Home
Search
Collections
Journals
About
Contact us
My IOPscience
Stress-dependent band gap shift and quenching of defects in Al-doped ZnO films
This article has been downloaded from IOPscience. Please scroll down to see the full text article.
2010 J. Phys. D: Appl. Phys. 43 465402
(http://iopscience.iop.org/0022-3727/43/46/465402)
View the table of contents for this issue, or go to the journal homepage for more
Download details:
IP Address: 124.124.247.140
The article was downloaded on 08/11/2010 at 12:52
Please note that terms and conditions apply.
IOP PUBLISHING
JOURNAL OF PHYSICS D: APPLIED PHYSICS
J. Phys. D: Appl. Phys. 43 (2010) 465402 (6pp)
doi:10.1088/0022-3727/43/46/465402
Stress-dependent band gap shift and
quenching of defects in Al-doped ZnO
films
Bhupendra K Sharma and Neeraj Khare
Department of Physics, IIT Delhi, New Dehi-110016, India
E-mail: [email protected] (N Khare)
Received 17 May 2010, in final form 3 October 2010
Published 4 November 2010
Online at stacks.iop.org/JPhysD/43/465402
Abstract
Al-doped ZnO (AZO) films were deposited on quartz substrates by the ultrasonically assisted
chemical vapour deposition technique. The undoped ZnO film was found to be subjected to a
stress which increases initially up to 3% Al doping, and then a slight decrease was observed
for 5% Al doping. The band gap of AZO shows a blue shift up to 3% of Al doping as
compared with the undoped ZnO. The blue shift in the band gap of the AZO films cannot be
understood in the framework of Burstein–Moss shift and has been attributed to an increase in
the stress present in the film. The photoluminescence spectrum of the undoped ZnO film
shows a wide peak in the visible region which is suppressed with a small red shift after Al
doping in the ZnO film. A detailed analysis of photoluminescence of ZnO and AZO films
indicates suppression of zinc interstitials (Zni ) and oxygen vacancies (VO ) and creation of
oxygen interstitial (Oi ) defects after Al doping in ZnO films. X-ray photoelectron
spectroscopy study also reveals suppression of oxygen vacancy-related defects after Al doping
in the ZnO film. The presence of Al in the ZnO matrix seems to change the defect equilibria
leading to a suppression of Zni and VO and enhancement of Oi defects. The suppression of Zni
defects is correlated with the increase in stress in Al-doped ZnO films.
(Some figures in this article are in colour only in the electronic version)
and molecular beam epitaxy [9]. There are reports on the
observation of blue shift in the band gap of Al-doped ZnO
(AZO) films, which has been attributed to the Burstein–Moss
(BM) shift [10–12]. The doping of Al in ZnO increases the
carrier concentration, n, which starts the filling of states at the
bottom of the conduction band leading to a broadening of the
band gap, called the BM effect [13].
Many studies have reported the change in stress
with deposition and post-heat treatment conditions and its
correlation with the optical band gap in ZnO films [14, 15]. In
ZnO, the presence of defects influences the photoluminescence
(PL) spectra [16]. It is expected that doping of extrinsic dopant
Al may influence the defect environment and stress present in
the film. Although in most of the studies related to Al-doped
ZnO films, the observed blue shift in the band gap is attributed
to the BM shift, there is a possibility that a change in stress
present in the films after Al doping can also cause the blue
shift in the band gap of Al-doped ZnO films. Therefore, it
1. Introduction
Zinc oxide (ZnO), a direct wide band gap (∼3.2 eV)
semiconductor with a large exciton binding energy (∼60 meV)
at room temperature, has attracted considerable attention in
recent years due to its potential for applications in light emitting
diodes,solar cells, transducers, varistors and photodetectors
[1, 2]. As-prepared ZnO naturally shows n-type conductivity
due to the presence of zinc interstitials (Zni ) and oxygen
vacancies (VO ). N-type conductivity in ZnO can also be
achieved by doping with group III elements (e.g. B, Al, Ga
and In) [2]. Among these dopants, Al has been found to be an
efficient n-type dopant for realizing high-quality samples with
strong ultraviolet/blue light emission and high transparency to
visible light [3, 4].
Al-doped ZnO (AZO) thin films have been prepared by
various techniques such as sol–gel [5], magnetron sputtering
[6], chemical vapour deposition [7], pulsed laser deposition [8]
0022-3727/10/465402+06$30.00
1
© 2010 IOP Publishing Ltd
Printed in the UK & the USA
J. Phys. D: Appl. Phys. 43 (2010) 465402
B K Sharma and N Khare
would be interesting to study the influence of Al doping on
stress and optical properties and find a correlation between the
stress, optical band gap and defects present in the film.
In this study, we have investigated the effect of Al doping
on the stress, blue shift in the band gap and the intrinsic defects
present in the ZnO film. The undoped ZnO film itself is
found to be under stress, which increases with Al doping and
results in a blue shift in the optical band gap. Suppression of
defects, which were initially present in pure ZnO, has also been
observed which seems to have a correlation with the increase
in stress.
2. Experimental details
ZnO and Al-doped ZnO films were deposited on quartz
substrates by the ultrasonically assisted chemical vapour
deposition technique. The precursor solutions used for
depositing ZnO and Al-doped ZnO (AZO) films were prepared
by dissolving zinc acetate dihydrate [Zn(CH3 COO)2 · 2H2 O]
and aluminium nitrate [Al(NO3 )3 · 9H2 O] in deionized water
in an appropriate ratio. Before depositing ZnO and AZO films,
the substrates were cleaned with acetone, deionized water and
finally with propanol in order as described using an ultrasonic
bath. A 0.1M aqueous solution of zinc acetate dihydrate in
deionized water was used as a spraying solution for depositing
the ZnO film. Pyrolysis of an ultrasonically created mist took
place at 450 ◦ C on quartz substrates. The deposition was
carried out for 45 min resulting in a ZnO film of thickness
∼500 nm. For depositing AZO films, we added a 0.01M
aqueous solution of aluminium nitrate [Al(NO3 )3 · 9H2 O] with
1%, 3% and 5% atomic ratio of [Al]/[Zn] in the aqueous
solution of zinc acetate dihydrate [Zn(CH3 COO)2 ·2H2 O]. The
resultant solution was vigorously stirred and used as a spraying
solution for the AZO films. The deposition parameters of the
AZO films were kept the same as those for the undoped ZnO
film. The AZO films with 1 at%, 3 at% and 5 at% atomic ratio
of [Al]/[Zn] were named AZO1, AZO3 and AZO5.
Structural characterizations were carried out by x-ray
diffraction (XRD). The step size for the XRD measurement
was kept as 0.001◦ and the scanning time per step was kept
as 0.5 s. The microstructures were examined by atomic
force microscopy. The optical absorption spectra were
recorded in the wavelength range 350–600 nm using a UV–Vis
spectrophotometer. The PL spectra were recorded in the
wavelength range 350–620 nm with a photoluminescence
spectrometer using 325 nm of excitation wavelength obtained
from a xenon lamp. The core level spectra of Zn and O for
the ZnO and AZO films were studied by x-ray photoelectron
spectroscopy (XPS).
Figure 1. XRD patterns of ZnO and Al-doped ZnO (AZO1, AZO3
and AZO5) films for 1%, 3% and 5% Al concentrations.
ZnO and AZO films can be due to the amorphous substrate
(quartz), high solution spray rate and relatively lower substrate
temperature. The 2θ values for the (0 0 2) peak for ZnO, AZO1,
AZO3 and AZO5 are 34.477◦ , 34.482◦ , 34.490◦ and 34.484◦ ,
and for the (1 0 0) peak for ZnO, AZO1, AZO3 and AZO5
are 31.815◦ , 31.818◦ , 31.828◦ and 31.822◦ , respectively. The
values of the c lattice parameter and a lattice parameter for
pure ZnO are found to be 5.198 nm and 3.245 nm which are
lower than the c lattice parameter (∼5.206 nm) and a lattice
parameter (∼3.249 nm) of stress-free bulk ZnO. Al doping in
ZnO films leads to a decrease in the c lattice parameter and a
lattice parameter values as compared with the pure ZnO film.
The values of lattice parameter decrease for the AZO1 and
AZO3 films, and for the AZO5 film, there is an increase in
its value as compared with that for AZO3 but it is still lower
than the lattice parameters of pure ZnO film. The change in
the c lattice parameter is found to be more as compared with
the change in a lattice parameter of the ZnO/AZO films with
respect to the stress-free bulk ZnO. The ionic radii of Al3+ and
Zn2+ are 0.53 Å and 0.60 Å, respectively. Thus, the addition
of Al atoms in the ZnO matrix is expected to shorten the c
lattice parameter if Al atoms get substituted at the Zn sites.
The difference in c value of ZnO and stress-free bulk ZnO
indicates the presence of stress in the as-grown ZnO film.
For hexagonal lattice the linear stress components (σx , σy
and σz ) are given by [17]
  
 
σx
C11 C12 C13
ea
  
 
(1)
σy  = C12 C11 C13  eb  ,
σz
C13 C13 C33
ec
where Cii are the elastic stiffness constant (C11 = 2.1 ×
1011 N m−2 , C12 = 1.2 × 1011 N m−2 , C13 = 1.05 ×
1011 N m−2 and C33 = 2.1 × 1011 N m−2 ) and ea , eb and ec
are the linear strain along the a-, b- and c-axis, respectively.
For hexagonal lattice a = b therefore, strain along the a-axis
will be equal to the strain along the b-axis i.e. ea = eb .
The total stress in the film can be expressed as
3. Results and discussion
3.1. Structural properties
Figure 1 shows the XRD patterns of the ZnO and AZO films, in
which peaks corresponding to the (1 0 0), (0 0 2), (1 0 1), (1 0 2)
and (1 1 0) planes are observed which confirm the wurtzite
structure of ZnO. The observed polycrystalline nature of the
σ = (σx + σy + σz ),
2
(2)
J. Phys. D: Appl. Phys. 43 (2010) 465402
B K Sharma and N Khare
Figure 2. Variation of stress as a function of different wt% of Al
doping in ZnO films.
where σx , σy and σz can be calculated using equations
σx = C11 ea + C12 eb + C13 ec ,
(3)
σy = C12 ea + C11 eb + C13 ec ,
(4)
σz = C13 ea + C13 eb + C33 ec .
(5)
Strains along the a-, b- and c-axis can be calculated by
a − a0
ea = eb =
a0
and
ec =
c − c0
c0
(6)
Figure 3. (a) Optical absorption spectra of ZnO and AZO films, (b)
plot of (αhν)2 versus photon energy (hν) for ZnO and AZO films.
Inset shows the variation of band gap of ZnO films with different Al
concentrations.
,
(7)
3.2. Optical absorption studies
where a0 and c0 are the lattice parameters of stress-free bulk
ZnO whereas a and c are the lattice parameters of the ZnO or
AZO thin films. Using equations (2)–(7), the stress in the ZnO
and AZO films was calculated. The variation of stress (σ ) with
the increase in Al doping in ZnO is shown in figure 2. The value
of σ for the pure ZnO film is obtained as 17.98 × 108 N m−2 .
For the AZO1 and AZO3 films the stress value increases while
for AZO5 it decreases, but it was still higher as compared with
the stress in the undoped ZnO film. It seems that there is a
critical value for Al% in ZnO up to which stress increases and
above which it decreases. Above this critical value, the Al
atoms, rather than substituting at the Zn sites, may go to the
interstitial sites and the c lattice parameter shows an increase
as compared with the AZO3 film. The total stress in the film
mainly consists of two components: one is the intrinsic stress
introduced by the doping and defects during the growth, and
the other is the extrinsic stress introduced by the mismatch in
lattice constants and thermal expansion coefficients of the film
and the substrate. In the present case, the growth temperature
of the substrates is kept the same for all the ZnO and AZO
films, therefore the intrinsic stress originating from the thermal
mismatch between films and substrates is expected to have the
same magnitude for all the films. Thus, the observed change
in stress in the Al-doped ZnO films is mainly from the doping
of Al and change in the defects of the ZnO films.
Figure 3(a) shows the optical absorption spectra of pure
ZnO and Al-doped ZnO films in the wavelength range 350–
600 nm. The decrease in absorption for the Al-doped ZnO
films as compared with the undoped ZnO film indicates that the
transparency of the ZnO film increases due to the incorporation
of Al in ZnO. The fundamental absorption corresponding
to the direct electronic transition from the valence band to
the conduction band is usually used to determine the optical
band gap for direct band gap semiconductors. The value of
absorption coefficient (α) for a direct band gap semiconductor
is expressed as
A(hν − Eg )1/2
,
(8)
α=
hν
where A is a constant, hν is the photon energy and Eg is the
band gap for the direct band gap semiconductor. The optical
band gap for the ZnO and Al-doped ZnO films is determined
by plotting (αhν)2 as a function of photon energy hν, and
extrapolating the linear region of (αhν)2 to the energy axis
where (αhν)2 corresponds to zero. Figure 3(b) shows the
plot of (αhν)2 as a function of photon energy (hν) for the
ZnO and AZO films. From these plots, the value of band gap
for the ZnO, AZO1, AZO3 and AZO5 films are obtained as
3.19 eV, 3.26 eV, 3.27 eV and 3.24 eV, respectively. The inset
of figure 3(b) shows the variation of band gap of the AZO films
3
J. Phys. D: Appl. Phys. 43 (2010) 465402
B K Sharma and N Khare
Figure 4. Normalized PL spectra of ZnO and AZO films. Inset (a)
shows the blow-up part of UV emission and inset (b) shows the
blow-up part of defect peaks for AZO films.
as a function of Al% in the ZnO film. Initially, the value of band
gap increases for the Al-doped ZnO films up to 3% of Al doping
and then it decreases for 5% Al doping. The observation of
blue shift in the band gap of Al-doped ZnO films has been
reported earlier also [10–12] and this has been attributed to
the BM shift. The carrier concentration for observing the BM
shift [18] should be 3.0 × 1019 cm−3 . In the present case,
using Hall measurements, we found the carrier concentrations
for the AZO1, AZO3 and AZO5 films as ≈1.7 × 1017 cm−3 ,
≈2.0 × 1018 cm−3 and 6.3 × 1017 cm−3 , respectively, which
is much lower than the critical carrier concentration required
for producing BM shift in the band gap of AZO films. Thus
in the present case, the observed blue shift in the band gap of
the AZO films cannot be attributed to the BM shift. We notice
that the trend of change in Eg with Al doping is similar to the
change in stress in the AZO films. The stress in the AZO films
increases as the Al% is increased up to 3% and it decreases as
the Al% is further increased to 5%. The band gap of the AZO
films also shows an increase as the Al% is increased up to 3%,
and afterwards it shows a decrease. Thus, the increase in the
band gap of the AZO films seems to be related to the enhanced
stress.
Figure 5. (a) PL spectra of pure ZnO film in the visible region and
three deconvoluted Gaussian curves (shown by dotted lines). Open
circles show the experimental points and solid lines show the sum of
the three deconvoluted Gaussian curves, (b) variation of relative
contributions of Zni , VO and Oi defects with different Al%.
causes red shifts in the defect-related peak. The red shift in the
defect-related PL peak of the Al-doped ZnO films indicate the
variation of relative contribution of different possible defects
in the AZO films. In order to find out the relative contribution
of the possible defects in the PL spectra of the ZnO and AZO
films for the visible region, a deconvolution method has been
employed [19, 20]. The most probable defects in intrinsically
grown ZnO are zinc interstitials (Zni ) and oxygen vacancies
(VO ). These defects are responsible for n-type conductivity
in undoped ZnO. In addition to these, other defects such
as oxygen interstitials (Oi ), vacancies of zinc (VZn ), oxygen
sitting at Zn site (OZn ), are also possible.
The defect peak of ZnO is deconvoluted into three
Gaussian peaks centred at ∼480, ∼520 and ∼560 nm, as
shown in figure 5(a). The deconvoluted Gaussian peak centred
at ∼480 nm is attributed to the transition from Zni to VZn
which gives the luminescence corresponding to the wavelength
∼480 nm [21]. The deconvoluted Gaussian peak centred
at ∼520 nm is attributed to the luminescence originating
from the conduction band minimum to VO [22] and the
deconvoluted Gaussian peak centred at ∼560 nm is attributed
to the luminescence originating from the conduction band
minimum to Oi [23]. A similar deconvolution analysis was
done for the AZO1, AZO3 and AZO5 films. Figure 5(b)
shows the relative contributions of Zni , VO and Oi defects in
the visible region of the PL spectrum as a function of Al%. In
3.3. PL studies
Figure 4 shows the room temperature PL spectra of the ZnO
and AZO films. All the films exhibit an excitonic peak in
the ultraviolet (UV) region and a defect-related peak in the
visible region. The excitonic peak in pure ZnO film shows
a blue shift with Al doping (inset (a) of figure 4). In the
visible region, the pure ZnO film exhibits a broad peak centred
around ∼502 nm related to the defects which is suppressed
for the AZO films. In the AZO films, the defects are still
present but their concentrations are much less as compared
with the pure ZnO film. For the AZO1, AZO3 and AZO5 films,
the defect-related peaks are centred at 530 nm, 540 nm and
525 nm, respectively (inset (b) of figure 4). This indicates that
in addition to suppressing the defects, Al doping in ZnO also
4
J. Phys. D: Appl. Phys. 43 (2010) 465402
B K Sharma and N Khare
Figure 6. Gaussian fitted XPS spectra of Zn 2p3/2 core level of (a) ZnO and (b) AZO1 and deconvoluted XPS spectra of O 1s core level of
(c) ZnO and (d) AZO1 films.
the pure ZnO film, Zni and VO defects mainly contribute to the
defect-related PL peak. For a doping of Al up to 3%, the Zni
and VO defects decrease whereas the Oi defects increase. For
5% Al doping the contribution from Zni and VO defects is more
as compared with that in the AZO3 film while contribution
from Oi defects decreases. However, in the 5% Al-doped film
also the contribution of Zni and VO is less as compared with
the pure ZnO film. This indicates that doping of Al in ZnO
leads to a reduction of Zni and VO defects and increases the
Oi defects. It seems that the presence of Al in the ZnO matrix
changes the defect equilibria leading to the suppression of Zni
and VO and enhancement of Oi defects. The presence of Al in
the ZnO matrix may enhance the diffusion of oxygen due to
the presence of Oi defects [24] which can result in the change
in defect equilibria in AZO films.
The c lattice parameter of the ZnO film is related to the
Zni defects present in the film and the reduction of Zni defects
leads to a reduction in the c parameter [25]. In our case, we
have also observed that Al doping leads to a suppression of Zni
defects and this is expected to result in the reduction of the c
lattice parameter as we have observed in the XRD studies of
the Al-doped ZnO films. Thus, there seems to be a correlation
between the decrease in Zni defects and increase in stress in
the Al-doped ZnO films.
ZnO and AZO1 films. The core level peak of Zn 2p3/2 for
the ZnO and AZO1 films can be fitted by a single Gaussian
peak, which is centred at 1022.3 eV. The symmetric nature of
the fitted Gaussian peaks for Zn 2p3/2 indicates that in the
ZnO and AZO1 films most of the Zn atoms remain in the
same formal valence state of Zn2+ within the oxygen deficient
ZnO1−x matrix. The core level peak of O 1s for the ZnO
and AZO1 films is deconvoluted into two Gaussian peaks as
shown in figures 6(c) and (d), respectively. The deconvoluted
Gaussian peaks of the O 1s core level peak for the ZnO and
AZO1 films are centred at 530.1 eV and 531.6 eV which are
attributed to O2− ions bonded with Zn2+ in the ZnO phase [26]
and oxygen vacancies (VO ) related defects [27], respectively.
The relative contributions of O2− ions bonded with Zn2+ in
the ZnO phase and VO -related defects for the pure ZnO film
are found to be 60% and 40%, respectively. In the AZO1
film, the relative contributions of O2− ions bonded with Zn2+
ions in the ZnO phase and VO -related defects are found to be
70% and 30%, respectively. This indicates that the relative
contribution of VO -related defects decreases in the Al-doped
ZnO film. This is similar to what we have observed in the PL
study of the ZnO and AZO films (section 3.3).
3.4. XPS study
In conclusion, Al-doped ZnO thin films have been synthesized
using ultrasonically assisted chemical vapour deposition. The
presence of stress has been found in the undoped ZnO film,
which increases after Al doping up to 3% in ZnO films and
then decreases for 5% Al doping. A blue shift in the band gap
of Al-doped ZnO films is observed which has been attributed
4. Conclusions
XPS measurements of the ZnO and AZO1 films were carried
out to investigate the effect of Al doping on defects in ZnO.
Figures 6(a) and (b) show the core level peaks of Zn 2p3/2 and
figures 6(c) and (d) show the core level peaks of O 1s for the
5
J. Phys. D: Appl. Phys. 43 (2010) 465402
B K Sharma and N Khare
to an increase in stress present in ZnO after Al doping. The
presence of Al in the ZnO matrix changes the defect equilibria
leading to a suppression of Zni and VO and enhancement of
Oi defects. The enhancement of oxygen interstitial defects is
responsible for the red shift of the defect-related PL peak in the
AZO1 and AZO3 films. The increase in stress in the AZO films
may be attributed to the decrease in Zni defect concentrations
after Al doping in ZnO films.
[9] Makino T, Tamura K, Chia CH, Segawa Y, Kawasaki M,
Ohtomo A and Koinuma H 2002 Phys. Status Solidi b
229 853
[10] Lu J G et al 2007 J. Appl. Phys. 101 083705
[11] Li Q H, Zhu D, Liu W, Liu Y and Ma X C 2008 Appl. Surf.
Sci. 254 2922
[12] Lo S S, Huang D, Tu C H, Hou C H and Chen C C 2009
J. Phys. D: Appl. Phys. 42 095420
[13] Burstein E 1954 Phys. Rev. 93 632
[14] Gupta V and Mansingh A 1996 J. Appl. Phys. 80 1063
[15] Kumar R, Khare N, Kumar V and Bhalla G L 2008 Appl. Surf.
Sci. 254 6509
[16] Wu X L, Siu G G, Fu C L and Ong H C 2001 Appl. Phys. Lett.
78 2285
[17] Maniv S, Westwood W D and Colombini E 1982 J. Vac. Sci.
Technol. 20 162
[18] Roth A P, Webb J B and Williams D F 1981 Solid State
Commun. 39 1269
[19] Yang Y, Wang X, Sun C and Li L 2009 J. Appl. Phys.
105 094304
[20] Ye H B, Kong J F, Shen W Z, Zhao J L and Li X M 2007 J.
Phys. D: Appl. Phys. 40 5588
[21] Wen F, Li W, Moon J H and Kim J H 2005 Solid State
Commun. 135 34
[22] Vanheusden K, Warren W L, Seager C H, Tallant D R,
Voigt J A and Gnade B E 1996 J. Appl. Phys. 79 7983
[23] Wei X, Man B, Xue C, Chen C and Liu M 2006 Japan. J. Appl.
Phys. 45 8586
[24] Ryoken H, Sakaguchi I, Ohashi N, Sekiguchi T, Hishita S and
Hneda H 2005 J. Mater. Res. 20 2866
[25] Driscoll J L M, Khare N, Liu Y and Vickers M E 2007 Adv.
Mater. 19 2925
[26] Chen M, Wang X, Yu Y H, Pei Z L, Bai X D, Sun C,
Huang R F and Wen L S 2000 Appl. Surf. Sci. 158 134
[27] Szorenyi T, Laude L D, Bertoti I, Kantor Z and Geretovszky Z
1995 J. Appl. Phys. 78 6211
Acknowledgment
The authors are grateful to Dr M N Kamalasana (NPL, New
Delhi, India) for providing photoluminescence and UV–vis
spectrophotometer facilities. Financial support from the CSIR
is gratefully acknowledged. One of the authors (B K Sharma)
is grateful to the CSIR, New Delhi, India, for awarding Senior
Research Fellowship.
References
[1] Mende L S and Driscoll J L M 2007 Mater. Today 10 40
[2] Ozgur U, Alivov Y I, Liu C, Teke A, Reshchikov M A,
Dogan S, Avrutin V, Cho S J and Morkoc H 2005 J. Appl.
Phys. 98 041301
[3] Hichou A E, Addou M, Bougrine A, Dounia R, Ebothe J,
Troyon M and Amrani M 2004 Mater. Chem. Phys. 83 43
[4] Park S M, Ikegami T and Ebihara K 2005 Japan. J. Appl. Phys.
44 8027
[5] Lin J P and Wu J M 2008 Appl. Phys. Lett. 92 134103
[6] Lee J H and Song J T 2008 Thin Solid Films 516 1377
[7] Martin A, Espinos J P, Justo A, Holgado J P, Yubero F and
Gonzalez-Elipe A R 2002 Surf. Coat. Technol. 151/152 289
[8] Mass J, Bhattacharya P and Katiyar R S 2003 Mater. Sci.
Eng. B 103 9
6