Nonlinear optical effects in In 2 O 3 :Sn glass nano-interfaces

Home

Search

Collections

Journals

About

Contact us

My IOPscience

Nonlinear optical effects in In2O3:Sn–glass nano-interfaces

This article has been downloaded from IOPscience. Please scroll down to see the full text article. 2003 J. Opt. A: Pure Appl. Opt. 5 61 (http://iopscience.iop.org/1464-4258/5/1/309) View the table of contents for this issue, or go to the journal homepage for more

Download details: IP Address: 194.57.151.153 The article was downloaded on 11/06/2012 at 14:53

Please note that terms and conditions apply.

INSTITUTE OF PHYSICS PUBLISHING

JOURNAL OF OPTICS A: PURE AND APPLIED OPTICS

J. Opt. A: Pure Appl. Opt. 5 (2003) 61–65

PII: S1464-4258(03)53034-3

Nonlinear optical effects in In2O3:Sn–glass nano-interfaces I V Kityk1 , J Eboth´e2 , A El Hichou2 , B El Idrissi3 , M Addou3 and J Krasowski1 1 Institute of Computer Modelling, Krakow University of Technology, ul. Warszawska 24, Krakow, Poland 2 Universit´e de Reims, UTAP-LMET, EA no 2061, UFR Sciences, BP 138, 21 rue Clement Ader, 51685 Reims Cedex 02, France 3 Universit´e Ibn Tofail, LOEPCM, Faculte des Sciences, BP 133, K´enitra, Maroc

Received 27 August 2002, in final form 7 November 2002 Published 13 December 2002 Online at stacks.iop.org/JOptA/5/61 Abstract Nano-sized thin layers (1–2 nm) between In2 O3 :Sn (ITO) crystalline films and glass substrates were studied using photo-induced optical and second-order nonlinear optical (second harmonic generation) effects. Photo-induced changes in the effective energy gap were found for the first time in In2 O3 films doped using different amounts of Sn deposited on the glass substrates. The photo-induced second-order nonlinear optical susceptibilities show a good correlation with the behaviour of the fundamental absorption edge. The maximal response of the photo-induced signal was observed for pump–probe delay times equal to approximately 26 ps. The experimental measurements performed indicate that the observed effects are stimulated by two factors: the first one is connected with the interface potential gradients on the boarder glass–ITO film. The second one is a consequence of the additional polarization due to the insertion of differing amounts of Sn atoms. The observed phenomenon may be proposed as a sensitive tool for the investigation of thin semiconducting–glass interface layers. Keywords: Photoinduced nonlinear optical effects, second harmonic

generation, ITO films (Some figures in this article are in colour only in the electronic version)

1. Introduction There are a lot of methods used to manufacture In2 O3 :Sn (ITO) thin films: we feel it is necessary to emphasize zone-confining evaporation [1]; r.-f. sputtering [2]; facingtarget sputtering [3] etc. Among the different techniques which exist for the creation of low-cost devices, the most appropriate seems to be the spray pyrolysis technique [4– 7]. One of the complications restricting the widespread use of this technique for optoelectronic deposition is the absence of a fast non-destructive method of monitoring the film and substrate structures. The traditional methods, including electron microscope control of the film’s morphology, x-ray diffraction [7, 8], XPS spectra, electrical measurement [9] etc, require sophisticated analyses leading to a great deal of

time before they lead to probable additional disturbances in the manufactured films. The traditional optic and spectral methods (such as transmission or reflection) are not sensitive enough to provide sufficient information [10] regarding the film–substrate parameters. However, a more sensitive method for monitoring of the boarder layers separating glasses and semiconductors may result from the photo-induced effects that stimulate additional polarization of the interfaces. This phenomenon can be substantially larger than the individual contribution of both the proper glasses and appropriate semi-conducting crystallites. Because of the many usual trapping levels occurring at the interfaces, one can expect the appearance of a large polarizability which determines the key opto-electronic features. Consequently, in this paper we propose a substantially new thin layer detection method that is

1464-4258/03/010061+05$30.00 © 2003 IOP Publishing Ltd Printed in the UK

61

I V Kityk et al

(a) (222)

(400)

15

20

Figure 2. The principal set-up used for the measurement of the photo-induced absorption.

(440)

(211)

25

30

35

40

45

50

55

(b) (222)

(400)

(440)

(211)

15

20

25

30

35

40

45

50

55

2θ

Figure 1. Diffraction patterns of tin-doped In2 O3 prepared using different partial pressures of oxygen: (a) 10−4 mbar; (b) 5 × 10−4 mbar.

based on the gain of the photo-induced polarization within the mentioned interface layers. In [11] we showed that one of the more sensitive ways of controlling the ITO- film–glass interfaces may be the photoinduced second-harmonic generation (SHG) method. In this work we propose the use of the method introduced earlier by us for the investigation of ITO films deposited on a glass substrate [11] in correlation with the Sn content. Using band structure and molecular dynamics simulations we reported that an important role in the electronic properties of such films is played by a thin interface layer with a thickness approximately equal to several nanometres [12]. In [12] we showed (for the case of ZnO films) that for such kinds of effects the key role belongs to the thin interfaces separating the glass and crystallite films. Besides, additional doping could lead to further noncentrosymmetry interface charge density which is responsible for the second-order non-linear optical effects (like a SHG). To clarify the origin of the observed phenomenon we performed simultaneous measurements of the photo-induced absorption (to estimate E g ) and second-order NLO susceptibility.

2. The specimen preparation technique The polycrystalline ITO films investigated in this work were obtained using a spraying pyrolysis (SP) technique whose deposition bath consisted of a solution of 0.05 M active ion species. An aqueous solution of the [In3+ ] species obtained from FLUKA InCl3 product is concerned for the formation of pure In2 O3 specimens. The Sn-doped In2 O3 samples we obtained from a mixed deposition bath consisting of two active species, [In 3+ ] and [Sn2+ ], keeping the same total concentration as above. The [Sn2+ ] species were prepared from SnCl2 (produced by FLUKA) diluted in ethanol solvent which was added to the [In 3+ ] aqueous solution, varying their 62

concentrations in order to control the composition of the resulting material. The flow of the deposition bath was directed towards a substrate made of bare glass, its rate was fixed at 3.50 ml min−1 . The substrate temperature was maintained at a constant value of Ts = 500 ◦ C. We deal, in every case, with the same film thickness (about 0.50 µm), estimated from the deposition time. Good agreement was obtained with the value deduced from the cross-sectional images of the samples obtained using a LEO 982 scanning electron microscopy (SEM). The material composition was additionally examined by x-ray spectroscopy using an EDAX 9100 analyser. The Sn dopant content in our investigated In2 O3 –Sn films ranged between 1 and 12 at.%. The film’s structure was investigated by x-ray diffraction using Cu Kα radiation (see figure 1). The diffraction patterns of the tin-doped In2 O3 films show that the films feature a cubic structure with a = 1.003 nm. As recently proposed for the sprayed specimens of comparable thickness [4], the reflection from 400 plan parallel to the deposition substrate is always predominant here. The SEM measurements confirmed the lateral growth orientation of the obtained films.

3. Experimental set-up The principal scheme used to perform the photo-induced absorption measurements is presented in figure 2. The light beam from the photo-inducing Q-switched nitrogen laser (λ = 337 nm, power 23 MW, time duration 26 ps) using a system of mirrors M and a polarizer P2 was incident to the specimen S causing additional charge density non-centrosymmetry in the investigated medium. As a probing beam for investigation of the photo-induced changes we used monochromatic light from the Shimadzu grating spectrophotometer (SP) operating within the 200–800 nm wavelength region (with a spectral resolution equal to about 0.7 nm mm−1 ). The diameter of the UV-photoinducing beam was equal to about 2–4 mm and varying it we were able to operate with a photo-induced beam power within the 0.1–1.5 GW cm−2 range. Polarizer P3 defines the desired laser polarization. The external photomultipliers PM were connected with a PC-controller. All the equipment was temporarily synchronized. The probing beam was operated via a delaying line (DL) and polarizer P3. The measurement set-up used for the detection of the SHG and simultaneous ellipsometric monitoring of the surfaces is shown in figure 3. The Q-switched nitrogen UV laser (λ = 0.337 µm), with the same parameters as given earlier, was used to produce the photo-inducing changes. Its power parameters were detected using a photomutiplier PM4 connected with a boxcar with a gate value of about 440 ps. The polarizer P1 together with the dielectric M1 mirrors define the polarization

Nonlinear optical effects in ITO–glass nano-interfaces

directions of the photo-inducing beam. The incident angle was varied within the 2.3◦ –9.4◦ range. The beam spot of the laser possessed Gaussian-like form and its diameter was varied within the 70–845 µm range. The system of mirrors M1 allows one to continuously vary the beam spot diameter and to operate by the corresponding photo-inducing beam power density within the range 0.1–2.1 GW cm−2 per pulse. As a probing laser beam we used a YAB–Gd laser (λ = 1.76 µm) which works like a laser temporarily synchronized with the photoinducing laser beam. This gives the possibility of generating short laser pulses within the 7–14 ps range with a power of about 12 MW, and with a frequency of approximately 11 Hz. Its beam diameter was equal to about 24 µm and was felt fully within the photo-inducing laser beam diameter. The time duration of the pulses was chosen in order to avoid overheating of the surface specimens. The light of the probing laser beam was perpendicular to the specimen’s surfaces. DL was performed from the electro-optically operated KDP crystals in order to study the influence of the delay time between the pumping and probing beams. At the same time the duration of the laser pulses was chosen in such a way to avoid specimen overheating. The intensity of the probing laser power beam was monitored using the photomultiplier PM3 and its polarization was defined by the polarizer P1. Polarizer P3 was used to investigate different light polarization’s that were propagated through the specimens (Spec). A grating spectrophotometer (SP) (with a spectral resolution of about 0.7 nm mm−1 ) was used to investigate the doubled-frequency signal (possessing a wavelength of λ = 0.88 µm). DL, pumping and probing laser pulses were connected through the boxcar that was also connected to the photomultipliers PM1 and PM2 monitoring the output doubled-frequency signal as well the scattered background. The specimen was rotated about the axes in order to produce angle-dependent SHG behaviour which corresponds to the Maker fringence maxima. At the same time the surface processes were monitored independently by photo-induced ellipsometry which was performed using an ellisometric set-up (indicated by El.1 and El.2). Using this measurement schema we performed the measurements for the pure glasses, investigated crystals (for both cubic and wurtzite-like structures) as well for the studied

Figure 4. Typical absorption edges of the In2 O3 samples doped by Sn(6%) at different photo-inducing power densities: 1—0 GW cm−2 ; 2—1.0 GW cm−2 ; 3—1.75 GW cm−2 . 4.3

3.8 Eg (eV)

Figure 3. The principal measurement set-up used for the photo-inducing SHG.

3.3

2.8 0.1

0.9

1.7

2.5

PI (GW/cm2)

Figure 5. The dependence of effective energy gap E g of In2 O3 versus the photo-induced power : (x)—pure; ( )—2% of Sn; ()—10%; ()—6%;
—8%.

films. We found that the output SHG signal for the composites (film–glass specimens) is at least half an order of magnitude larger compared to the other specimens.

4. Results and discussion The measurements of the absorption edges (see figure 4) confirm indirect features of the fundamental absorption edge for the ITO sample. One can see that with increasing photoinduced light power we observe substantial shifts of the absorption edge towards longer wavelengths. We only present the more obvious shift that is observed for the specimens doped with Sn (6%). One can see that the effective energy gap E g (see figure 5) is minimal for the specimens with an Sn content equal to approximately 8%. With the following increase of the Sn content, E g once again begins to increase. This is a consequence of the occurrence of the bulk solid state alloy (in contrast to the local doping). In this case we have an increase with the following saturation of the dipole transition moments versus the Sn content which is presented in table 1 and is obtained from quantum chemical calculations. 63

I V Kityk et al

d222 (pm/V)

0.35

0.3

0.25 0.1

0.7

1.3

1.9

2.5

PI (GW/cm2)

Figure 6. Photo-induced changes of the NLO susceptibilities d222 for the In2 O3 films with different amounts of Sn: —8%; ♦—6%; —10%; ∗—2%; —pure. 

◦

Table 1. Transition dipole moments of ITO films doped by different amounts of Sn. The content of x (in weight. units)

Transition dipole moments (in Debye)

Pure 2% 6% 8% 10%

6.7 11.2 27 32 24

Simultaneously the photo-induced behaviour of E g demonstrates a step-like feature versus the UV-photo-induced beam power. Emerging from general expressions connecting the macroscopic second-order non-linear optical susceptibility di j k and microscopic hyperpolarizability βi j k we can write χi(ω,ω,ω) = f i(ω) f j(ω) f k(ω) βi(ω,ω,ω) , jk jk

(1)

where χi(ω,ω,ω) is the macroscopic second-order non-linear jk susceptibility responsible for the LEE; f i,(ω) j,k are the i, j, k components of the local Lorentz fields. Using the oversimplified expression we can present the microscopic hyperpolarizability in the form ∼ βi(ω,ω,ω) = jk

µ 2gr µ tr ( j → l) |E j − E l |3

(2)

where µgr is the ground state dipole moment and µgr ( j → l) the transition dipole moments between the j and l band states. As a result of the measured dependencies of the SHG signals we have performed the second-order nonlinear susceptibility calculations. In particular, we have carried out the calculations for the d222 tensor components. In our case we have chosen the tensor index 2 parallel to the specimen’s surface. In figure 6 we present the photo-induced dependencies of d222 for the In2 O3 specimens with different Sn content values. One can observe a striking correlation with the absorption edge behaviour (see figure 5) described by equations (1) and (2) and the dipole transition moments presented in table 1. The key role belongs to narrowing of the energy gap which is inversely proportional to the corresponding hyperpolarizability. 64

Figure 7. Principal schema of the trapping levels separating glasses and crystalline ITO films. The ground state dipole moments are indicated by µ.

The observed behaviours of the photo-induced secondorder nonlinear optical susceptibilities are inversely proportional to E g3 . Consequently we were able to confirm that the behaviour of the energy gap also defines the photo-induced behaviour of the second-order NLO susceptibilities. For convenience we consider the tensor component d222 . For this parameter even the step-like features are similar. These anomalies are caused by the occurrence of the trapping levels within the thin nano-layers separating the glass and the ITO films. Using an approach developed previously for semiconducting large-sized nanocrystallites [13] we have introduced the excitonic-like quantum-confined energy terms separating the glass and the ITO. The approach developed previously may also be applied to ITO films because the surrounding amorphous-like background plays a crucial role for the quantum wells within which are incorporated the investigated films (such as the quantum-well heterojunction). Consequently, there occurs a large number of trapping levels at the interfaces (see figure 7) possessing relatively large ground state dipole moments (up to 350 D). As a consequence, under the influence of external photoinducing light, the role played by the interface states will be enhanced, and using the method described above the corresponding effect will also be enhanced.

5. Conclusions The results presented here unambiguously show that photoinduced SHG may be a sensitive tool for investigating the thin nanolayers separating doped crystalline semiconductors and glasses. We have shown experimentally that the observed effects are caused, in the main, by the interface trapping levels that are responsible for the large dipole moment value. Photo-induced changes of the effective energy gap were observed for the first time in the indium oxide films doped by different amounts of tin and deposited on the glass substrates. The photo-induced second-order nonlinear optical susceptibilities show a good correlation with the absorption edge. The experimental investigations performed indicate that the observed effects are a consequence of two factors: the first one is connected with the interface (film–glass) potential

Nonlinear optical effects in ITO–glass nano-interfaces

gradients on the boarder glass–ITO film; the second one is a consequence of the additional polarization due to the insertion of the Sn impurities. The observed phenomenon may be proposed as a sensitive tool for investigation of thin interface layers separating glass-like substrates and semiconducting films.

References [1] Rauf L A 1993 Mater. Lett. 18 123 [2] Bender M, Trube J and Satollenwerk J 1999 Thin Solid Films 354 100 [3] Tominaga K, Ueda T, Ao T, Kataoka M and Mori I 1996 Thin Solid Films 281/282 194 [4] Manifacier J C, Szepessy L, Bresse J F, Peroten M and Stuck R 1979 Mater. Res. Bull. 14 109 [5] Shigesat Y, Koshi-ishi R, Kawashima T and Ohsako J 2000 Vacuum 59 614

[6] Kachonane A, Addou M, Bougrine A, Elidrissi B, Messousi R, Regragui M and Bernede J C 2001 Mater. Chem. Phys. 70 285 [7] Korotcenkov G, Brinzari V, Cerneavchi A, Cornet A, Moraute J, Cabot A and Arbiol J 2002 Sensors Actuators B 4212 1 [8] Bhira L, Ben Nasrallah T, Benede J C and Belgacem S 2001 Mater. Chem. Phys. 72 320 [9] Margel D, Schenkel M, Grabre M and Sulkowski M 2001 Thin Solid Films 392 91 [10] Adurodija F O, Izumi H, Ishihara T, Yoshioka H and Matoyoura M 2002 Sol. Energy Mater. Sol. Cells 71 1 [11] Olesik Z, Kityk I V, Kasperczyk J, Olesik J and Mefleh A 2000 Thin Solid Films 358 114 [12] Kityk I V, Ebothe J and Elhichou A 2001 Nonlinear Opt. J. 28 121 [13] Kityk I V, Kassiba A, Plucinski K J and Berdowski J 2000 Phys. Lett. A 265 403

65