Materials Science in Semiconductor Processing 89 (2019) 149–153

Contents lists available at ScienceDirect

Materials Science in Semiconductor Processing

journal homepage: www.elsevier.com/locate/mssp

Study of defects in Li-doped ZnO thin films T

a a,⁎ b b
M. Hjiri , M.S. Aida , O.M. Lemine , L. El Mir
a
Department of Physics, College of Sciences, King Abdulaziz University, 21589 Jeddah, Saudi Arabia
b
Al Imam Mohammad Ibn Saud Islamic University (IMSIU), College of Sciences, Department of Physics, Riyadh 11623, Saudi Arabia

A R T I C LE I N FO A B S T R A C T

Keywords: Li-doped ZnO films were deposited on glass substrate by sol-gel spin coating technique. In order to investigate
ZnO the effect of Li doping on ZnO thin films properties, we have varied the Li doping ratio in the range of 0–10%.
Thin films The structural and optical properties of samples have been investigated by X-ray diffraction and UV–Vis spec-
Sol-gel troscopy. XRD analysis reveals that doped and undoped films have an hexagonal Wurtzite structure. The optical
Photoluminescence
measurements indicate that Li doping causes the red shift of the absorption edge. Films photoluminescence
spectra do not contain any defects peak related to Li atoms. Being too close to the valence band, the Li defect
level is buried in the band tail width states; this explains the difficulty of its detection. However, the effect of Li
doping, manifests through its oxygen vacancy defects promotion to the detriment of oxygen interstitial oxygen
defect.

1. Introduction produce p type is still controversial. Recently, Ardyanian et al. [21]

Zinc oxide thin films have attracted an increasing interest due to
their excellent semiconducting properties [1–3]: a direct wide band gap
of 3.37 eV and large excitation binding energy (60 meV) at room tem-
perature [4]. ZnO has received a considerable attention as a promising
material for electronics, optics and optoelectronics devices fabrication.
Some of these applications involve the photoluminescence and elec-
troluminescence properties of ZnO. Therefore, a large number of in-
vestigations have been devoted to the study of ZnO photoluminescence
[5]. The visible emission of ZnO is close related to the intrinsic defects
such as oxygen vacancies, zinc vacancies or oxygen interstitial and
doping atoms. Several emissions in ZnO have been observed green [6],
red-orange [7] and yellow [8].
Doping ZnO with atom causing the formation of shallow level is
among routes to enhance the optical emission of ZnO [9]. Therefore, an
effort devoted to ZnO doping has been carried out [10–12]. Introducing
extrinsic doping atoms can alter the emission wavelength spectra of
ZnO films. Therefore, several impurity atoms effect has been in-
vestigated such as Li, Al [13,14], Mg [15] and Eu [16]. Beside the ex-
trinsic defects, the effect of native ones on the photoluminescence
properties of ZnO structures has been extensively studied in the lit-
erature [17–19].
The most promising doping atoms for p-type ZnO production are the
I and V elements group, while, theory suggests some difficulties in
achieving shallow acceptor level [20]. Therefore, Li doping of ZnO to
indicated that the formation of deep acceptor levels implies not to
contribute significantly to p-type conduction. Despite that ZnO photo-
luminescence is extensively studied, Li doping effect on the ZnO pho-
toluminescence is less investigated. Introduction of doping atoms in the
semiconductor network is usually accompanied by the creation of de-
fect levels within the material band gap. The present study aims the
investigation of the defects related to Li doping atoms using photo-
luminescence technique.
In the present work, we have prepared Li-doped ZnO thin films with
different Li doping ratio by sol-gel technique. The effect of Li doping
ratio on ZnO films structural, optical and photoluminescence properties
are investigated.
2. Experimental details
2.1. Synthesis of pure and Li-doped ZnO thin films
Sol-gel spin coating technique was utilized to prepare Li-doped ZnO
thin films. The solution was prepared using zinc acetate dehydrate (Zn
(CH3COO)2·2H2O) as Zn precursor, lithium-chloride (LiCl3) as a dopant
source, mono-ethanolamine (MEA, C2H7N14) as a stabilizer and 2-
methoxyethanol as a solvent. The molar ratio of MEA to zinc acetate
was fixed to 1 and the concentration of lithium-chloride was varied
from 2 to 10 at%. The solution was stirred at 60 °C for 2 h before aged at
room temperature for 24 h to yield clear and homogeneous solution.

⁎
Corresponding author.
E-mail address: meida@kau.edu.sa (M.S. Aida).

https://doi.org/10.1016/j.mssp.2018.09.010
Received 28 June 2018; Received in revised form 13 August 2018; Accepted 13 September 2018
Available online 20 September 2018
1369-8001/ © 2018 Elsevier Ltd. All rights reserved.
M. Hjiri et al. Materials Science in Semiconductor Processing 89 (2019) 149–153

The solution was then used for the deposition process by spin-coating
technique. The used substrates were microscope glass. The spin coating
was performed at room temperature using 10 drops of solution per
deposition with a speed of 2000 rpm for 30 s and 3000 rpm for 10 s.
After the deposition process, the thin films were preheated in atmo-
sphere ambient at 300 °C for 10 min to evaporate the solvent. After
repeating the coating procedure 7 times, the films were annealed at
500 °C for 1 h in air ambient using a hot plate.
Sample code are named as L0ZO, L2ZO, L3ZO, L4ZO, L5ZO and
L10ZO, according the nominal Li loading of each sample.

2.2. Characterizations

The crystalline phases of the obtained thin films were identified by

X-ray diffraction (XRD) using a Bruker D5005 powder X-ray dif-
Fig. 2. Variation of the intensity and the position of the diffraction peak (002).
fractometer using a CoKα source (1.78901 Å radiation). The average
crystallite size (G, in Å) was estimated from the Scherer's equation [22]:
0.9λ
G=
B cos θB (1)

where λ is the X-ray wavelength (1.78901 Å), θB is the maximum of the

Bragg diffraction peak (in radians) and B is the line width at half
maximum.
The optical transmittance of the powders was determined using a
Schimadzu UV-3101 PC spectrophotometer with integrating sphere in
the wavelength range 200–900 nm.
PL measurements were performed under excitation of the 325 nm
HeCd laser line.

3. Results and discussion

3.1. Structural properties

Fig. 1 shows the XRD diffraction patterns of undoped and Li doped
ZnO thin films. As can be deduced from these patterns, films have a
hexagonal Wurtzite structure with a strong orientation along the (002)
plan. Regardless the deposition technique, this preferential orientation
have been reported by several authors working on ZnO. This is due to
the fact that the plan (002) requires less energy to be formed [23].
In Fig. 2 we have reported the variation of (002) peak intensity and
position in the same plot. At low doping range from 0% to 3%, the peak
intensity increases with Li concentration, this is consistent with crys-
tallite size increase as shown in Fig. 3. This can be attributed to the
modification of the nanocrystalline growth as suggested by Chu et al.
[24]. It is well known, in thin film growth process, that any foreign
atom such as doping atom acts as a new nucleation center; this may
explain the increase in the peak intensity and crystallite size with in-
creasing the doping ratio until 3%. However, above 3%, due to more Li
Fig. 1. X-ray diffraction patterns of pure and Li doped ZnO thin films.
Fig. 3. Variation of average crystallite size as a function of Li concentration.
atoms incorporation in ZnO network, further increase in the nucleation
center hinders the continuous crystallite growth [25]. Consequently,
the crystallite size enlargement is stopped due to the completion be-
tween adjacent growing nanocrystals. This is consistent with the re-
duction of both peak intensity (Fig. 2) crystallite size as a function of Li
doping (Fig. 3). According to these results, 3% doping level seems to be
a critical one, where Li atoms start altering the host network beyond
this ratio.
As shown in Fig. 2, the (002) peak position remains unchanged at
low doping ratio (less than 3%), while above this ratio, the peak slightly
shifts gradually towards higher angles with increasing the doping ratio.
The peak shift has been reported by several authors working on Li
doped ZnO [21,26,27]. The shift towards higher angle can be attributed
to stress increase with Li incorporation in film network as suggested by
[26]. However, Ardyanian et al. [21] has reported an alternating shift
to higher and lower angles, they implement this behavior to the com-
pletion between tensile and compressive stress in the lattice, due to the
Li substitution at the interstitial and Zn sites respectively. In our case
the observed shift towards the higher angles, as shown in Fig. 2, sug-
gests that Li is mainly incorporated in the interstitial sites.
In Fig. 4 we have regrouped the transmittance spectra of different Li
doped ZnO thin films. All films exhibit a large transmittance in the
visible range, mostly in the order of 90%. As can be seen, Li doping do
not alter the transmittance level in the thin films.
The optical band gap is estimated by using Tauc plot [28]. It is
based on the determination of the intercept of the linear extrapolation
of (αhν)2 as a function of photon energy plot with the X- axis. The
variation of the optical band gap as a function of Li doping is reported
in Fig. 5. In general, Li doping causes the optical gap shrinking. It varies
from 3.3 eV in the undoped film to 3.27 eV in 3% Li doped one. Salah

150
M. Hjiri et al.
Fig. 4. Optical transmission spectra of LZO thin films.
Fig. 5. Optical band gap of LZO thin films.
et al. [29] have noticed the slight decrease in the optical gap after Li
introduction in ZnO. Even at high Li doping, Ardyanian et al. [21] have
observed a band gap narrowing about 0.02 eV by Li incorporation. This
is close to the 0.03 eV variation in our case.
The band gap reduction can be attributed to the increase in the
network disorder due to Li introduction in the film network as shown in
Fig. 6. The disorder, known as the deviation from the standard bond
length and angle in the crystal create continuous defects near the band
edges. The disorder is known also as band tail width or Urbach energy.
The Urbach energy can be easily estimated from the slope of the plot of
ln (αhν) as a function of photon energy using the following formula
[30]:
Fig. 6. Disorder in thin films with different Li loading.
151
Materials Science in Semiconductor Processing 89 (2019) 149–153
Fig. 7. PL spectra of pure and doped ZnO thin films.
α = α 0exp (h ν/ Eu ) (2)
where α and Eu are respectively the absorption coefficient and Urbach
energy.
In Fig. 6, we have plotted the variation of the disorder in films with
Li doping ratio. The disorder increase and material cristallinity de-
gradation usually accompanies the material doping as reported by
several authors [31–33].
Fig. 7 presents the PL spectra of undoped and Li doped ZnO thin
films. As seen in Fig. 7.a, the spectrum exhibits two peaks: a low near
band edge (NBE) located in the UV region and a dominant and broad
peak in the visible region. The visible peak originating from the emis-
sion of different centers located in the forbidden band gap region of
ZnO are reported in Fig. 7.b. The NBE peak remains almost constant
with Li doping. It varies slightly with Li doping in contrary to the green
peak increases with Li doping. However, as shown in Fig. 7.b, the po-
sition and intensity of peaks related to the defects are sensitive to Li
doping, this indicates that Li doping alters the defect in forbidden gap of
ZnO films.
The peak position varies with Li doping as shown in Fig. 8. The
green peak is originated from levels in the band gap assigned to the
defects mainly related to the oxygen in film network. This indicates that
Li doping may later the defects in ZnO material. Similarly, to the
structural analysis conclusion and according to the peak position and
intensity variation with Li doping, 3% ratio seems to be a critical one.
As can be seen, the film prepared with 3% of Li doping exhibits the
higher intensity accompanied by a red shift by comparison to the other
samples. Wang et al. [34] have also remarked a red shift of the visible
band as Li atoms is incorporated in ZnO.
In order to identify defects and their related level position in the
bang gap, we have remedied to the deconvolution of the broad green
band. A typical deconvolutions result shown in Fig. 9 for different Li
doped films are used as example. The deconvolution reveals that the PL
M. Hjiri et al. Materials Science in Semiconductor Processing 89 (2019) 149–153

Fig. 8. Variation of peaks position and peaks intensity with Li loading.

Fig. 10. Variation of the two defects Vo and Oi as a function of Li concentration.

broad peak is mainly composed with two peaks: L1 located at 2.06 eV
and L2 at 2.20 eV. Both are related to the oxygen defects, the former is
network. Since, the disorder value varies from 80 to 130 meV (Fig. 6),
assigned to the oxygen interstitial Oi, it is due to the transition starting
the position of Li defect is then located in the band tail width. This may
from Zni level to Oi level. The latter is assigned to the oxygen vacancy
explains the difficulty of Li related defect observation climbed by sev-
Vo, it is originated from the transition from conduction band to Vo level.
eral authors [29]. This is consistent with the theory prediction of some
At high doping level (10%), the PL spectrum is composed only with the
difficulties in achieving shallow acceptor level in ZnO when doped with
L2 level assigned to oxygen interstitial defect.
group-I elements such as Li [35–37], Na, K, Cu [37] or Ag [38]. While,
No defects related directly to Li were observed. Indeed, theoretical
the created deep acceptors levels do not significantly contribute to p-
and experimental studies have shown that Li acceptor level, LiZn possess
type conduction. However, it is believed that the most promising do-
a shallow level energy as small as 90 meV relative to the maximum of
pants for p-type ZnO are the group-V elements such as N, P, and As
the valence band. Being too close to valence band, the Li related defect
[39].
could be imbedded in the band tail width due to the disorder in the film
In Fig. 10, we have shown the variation of the two defects Vo and Oi

Fig. 9. A typical deconvolution results for different doped samples.

152
M. Hjiri et al.
as a function of Li. The introduction of Li has a significant effect on
these defects concentration. With increasing Li doping ratio, the oxygen
vacancy increases to the detriment of interstitial oxygen. At low Li
ratio, Oi defect concentration is larger than Vo defect. However, at high
Li doping level the situation is reversed the Vo becomes larger. This is
consistent with the results of several authors [40–42]. Where, it is
outlined that Li indirectly alters the PL spectra through its influence on
oxygen defect. Wang et al. [43] suggested that Li cannot cause any
emission in ZnO, but serves as defect mediator by changing the relative
concentration of the intrinsic defect.
The same conclusion has been outlined by Rakkech et al. [44]. In
their study of Li doped ZnO they concluded that Li incorporation in-
duces oxygen enrichment at ZnO film surface. This is consistent with
our result indicating the increase in oxygen vacancy due its migration
towards the surface.
To maintain the charge balance of the lattice, Vo and Oi defects
could be created through the following the reactions [34]:
Li2 O → 2Li+Oi
Li2 O→LiZn + Vo
Thereafter, according to the above relations, Li in substitutional
sites is accompanied with the oxygen vacancies. While, Li in interstitial
sites yields to oxygen interstices. It is worth noting that at low Li doping
ratio the most probable occupying sites of Li is the interstitial ones,
hence Oi defect formation is more favorable as shown in Eq. (3). While,
with increasing the Li doping the probability of finding Li in substitu-
tional sites becomes larger. Therefore, more oxygen vacancy defects Vo
are created according to the relation (4).
4. Conclusion
In conclusion, Li doped ZnO thin films have been successfully pre-
pared by spin coating technique. The influence of Li doping ratio on the
structural, optical films properties is investigated. The results indicated
that Li atom alters the ZnO film properties. The ratio 3% seems critical.
Bellow this ratio, the crystalline structure is improved; this is suggested
by the increase in the diffraction peak intensity. Above this value the
films cristallinity is deteriorated. However, the optical gap slightly
varies with the Li doping, a small red shift with increasing Li doping is
noticed. The most interesting feature concluded from PL analysis is the
absence of any related defects to Li. This might be due to the burring of
the shallow defect created by Li atoms in the band tail width caused by
the disorder in the film network. In addition, Li atom doping promotes
the oxygen related defects, at low Li doping the interstitial oxygen
defect are relatively larger. However, with increasing Li doping above
3% the oxygen vacancy defect become dominant.
References
[1] X.L. Xu, S.P. Lau, J.S. Chen, Z. Sun, B.K. Tay, J.W. Chai, Mater. Sci. Semicond.
Materials Science in Semiconductor Processing 89 (2019) 149–153
Process. 4 (2001) 617.
[2] U. Ozgur, Y.I. Alivov, C. Liu, A. Teke, M.A. Reshchikov, S. Dogan, V. Avrutin, S.-
J. Cho, H. Morkoç, J. Appl. Phys. 98 (2005) 041301.
[3] Z. Yang, S. Chu, W.V. Chen, L. Li, J. Kong, J. Ren, P.K.L. Yu, J. Liu, Appl. Phys.
Express 3 (2010) 032101.
[4] J.P. Kim, S.A. Lee, J.S. Bae, S.K. Park, U.C. Choi, C.R. Cho, Thin Solid Films 516
(2008) 5223.
[5] K. Petruf, O. Panak, J. Valenta, P. Neme, Ceram. Intern. 43 (2017) 5223.
[6] S.S. Kurbanov, G.N. Panin, T.W. Kim, T.W. Kang, J. Lumin. 129 (2009) 1099.
[7] X. Xiu, X. Wu, H. Cao, R.P.H. Chang, J. Appl. Phys. 95 (2004) 3147.
[8] Y.W. Heo, D.P. Norton, S.J. Pearton, J. Appl. Phys. 98 (2005) 073502.
[9] J.C. Johnson, H. Yan, R.D. Schaller, L.H. Haber, R.J. Saykally, P. Yang, J. Phys.
Chem. B 105 (2001) 11388.
[10] D.C. Look, D.C. Reynolds, C.W. Litton, R.L. Jones, D.B. Eason, G. Cantwell, Appl.
Phys. Lett. 81 (1830) (2002) 9.
[11] P. Wang, N. Chena, Z. Yin, F. Yang, C. Peng, J. Cryst. Growth 290 (56) (2006) 10.
[12] S.Y. Bae, C.W. Na, J.H. Kang, J. Park, J. Phys. Chem. B 109 (2005) 2526.
[13] A. Mondal, S. Pala, A. Sarkar, T.S. Bhattacharya, A. Das, N. Gogurla, S.K. Ray,
P. Kumar, D. Kanjilal, K.D. Devi, A. Singha, S. Chattopadhyay, D. Jana, Mater. Sci.
Semicond. Process. 80 (2018) 111.
[14] N. Ohashi, T. Nakata, T. Sekiguchi, H. Hosono, M. Mizuguchi, T. Tsurumi,
J. Tanaka, H. Haneda, Jpn. J. Appl. Phys. 38 (1999) L113.
[15] S. Fujihara, Y. Ogawa, A. Kasai, Chem. Mater. 16 (2004) 2965.
(3) [16] Y.-K. Park, J.-I. Han, M.-G. Kwak, H. Yang, W.-S. Cho, J. Lumin. 78 (1998) 87.
[17] Q. Wang, J. Zhang, J. Lumin. 195 (2018) 371.
(4) [18] S.A. Studenikin, N. Golego, M. Cocivera, J. Appl. Phys. 84 (1998) 2287.
[19] Z.K. Tang, G.K.L. Wong, P. Yu, M. Kawasaki, A. Ohtomo, H. Koinuma, Y. Segawa,
Appl. Phys. Lett. 72 (1998) 3270.
[20] T. Chiba, Y.J. Pu, M. Hirasawa, A. Masuhara, H. Sasabe, J. Kido, ACS Appl. Mater.
Interfaces 4 (2012) 6104.
[21] M. Ardyanian, N. Sedigh, Bull. Mater. Sci. 37 (2014) 1309.
[22] B.D. Cullity, Elements of X-ray Diffraction 102 Addison-Wesley, Reading, MA, 1978.
[23] M.H. Aslan, A.Y. Oral, E. Menşur, A. Gül, E. Başaran, Sol. Energy Mater. Sol. Cells
82 (2004) 543.
[24] H. Chu, L. Wei, R. Cui, J. Wang, Y. Li, Coord. Chem. Rev. 254 (2010) 1117.
[25] V. Bilgin, J. Electron. Mater. 38 (2009) 715.
[26] M. Salah, S. Azizi, A. Boukhachem, C. Khaldi, M. Amlouk, Int. J. Contr. Energy
Electr. Engine 3 (2016) 62.
[27] R.A. Rakkesh, S. Balkumar, Process Appl. Ceram. 8 (2014) 7.
[28] J.J. Tauc, Amorphous and Liquid Semiconductors, Plenum Press, New York, 1974.
[29] M. Salah, S. Azizi, A. Boukhachem, C. Khaldi, M. Amlouk, J. Lamloumi, J. Mater.
Sci. 52 (2017) 10439.
[30] F. Urbach, Phys. Rev. 92 (1953) 1324.
[31] T. Jannanea, M. Manoua, A. Liba, N. Fazouan, A. El Hichou, A. Almaggoussi,
A. Outzourhit, M. Chaik, J. Mater. Environ. Sci. 8 (1) (2017) 160.
[32] S.M. Salaken, E. Farzana, J. Podder, J. Semicond. 34 (2013) 07300–07301.
[33] S.N.S. Lekshmy, V.S.N. Anitha, P.K.V. Thomas, K. Joy, J. Am. Ceram. Soc. 97 (2014)
3184.
[34] M. Wang, E.J. Kim, S.H. Hahn, J. Lumin. 131 (2011) 01428.
[35] O.F. Schirmer, J. Phys. Chem. Solids 29 (1968) 1407.
[36] A. Valentini, F. Quaranta, M. Rossi, G. Battaglin, J. Vac. Sci. Technol. A 9 (1991)
286.
[37] Y. Kanai, Jpn. J. Appl. Phys. 30 (1991) 703.
[38] Y. Kanai, Jpn. J. Appl. Phys. 30 (1991) 2021.
[39] C.H. Park, S.B. Zhang, S.-H. Wei, Phys. Rev. B 66 (2002) 073202.
[40] M. Willander, O. Nur, J.R. Sadaf, M.I. Qadir, S. Zaman, A. Zainelabdin, N. Bano,
I. Hussain, Materials 3 (2010) 2643.
[41] B. Abdallah, A.K. Jazmati, R. Refaai, Mat. Res. 20 (2017) 1516.
[42] A. Janotti, C.G. Van de Walle, Rep. Prog. Phys. 72 (2009) 126501.
[43] M. Wang, E.J. Kim, S.H. Hahan, J. Lumin. 131 (2011) 1428–1433.
[44] R.A. Rakkech, S. Balakumae, Process. Appl. Ceram. 8 (2014) 7.
153