NAN O E X P R E S S Open Access
Effects of Al Doping on the Properties of
ZnO Thin Films Deposited by Atomic Layer
Deposition
Chen-Hui Zhai
1
, Rong-Jun Zhang
1*
, Xin Chen
2*
, Yu-Xiang Zheng
1
, Song-You Wang
1
, Juan Liu
3
, Ning Dai
2
and Liang-Yao Chen
1
Abstract
The tuning of structural, optical, and electrical properties of Al-doped ZnO films deposited by atomic layer deposition
technique is reported in this work. With the increasing Al doping level, the evolution from (002) to (100) diffraction
peaks indicates the change in growth mode of ZnO films. Spectroscopic ellipsometry has been applied to study the
thickness, optical constants, and band gap of AZO films. Due to the increasing carrier concentration after Al doping,
a blue shift of band gap and absorption edge can be observed, which can be interpreted by Burstein-Moss effect.
The carrier concentration and resistivity are found to vary significantly among different doping concentration, and
the optimum value is also discussed. The modulations and improvements of properties are important for Al-doped
ZnO films to apply as transparent conductor in various applications.
Keywords: Al-doped ZnO thin films, Atomic layer deposition, Optical properties, Spectroscopic ellipsometry, Electrical
properties
Background
A transparent conductiv e oxide (TCO) has received con-
siderable attentions and been widely used in electronic
and optoelectronic devices [1], such as solar cells [2],
liquid crystal [3], and high-d efinition displays [4], due to
their low resistivity and high transmitt ance. There are
various TCO materials, including In, Sb, Zn, Cd, Sn
metal oxides, and their composite oxides. Among them,
indium tin oxide (ITO) film is the most widely used
TCO material [5]. However, the scarce and toxic nature
of indium and instability of ITO are the main obstacles
for its further development, which arouses the interests
of researchers to explore alternative TCO materials for
ITO. A s a candidate for TCO, ZnO films doped with
trivalent metal cations have attracted considerable atten-
tions [6–9]. Thereinto, Al-doped ZnO (AZO) film is one
of the most promising candidates [10], since it has many
advantages, such as low cost, abundant resource, non-
toxicity, and good stability in hydrogen plasma. Impor-
tantly, the optical and electrical behaviors of AZO films
can be improved or modified by controlling their doping
level [11], which is critical to achieving functionalization
and tunability of TCO-based devices. Therefore, it is
useful to investigate the correlation between the proper-
ties of AZO films and the concentration of Al doping.
Various methods have been used to prepare AZO films,
including atomic layer deposition (ALD) [12, 13], chemical
vapor deposition (CVD) [14, 15], magnetron sputtering
[16, 17], and pulsed laser deposition (PLD) [18]. Compar -
ing to other techniques, ALD is an excellent deposition
technique based on self-limiting surface chemical reactions
that can be used to prepare highly uniform and smooth
films while their thickness can be precisely controlled.
Various properties of ALD-based AZO films have been
reported by many research groups [19–22]. Among
these properties, optical properties are generally studied
and analyzed based on transmission and photolumines -
cence spectra [21, 22]. However, there are few reports
on the properties evaluated by spectroscopic ellipsome-
try (SE) analysis [17]. SE known for its precision and
* Correspondence: rjzhang@fudan.edu.cn; xinchen@mail.sitp.ac.cn
1
Department of Optical Science and Engineering, Ministry of Education, Key
Laboratory of Micro and Nano Photonic Structures, Fudan University, 220
Handan Road, Shanghai 200433, China
2
National Laboratory for Infrared Physics, Shanghai Institute of Technical
Physics, Chinese Academy of Sciences, Shanghai 200083, China
Full list of author information is available at the end of the article
© 2016 The Author(s). Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made.
Zhai et al. Nanoscale Research Letters (2016) 11:407
DOI 10.1186/s11671-016-1625-0
non-destructiveness is a useful tool for the optical
characterization of nanostructures [23 , 24]. Thickness,
optical constants, and band gap energy information can
be determined accurately by using SE. With increasing
the Al doping level, the modulation of the optical pro-
perties provides references to the change of electrical
characteristics.
In this work, we investigated the structural, optical,
and electrical properties of AZO thin films deposited by
ALD technique with a wide range of doping levels. The
effects of concentration of Al doping on properties of
AZO thin films were discussed in detail. The thickness,
optical constants , and band gap of AZO samples were
calculated by fitting ellipsometry data in a broadband
spectral region. The modulated transmittance of AZO
thin films was shown by transmission spectrum mea-
surement. The blue shift of band gap and absorption
edge were observed and discussed in light of Burstein-
Moss effect. The electrical properties of the films were
measured by using a Hall effect measurement system.
The optimum concentration of Al for resistivity was
studied with the structural and optical properties keep-
ing excellent as well.
Methods
Both pure ZnO and Al-doped ZnO thin films were
prepared on the Si and quartz substrates throu gh a
custom-made ALD reactor. The deposition procedure
was at a temperature of 150 °C and a working pressure
of 80 Pa. Diethylzinc [DEZ; Zn(C
2
H
5
)
2
] and trimethyla-
luminum [TMA; Al(CH
3
)
3
] were used as the precursors
for Zn and Al, respectively, and deionized water (H
2
O)
was used as the oxidant reactant. High purity nitrogen
(N
2
) with a gas flow rate of 50 sccm (standard cubic cen-
timeters per minute) was used as the carrier to deliver
precursors into the chamber and purging gas to take the
needless products away from the chamber. During the de-
position process, the DEZ and H
2
O were alternatively in-
troduced into the chamber to grow the ZnO films
through DEZ-H
2
O cycles (DEZ/exposure/N
2
/H
2
O/expo-
sure/N
2
), with pulse times of 0.03/3/15/0.03/5/15 s. For
Al-doping into the ZnO films, TMA-H
2
OcyclesofAl
2
O
3
were introduced with the same process as the DEZ-H
2
O.
The ZnO and Al
2
O
3
monolayers grow through their
surface reactions [25]. The structure diagram of the inves-
tigated AZO film is shown in Fig. 1. The film deposition
consists of several super cycles while one super cycle
consists of one monolayer of Al
2
O
3
and n monolayers of
ZnO. Different Al content of AZO films was obtained by
varying the number of ZnO monolayers during one super
cycle. Hence, various numbers of DEZ-H
2
Ocyclesand
one TMA-H
2
O cycle were repeatedly carried out for film
depositions, as the cycle ratio given in Table 1. Same
number of ZnO monolayers (N2) was adopted for each
sample to fix the thickness of the main body of the films
which could avoid the thickness effects on the properties
of samples. According to N2 and the cycle ratio of
DEZ-H
2
O and TMA-H
2
O, the number of super cycle or
Al
2
O
3
monolayers (N1) and the total numbers of cycles
(N, N = N1 + N2) is also determined.
The chemical compositions and chemical bond states
of the thin films were characterized using X-ray photo-
electron spectroscopy (XPS). XPS analysis was carried
out through a scientific spectrometer (Ax is Ultra DLD)
with an Al KR X-ray source (1486.6 eV). The phase and
crystallinity of samples were mea sured by X-ray diffrac-
tion (XRD; Bruker D8 ADVANCE) with Cu Kα radiation
(λ = 1.5418 Å). The images of atomic force microscopy
(AFM) of samples were obtained by using a Bruker
Dimension Icon microscope VT-1000 System operated in
tapping mode. The spectroscopic ellipsometry measure-
ment was carried out by a vertical variable-angle SE
(V-VASE; J.A. Woollam Co., Inc.) in the wavelength range
of 200–1000 nm with a spectral resolution of 5 nm. The
incident angle was selected as 65, 70, and 75° to insure the
reliability of fitting results. Optical transmission spectra
were measured with a double beam UV-VIS-NIR spectro-
photometer (Shimadzu UV-3600) in the wavelength range
of 250–1000 nm. And the electrical properties of the films
were measured by using thevan der Pauw method with
a Hall effect measurement system (Ecopia HMS3000). All
these measurements were carried out at room temperature.
Fig. 1 The structure diagram of the AZO films
Table 1 Deposition parameters for samples, Al composition of
AZO films via XPS, and film thicknesses fitted by SE
Samples DEZ-H
2
O/TMA-H
2
O
cycle ratio
N1 N2 N at. %Al Thicknesses
(nm)
ZnO ––200 200 – 41.2
AZO 50:1 50:1 4 200 204 3.7 % 41.7
AZO 20:1 20:1 10 200 210 4.9 % 44.3
AZO 10:1 10:1 20 200 220 7.1 % 46.8
AZO 5:1 5:1 40 200 240 12.7 % 48.6
Zhai et al. Nanoscale Research Letters (2016) 11:407 Page 2 of 8
Results and Discussion
Composition and Structure Analysis
To verify the concentration of Al doping, the XPS mea-
surement was carried out on Al-doped ZnO films. Figure 2
reveals the XPS spectra of the AZO films after calibration
with carbon peak. In Fig. 2a, the high symmetry energy
peak of Zn 2p
3/2
is located at 1021.3 ± 0.1 eV, which is
approximately equal to the value of Zn in bulk ZnO [26].
It indicates that Zn in AZO films exists in the oxidized
states. From Fig. 2b, the main peak located at 530.0 ±
0.1 eV is assigned to lattice oxygen bonded as O
2−
ions in
the ZnO matrix [27]. Here, carbon composition has been
removed from those spectra, so no peak of carbon com-
position can be observed in Fig. 2b. As for Fig. 2c, the
energy peaks of Al 2p exhibit a symmetry feature, and is
located at around 73.5 ± 0.1 eV which is corresponding to
the characteristic peak of Al
2
O
3
[28]. The growth of peak
intensity indicates the increase of Al concentration in the
films. The concentration of Al is calculated from the ratio
of Zn, Al, and O atoms, as shown in Table 1. With varying
the cycle ratio of DEZ-H
2
O and TMA-H
2
Ofrom50:1to
5:1, the Al concentration increases from 3.7 to 12.7 %.
The X-ray diffraction patterns of samples grown on Si
are shown in Fig. 3. To obtain clearer phenomenon of
crystallinity, the samples were annealed at 400 °C in N
2
atmosphere for 1 h. The crystalline structure of the films
exhibits a hexagonal wurtzite stru cture with growth
directions of ZnO (100) or (002). No clear peaks of
Al
2
O
3
manifests the deposited Al
2
O
3
is amorphous
under the growth conditions in this work. The crystal-
line state and crystal orientation of these films are found
to change with the increasing doping concentration of
Al
3+
. The pure ZnO film show s a preferred growth with
ZnO (002) direction, and the peak position loca tes at
about 34.5°. Howe ver, the (002) peak disappears and new
ZnO (100) peak becomes dominant after Al doping. The
evolution of ZnO peaks demonstrates that Al-doping
affects the growth mode of ZnO films, which is similar
to the results reported by Banerjee et al. [29]. With
further increasing Al doping, no peak is observed for
AZO 5:1 sample, which indicates the AZO film at 12.7 at.%
Al is amorphous. Since the Al
2
O
3
layer by ALD is amor-
phous in our growth conditions, the Al
2
O
3
doping layers
Fig. 2 XPS spectra of AZO films grown on Si substrates: (a)Zn2p
peaks; (b) O 1s peaks; (c) Al 2p peaks
Fig. 3 XRD patterns of samples grown on Si substrates after annealing
Zhai et al. Nanoscale Research Letters (2016) 11:407 Page 3 of 8
destroy the crystal quality of the AZO 5:1 films, which
causesthedisappearanceofits(100)peak.FromFig.3,it
can also be observed that the ZnO (100) peak shifts to a
larger diffraction angle with the increasing Al
3+
content.
Here, the Al
3+
ions (ion radius 0.53 Å) is smaller than Zn
2+
ions (ion radius 0.74 Å), so the increasing Al concentration
will reduce the lattice constant of samples by substitutions
of Zn
2+
ions with Al
3+
[30].
The surface morphologies of the samples grown on Si
substrates are analyzed by using the AFM method with a
scanning area of 5 × 5 μm
2
. The 3D AFM images of the
samples are shown in Fig. 4. All the films show good
uniformity over the whole scanning area. The hill-shaped
features with average lateral dimensions 50–100 nm can
be observed on the surface of samples. The root-mean-
square roughness (R
q
) of each samples calculated from
AFM data is less than 1.00 nm, as given unde r their
AFM images, respectively. It is clear that all samples
present smooth surface with inessential surface roughness.
Therefore, the surface scattering is weak enough for SE
analysis, which makes the analysis more reliable [31].
Optical Properties of Samples
Spectroscopic ellipsometry is generally applied for the
investigation of the thickness, optical properties of
samples, which is based on measuring the change in the
polarization state of a linearly polarized light reflected
from the sample surface. In order to get more accurate
information, Si is selected as the substrate to provide
enough reflected light in SE mea surement. For SE
analysis, an optical model of our samp les is firstly con-
structed, which consists of a semi-infinite Si substrate/
pure ZnO or AZO film/air ambient structure. No rough-
ness layer is introduced, due to the smooth surfa ce of
samples revealed by AFM measurements. The obtained
ellipsometry spectra (Ψ and Δ at range of 200–1000 nm)
of the films are fitted by using the Forouhi-Bloomer
dispersion model (Additional file 1)[32]. This dispersion
Fig. 4 AFM 3D images of pure ZnO and AZO thin films grown on Si substrates
Zhai et al. Nanoscale Research Letters (2016) 11:407 Page 4 of 8
model is widely used to describe the optical properties
in the spectral region which is dominated by inter-band
transitions and contains the inform ation of the band gap
[32]. The thickness, optical constants, and band gap of
the films are evaluated in a fitting procedure by minim-
izing the root-mean-square error (RMSE) which defined
as follows [33]:
RMSE ¼
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1
2x−y−1
X
i¼1
x
Ψ
cal
i
−Ψ
exp
i
2
þ Δ
cal
i
−Δ
exp
i
2
hi
s
ð1Þ
Here, x is the number of data points in the spectrums,
y is the number of variable parameters in the model, and
“exp” and “cal” represent the experimental and the
calculated data , respe ctively.
The fitted thickness of samples is shown in Table 1.
With the increasing concentration of Al, the thickness
of the film displays a growing trend due to the increased
total numbers of cycles. Figure 5 illustrates the refractive
index n and extinction coeffici ent k of the films wi th
various doping levels at t he wavelength range of 200–
1000 nm. The calculated o ptical constant s of the pure
ZnO films obtained by ALD are in good agreement with
our previous work [34]. Due to the Al doping is not deep
enough, optical constants of AZO 50:1 is very close to
pure ZnO. Figure 5a demonstrates the refractive index n
of the films with various doping levels. Obviously, the
refractive index of films decreases gradually with doping
level increasing, since Al impurity can act as effective
n-type donors to generate free carriers. The doping of
Al
2
O
3
increases the free carrier concentration in films,
which results in the decrease of the refractive index of the
films [35]. So, the refractive index can be modulated by Al
doping level. Figure 5b describes the extinction coefficient
k of the films with various doping levels. It can be seen
that the k of all films is close to zero infinitely in the wave-
length range of 400–1000 nm, which indicates that the
films are nearly transparent in this wavelength region.
Besides,ablueshiftoftheabsorptionedgecanbe
obser ved with the increasing doping level.
In order to better understand the blue shift of absorp-
tion edge, Tauc method is used to calculate the band gap
of the sample by using formulas as follows [35]:
αhνðÞ
2
¼ AE−E
g
ð2Þ
α ¼
4πk
λ
ð3Þ
where A is a constant, E
g
is the optical band gap energy,
α is the optical absorption coe fficient which can be
calculated from extinction coefficient (k) and wavelength
of SE results. In order to simplify the calculation, Tauc
extrapolation is utilized to obtain the band gap energy of
samples. As shown in Fig. 6, we made a plot of (αhν)
2
versus the photon energy. The optical band gap can be
determined by the linear fitting. Through extrapolation,
the point value of fitted line and x axis is E
g
, as revealed
in the inset figure of Fig. 6. The band gap of the ZnO
film is 3.3 eV, which agrees well with the ideal band gap
of pure ZnO. And the E
g
displays a growing trend with
the increasing concent ration of Al. The tendency is simi-
lar to that of other elements doped ZnO and In
2
O
3
:Sn
films (TCO materials) [36]. It can be interpreted by the
Burstein-Moss effect [36, 37]. ZnO is a n type semicon-
ductor material with direct transition, and its Fermi level
will enter into the conduction band when it is heavily
doped. The state below Fermi level is occupied by elec-
trons. The absorption transition process of light can only
exist between the valence band and the vicinity to Fermi
level. It results the optical band gap of films moves to
the high energy region. Moreover, the Burstein-Moss
effect is related to the carrier density. Extrinsic Al
3+
are
substituted for Zn
2+
in the AZO films, so the spare elec-
trons from Al
3+
can increase the concentration of free
carriers in films, resulting in the growth of optical band
Fig. 5 The optical constants of samples grown on Si substrates: a the refractive index n. b The extinction coefficient k
Zhai et al. Nanoscale Research Letters (2016) 11:407 Page 5 of 8