ZnO thin films with very low surface roughness and bulklike electron density were grown on Si and SiO2 by atomic layer deposition. The real and imaginary parts of the complex dielectric function of ZnO on Si show monotonically decreasing values with decreasing film thickness at and below a threshold of about 20 nm. On the other hand, x-ray reflectivity results show that the electron density of our ZnO films is close to that of bulk ZnO and does not vary considerably with film thickness. While the reduction of the dielectric function cannot be explained by the electron density of our ZnO films, the Tanguy–Elliott amplitude prefactor governing the strength of optical interband transitions can explain our results consistently through the lowering of the electron–hole overlap factor at the ZnO/Si interface. In the case of ZnO/Si, a staggered type-II (spatially indirect) quantum well, holes are scattered into the Si substrate, causing a lowering of the electron–hole overlap factor and thus the reduction of excitonic absorption, consequently a decrease in the real and the imaginary parts of the dielectric function. This hypothesis was confirmed with ZnO films grown on SiO2, where a thin type-I quantum well, consisting of a narrower-bandgap semiconductor grown on a wider-bandgap (insulator) substrate, in which both the electron and the hole are confined in the ZnO thin film, leads to an increase in the electron–hole overlap matrix element with decreasing film thickness due to confinement, resulting in enhancement of the excitonic absorption in thinner ZnO films on SiO2.

In recent years, the optical properties of zinc oxide (ZnO), a direct and wide band gap semiconductor (Eg ∼ 3.37 eV at room temperature) with large exciton binding energy (60 meV), have gained substantial interest for potential applications in optoelectronics and photonics.1–3 The large exciton binding energy ensures efficient excitonic emission in ZnO at room temperature and even higher,4 which makes ZnO a promising material for short-wavelength optoelectronic devices, especially for ultraviolet light-emitting diodes and laser diodes.5 The thermal and chemical stabilities of ZnO, along with its excellent optical/opto-electronic properties,6 motivated extensive research for its use in a number of applications such as ultraviolet light emitters,6 thin film transistors,7 solar cells,8 transparent conducting oxide,9 gas sensors,10 and nanogenerators.11 ZnO has also attracted a lot of interest for being used in hybrid solar cells because it is less toxic than many other II–VI semiconductors and is relatively easy to synthesize in large quantities with a variety of techniques.12 Furthermore, ZnO as an n-type semiconductor with a wide band gap and a very good optical transmittance remains a potential candidate for an efficient cathode buffer layer or a transparent electrode in the organic and hybrid solar cells.12 In view of the above facts, tuning of the structural and optical properties of ZnO films is highly important and desired for optimal device performance.13 Studies showed that the thickness of ZnO films is a key parameter in photovoltaic devices where thicknesses below 300 nm are of specific interest in terms of optimization of device performance.11,14 The quantum size effects in ZnO nanostructures, such as nanocrystals,15,16 quantum wells,17 nanorods,18,19 nanowires,20 quantum dots,21,22 etc., have been extensively investigated by many research groups. However, few reports are available on the thickness dependence of properties of ZnO thin films.23–25 Experimentally, it has been observed that the quantum effects in ZnO thin films have become apparent in their optical properties specifically in the dimensions of 20 nm or less.23–25 The systematic quantitative thickness dependent study of the optical properties of ZnO thin films in this regime is highly important, and a convincing scientific explanation of the observed phenomena needs to be found.

In recent years, atomic layer deposition (ALD) has drawn significant interest for growing ZnO thin films using sequential, self-limiting surface reactions.26–29 Over other thin film deposition techniques, ALD offers several advantages such as excellent conformality, precise thickness control down to atomic level, pinhole-free films, and high degree of reproducibility.30–32 Also, ALD grown ZnO thin films generally show n-type conductivity which makes it an important candidate for use as a transparent conductive oxide material.6,27,33

Herein, we demonstrate the thickness induced change in optical properties of ALD-grown smooth, pinhole-free ZnO thin films on silicon (Si) and SiO2 substrates. We show clearly that in thin ZnO films (at and below 20 nm), with decreasing film thickness, there is a significant drop in the real and imaginary parts of the dielectric function for ZnO/Si system. The reduction of the complex dielectric function is explained by the Tanguy–Elliot amplitude prefactor through the effect of delocalization of holes in Si from ZnO at the interface, which leads to the decrease in the overlap of electron and hole wave functions, thus the deconfinement of the exciton, causing the drop in dielectric function.34,35 This hypothesis is consistent with the case of the ZnO/SiO2 system, where both the electron and the hole are confined in the ZnO thin film, which leads to an increase in the electron–hole overlap matrix element with decreasing film thickness, resulting in the enhancement of excitonic absorption in thinner ZnO films on SiO2.

ZnO thin films were grown on silicon [Si (100)] and fused quartz (SiO2) substrates at 200 °C deposition temperature (or process temperature) using a BENEQ TFS-200 ALD reactor. Before loading the substrates into the ALD reactor, all the substrates were ultrasonically cleaned in acetone and ethanol in sequence for 2 min, and then kept in deionized (DI) water and finally dried in nitrogen (N2, 99.999% purity). The precursors for zinc and oxygen were diethyl zinc [DEZn, Zn(C2H5)2, Sigma-Aldrich] and DI water, respectively. Nitrogen (N2, 99.999% purity) was used both as a carrier and the purging gas. The precursors were alternately introduced into the reactor using their intrinsic vapor pressures from external containers kept at 18 °C. The reactor pressure during the growth was kept at 1.75 mbar. One ALD reaction cycle consisted of a 0.2 s exposure to DEZ, followed by a 0.75 s for N2 purge, a 0.2 s exposure to H2O, and then another 0.75 s N2 purge. For the growth studies, the number of ALD cycles ranged from 30 to 410. We achieved a growth rate of about 1.6 Å per cycle which is similar to the growth rate reported by the other groups.36,37 ZnO films with different thickness (5–70 nm) were grown on Si and fused quartz (SiO2) substrates to explore the dependence of the optical properties on ZnO thickness and substrate choice.

The crystalline structures of the as-grown ZnO films were investigated by x-ray diffraction (XRD) with CuKα radiation (λ = 1.54 Å) using a Rigaku SmartLab automated multipurpose x-ray diffractometer.

The thickness of the films was estimated by using x-ray reflectivity (XRR) measurements. X-ray reflectivity data of ZnO films were collected with CuKα radiation using a Rigaku SmartLab automated multipurpose x-ray diffractometer. Specular reflectivity scans, i.e., scans in the plane containing the incident beam and the normal to the sample surface, with the incident angle θin equal to the scattering angle θsc, were performed with θin varying from 0° to 3°. If q = ks− ki is the momentum transfer vector and ks and ki the scattered and incident x-ray wave vectors, respectively, then this geometry makes the components in the sample plane, qx = qy = 0, and the value of qz (= (4π/λ) sin θin), the component normal to the sample surface, varies from 0 to 0.43 Å−1. The reflectivity data of ZnO films were analyzed by the Parratt formalism.38–42 This scheme recursively solves the Fresnel equations at each interface, i.e., the change in electron density (ρ) within any film. More precisely, in this method, we recursively solve the equation for the reflectance rn–1,n at the interface between the nth and the (n − 1)th layers, with the nth layer having thickness dn and electron density ρn, given as38 

rn1,n=[(rn,n+1+Fn1,n)(1+rn,n+1Fn1,n)]exp(iqz,n1dn1),
(1)

where

Fn1,n=(qz,n1qz,n)(qz,n1+qz,n),
(2)

with

qz,n=(qz,02qc,n2)1/2.
(3)

qz,0 is the z component (normal to the surface) of the momentum transfer in vacuum or air, and qc,n2=16πρnre is the critical value of qz for total external reflection for the nth layer. The constant re is the classical electron radius (=2.8 × 10−6 nm). Equation (1) then yields the reflectivity as

Rn1,n=|rn1,n|2.
(4)

This expression for reflectivity, valid for an ideally smooth interface, is then modified in the presence of the Gaussian interface width or “roughness” (σn–1,n) as

Rrough(qz)=R(qz)exp(qz2σ2).
(5)

The values of film thickness, electron density, and interface width or roughness extracted from the best fit of reflectivity data using Eqs. (1)–(5) have constructed the electron density profiles (EDPs), i.e., the electron density as a function of film depth from the top for ZnO films, after convoluting the profile with the interface widths.

The surface morphologies of ZnO films were examined by field emission scanning electron microscopy (FESEM, JEOL JSM-7600 F) and atomic force microscopy (AFM) in the tapping mode using a Bruker Dimension icon with ScanAsyst (Bruker, Santa Barbara, CA).

Spectroscopic ellipsometry (SE) measurements were performed on ZnO thin films deposited on Si substrates with an ellipsometer (J.A. Woollam, Co., Lincoln, NE, Model: VASE) in the photon energy range of 0.8–6.5 eV at three incident angles of 60°, 65°, and 70°, similar to the technique described elsewhere.43 Ellipsometry techniques have been extensively used to explore the optical properties of the ZnO films, namely, complex dielectric function, absorption coefficient, and band gap along with the film thickness and roughness, using the wvase (J.A. Woollam, Co.) software.

UV-VIS absorption spectra of ZnO films, deposited on fused quartz substrates (SiO2), were measured using a UV-VIS spectrophotometer in transmission mode (Perkin Elmer, lambda-35) in the wavelength (λ) range of 190–1100 nm. The absorption coefficient α can be calculated from the Beer-Lambert's law as44 

α=2.303Abs(λ)d,
(6)

where d and Abs are the film thickness and film absorbance, respectively.

The XRR technique provides information about the film thickness, the EDP, and the surface and interface roughness of thin films. Figure 1(a) shows x-ray reflectivity data for ZnO/Si samples with different ZnO film thickness, where circles and lines represent experimental data and theoretical fit, respectively. The x-ray reflectivity data have been fitted by the well-known Parratt's exact recursive method.38–42 X-ray reflectivity profile shows clear Kiessig fringes, and the corresponding extracted electron density indicates the formation of ZnO thin films with bulklike electron density, on silicon (Si) substrates, with very low surface and interface roughness. Figure 1(b) shows the representative EDPs (ρ) of the respective films extracted from XRR fits. Results show that the electron densities of ALD-grown ZnO films are close to its bulk electron density, except that at below about 10 nm film thickness, it drops slightly (by less than 7%). This signifies the formation of high quality pinhole-free ZnO films, with nearly complete surface coverage.

Fig. 1.

(Color online) (a) XRR for ZnO/Si with film thickness: (i) 5 nm, (ii) 9 nm, (iii) 19 nm, and (iv) 38 nm; circles and lines represent experimental data and theoretical fit, respectively; and (b) corresponding extracted electron density (ρ) profile from XRR fits.

Fig. 1.

(Color online) (a) XRR for ZnO/Si with film thickness: (i) 5 nm, (ii) 9 nm, (iii) 19 nm, and (iv) 38 nm; circles and lines represent experimental data and theoretical fit, respectively; and (b) corresponding extracted electron density (ρ) profile from XRR fits.

Close modal

It should be noted that ZnO films grown on fused quartz (SiO2) substrates (ZnO/SiO2 samples) show similar film thickness, electron density profile, and surface roughness, as observed for ZnO/Si systems. The thickness of our ALD-grown ZnO films ranges from about 5 to 70 nm. The growth rate of ZnO determined with XRR from the deposited films on Si and SiO2 was about 1.6 Å/cycle (at 200 °C), which is in agreement with previous studies of ZnO growth by ALD.36,37 Film thicknesses, corresponding growth parameters, surface roughness, and other relevant parameters are summarized in Table I.

Table I.

ALD grown ZnO films deposited at 200 °C process temperature, on Si and SiO2 substrates.

ZnO/SiZnO/SiO2
No. of ALD cyclesZnO film thickness from XRR (Å)Growth rate (Å/cycle)SiO2 thickness at ZnO/Si interface from XRR (Å)Surface roughness from XRR (Å)RMS of Surface roughness from AFM (Å)Mean grain size from XRD (Å)ZnO film thickness from XRR (Å)Growth rate (Å/cycle)Surface roughness from XRR (Å)RMS of surface roughness from AFM (Å)Mean grain size from XRD (Å)
30 50 1.65  <50      
60 89 1.49 11 56 87 1.45 11 ∼30 
120 189 1.58 11 15 15 123 188 1.56 16 14 133 
230 379 1.65 20  208 375 1.63 25  218 
300 515 1.72 10 19  194 520 1.73 21  212 
410 689 1.68 17 18 212 685 1.67 21 19 214 
ZnO/SiZnO/SiO2
No. of ALD cyclesZnO film thickness from XRR (Å)Growth rate (Å/cycle)SiO2 thickness at ZnO/Si interface from XRR (Å)Surface roughness from XRR (Å)RMS of Surface roughness from AFM (Å)Mean grain size from XRD (Å)ZnO film thickness from XRR (Å)Growth rate (Å/cycle)Surface roughness from XRR (Å)RMS of surface roughness from AFM (Å)Mean grain size from XRD (Å)
30 50 1.65  <50      
60 89 1.49 11 56 87 1.45 11 ∼30 
120 189 1.58 11 15 15 123 188 1.56 16 14 133 
230 379 1.65 20  208 375 1.63 25  218 
300 515 1.72 10 19  194 520 1.73 21  212 
410 689 1.68 17 18 212 685 1.67 21 19 214 

Figure 2(a) shows representative x-ray diffraction patterns of ZnO thin films of different thicknesses, grown on Si substrates. The diffraction peaks were matched with the standard diffraction pattern of hexagonal wurtzite ZnO with a space group of P63mc (JCPDS PDF Card No. 01-079-2205), illustrated in Fig. 2(b).

Fig. 2.

(Color online) (a) X-ray diffraction patterns of ZnO thin films with different thicknesses, on Si substrates. (b) Reference data. Standard diffraction pattern of hexagonal wurtzite ZnO with a space group of P63mc (JCPDS PDF Card No. 01-079-2205). (c) Average or mean grain size, derived from Scherrer's formula, for different ZnO film thicknesses, on Si substrates.

Fig. 2.

(Color online) (a) X-ray diffraction patterns of ZnO thin films with different thicknesses, on Si substrates. (b) Reference data. Standard diffraction pattern of hexagonal wurtzite ZnO with a space group of P63mc (JCPDS PDF Card No. 01-079-2205). (c) Average or mean grain size, derived from Scherrer's formula, for different ZnO film thicknesses, on Si substrates.

Close modal

The XRD pattern of our ZnO films shows a much stronger ZnO (002) peak compared to ZnO (101), whereas the standard powder XRD pattern of bulk ZnO (hexagonal wurtzite bulk ZnO) shows the maximum intensity for the ZnO (101) peak (JCPDS PDF Card No. 01-079-2205). This clearly indicates that our ZnO films on Si exhibited a preferred orientation along the ⟨0002⟩ direction with the c axis predominantly perpendicular to the substrate surface. Average grain sizes (out-of-plane) were estimated using Scherrer's formula,45 as shown in Fig. 2(c). The average grain sizes are about 20 nm for thicker ZnO films, but thin ZnO films (at and below 20 nm) show lower grain sizes with decreasing film thickness. It should be noted that our ZnO films grown on SiO2 show similar XRD pattern as ZnO/Si. The summary of the extracted results for both the sample types is shown in Table I.

The surface morphologies of ZnO films were examined by FESEM and AFM. Two representative SEM micrographs are shown in Figs. 3(a) and 3(c) for 69 and 9 nm ZnO films on Si, respectively. Corresponding AFM images are displayed in Figs. 3(b) and 3(d). The 69 nm ZnO films showed wedge-shaped grains (in-plane size along major and minor axes are ∼40 and ∼20 nm, respectively) in their surface morphology, which is consistent with earlier reported findings on higher temperature ALD-grown ZnO films.37 On the other hand, SEM and AFM images show smooth and uniform surface morphology for our 9 nm ZnO film. The root-mean-square (RMS) roughness was estimated for characteristic ZnO film thicknesses from AFM images, which were found to decrease in thinner films (at and below 20 nm), as shown in Table I. The results were consistent with the surface roughness obtained from XRR measurements.

Fig. 3.

(Color online) (a) and (b) Top-view SEM and AFM images of 69 nm ZnO/Si. (c) and (d) Top-view SEM and AFM images of 9 nm ZnO/Si.

Fig. 3.

(Color online) (a) and (b) Top-view SEM and AFM images of 69 nm ZnO/Si. (c) and (d) Top-view SEM and AFM images of 9 nm ZnO/Si.

Close modal

The surface roughness of ZnO films of thickness below ∼20 nm is found to be less than that of ZnO film of 69 nm and also observed to decrease with film thickness. This is further corroborated by the smaller grain sizes obtained from XRD data for thinner ZnO films with thickness below ∼20 nm, as shown in Table I and Fig. 2(c).

The optical properties of as-grown ZnO thin films on Si, namely, the complex dielectric function, absorption coefficient, and band gap, were determined using SE in the 0.8–6.5 eV photon energy range. The main advantages of the SE technique are its precision and nondestructiveness46,47 and, particularly, the ability to measure simultaneously the thicknesses and the optical constants of the system.47 Spectroscopic ellipsometry measures the Jones ratio, J, as a function of photon energy (E) and angle of incidence (ϕ), described by the equation48–50 

J(E,ϕ)=(rp/rs)=(tanψ)eiΔ,
(7)

where rp and rs are the complex Fresnel reflectance ratios for p and s polarized light, respectively. ψ and Δ are the ellipsometric angles corresponding to the amplitude ratio and the relative phase change, respectively.49,51

Figures 4(a) and 4(b) show the ellipsometric angles, ψ and Δ, for as-deposited 69 and 19 nm ZnO films on Si, respectively. We found that the calculated ψ and Δ are in good agreement with the experimental data. Excellent fits with mean-squared-error below 5 were achieved for all our samples. In this study, for ZnO film thickness of 5, 9, and 19 nm, a three layer model (i.e., air/ZnO layer/Si substrate) and for ZnO film thickness of 38, 52, and 69 nm, a four layer model (i.e., air/surface-roughness layer/ZnO layer/Si substrate) have been used to extract the optical constants of the films, while the optical constants of Si are well known.52 The Si substrate is about 1 mm thick and has been treated as infinite. The surface roughness factor is omitted from the model for the thinner films because the SE is not sensitive enough to detect the surface roughness in very thin films.

Fig. 4.

(Color online) Ellipsometric angles ψ and Δ as a function of photon energy for ZnO films of (a) 69 nm, and (b) 19 nm film thickness, grown on Si substrates, acquired with incidence angles ranging from 60° to 70°; circles and lines represent experimental data and theoretical fit, respectively.

Fig. 4.

(Color online) Ellipsometric angles ψ and Δ as a function of photon energy for ZnO films of (a) 69 nm, and (b) 19 nm film thickness, grown on Si substrates, acquired with incidence angles ranging from 60° to 70°; circles and lines represent experimental data and theoretical fit, respectively.

Close modal

ZnO is a direct band gap semiconductor with a fundamental absorption edge at about 3.37 eV, which corresponds to the direct transition.53 It is known that the excitonic interaction close to the fundamental band edge strongly influences the optical properties of ZnO.53,54 For the parameterization of the optical properties of ZnO, a Tauc-Lorentz (T-L) model has been employed in this work. The T-L oscillator was developed by Jellison and Modine.55,56 The complex dielectric function (ε = ε1 + 2) of the ZnO thin films as a function of the photon energy can be written in the following functional form known as the T-L model:47,55,56

ε2(E)=AE0C(EEg)2(E2E02)2+C2E21E(E>Eg)=0(EEg),
(8)
ε1(E)=ε+2Pπξε2(ξ)ξ2E2dξ.
(9)

The T-L oscillator model is based on the Tauc joint density of states and the Lorentz oscillator; the fit parameters are A, E0, C, Eg, and ε.47,55 The parameter A stands for the transition matrix element, which is proportional to the magnitude of the real and imaginary part of complex dielectric constant.47 E0 corresponds to the peak transition energy, which is related to the Penn gap which represents an average separation between the valence band and conduction band.47 C is the broadening term, related to the degree of disorder in the material.47Eg stands for the optical band gap, ε is the high frequency dielectric constant, and P stands for the principal part of the integral.47,55

The calculated ε1 and ε2 using the best-fit parameters for ZnO thin films on Si with various thicknesses are shown in Fig. 5. As can be seen in Fig. 5, the film thickness has a significant effect on the complex dielectric functions of ZnO films on Si. The magnitudes of both the real and imaginary parts of dielectric functions near the band edge significantly drop in thin ZnO films on Si (i.e., at and below 20 nm film thickness). As compared to 69 nm film, the ZnO film on Si with a thickness of 9 nm shows a reduction of ∼34% and ∼57% in the real and imaginary parts of the complex dielectric function at the photon energy of ∼3.3 eV, respectively, while the ZnO film on Si with a thickness of 5 nm shows a drastic reduction of ∼46% and ∼73%, respectively. The peaks of ε1 and ε2 also show a blue shift with the decreasing film thickness (at and below ∼20 nm).

Fig. 5.

(Color online) Evolution of (a) real part (ε1) and (b) imaginary part (ε2) of the complex dielectric function with ZnO film thickness. These ZnO films were grown on Si.

Fig. 5.

(Color online) Evolution of (a) real part (ε1) and (b) imaginary part (ε2) of the complex dielectric function with ZnO film thickness. These ZnO films were grown on Si.

Close modal

To explain the significant drop in the magnitudes of both the real and imaginary parts of the complex dielectric functions, we compare the extracted electron density coverage, ρcov (by fitting the XRR data using Parratt formalism39), with the relative (ε − 1) (which is proportional to valence electron density, unless the transition matrix element changes57) of ZnO, versus ZnO film thickness in Fig. 6. The relative electron density (ρrel) was estimated using the following equation:

ρrel=electron density of a given ZnO filmbulk ZnO electron density×100.
(10)

ε was calculated at a photon energy of 0.8 eV. The relative (ε − 1) was estimated using the following equation:

(ε1)rel=(ε1)ofagivenZnOfilm(ε1)of69nmZnOfilm×100.
(11)

Figure 6 shows that for the same thickness range, the valence electron density, proportional to (ε − 1), starts decreasing more sharply at and below 20 nm film thickness, and drops by ∼45% in the thinnest film. On the other hand, XRR results of ZnO films on Si indicate that the electron density of ZnO films remains constant down to 20 nm thickness, and then decreases only a little, by less than 7%. A slight decrease of the electron density in our thinnest films, obtained from XRR, is not sufficient to explain the following phenomena: (1) the significant drop in (ε − 1) and (2) the significant reduction in near-band gap absorption.

Fig. 6.

(Color online) Plot of the extracted relative electron density, ρrel (by fitting the XRR data using the Parratt formalism), and relative (ε − 1), (ε − 1)rel of ZnO, vs ZnO film thickness. Si was used as the substrate for these films.

Fig. 6.

(Color online) Plot of the extracted relative electron density, ρrel (by fitting the XRR data using the Parratt formalism), and relative (ε − 1), (ε − 1)rel of ZnO, vs ZnO film thickness. Si was used as the substrate for these films.

Close modal

The reductions in the complex dielectric functions could be explained by using the concept of the Tanguy–Elliot amplitude prefactor. According to Tanguy's theory for excitonic absorption, the complex dielectric function, (ε = ε1 + 2), can be written34,35 as

ε(E)=AR(E+iΓ)2{ga(ξ(E+iΓ))+ga(ξ(EiΓ))2ga(ξ(0))},
(12)

where the amplitude prefactor

A=2q22πε0m02(2μ2)32|eMcv(0)|2,
(13)

where |eMcv(0)| is the dipole matrix element which corresponds to the overlap of electron and hole.34,R is the energy of the fundamental exciton bound state, Eg is the band gap energy, μ is the reduced mass of the exciton, and Γ is the broadening of the energy levels.34 

ga(ξ)=2lnξ2πcot(πξ)2ψ(ξ)1/ξ,
(14)
ξ(z)=REgz,
(15)
ψ(z)=dlnΓ(z)dz.
(16)

In our study, in thinner ZnO films on Si (at or below 20 nm film thickness), the exciton is deconfined because of the staggered band alignment. The electron remains mostly confined in the ZnO conduction band quantum well (at least at low temperatures, where thermal excitation of the electron across the barrier into the Si substrate can be ignored), while the photoexcited ZnO hole will quickly relax across the interface into the Si substrate. This reduces the overlap of the electron and hole wave functions and thus the amplitude prefactor, A, in the Tanguy dielectric function [Eq. (12)]. For thinner films, where the thickness is near or below the excitonic Bohr radius, the delocalization of excitons near the interface plays an important role. These excitons are effectively probed by optical techniques such as ellipsometry, where the excitonic absorption strength modulates the magnitude of the real and imaginary parts of the dielectric function.

However, for thicker ZnO films, with film thickness above ∼20 nm, the delocalization phenomenon will still occur at the ZnO/Si interface, but the contribution of the probed interface is much less, resulting in negligible effect in dielectric function.

To understand the phenomenon clearly, the absorption coefficients of ZnO/Si systems, shown in Fig. 7(a), were compared with those of the ZnO/SiO2 shown in Fig. 7(b). Figure 7(a) shows that wider-band gap ZnO deposited on narrower-band gap Si, with staggered type alignment, absorbs less light with decreasing film thickness, at and below ∼20 nm. Whereas in the case of ZnO deposited on wider band gap SiO2 corresponding to type-I quantum well, the thinner ZnO films show higher absorption coefficient at and above the band edge, shown in Fig. 7(b).

Fig. 7.

(Color online) (a) Absorption coefficients α (determined from Ellipsometry) of ZnO films of different thicknesses grown on Si at 200 °C. The inset of (a) shows wider-band gap ZnO deposited on narrower-band gap Si, with staggered type-II alignment and (b) absorption coefficients α (determined from UV-VIS) of ZnO films of equivalent thickness grown on fused Quartz (SiO2) at 200 °C. The inset of (b) shows ZnO deposited on wider band gap SiO2 corresponding to type-I quantum well.

Fig. 7.

(Color online) (a) Absorption coefficients α (determined from Ellipsometry) of ZnO films of different thicknesses grown on Si at 200 °C. The inset of (a) shows wider-band gap ZnO deposited on narrower-band gap Si, with staggered type-II alignment and (b) absorption coefficients α (determined from UV-VIS) of ZnO films of equivalent thickness grown on fused Quartz (SiO2) at 200 °C. The inset of (b) shows ZnO deposited on wider band gap SiO2 corresponding to type-I quantum well.

Close modal

In the case of ZnO deposited on Si, which is a staggered type-II quantum well, shown in inset of Fig. 7(a), the photogenerated holes at ZnO are deconfined to the Si substrate at the interface, whereas electrons are confined in ZnO thin film, causing a lowering of electron–hole overlap factor and thus exhibits low excitonic absorption in thinner films, consequently bringing about a decrease in real and imaginary parts of the dielectric function.

The inset of Fig. 7(b) illustrates the ZnO/SiO2 system, which is a thin type-I quantum well, consisting of a narrower-bandgap semiconductor grown on a wider-bandgap substrate. In this case, both the electron and the hole are confined in the ZnO film, which leads to an increase in the electron–hole overlap matrix element in thin ZnO films, resulting in higher absorption coefficient at the band edge. In our study, the absorption coefficients for ZnO/Si samples were calculated from ellipsometry data, while the absorption coefficients for ZnO/SiO2 samples were estimated from UV-VIS absorption spectra, measured in the transmission mode. The absorption coefficients of thicker ZnO films (∼69 nm) for both the systems, ZnO/Si and ZnO/SiO2, show similar values (∼1.5 × 105 cm−1), indicating the negligible effect of the interface (ZnO/substrate). With decreasing film thickness, the absorption coefficient of thinner ZnO films (at and below ∼20 nm) was found to decrease for ZnO/Si samples whereas that increases for ZnO/SiO2 samples. These observations show that the thickness dependent variations of the optical properties of ZnO thin films are also influenced by the choice of the substrates.

ZnO thin films of different film thicknesses, ranging from ∼70 to ∼5 nm, were grown by ALD at 200 °C with DEZn and water precursors. The films were grown on two different substrates, namely, Si (100) and fused quartz (SiO2). The thickness dependent optical properties of the ALD grown thin films and the effect of the substrates were studied in depth. Their thickness, structure, and morphology were characterized by XRR, XRD, SEM, and AFM. We have demonstrated from SE and UV-VIS absorption spectroscopy studies that for thinner ZnO films (at and below 20 nm) on Si, the absorption coefficient drops, while it increases for similar ZnO films on SiO2. These phenomena have been explained consistently in terms of Tanguy–Elliott amplitude prefactor. For ZnO on Si with its type-II band alignment, the exciton disintegrates near the interface, and therefore, the Tanguy–Elliott overlap factor drops, leading to a drop in the absorption coefficient. For ZnO on SiO2, on the other hand, the type-I band alignment strongly confines the excitons in thin ZnO films, increases the overlap factor, and therefore the absorption coefficient increases.

Our study establishes that for ZnO thin films, where the thickness is near or below the excitonic Bohr radius, the delocalization and localization of excitons near the interface (ZnO/substrate) based on the selection of the substrate play an important role in systematically tuning its optical properties.

The authors would like to acknowledge IIT Indore for all kinds of support to this work. The authors would also like to thank the Department of Science and Technology (DST), Government of India, and Sophisticated Analytical Instrument Facility (SAIF), IIT Bombay, for the SEM measurements, and the Nanoscale Research Facility (NRF) at IIT Delhi for the AFM measurements. The work at NMSU was supported by the National Science Foundation (DMR-1505172). This work was partially supported by the Department of Science and Technology (DST), India, Project No. SB/S2/CMP-077/2013 and Council of Scientific and Industrial Research (CSIR), India, Project No. 03 (1310)/14/EMR-II.

1.
J.
Anderson
and
C.
Van de Walle
,
Rep. Prog. Phys.
72
,
126501
(
2009
).
2.
A. B.
Djurišić
,
A. M. C.
Ng
, and
X. Y.
Chen
,
Prog. Quantum Electron.
34
,
191
(
2010
).
4.
D.
Bagnall
,
Y.
Chen
,
Z.
Zhu
,
T.
Yao
,
S.
Koyama
,
M. Y.
Shen
, and
T.
Goto
,
Appl. Phys. Lett.
70
,
2230
(
1997
).
5.
S. T.
Tan
,
B. J.
Chen
,
X. W.
Sun
,
W. J.
Fan
,
H. S.
Kwok
,
X. H.
Zhang
, and
S. J.
Chua
,
J. Appl. Phys.
98
,
013505
(
2005
).
6.
E.
Guziewicz
 et al,
J. Appl. Phys.
103
,
033515
(
2008
).
7.
P. F.
Carcia
,
R. S.
McLean
, and
M. H.
Reilly
,
Appl. Phys. Lett.
88
,
123509
(
2006
).
8.
A. M.
Peiro
,
P.
Ravirajan
,
K.
Govender
,
D. S.
Boyle
,
P.
O'Brien
,
D. D. C.
Bradley
,
J.
Nelson
, and
J. R.
Durrant
,
J. Mater. Chem.
16
,
2088
(
2006
).
9.
P.
Banerjee
,
W.-J.
Lee
,
K.-R.
Bae
,
S. B.
Lee
, and
G. W.
Rubloff
,
J. Appl. Phys.
108
,
043504
(
2010
).
10.
M.
Suchea
,
S.
Christoulakis
,
K.
Moschovis
,
N.
Katsarakis
, and
G.
Kiriakidis
,
Thin Solid Films
515
,
551
(
2006
).
11.
B. N.
Illy
 et al,
J. Mater. Chem.
21
,
12949
(
2011
).
12.
J.
Huang
,
Z.
Yin
, and
Q.
Zheng
,
Energy Environ. Sci.
4
,
3861
(
2011
).
13.
M.-J.
Jin
,
J.
Jo
,
G. P.
Neupane
,
J.
Kim
,
K.-S.
An
, and
J.-W.
Yoo
,
AIP Adv.
3
,
102114
(
2013
).
14.
N. O. V.
Plank
,
M. E.
Welland
,
J. L.
MacManus-Driscoll
, and
L.
Schmidt-Mende
,
Thin Solid Films
516
,
7218
(
2008
).
15.
T.-B.
Hur
,
Y.-H.
Hwang
, and
H.-K.
Kim
,
Appl. Phys. Lett.
86
,
193113
(
2005
).
16.
A. L.
Schoenhalz
,
J. T.
Arantes
,
A.
Fazzio
, and
G. M.
Dalpian
,
J. Phys. Chem. C
114
,
18293
(
2010
).
17.
B. P.
Zhang
,
N. T.
Binh
,
K.
Wakatsuki
,
C. Y.
Liu
,
Y.
Segawa
, and
N.
Usami
,
Appl. Phys. Lett.
86
,
032105
(
2005
).
18.
Y.
Gu
,
I. L.
Kuskovsky
,
M.
Yin
,
S.
O'Brien
, and
G. F.
Neumark
,
Appl. Phys. Lett.
85
,
3833
(
2004
).
19.
K.
Kitamura
,
T.
Yatsui
, and
M.
Ohtsu
,
Appl. Phys. B
97
,
825
(
2009
).
20.
H. T.
Ng
,
B.
Chen
,
J.
Li
,
J.
Han
,
M.
Meyyappan
,
J.
Wu
,
S. X.
Li
, and
E. E.
Haller
,
Appl. Phys. Lett.
82
,
2023
(
2003
).
21.
J. G.
Lu
,
Z. Z.
Ye
,
Y. Z.
Zhang
,
Q. L.
Liang
,
S.
Fujita
, and
Z. L.
Wang
,
Appl. Phys. Lett.
89
,
023122
(
2006
).
22.
K.-F.
Lin
,
H.-M.
Cheng
,
H.-C.
Hsu
,
L.-J.
Lin
, and
W.-F.
Hsieh
,
Chem. Phys. Lett.
409
,
208
(
2005
).
23.
J. C.
Nie
,
J. Y.
Yang
,
Y.
Piao
,
H.
Li
,
Y.
Sun
,
Q. M.
Xue
,
C. M.
Xiong
,
R. F.
Dou
, and
Q. Y.
Tu
,
Appl. Phys. Lett.
93
,
173104
(
2008
).
24.
A. A.
Mosquera
,
D.
Horwat
,
A.
Rashkovskiy
,
A.
Kovalev
,
P.
Miska
,
D.
Wainstein
,
J. M.
Albella
, and
J. L.
Endrino
,
Sci. Rep.
3
,
1714
(
2013
).
25.
X. D.
Li
,
T. P.
Chen
,
P.
Liu
,
Y.
Liu
,
Z.
Liu
, and
K. C.
Leong
,
J. Appl. Phys.
115
,
103512
(
2014
).
26.
T.
Tommi
and
K.
Maarit
,
Semicond. Sci. Technol.
29
,
043001
(
2014
).
27.
M.
Godlewski
 et al,
Thin Solid Films
518
,
1145
(
2009
).
28.
E.
Guziewicz
 et al,
J. Appl. Phys.
105
,
122413
(
2009
).
29.
Z.
Shi
and
A. V.
Walker
,
J. Phys. Chem. C
119
,
1091
(
2015
).
30.
R. W.
Johnson
,
A.
Hultqvist
, and
S. F.
Bent
,
Mater. Today
17
,
236
(
2014
).
31.
S. M.
George
,
Chem. Rev.
110
,
111
(
2010
).
32.
S. M.
George
,
A. W.
Ott
, and
J. W.
Klaus
,
J. Phys. Chem.
100
,
13121
(
1996
).
33.
N.
El-Atab
,
S.
Alqatari
,
F. B.
Oruc
,
T.
Souier
,
M.
Chiesa
,
A. K.
Okyay
, and
A.
Nayfeh
,
AIP Adv.
3
,
102119
(
2013
).
35.
C.
Tanguy
,
Phys. Rev. Lett.
76
,
716
(
1996
).
36.
Z.
Baji
,
Z.
Lábadi
,
Z. E.
Horváth
,
M.
Fried
,
B.
Szentpáli
, and
I.
Bársony
,
J. Therm. Anal. Calorim.
105
,
93
(
2011
).
37.
S.
Jeon
,
S.
Bang
,
S.
Lee
,
S.
Kwon
,
W.
Jeong
,
H.
Jeon
,
H. J.
Chang
, and
H.-H.
Park
,
J. Electrochem. Soc.
155
,
H738
(
2008
).
38.
S.
Chattopadhyay
and
A.
Datta
,
Macromolecules
40
,
3313
(
2007
).
39.
L. G.
Parratt
,
Phys. Rev.
95
,
359
(
1954
).
40.
S.
Chattopadhyay
and
A.
Datta
,
Phys. Rev. B
72
,
155418
(
2005
).
41.
S.
Chattopadhyay
,
A.
Uysal
,
B.
Stripe
,
G.
Evmenenko
,
S.
Ehrlich
,
E. A.
Karapetrova
, and
P.
Dutta
,
Phys. Rev. Lett.
103
,
175701
(
2009
).
42.
S.
Chattopadhyay
,
A.
Uysal
,
B.
Stripe
,
Y.-G.
Ha
,
T. J.
Marks
,
E. A.
Karapetrova
, and
P.
Dutta
,
Phys. Rev. Lett.
105
,
037803
(
2010
).
43.
C. V.
Weiss
,
J.
Zhang
,
M.
Spies
,
L. S.
Abdallah
,
S.
Zollner
,
M. W.
Cole
, and
S. P.
Alpay
,
J. Appl. Phys.
111
,
054108
(
2012
).
44.
N.
Ghobadi
,
Int. Nano Lett.
3
,
2
(
2013
).
45.
B. M.
Han
,
S.
Chang
, and
S. Y.
Kim
,
Thin Solid Films
338
,
265
(
1999
).
46.
Q. H.
Li
,
D.
Zhu
,
W.
Liu
,
Y.
Liu
, and
X. C.
Ma
,
Appl. Surf. Sci.
254
,
2922
(
2008
).
47.
J. P.
Zhang
,
G.
He
,
L. Q.
Zhu
,
M.
Liu
,
S. S.
Pan
, and
L. D.
Zhang
,
Appl. Surf. Sci.
253
,
9414
(
2007
).
48.
H.
Fujiwara
,
Spectroscopic Ellipsometry: Principles and Applications
(
Wiley
,
West Sussex, England
,
2007
).
49.
R. M.
Azzam
and
N. M.
Bashara
,
Ellipsometry and Polarized Light
(
North-Holland/Elsevier Science
,
Amsterdam, Netherlands
,
1987
).
50.
H. G.
Tompkins
and
W. A.
McGahan
,
Spectroscopic Ellipsometry and Reflectometry: A User's Guide
(
Wiley
,
New York
,
1999
).
51.
H.
Tompkins
and
E. A.
Irene
,
Handbook of Ellipsometry
(
William Andrew
,
New York
,
2005
).
52.
C. M.
Herzinger
,
B.
Johs
,
W. A.
McGahan
,
J. A.
Woollam
, and
W.
Paulson
,
J. Appl. Phys.
83
,
3323
(
1998
).
53.
S.
Logothetidis
,
A.
Laskarakis
,
S.
Kassavetis
,
S.
Lousinian
,
C.
Gravalidis
, and
G.
Kiriakidis
,
Thin Solid Films
516
,
1345
(
2008
).
54.
G. E.
Jellison
and
L. A.
Boatner
,
Phys. Rev. B
58
,
3586
(
1998
).
55.
G. E.
Jellison
and
F. A.
Modine
,
Appl. Phys. Lett.
69
,
371
(
1996
).
56.
G. E.
Jellison
and
F. A.
Modine
,
Appl. Phys. Lett.
69
,
2137
(
1996
).
57.
A. M.
Fox
,
Optical Properties of Solids
(
Oxford University
,
New York
,
2001
), Vol.
3
.