Highly textured ZnO thin films were successfully grown on Si(111) by atomic layer deposition using an epitaxial AlN buffer layer at deposition temperatures between 100 and 300 °C. X-ray diffraction analysis proves an epitaxial relationship of ZnO[0001]//AlN[0001] and ZnO[112¯0]//AlN[112¯0]. Omega scans of the (0002) and (101¯0) reflections of ZnO demonstrate an improving crystalline quality for increasing deposition temperatures. An additional thermal postannealing step at 800 °C is found to be beneficial to further improve the crystal structure.

Due to its wide and direct bandgap of 3.37 eV, an exciton binding energy of 59 meV, and its wurtzite structure with a high piezoelectric coefficient, ZnO is an attractive candidate for future optoelectronic devices as well as for piezoelectric applications. Devices based on ZnO have attracted huge interest in the field of transparent conductive oxides,1 light emitting diodes,2 lasers,3 UV detectors,4 and surface acoustic wave filters.5 The deposition of high-quality semiconductor thin films is essential for many modern devices and depends particularly on the substrate. The heteroepitaxial growth of ZnO films with good crystalline quality was predominantly reported directly on sapphire6–8 and on epitaxial GaN films grown on sapphire.9,10

The use of sapphire as a substrate has several drawbacks in contrast to Si. It is expensive, has an insulating behavior, and its diameter is limited. The direct heteroepitaxial growth of ZnO on Si, however, has not been accomplished as the Si substrate is easily oxidized during the deposition process, which results then in an amorphous or polycrystalline ZnO film. As GaN and ZnO have the same wurtzite crystal structure and similar lattice constants, GaN would be a promising buffer layer for the epitaxial growth of ZnO on Si. Nahhas et al.11 reported on the epitaxial growth of ZnO on Si by RF magnetron sputtering using a GaN buffer layer, but comparable results on the crystalline quality of the films in terms of omega scans (ω-scans) were not given. However, the direct heteroepitaxial growth of GaN on Si substrates remains challenging due to different thermal expansion coefficients combined with high growth temperatures as well as a destructive reaction between Ga and Si.12 Therefore, an AlN buffer layer is commonly used to prevent the formation of a Ga-Si alloy and to grow high-quality GaN films.13 The direct epitaxial growth of AlN on Si(111), on the other hand, is well established. Since AlN has the same wurtzite crystal structure and only a small lattice mismatch to ZnO of 4%, an AlN buffer layer could enable to integrate ZnO thin films in Si-based MEMS technology. The growth of ZnO films on AlN has already been demonstrated using metal organic chemical vapor deposition (MOCVD),14 pulsed-laser deposition (PLD),15 sputtering,16–19 and atomic layer deposition (ALD).20,21 The growth of ZnO at high temperatures using MOCVD (680 °C) and PLD (750 °C) resulted in a high crystalline quality.14,15 The sputtering of ZnO films on AlN was reported at much lower temperature, but the crystalline quality is significantly reduced.16–19 The reports on the ALD of ZnO films on AlN, however, lack information about the crystalline quality and the in-plane orientation. Zhu et al.20 demonstrated the growth of polycrystalline ZnO by ALD on Ti using an AlN interlayer but focused mainly on the piezoelectric performance. The ZnO films grown by ALD on an molecular beam epitaxy grown AlN buffer layer by Wang et al.21 showed a preferred c-oriented growth but comparable information about the crystal quality or the in-plane orientation was not given.

ALD is a particularly promising method, since the uniform and conformal growth of thin films is ensured over large areas at relatively low temperatures. Due to its self-limiting nature, ALD is an ideal technique to deposit thin films with precise thickness control even on nanoscale surfaces with a high aspect ratio. However, no reports on the heteroepitaxial growth of ZnO on AlN by ALD have been published so far. Furthermore, although the epitaxial ALD of ZnO was already demonstrated on sapphire and GaN,8,10 a comprehensive study on the influence of deposition temperature and an additional postannealing treatment on the crystalline quality and on the in-plane orientation of the ZnO films is still missing.

In this work, we describe the epitaxial growth and the detailed characterization of highly textured ZnO thin films on Si(111) by ALD using an AlN buffer layer. The influence of the deposition temperature and an additional postannealing step on the structural and optical quality was investigated in detail by xX-ray diffraction (XRD) and photoluminescence (PL) measurements. The results are compared with data available from the literature and critically evaluated.

The Si(111) substrates with a 200 nm thick epitaxial AlN buffer layer grown by MOCVD were cleaned in acetone and isopropanol using an ultrasonic bath for 5 min. Afterward, the ZnO thin films were grown by thermal ALD in an OPAL reactor from Oxford Instruments. Diethylzinc and de-ionized H2O were used as precursors with dose times of 50 and 30 ms, respectively. The growth chamber was purged with 25 SCCM N2 for 10 s after each dose step. All samples were grown with 500 cycles, while the substrate temperature was varied (100 °C, 200 °C, and 300 °C) resulting in a film thickness of 72, 95, and 67 nm for the respective as-grown films.

The effect of postannealing was studied by heating the samples to 800 °C in a horizontal quartz tube furnace under vacuum conditions (4.0 × 10−5 mbar) for 1 h. To remove residual gases, the system was pumped to vacuum followed by a purging step with Ar gas twice.

The surface morphology of the as-grown and annealed ZnO thin films was investigated by SEM using a FEI Nova NanoSEM 430. The structural quality of the respective films was characterized by XRD using a PANalytical X’Pert3 MRD with Cu-Kα1 radiation and a Rigaku Smartlab μHR for in-plane ω-scans. The in-plane epitaxial relationship between ZnO film and AlN buffer layer was determined using phi-scans. The ZnO films were optically characterized using room-temperature photoluminescence (PL) spectroscopy by exciting the samples with a HeCd laser (λ = 325 nm) and a power density of 100 mW cm−2. All spectra were corrected for the spectral response of the PL measurement system.

Figure 1 compares the surface morphology of as-grown and post-annealed ALD ZnO thin films deposited at substrate temperatures of 100 °C [(a), and (d)], 200 °C [(b) and (e)], and 300 °C [(c) and (f)]. The as-grown ALD films have a grainy surface similar to other reports. The average grain size increases with increasing deposition temperature. The postannealing step at 800 °C provides enough thermal energy for recrystallization of the ZnO films and the coalescence of grains. Coalescent grains are most pronounced for the ZnO film deposited at a substrate temperature of 300 °C. The SEM images also indicate voids in the postannealed ZnO films, particularly for the sample deposited at 200 °C. Those voids could be avoided by using thicker ZnO films or by thermal postannealing at a lower temperature.

FIG. 1.

Top-view and cross-sectional (inset) SEM images of as-grown and post-annealed ZnO thin films. As-grown samples were deposited at (a) 100 °C, (b) 200 °C, and (c) 300 °C, and their 800 °C annealed counterparts are shown below in (d), (e), and (f).

FIG. 1.

Top-view and cross-sectional (inset) SEM images of as-grown and post-annealed ZnO thin films. As-grown samples were deposited at (a) 100 °C, (b) 200 °C, and (c) 300 °C, and their 800 °C annealed counterparts are shown below in (d), (e), and (f).

Close modal

The effect of different deposition temperatures and the additional postannealing treatment on the structural quality of the ZnO films was studied by XRD. The peaks in the symmetric θ/2θ scan provided in Fig. 2 were identified as Si(111) of the Si substrate, AlN(0002) of the 200 nm thick AlN buffer layer, and ZnO(0002) of the ALD ZnO film. The results indicate an out-of-plane relationship of ZnO[0001]//AlN[0001]//Si[111]. No other peaks were observed.

FIG. 2.

Symmetric θ/2θ XRD patterns of as-grown (100, 200, and 300 °C) and post-annealed ZnO thin films grown on AlN/Si substrates.

FIG. 2.

Symmetric θ/2θ XRD patterns of as-grown (100, 200, and 300 °C) and post-annealed ZnO thin films grown on AlN/Si substrates.

Close modal

The crystalline quality of the ZnO films was characterized in more detail using ω-scans of the out-of-plane (0002) and the in-plane (101¯0) reflection. The ω-scans of all samples are provided in the supplementary materia22 and were analyzed in detail in terms of their full width at half maximum (FWHM).22 The FWHM values of both out-of-plane (0002) and in-plane (101¯0) ω-scans of as-grown and postannealed ZnO films are compared as a function of the deposition temperature in Fig. 3. The FWHM of an ω-scan indicates the degree of alignment of the respective plane. As shown in Fig. 3, the FWHM of the ω-scan of the (0002) reflection of the as-deposited ZnO films decreases significantly from 3.83° to 1.22° as the deposition temperature was increased from 100 to 200 °C. A change in the deposition temperature to 300 °C causes a slight further reduction to 1.21°. The crystalline quality of the ZnO films improves further by an additional postannealing step at 800 °C. After annealing, the FWHM of the (0002) ω-scans were calculated to be 1.37°, 0.96°, and 0.75° for the ZnO films deposited at 100, 200, and 300 °C, respectively.

FIG. 3.

FWHM values of the ω scans of the (0002) and (101¯0) reflections of as-grown and post-annealed ZnO as a function of deposition temperature (note that the lines are just a guide to the eye). The FWHM of the AlN buffer layer are marked by dotted lines.

FIG. 3.

FWHM values of the ω scans of the (0002) and (101¯0) reflections of as-grown and post-annealed ZnO as a function of deposition temperature (note that the lines are just a guide to the eye). The FWHM of the AlN buffer layer are marked by dotted lines.

Close modal

A similar trend was observed for the ω-scans of the in-plane (101¯0) reflection. The in-plane orientation improves drastically from 7.39° to 2.41° and further to 1.80° for a deposition temperature of 100, 200, and 300 °C, respectively. The additional postannealing treatment is also beneficial for the in-plane quality. The FWHM of the ω-scans of the (101¯0) reflection were 2.09°, 1.47°, and 1.24° for the postannealed ZnO films deposited at 100, 200, and 300 °C, respectively.

Although the postannealing temperature is significantly higher than the deposition temperature, the latter has a significant influence on the crystalline quality of the postannealed samples. The crystalline quality of the postannealed ZnO film deposited at 300 °C approaches the FWHM of the ω-scan of the (0002) reflection of the MOCVD grown AlN buffer layer of 0.52° as well as of the respective in-plane value of 0.92°. Hence, the quality of the ZnO films reported here might be limited by the quality of the AlN buffer layer.

Wang et al.14 reported a FWHM of the (0002) ω-scan of 410 arc sec (0.11°) for a ZnO film deposited by MOCVD on Si(111) with a 200 Å thick AlN layer. The growth of high-quality ZnO films was also demonstrated by PLD on sapphire with a FWHM of 0.09° using an AlN buffer layer.15 The reported substrate temperatures, however, were 680 and 750 °C for MOCVD and PLD, respectively. Up to now, ZnO deposition on AlN at lower temperatures was mainly achieved by sputtering. A FWHM of 4.5° was reported for ZnO grown with DC sputtering at a substrate temperature of 400 °C using an AlN buffered sapphire substrate.16 However, ZnO deposition by RF magnetron sputtering showed a better crystalline quality. ZnO sputtered on AlN buffer layers on Si(100) were reported with a FWHM of 2.4° and 2.8° at temperatures smaller than 100°C17,18 and 1.88° for AlN on Si(111) at 300 °C.19 

Our results prove the feasibility of the deposition of high-quality ZnO films at low temperatures. The low temperature ALD of as-grown ZnO on an AlN buffer layer demonstrated here has already a better crystalline quality than sputtering for deposition temperatures higher than 200 °C. In addition, if needed, the additional postannealing step can improve the crystalline quality even further. We have to note here that the quality of ZnO grown with high temperature processes like PLD or MOCVD was not achieved. However, because the crystallinity of the ZnO films after annealing was comparable to that of the AlN layer, it might have been limited by the quality of the AlN layer. Thus, further improvements might be possible.

The epitaxial in-plane orientation was studied by phi scans of the (101¯0) reflection. The azimuthal scan for the sample deposited at 300 °C with an additional postannealing step at 800 °C is shown in Fig. 4. The typical sixfold symmetry of wurtzite materials with six well defined peaks with 60° distance was observed for both AlN and ZnO. This confirms an in-plane oriented growth of the ZnO film. The epitaxial relationship of the ZnO film to the AlN buffer layer is ZnO[112¯0]//AlN[112¯0]. All six samples have a six-fold in-plane symmetry with an identical epitaxial relationship to the AlN layer (not shown).

FIG. 4.

Phi scan of the (101¯0) reflection of the ZnO film and the AlN buffer layer indicating the typical sixfold symmetry of wurtzite materials and in-plane oriented growth.

FIG. 4.

Phi scan of the (101¯0) reflection of the ZnO film and the AlN buffer layer indicating the typical sixfold symmetry of wurtzite materials and in-plane oriented growth.

Close modal

The optical properties of as-grown and postannealed ZnO films were characterized using room-temperature PL. Please note that, during the PL measurement of the substrate without the ZnO film, no luminescence of the AlN/Si substrate was observed in the measured wavelength range. The measured spectra of the ZnO films grown on the AlN/Si substrate are illustrated in Fig. 5. The excitation power density was 100 mW cm−2 for all measurements. The PL spectra of all ZnO films show the typical near-band-edge (NBE) emission peak of ZnO centered at around 388 nm and a broad emission in the visible range with two maxima at 500 nm and 680 nm due to different defect-related transitions. According to Leiter et al.,23 oxygen vacancies might be responsible for the green emission peaking at 500 nm, while the red emission is often associated with Zn vacancies in the crystal lattice.24 For the as-grown samples, the ZnO film deposited at 200 °C has the highest NBE emission intensity. The intensity decreases if the samples were deposited at 100 or 300 °C. A similar trend has also been observed by other groups, but a sufficient explanation is still lacking. The ZnO film grown at 100 °C has the largest contribution of defect-related emission, whereas the defect-related emission diminishes for the sample deposited at 200 °C. The additional postannealing step increases the NBE emission intensity for all samples significantly. While the postannealed samples deposited at 100 °C and 200 °C have a similar NBE emission intensity, the intensity of the postannealed sample grown at 300 °C is more than twice as high. The increase in NBE emission intensity after postannealing could be related to the decrease in nonradiative recombination due to the improvement of the crystalline quality and the coalescence of grains (see Fig. 1). The ratio of the NBE to the defect-related emission intensity is often cited as a figure of merit. However, this ratio depends on the excitation power density as interband transitions are more efficient than defect-related ones.25 Despite the low excitation power density of 100 mW cm−2, the intensity of the NBE-related emission is higher than the defect-related emission for all samples. This indicates the high optical quality of all samples.

FIG. 5.

Room-temperature PL spectra of as-grown and post-annealed ZnO films (please note the different intensity scale for the samples deposited at 300 °C).

FIG. 5.

Room-temperature PL spectra of as-grown and post-annealed ZnO films (please note the different intensity scale for the samples deposited at 300 °C).

Close modal

We successfully deposited highly textured ZnO films on Si(111) using an epitaxial AlN buffer layer by ALD. This method demonstrates the feasibility of growing ZnO films epitaxially on Si with an AlN buffer layer even at low temperatures. The crystalline quality determined by out-of-plane (0002) and in-plane (101¯0) ω-scans improves with increasing deposition temperature. An additional postannealing step at 800 °C improves the crystallinity further with an FWHM of the ω-scan of 0.75° for the (0002) and 1.24° for the (101¯0) reflection of the postannealed sample deposited at 300 °C. Phi scans confirmed the sixfold in-plane symmetry of all samples which is typical for wurtzite materials. These results prove an in-plane oriented growth of all ZnO thin films on the AlN buffer layer with an epitaxial relationship of ZnO[112¯0]//AlN[112¯0] even for the lowest deposition temperature of 100 °C. Room-temperature PL measurements highlight the influence of the deposition temperature and the additional postannealing treatment. The highest NBE emission intensity for the as-grown ZnO films was measured for the sample deposited at 200 °C. The NBE emission intensity increases further with the additional postannealing treatment and is highest for the sample deposited at 300 °C.

M.K. acknowledges funding by the Graduate School of Robotics, University of Freiburg; J.B. gratefully acknowledges support from the Deutsche Forschungsgemeinschaft in the framework of Major Research Instrumentation Program No. INST 272/266-1. The authors are grateful to the Fraunhofer IAF for providing the epitaxially grown AlN buffer layer.

The data that support the findings of this study are available from the corresponding author upon reasonable request.

1.
S.
Faÿ
,
J.
Steinhauser
,
S.
Nicolay
, and
C.
Ballif
,
Thin Solid Films
518
,
2961
(
2010
).
2.
J.-H.
Lim
,
C.-K.
Kang
,
K.-K.
Kim
,
I.-K.
Park
,
D.-K.
Hwang
, and
S.-J.
Park
,
Adv. Mater.
18
,
2720
(
2006
).
3.
D. M.
Bagnall
,
Y. F.
Chen
,
Z.
Zhu
,
T.
Yao
,
S.
Koyama
,
M. Y.
Shen
, and
T.
Goto
,
Appl. Phys. Lett.
70
,
2230
(
1997
).
4.
Y.
Liu
,
C. R.
Gorla
,
S.
Liang
,
N.
Emanetoglu
,
Y.
Lu
,
H.
Shen
, and
M.
Wraback
,
J. Electron. Mater.
29
,
69
(
2000
).
5.
N. W.
Emanetoglu
,
C.
Gorla
,
Y.
Liu
,
S.
Liang
, and
Y.
Lu
,
Mater. Sci. Semicond. Process.
2
,
247
(
1999
).
6.
C. R.
Gorla
,
N. W.
Emanetoglu
,
S.
Liang
,
W. E.
Mayo
,
Y.
Lu
,
M.
Wraback
, and
H.
Shen
,
Jpn. J. Appl. Phys.
85
,
2595
(
1999
).
7.
P.
Fons
,
K.
Iwata
,
S.
Niki
,
A.
Yamada
, and
K.
Matsubara
,
J. Cryst. Growth
201
,
627
(
1999
).
8.
S.
Yang
,
B. H.
Lin
,
W.-R.
Liu
,
J.-H.
Lin
,
C.-S.
Chang
,
C.-H.
Hsu
, and
W. F.
Hsieh
,
Cryst. Growth Des.
9
,
5184
(
2009
).
9.
R. D.
Vispute
 et al.,
Appl. Phys. Lett.
73
,
348
(
1998
).
10.
C.-W.
Lin
,
D.-J.
Ke
,
Y.-C.
Chao
,
L.
Chang
,
M.-H.
Liang
, and
Y.-T.
Ho
,
J. Cryst. Growth
298
,
472
(
2007
).
11.
A.
Nahhas
,
H. K.
Kim
, and
J.
Blachere
,
Appl. Phys. Lett.
78
,
1511
(
2001
).
12.
A.
Krost
and
A.
Dadgar
,
Mater. Sci. Eng. B
93
,
77
(
2002
).
13.
A.
Dadgar
 et al.,
J. Cryst. Growth
248
,
556
(
2003
).
14.
L.
Wang
,
Y.
Pu
,
Y. F.
Chen
,
C. L.
Mo
,
W. Q.
Fang
,
C. B.
Xiong
,
J. N.
Dai
, and
F. Y.
Jiang
,
J. Cryst. Growth
284
,
459
(
2005
).
15.
H.
Xiong
,
J. N.
Dai
,
X.
Hui
,
Y. Y.
Fang
,
W.
Tian
,
D. X.
Fu
,
C. Q.
Chen
,
M.
Li
, and
Y.
He
,
J. Alloy. Compd.
554
,
104
(
2013
).
16.
K.
Kondo
,
M.
Harada
, and
N.
Shibata
,
J. Ceram. Soc. Jpn.
110
,
343
(
2002
).
17.
L.
Li
,
L.
Sha
, and
Y.
Yuan
,
Acta Phys. Pol. A
137
,
17
(
2020
).
18.
S.
Rahmane
,
B.
Abdallah
,
A.
Soussou
,
E.
Gautron
,
P.-Y.
Jouan
,
L.
Le Brizoual
,
N.
Barreau
,
A.
Soltani
, and
M. A.
Djouadi
,
Phys. Status Solidi A
207
,
1604
(
2010
).
19.
J.-P.
Jung
,
J.-B.
Lee
,
J.-S.
Kim
, and
J.-S.
Park
,
Thin Solid Films
447
,
605
(
2004
).
20.
L.-Y.
Zhu
,
J.-G.
Yang
,
K.
Yuan
,
H.-Y.
Chen
,
T.
Wang
,
H.-P.
Ma
,
W.
Huang
,
H.-L.
Lu
, and
D. W.
Zhang
,
APL Mater.
6
,
121109
(
2018
).
21.
W.
Wang
,
C.
Chen
,
G.
Zhang
,
T.
Wang
,
H.
Wu
,
Y.
Liu
, and
C.
Liu
,
Nanoscale Res. Lett.
10
,
91
(
2015
).
22.
See supplementary material at http://dx.doi.org/10.1116/6.0000793 for the ω-scans of all samples.
23.
F. H.
Leiter
,
H. R.
Alves
,
A.
Hofstaetter
,
D. M.
Hofmann
, and
B. K.
Meyer
,
Phys. Status Solidi B
226
,
R4
(
2001
).
24.
K. E.
Knutsen
,
A.
Galeckas
,
A.
Zubiaga
,
F.
Tuomisto
,
G. C.
Farlow
,
B. G.
Svensson
, and
A. Y.
Kuznetsov
,
Phys. Rev. B
86
,
121203
(
2012
).
25.
H.
Beh
,
D.
Hiller
,
J.
Salava
,
F.
Trojánek
,
M.
Zacharias
,
P.
Malý
, and
J.
Valenta
,
J. Lumin.
201
,
85
(
2018
).

Supplementary Material