Faerber, J. Appl. Phys. 90 (2001) 456. [231] Z. Zhang, P.A. VanRompay, J.A. Nees, C.A. Stewart, X.Q. Pan, L. Fu, P.P. Pronko, Proc. SPIE 3935 (2000) 86. [232] E. Millon, J. Perrie`re, R.M. Defourneau, D. Defourneau, O. Albert, J. Etchepare, Appl. Phys. A 77 (2003) 73. [233] M. Okoshi, K. Higashikawa, M. Hanabusa, Appl. Surf. Sci. 154e155 (2000) 424. [234] E. Millon, O. Albert, J.C. Loulergue, J. Etchepare, D. Hulin, W. Seiler, J. Perrie`re, J. Appl. Phys. 88 (2000) 6937. [235] J. Perrie`re, E. Millon, R.M. Defourneau, D. Defourneau, O. Albert, J. Etchepare, J.C. Loulergue, W. Seiler. in: M. Guilloux-Viry, A. Perrin (Eds.), Crystal Growth in Thin Solid Films, Research Signpost, Kerala, India, 2003, p. 153. [236] O. Albert, S. Roger, Y. Glinec, J.C. Loulergue, J. Etchepare, C. Boulmer-Leborgne, J. Perrie`re, E. Millon, Appl. Phys. A 76 (2003) 319. [237] S. Zhu, D.H. Lowndes, J.D. Budai, D.P. Norton, Appl. Phys. Lett. 65 (1994) 2012. [238] R.P. Wang, S.H. Pan, Y.L. Zhou, Solid State Comm. 114 (2000) 613. [239] A. Basillais, PhD thesis, Universite´ d’Orle´ans, France, December 2000. [240] A. Ohtomo, M. Kawasaki, Y. Sakurai, Y. Yoshida, H. Koinuma, P. Yu, Z.K. Tang, G.K.L. Wong, Y. Segawa, Mater. Sci. Eng. B 54 (1998) 24. [241] H. Cao, Y.G. Zhao, H.C. Hong, S.T. Ho, J.Y. Dai, J.Y. Wu, R.P.H. Chang, Appl. Phys. Lett. 73 (1998) 3656. [242] H. Cao, Y.G. Zhao, S.T. Ho, E.W. Seelig, Q.H. Wang, R.P.H. Chang, Phys. Rev. Lett. 82 (1999) 2278.

136

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

[243] W. Zhang, H. Wang, K.S. Wong, Z.K. Tang, G.K.L. Wong, J. Ravinder, Appl. Phys. Lett. 75 (1999) 3321. [244] S. Lettieri, P. Maddalena, J. Appl. Phys. 91 (2002) 5564. [245] W. Shi, Z. Chen, T. Zhao, H. Lu, Y. Zhou, G. Yang, J. Am. Opt. Soc. B 17 (2000) 235. [246] G. Wang, G.T. Kiehne, G.K.L. Wong, J.B. Ketterson, X. Liu, R.P.H. Chang, Appl. Phys. Lett. 80 (2002) 401. [247] J.E. Dominguez, L. Fu, X.Q. Pan, Appl. Phys. Lett. 79 (2001) 614. [248] M. Joseph, H. Tabata, T. Kawai, Jpn. J. Appl. Phys. 38 (1999) L1205. [249] M. Sumya, S. Fuke, A. Tsukazaki, K. Tamura, A. Ohtomo, M. Kawasaki, H. Koinuma, J. Appl. Phys. 93 (2003) 2562. [250] A. Tsukazaki, H. Saito, K. Tamura, M. Ohtani, H. Koinuma, M. Sumya, S. Fuke, T. Fukumura, M. Kawasaki, Appl. Phys. Lett. 81 (2002) 235. [251] H. Tabata, M. Saeki, S.L. Guo, J.H. Choi, T. Kawai, Physica B 308e310 (2001) 993. [252] Y.W. Heo, S.J. Park, K. Ip, S.J. Pearton, D.P. Norton, Appl. Phys. Lett. 83 (2003) 1128. [253] Y.R. Ryu, S. Zhu, D.C. Look, J.M. Wrobel, H.M. Jeong, H.W. White, J. Cryst. Growth 216 (2000) 330. [254] Y.R. Ryu, T.S. Lee, H.W. White, Appl. Phys. Lett. 83 (2003) 1. [255] M.A.L. Johnson, S. Fujita, W.H. Rowland Jr., W.C. Hughes Jr., J.W. Cook, J.F. Schetzina, J. Electron. Mater. 25 (1996) 855. [256] Y. Chen, D.M. Bagnall, Z. Zhu, T. Sekiuchi, K.-T. Park, K. Hiraga, T. Yao, S. Koyama, M.Y. Shen, T. Goto, J. Cryst. Growth 181 (1997) 165. [257] H.-B. Kang, K. Nakamura, K. Yoshida, K. Ishikawa, Jpn. J. Appl. Phys. 36 (1997) L933. [258] S. Yamauchi, H. Handa, A. Nagayama, T. Hariu, Thin Solid Films 345 (1999) 12. [259] P. Fons, K. Iwata, S. Niki, A. Yamada, K. Matsubara, J. Cryst. Growth 201/202 (1999) 627. [260] K. Sakurai, M. Kanehiro, K. Nakahara, T. Tanabe, Sz. Fujita, Sg. Fujita, J. Cryst. Growth 209 (2000) 522. [261] Y. Chen, H.-J. Ko, S.-K. Hong, T. Yao, Appl. Phys. Lett. 76 (2000) 559. [262] K. Ogata, K. Sakurai, Sz. Fujita, Sg. Fujita, K. Matsushige, J. Cryst. Growth 214/215 (2000) 312. [263] K. Sakurai, M. Kanehiro, K. Nakahara, T. Tanabe, Sz. Fujita, Sg. Fujita, J. Cryst. Growth 214/215 (2000) 92. [264] K. Iwata, P. Fons, A. Yamada, K. Matsubara, S. Niki, J. Cryst. Growth 209 (2000) 526. [265] S. Yamauchi, T. Ashiga, A. Nagayama, T. Hariu, J. Cryst. Growth 214/215 (2000) 63. [266] Y.F. Chen, H.J. Ko, S.K. Hong, K. Inaba, Y. Segawa, T. Yao, J. Cryst. Growth 227/228 (2001) 917. [267] F. Vigue´, P. Venne´gue`s, S. Ve´zian, M. Lau¨gt, J.-P. Faurie, Appl. Phys. Lett. 79 (2001) 194. [268] S.-H. Lim, D. Shindo, H.-B. Kang, K. Nakamura, J. Cryst. Growth 225 (2001) 208. [269] F. Vigue´, P. Venne´gue`s, C. Deparis, S. Ve´zian, M. Lau¨gt, J.-P. Faurie, J. Appl. Phys. 90 (2001) 5115. [270] Y. Chen, N.T. Tuan, Y. Segawa, H.-J. Ko, S.-K. Hong, T. Yao, Appl. Phys. Lett. 78 (2001) 1469. [271] H.-J. Ko, T. Yao, Y. Chen, S.-K. Hong, J. Appl. Phys. 92 (2002) 4354. [272] S. Shirakata, K. Saeki, T. Terasako, J. Cryst. Growth 237/239 (2002) 528. [273] H.-B. Kang, K. Nakamura, S.-H. Lim, D. Shindo, Jpn. J. Appl. Phys. 37 (1998) 781. [274] P. Fons, K. Iwata, A. Yamada, K. Matsubara, S. Niki, K. Nakahara, H. Takasu, Appl. Phys. Lett. 77 (2000) 1801. [275] P. Fons, K. Iwata, S. Niki, A. Yamada, K. Matsubara, M. Watanabe, J. Cryst. Growth 209 (2000) 532. [276] K. Nakahara, T. Tanabe, H. Takasu, P. Fons, K. Iwata, A. Yamada, K. Matsubara, R. Hunger, S. Niki, Jpn. J. Appl. Phys. 40 (2001) 250. [277] S.-H. Lim, D. Shindo, H.-B. Kang, K. Nakamura, J. Cryst. Growth 225 (2001) 202. [278] P. Fons, K. Iwata, A. Yamada, K. Matsubara, S. Niki, K. Nakahara, T. Tanabe, H. Takasu, J. Cryst. Growth 227/228 (2001) 911. [279] K. Nakahara, H. Takasu, P. Fons, K. Iwata, A. Yamada, K. Matsubara, R. Hunger, S. Niki, J. Cryst. Growth 227/228 (2001) 923. [280] N. Nakahara, H. Takasu, P. Fons, A. Yamada, K. Iwata, K. Matsubara, R. Hunger, S. Niki, Appl. Phys. Lett. 79 (2001) 4139. [281] A.B.M.A. Ashrafi, I. Suemune, H. Kumano, S. Tanaka, Jpn. J. Appl. Phys. 41 (2002) L1281.

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138

137

[282] S.F. Chichibu, T. Sota, P.J. Fons, K. Iwata, A. Yamada, K. Matsubara, S. Niki, Jpn. J. Appl. Phys. 41 (2002) L935. [283] A.B.M.A. Ashrafi, I. Suemune, H. Kumano, K. Uesugi, Phys Stat. Sol. (A) 192 (2002) 224. [284] A.B.M.A. Ashrafi, I. Suemune, H. Kumano, Jpn. J. Appl. Phys. 41 (2002) 2851. [285] T. Ohgaki, N. Ohashi, H. Kakemoto, S. Wada, Y. Adachi, H. Haneda, T. Tsurumi, J. Appl. Phys. 93 (2003) 1961. [286] K. Iwata, P. Fons, S. Niki, A. Yamada, K. Matsubara, K. Nakahara, H. Takasu, Phys. Stat. Sol. (A) 180 (2000) 287. [287] X. Wang, H. Iwaki, M. Murakami, X. Du, Y. Ishitani, A. Yoshikawa, Jpn. J. Appl. Phys. 42 (2003) L99. [288] A. Yoshikawa, X. Wang, Y. Tomita, O.-H. Roh, H. Iwaki, Y. Ishitani, Phys. Stat. Sol. (C), in press. [289] X. Wang, H. Iwaki, Y. Ishitani, A. Yoshikawa, Phys. Stat. Sol. (C), in press. [290] K. Nakamura, T. Shoji, H.-B. Kang, Jpn. J. Appl. Phys. 39 (2000) L534. [291] H. Kumano, A.A. Ashrafi, A. Ueta, A. Avramescu, I. Suemune, J. Cryst. Growth 214/215 (2000) 280. [292] T. Matsumoto, K. Nishimura, Y. Nabetani, T. Kato, Phys. Stat. Sol. (C), in press. [293] K. Iwata, P. Fons, S. Niki, A. Yamada, K. Matsubara, K. Nakahara, T. Tanabe, H. Takasu, J. Cryst. Growth 214/215 (2000) 50. [294] M. Fujita, N. Kawamoto, T. Tatsumi, K. Yamagishi, Y. Horikoshi, Jpn. J. Appl. Phys. 42 (2003) 67. [295] H.J. Ko, Y. Chen, S.K. Hong, T. Yao, J. Cryst. Growth 209 (2000) 816. [296] S.-K. Hong, H.-J. Ko, Y. Chen, T. Hanada, T. Yao, J. Cryst. Growth 214/215 (2000) 81. [297] S.-K. Hong, T. Hanada, H.-J. Ko, Y. Chen, T. Yao, Appl. Phys. Lett. 77 (2000) 3571. [298] H.J. Ko, Y.F. Chen, S.K. Hong, H. Wenisch, T. Yao, Appl. Phys. Lett. 77 (2000) 3761. [299] H.J. Ko, Y.F. Chen, T. Yao, K. Miyajima, A. Yamamoto, T. Goto, Appl. Phys. Lett. 77 (2000) 537. [300] S.K. Hong, T. Hanada, H.J. Ko, Y. Chen, T. Yao, Phys. Rev. B 65 (2002) 1. [301] H. Kato, M. Sano, K. Miyamoto, J. Appl. Phys. 92 (2002) 1960. [302] H.J. Ko, Y.F. Chen, Z. Zhu, T. Yao, I. Kobayashi, K. Uchiki, Appl. Phys. Lett. 76 (2000) 1905. [303] D.C. Look, D.C. Reynolds, C.W. Litton, R.L. Jones, D.B. Eason, G. Cantwell, Appl. Phys. Lett. 81 (2002) 1830. [305] H. Kato, M. Sano, K. Miyamoto, T. Yao, Phys. Stat. Sol. (C), in press. [306] B.K. Meyer, A. Zeuner, B. Struck, H. Alves, D.M. Hofmann, M. Heuken, Second International Workshop on ZnO, Dayton, USA, October 2002. [307] K. Ogata, K. Koite, T. Tanite, T. Komuro, F. Yan, S. Sasa, M. Inoue, M. Yano, J. Cryst. Growth 251 (2003) 623. [308] Sz. Fujita, T. Takagi, H. Tanaka, Sg. Fujita, Phys. Stat. Sol. (C), in press. [309] K. Sakurai, T. Kubo, D. Kajita, T. Tanabe, H. Takasu, Sz. Fujita, Sg. Fujita, Jpn. J. Appl. Phys. 39 (2000) L1146. [310] K. Sakurai, T. Takagi, T. Kubo, D. Kajita, T. Tanabe, H. Takasu, Sz. Fujita, Sg. Fujita, J. Cryst. Growth 237/239 (2002) 514. [311] T. Minami, H. Nanto, S. Takata, Jpn. J. Appl. Phys. 23 (1984) L280. [312] T. Minami, H. Sato, H. Nanto, S. Takata, Jpn. J. Appl. Phys. 24 (1985) L781. [313] T. Minami, H. Sato, H. Nanto, S. Takata, Jpn. J. Appl. Phys. 25 (1986) L776. [314] Z.-C. Jin, C.G. Granqvist, Appl. Opt. 26 (1987) 3191. [315] T. Minami, K. Oohashi, S. Takata, Thin Solid Films 193/194 (1990) 721. [316] T. Yamamoto, T. Shiosaki, A. Kawabata, J. Appl. Phys. 51 (1980) 3113. [317] K. Ohji, O. Yamazaki, K. Wasa, S. Hayakawa, J. Vac. Sci. Technol. 14 (1978) 1601. [318] O. Yamazaki, T. Mitsuyu, K. Wasa, IEEE Trans. SU-27 (1980) 369. [319] Y. Manabe, T. Mitsuyu, Jpn. J. Appl. Phys. 29 (1990) 334. [320] M. Kadota, T. Miura, M. Minakata, J. Cryst. Growth 237/239 (2002) 523. [321] F. Quaranta, A. Valentini, F.R. Rizzi, G. Casamassina, J. Appl. Phys. 74 (1993) 244. [322] K. Yamada, Y. Yamaki, H. Nakanishi, S. Chichibu, App. Phys. Lett. 72 (1998) 235. [323] S.F. Chichibu, T. Yoshida, T. Onuma, J. Appl. Phys. 91 (2002) 874. [324] P.M. Verghese, D.R. Clarke, J. Mater. Res. 14 (1999) 1039. [325] A.J. Drehman, S.-Q. Wang, P.W. Yip, Mater. Res. Symp. Proc. 482 (1998) 289.

138 [326] [327] [328] [329] [330] [331] [332] [333] [334] [335] [336] [337] [338] [339] [340] [341] [342] [343] [344] [345] [346] [347] [348] [349] [350] [351] [352]

R. Triboulet, J. Perrie`re / Progress in Crystal Growth and Characterization of Materials 47 (2003) 65e138 Shen Zhu, C.-H. Su, S.L. Lehoczky, M.A. George, D.H. Lowndes, J. Mater. Res. 15 (2000) 1125. P. Sharma, A. Mansingh, K. Sreenivas, Appl. Phys. Lett. 80 (2002) 553. P. Sharma, K. Sreenivas, K.V. Rao, J. Appl. Phys. 93 (2003) 3963. A. Hachigo, H. Nakahata, K. Higaki, S. Fujii, S.-I. Shikata, Appl. Phys. Lett. 65 (1994) 2556. A. Hachigo, H. Nakahata, K. Higaki, S. Fujii, S.-I. Shikata, Appl. Phys. Lett. 65 (1994) 2556e2558. K. Matsubara, P. Fons, A. Yamada, M. Watanabe, S. Niki, Thin Solid Films 347 (1999) 238. J. Moralius, J. Kaitila, T. Pensala, M. Ylilammi, J. Mater. Sci. 14 (2003) 431. S.-H. Jeong, B.-S. Kim, B.-T. Lee, Appl. Phys. Lett. 82 (2003) 2625. S.-H. Jeong, B.-S. Kim, B.-T. Lee, J. Phys. D: Appl. Phys. 36 (2003) 2017. Y. Zhang, G. Du, D. Liu, X. Wang, Y. Ma, J. Wang, J. Yin, X. Yang, X. Hou, S. Yang, J. Cryst. Growth 243 (2002) 439. H.K. Kjro, A. Nahhas, E. Kjro and B. Lee, Oral communication in the International Workshop on ZnO, Dayton, USA, October 7e8 1999. S.-Y. Chu, W. Water, J.-T. Liaw, J. Euro. Cer. Soc. 23 (2003) 1593. M.K. Ryu, S.H. Lee, M.S. Jang, G.N. Panin, T.W. Kang, J. Appl. Phys. 92 (2002) 154. S.J. Chang, Y.K. Su, Y.P. Shei, J. Vac. Sci. Technol. A 13 (1995) 381. S.J. Chang, Y.K. Su, Y.P. Shei, J. Vac. Sci. Technol. A 13 (1995) 385. K.-K. Kim, J.-H. Song, H.-J. Jung, W.-K. Choi, J. Appl. Phys. 87 (2000) 3573. K.-K. Kim, J.-H. Song, H.-J. Jung, S.-J. Park, J.-H. Song, J.-Y. Lee, J. Vac. Sci. Technol. A 18 (2000) 2864. A.J. Drehman, L.O. Bouthillette, P.W. Yip, S-Q. Wang, K. Vaccaro, S.M. Spaziani, C.L. Woods, Oral communication in the International Workshop on ZnO, Dayton, USA, October 7e8 1999. S. Zhu, C.-H. Su, S.L. Lehoczky, M.A. George, P. McCarty, Oral communication in the International Workshop on ZnO, Dayton, USA, October 7e8 1999. K.-K. Kim, H.-S. Kim, D.-K. Hwang, J.-H. Lim, S.-J. Park, Appl. Phys. Lett. 83 (2003) 63. A.V. Singh, R.M. Mehra, A. Wakahara, A. Yoshida, J. Appl. Phys. 93 (2003) 396. Z.-Z. Ye, J.-G. Lu, H.-H. Chen, Y.-Z. Zhang, L. Wang, B.-H. Zhao, J.-Y. Huang, J. Cryst. Growth 253 (2003) 258. J. Huang, Z. Ye, H. Chen, B. Zhao, L. Wang, J. Mater. Sci. Lett. 22 (2003) 249. J. Lu, Y. Zhang, Z. Ye, L. Wang, B. Zhao, J. Huang, Mater. Lett. 57 (2003) 3311. C. Wang, Z. Ji, K. Liu, Y. Xiang, Z. Ye, J. Cryst. Growth 259 (2003) 279. D. Rogers, F. Teherani, Oral communication in the SPIE Electronic Imaging Symposium, San Jose, 18e22 January 2004. E. Ohshima, H. Ogino, I. Niikura, K. Maeda, M. Sato, M. Ito, T. Fukuda, J. Cryst. Growth 260 (2004) 166.