Chapter 2 Characterizations and analyses of the pressure- and doping-
2.2 Motivation and investigation procedures
In this chapter, the pressure- and doping-dependent ZnO:B thin flims grown on glass substrates by a LPCVD were investigated.
Fig. 2.2 shows the experimental procedures of this study. The growth methods and conditions of ZnO:B thin films on glass substrates will be described in the section 2.3. The surface morphologies of samples are measured by the atomic force microscopy (AFM) and scanning electron microscope (SEM) measurements. The film quality and optical properties of
the samples are studied by the cathodoluminescence (CL). The crystallization of ZnO grains and their properties are determined by X-ray diffraction (XRD) measurement. The strains of ZnO:B grains are analyzed by Raman scattering measurement. Analyzing the results of these measurements, we will discuss the characteristics and properties of the pressure- and doping-dependent ZnO:B thin flims grown on glass substrates. The experimental results and their theoretical explanations will be discussed in the Section 2.4. A brief conclusion will be summarized in the Section 2.7.
2.3 Sample structures and growth conditions of the pressure-dependent ZnO:B films grown on glass substrates
Fig. 2.3 shows the sample structures of the pressure-dependent ZnO:B grown on glass substrates by a low pressure chemical vapor deposition (LPCVD). Glass substrates are prepared and cut into 22cm2 pieces. Two ZnO:B films with 0.4 and 0.6 torr growth pressures and 1.209 and 1.013 μm in thickness (denoted as the samples BZO0.4, and BZO0.6, respectively) were deposited on glass substrates at 170oC temperature by a LPCVD. The deposition time is 70 mins. Diethylzinc (DEZ) and water (H2O) vapors carried by argon gas were used as precursors, and their flows were set to 500 and 550 sccm, respectively. The boron doping is accomplished by controlling the flow rate of B2H6 to be 5 sccm.
2.4 Characterizations and analyses of the pressure-dependent ZnO:B films grown on glass substrates
2.4.1 X-Ray diffraction and lattice strain
The sample structures are analyzed by using the XRD. The results are shown in Fig.
2.4(a) There are three main diffraction peaks from the crystalline ZnO:B grains: (100), (002), and (110) peaks. Generally, the percentages of grains associated with different peaks are proportional to their correspondent XRD integrated intensities. As a result, the percentages of
(100), (002), and (110) grains obtained from XRD data are shown in Fig. 2.4(b). The peak integrated intensities, full width at half maximum (FWHM), and grain percentages of the (100), (002), and (110) peaks associated with three different crystalline orientations of ZnO:B grains are listed in the Table 2.1. Lattice constants a and c obtained from the XRD peaks of the BZO0.4
and BZO0.6 samples are shown in Table 2.2.
Fig. 2.5 shows the peak intensities, FWHMs, and grain percentages of the (100), (002), and (110) XRD peaks of the BZO0.4 and BZO0.6 samples. In addition, Fig. 2.6 shows the grain sizes, strains, and averaged grain sizes obtained from (100), (002), and (110) XRD peaks of the two samples.
The ZnO:B grains on glass substrates have three major crystalline orientations: (100), (002), and (110). Their grain sizes D can be calculated from the following formula [3]
diffraction angle of X-rays, respectively. As a result, the average grain sizes associated with different peaks can be calculated by using this formula. As shown in Table 2.1, the average grain size of the BZO0.4 sample on the glass substrate are larger than that of the BZO0.6 sample.
These results can also be somehow verified by the SEM images which also show that the surface granular textures on the BZO0.4 sample are slightly larger than those on the BZO0.6
sample.
The XRD data can be analyzed to obtain not only the percentages of different grain crystal orientations but also the strains associated with them. The strains associated with the diffraction peaks can be calculated by the [3]
From the results of the XRD, the average grain sizes of the zinc oxides decrease due to the increases of their average strains. The strains of the BZO0.4 samples are slightly smaller than those of the BZO0.6 sample.
2.4.2 Scanning electron microscope (SEM), cathodo-luminescence (CL), and atomic force microscopy (AFM) studies
The surface structures from the results of the SEM are shown in the Fig. 2.7. The relations between microstructures and crystal growth orientations of ZnO grown on glass substrates are schematically illustrated in Fig. 2.8 [4,5]. ZnO(100) grains expose polygon-like structures with c-axis parallel to the substrate, while ZnO(110) grains will disclose ridge-like structures with c-axis parallel to the substrate. In contrast, ZnO(002) grains, with c-axis perpendicular to the substrate, will reveal their hexagonal cylinder with or without the pyramidal tip, or pyramidal tip without the hexagonal cylinder. In short, ZnO(110), ZnO(100), and ZnO(002) grains will show ridge-like, polygon-like, and hexagonal cylinder surface structures, respectively. As a result, it is very difficult to estimate the percentage of these two different structures purely from the SEM images.
Fig. 2.9 shows the AFM images of the 5 μm × 5 μm samples of (a) BZO0.4 (RQ 24.092nm) and (b) BZO0.6 (RQ 15.813nm). Surface roughness of each sample, Rq, is shown in the parentheses. From the AFM images, the surface roughness of the BZO0.4 is much sharper than that BZO0.6. Therefore, from the results of the AFM images, it can be conjectured that the BZO0.4 have more ridge-like structures than the hexagonal cylinder ones. These results certainly indicate that the ZnO(110) grains have the largest percentage on the BZO0.4 sample and thus show their ridge-like structures to the surface. From the SEM and AFM results, it can be concluded the thicker film of the BZO0.4 sample has the largest percentage the ZnO(110) grains with ridge-like surface structures.
Fig. 2.10 shows CL spectra of samples (a) BZO0.4 and (b) BZO0.6 with the excitations of 16, 18, and 20kV electron voltages at room temperature. In general, thicker films have stronger CL intensities and narrow FWHM, since thicker films have a better crystallization.
In addition, the average ZnO:B grain sizes can be indirectly verified from the results of the CL measurements. Since larger grain size will have more crystalline structures and thus less defects, and therefore stronger CL peak intensity, it can be expected that CL emission intensities are linearly proportional to their associated average grain size. The average ZnO:B grain sizes estimated from the XRD almost fully agree with those from the CL measurements.
2.4.3 Estimation of strain by Raman scattering measurement
The Raman scattering spectra, as shown in Fig. 2.11, is a useful tool to analyze the strain for crystalline materials by the equation as [6]
2 )
where the deformation potential constants a=-580cm-1 and b=-765cm-1 are E2-high mode [7].
The elastic stiffness constants C33 and C13 are 216 and 104 Gpa, respectively [8].
Raman frequency shift EZnO2(High) of E2(High) mode, strain , Rq, CL, and surface morphology of samples BZO0.4 and BZO0.6 are shown in Table 2.2. The strains on the BZO0.4
sample are slightly smaller than those on the BZO0.6 sample. The strains calculated from the Raman scattering strongly agree with those calculated from the XRD.
The lattice constants a and c of the BZO0.4 and BZO0.6 are shown in Fig. 2.12. For the BZO0.4 sample, the crystallization of ZnO(100) grains is compressive along the a-axis. As the growth pressure increases, their compressive strains along the a-axis will increase and thus their crystallization will become less probable. As a result, their crystallization and grain percentages will decrease with the increase of growth pressure. Also, ZnO(110) grains are
compressively strains along the a-axis slightly increase. This indicates that the crystallization of ZnO(110) grains will become more favorable as the pressure decreases.
Fig. 2.13 shows the transmittance spectra for the two samples. The ZnO transmittance decreases as the pressure decreases, particularly in the near infrared (NIR) region. Since a sample with larger grain size has more crystalline structures, it can be expected that a thicker sample shows a lower transmittance. Also, as the pressure decreases, the sharper absorption edge in the 360-380 nm spectral range is consistent with the more uniform distribution of grain size and the red-shifted absorption edge is consistent with the weaker quantum size effect. The bandgap energy of semiconductor, Eg, can be estimated by extrapolating the linear portion of the transmittance to zero. As shown in Table 2.2, for the low-pressure sample with a larger thickness, the red-shifted Eg is consistent with the weaker quantum size effect [9].
2.5 Sample structures and growth conditions of the doping-dependent ZnO:B films grown on glass substrates
Fig. 2.14 shows the sample structures of the doping-dependent ZnO:B films grown on glass substrates by a LPCVD with different film thicknesses. Glass substrates are prepared and cut into 22cm2 pieces. These glass substrates were dipped into 5% HF solutions to remove their native oxide layers and then rinsed in de-ionized water for 3 minutes. The total pressure is 0.6 torr and temperature is 170oC during deposition. The deposition time is 70 mins.
Diethylzinc (DEZ) and water (H2O) vapors carried by argon gas were used as precursors, and their flow rates were set to 500 and 550 sccm, respectively. The boron doping is accomplished by controlling the flow rates of B2H6 to be 1 and 3 sccms (denoted as the samples BZO1 and BZO3, respectively) with corresponding 1.589 and 1.563 μm in film thicknesses.
2.6 Characterizations and analyses of the doping-dependent ZnO:B films grown on glass substrates
2.6.1 X-Ray diffraction and lattice strain
As shown in Fig. 2.15(a), there are three main XRD diffraction peaks from the crystalline ZnO:B grains: (100), (002), and (110) peaks. Generally, the percentages of grains associated with different peaks are proportional to their correspondent XRD integrated intensities. As a result, the percentages of (100), (002), and (110) ZnO:B grains obtained from XRD result are shown in Fig. 2.15(b). The peak and integrated intensities, and the full width at half maximum (FWHM) of the (100), (002), and (110) peaks associated with three different crystalline orientations of ZnO:B grains are listed in the Table 2.3. Lattice constants a and c obtained from the XRD peaks of the BZO1 and BZO3 samples are shown in Table 2.4.
Fig. 2.16 shows the peak intensities, FWHMs, and grain percentages of the (100), (002), and (110) XRD peaks of the BZO1 and BZO3 samples. In addition, Fig. 2.17 (a), (b), and (c) shows the grain sizes, strains, and averaged grain sizes versus strain of the two samples, respectively.
The ZnO:B grains on glass substrates have three major crystalline orientations: (100), (002), and (110). Their grain sizes D can be calculated from the following formula [3]
diffraction angle of X-rays, respectively. As a result, the average grain sizes associated with different peaks can be calculated by using this formula. The calculated results are shown in Table 2.3. The average grain size of the BZO1 sample is larger than that of the BZO3 sample.
The XRD data can be analyzed to obtain not only the percentages of different grain crystal orientations but also the strains associated with them. The strains associated with the diffraction peaks can be calculated by the [3]
h k l
where
hkl FWHM
, and is x-ray diffraction. The strain of the BZO1 sample is slightly smaller than that of the BZO3 sample. From the results of the XRD, the average grain sizes of the zinc oxides decrease due to the increases of their average strains.2.6.2 Scanning electron microscope (SEM), cathodo-luminescence (CL), and atomic force microscopy (AFM) studies
The surface structures from the results of the SEM are shown in Fig. 2.18. As schematically illustrated in Fig. 2.8, the ZnO(100) grains will form a polygon structure with the c-axis parallel to the substrate surface. On the other hand, the ZnO(002) grains will form hexagonal cylinder with or without the pyramidal tip. In addition, the ZnO(110) grains will reveal ridge-like structures to the surface. The SEM images in Fig. 2.18 show that ZnO grains on glass substrates have many ridge-like as well as pyramid-like structures. These results can also be somehow verified by the SEM images which also show that surface granular textures the BZO1 sample on the glass substrate are slightly larger than those of the BZO3 sample.
Fig. 2.19 shows the AFM images of the 5 μm × 5 μm of the samples (a) BZO1 (Rq 25.856nm) and (b) BZO3 (Rq 19.386nm). The surface roughness of each sample, Rq, is indicated in the parentheses. However, from the AFM images, the surface roughness of the BZO1 is much sharper than that BZO3. Therefore, from the results of the AFM images, it can be conjectured that the BZO1 and BZO3 have more ridge-like structures than the hexagonal cylinder ones. These results certainly indicate that The ZnO(110) grains have the largest percentage on the BZO1 and BZO3 samples and thus show their ridge-like structures to the surface. From the SEM and AFM results, it can be concluded the thickest film of the BZO1 and BZO3 samples in this study has the largest percentage the ZnO(110) grains with ridge-like surface structures.
2.6.3 Estimation of strain by Raman scattering measurement
The Raman scattering spectra, as shown in Fig. 2.20, is a useful tool to analyze the strain for crystalline materials by the equation as [6]
2 )
where the deformation potential constants a=-580cm-1 and b=-765cm-1 are E2-high mode [7].
The elastic stiffness constants C33 and C13 are 216 and 104 Gpa, respectively [8].
Raman frequency shift EZnO2(High) of E2(High) mode, strain , Rq, CL, and surface morphology of samples BZO1 and BZO3 are shown in Table 2.4. The strains on the BZO1
sample are slightly smaller than those on the BZO3 sample. The strains calculated from the Raman scattering strongly agree with those calculated from the XRD.
The lattice constants a and c of the BZO1 and BZO3, as shown in Fig. 2.21. For the BZO1
sample, the crystallization of ZnO(110) grains is compressive along the a-axis. As the growth flow rate increases, their compressive strains along the a-axis will decrease and thus their crystallization will become more probable. On the contrary, ZnO(002) grains are tensilely strained along the c-axis for the BZO1 sample. As the growth flow rate increases, their tensile strains along the c-axis slightly increase. This indicates that the crystallization of ZnO(002) grains will become less favorable as the growth flow rate increases.
Fig. 2.22 shows the transmittance spectra for the two samples. The ZnO transmittance slightly increases as the doping increases. Since samples with larger grain size have more crystalline structures, it can be expected that a thicker sample shows a lower transmittance. The bandgap energy of semiconductor, Eg, can be estimated by extrapolating the linear portion of the transmittance to zero. As shown in Table 2.4, the red-shifted Eg for the low-doping sample is consistent with the weaker quantum size effect [9].
The pressures-dependent ZnO:B films grown on glass substrates by a LPCVD are analyzed by XRD, SEM, AFM, CL, and transmittance measurements. Two ZnO:B films with 0.4 and 0.6 torr growth pressures (denoted as the sample BZO0.4, and BZO0.6, respectively) and 1.209 and 1.013 μm in thickness were deposited on glass substrates at 170oC temperature. Both the BZO0.4 and BZO0.6 samples particularly favor the growth of ZnO(110) grain structures. As the pressure increases, the ZnO(110) grain percentage increases. The average grain size of the BZO0.4 sample are larger than that of the BZO0.6 sample. When the growth pressure increases, the average grain size decreases and average strain increases relatively. The grain size of BZO is proportional to the thickness, but inversely to strain. Transmittance spectra are inversely proportional to thickness. These XRD results are shown to strongly agree with the SEM, AFM, CL, and transmittance measurements as well.
The doping-dependent ZnO:B films grown on glass substrates by a LPCVD are analyzed by XRD, SEM, AFM, CL, and transmittance measurements. The boron doping is accomplished by controlling the flow rates of B2H6 to be 1 and 3 sccms (denoted as the samples BZO1 and BZO3, respectively) with corresponding 1.589 and 1.563 μm in film thicknesses. The XRD result shows that the ZnO:B films particularly favor the growth of ZnO(110) ridge-like grain structures. These results also indicate that the average grains of the BZO1 sample are larger than that of the BZO3 sample. When the growth flow rate increases, the average grain size decreases and average strain increases relatively. Therefore, the strains of the BZO1 sample are slightly smaller than those of the BZO3 sample. The grain size of BZO is proportional to the thickness, but inversely to strain. Transmittance spectra are inversely proportional to thickness.
These XRD results are shown to strongly agree with the SEM, AFM, CL, and transmittance measurements as well.
References
[1] U. H. Frank, J. Phys.: Conf. Ser. 764, 012001 (2016).
[2] K. Ellmer, A. Klein, B. Rech, Transparent Conductive Zinc Oxide, Springer Series in Material Science 104, 235 (2008).
[3] V. D. Mote, Y. Purushotham, B. N. Dole, J. Theor. Appl. Phys. 6 6 (2012).
[4] Y. H. Hu, Y. C. Chen, H. J. Xu, H. Gao, W. H. Jiang, F. Hu, Y. X. Wang, Engineering 2 973-978 (2010).
[5] T. Shinagawa, K. Shibata, O. Shimomura, M. Chigane, R. Nomura, M. Izaki, J. Mater.
Chem. C 2 2908-2917 (2014).
[6] T. Gruber, G. M. Prinz, C. Kirchner, R. Kling, F. Reuss, W. Limmer, and A. Waag, J.
Journal of Applied Physics 96, 289 (2004).
[7] G. Callsen, J. S. Reparaz, M. R. Wagner, R. Kirste, C. Nenstiel, A. Hoffmann, and M. R.
Phillips, Appl. Phys. Lett. 98, 061906 (2011).
[8] H. Morkoc and Ü . Ö zgür, Zinc Oxide: Fundamentals, Materials and Device Technology (Weinheim: Wiley-Vch, 2007).
[9] K. Y. Kou, Y. E. Huang, C. H. Chen, S. W. Feng, Beilstein J. Nanotechnol. 75 75-80 (2016)
BZO0.4 BZO0.6
Peak intensity ZnO(100) (arb. unit) 208 152 Peak intensity ZnO(002) (arb. unit) 280 66 Peak intensity ZnO(110) (arb. unit) 654 460 FWHM ZnO(100) (degree) 0.31135 0.36461 FWHM ZnO(002) (degree) 0.33396 0.45349 FWHM ZnO(110) (degree) 0.45612 0.56896 Integrated intensity ZnO(100) (arb. unit) 135.7 110.06 Integrated intensity ZnO(002) (arb. unit) 155.68 23.3 Integrated intensity ZnO(110) (arb. unit) 360.76 339.76 Grain percentage ZnO(100) (%) 0.11693 0.13269 Grain percentage ZnO(002) (%) 0.18046 0.03862 Grain percentage ZnO(110) (%) 0.7026 0.82868 Grain size ZnO(100) (nm) 46.33245 39.57873 Grain size ZnO(002) (nm) 43.46972 32.02652 Grain size ZnO(110) (nm) 34.57508 27.73996 Average grain size (nm) 37.55508 29.47645
Strain ZnO(100) 0.27273 0.31786
Strain ZnO(002) 0.27058 0.36553
Strain ZnO(110) 0.21085 0.2621
Average strain 0.22887 0.27349
Table 2.1 Peak intensities, FWHMs, and integrated intensities of XRD peaks and the corresponding grain percentages, average grain sizes, and average strains of the BZO0.4 and BZO0.6 samples.
Sample BZO0.4 BZO0.6 Thickness (nm) 1209 1013 Main grain (110) (110) Surface morphology Ridge-like Ridge-like Rq (nm) 24.092 15.813
D (nm) of ZnO: B(100) (nm) 46.332 37.578
D (nm) of ZnO: B(110) (nm) 43.469 32.026
D (nm) of ZnO: B(002) (nm) 34.575 27.739
Average grain size (nm) 37.555 29.476 Average strain 0.2288 0.2734
Lattice constant a from ZnO: B(100) (Å ) 3.2437 3.2505 Lattice constant a from ZnO: B(110) (Å ) 3.2416 3.2474 Average a (Å ) 3.2512 3.2499
Lattice constant c from ZnO: B(002) (Å ) 5.2293 --- CL intensity Stronger Weaker Eg (eV) 3.3494 3.3904 Raman shiftEZnO2(High)(nm-1)
-1.1379 -2.2288
Strain (%) 0.2687 0.5264
Table 2.2 Thickness, main grain, surface morphology, Rq of AFM, grain size (D), strain, lattice constants a and c, CL intensity, and energy gap (Eg) of the BZO0.4 and BZO0.6 samples.
BZO1 BZO3
Peak intensity ZnO(100) (arb. unit) 1811 1362 Peak intensity ZnO(002) (arb. unit) 2094 1684 Peak intensity ZnO(110) (arb. unit) 3462 3391 FWHM ZnO(100) (degree) 0.25056 0.26478
FWHM ZnO(002) (degree) 0.27544 0.27505
FWHM ZnO(110) (degree) 0.37351 0.37252
Integrated intensity ZnO(100) (arb. unit) 1046.52 1204.92 Integrated intensity ZnO(002) (arb. unit) 1218.22 1100.34 Integrated intensity ZnO(110) (arb. unit) 2112.08 2231.16 Grain percentage ZnO(100) (%) 0.17963 0.17077 Grain percentage ZnO(002) (%) 0.24425 0.21613 Grain percentage ZnO(110) (%) 0.57611 0.61311 Grain size ZnO(100) (nm) 57.56693 54.47649 Grain size ZnO(002) (nm) 52.70208 52.77675 Grain size ZnO(110) (nm) 42.20154 42.31998 Average grain size (nm) 47.52648 46.6559
Strain ZnO(100) 0.21981 0.23222
Strain ZnO(002) 0.22332 0.22301
Strain ZnO(110) 0.17304 0.17247
Average strain 0.19372 0.19394
Table 2.3 Peak intensities, FWHMs, and integrated intensities of XRD peaks and corresponding grain percentages, average grain sizes, and average strains of the BZO1 and BZO3 samples.
Sample BZO1 BZO3
Thickness (nm) 1,589 1,563 Main grain (110) (110)
Surface morphology Ridge-like Ridge-like Rq (nm) 25.856 19.386
D (nm) of ZnO: B(100) (nm) 57.566 54.476
D (nm) of ZnO: B(110) (nm) 42.201 42.319
D (nm) of ZnO: B(002) (nm) 52.702 52.395
Average grain size (nm) 47.526 46.655 Average strain 0.1937 0.1939
Lattice constant a from ZnO: B(100) (Å ) 3.2434 3.2471 Lattice constant a from ZnO: B(110) (Å ) 3.2410 3.2449 Average a (Å ) 0.27349 3.24589
Lattice constant c from ZnO: B(002) (Å ) 5.2293 5.2326 CL intensity Stronger Weaker Eg (eV) 3.3243 3.3288 Raman shiftEZnO2(High)(nm-1)
-0.0667 -1.1324
Strain (%) 0.0157
0.2674
Table 2.4 Thickness, main grain, surface morphology, Rq of AFM, grain size (D), lattice constants a and c, CL intensity, and energy gap (Eg) of the BZO1 and BZO3 samples.
Figure 2.1 Feature size, haze, film thickness, and sheet resistance as a function of diborane flow.
Figure 2.2 Experimental flow chart of this chapter.
To investigate transmittance of samples
Pressure- and doping- dependent BZO grown on glass substrates by a LPCVD
Sample preparations
To measure the surface morphology of samples To investigate strain and grain
size of samples
To calculate strain of samples Atomic Force Microscopy (AFM)
Measurement
Raman Measurement X-ray Diffraction (XRD)
Measurement
UV-Visible Transmittance Spectra Study
To investigate microstructures and nanophotonics of samples Scanning Electron Microscope
(SEM) and Cathodoluminescence (CL) Measurements
(a) (b)
Figure 2.3 Structures of the BZO0.4 and BZO0.6 samples grown on glass substrates. 0.4 and 0.6 mean growth pressures in torr with corresponding thicknesses 1,209 and 1,013 nm, respectively.
BZO BZO
0
FW HM (d eg ree)
ZnO (100) ZnO (002)ZnO (110)
25
Figure 2.6 (a) Grain size, (b) strains obtained from XRD peaks, and (c) averaged grain size versus XRD strain of the BZO0.4 and BZO0.6 samples.
Figure 2.7 SEM images of the (a) BZO0.4 and (b) BZO0.6 samples.
Figure 2.8 ZnO grown on glass substrates. (a) For ZnO(100), a polygon structure for the c-axis growth parallel to the substrate; (b) For ZnO(110), the ZnO grains will reveal ridge-like structures; (c) For ZnO(002), the ZnO grains will form hexagonal cylinder with or without the pyramidal tip; (d) For ZnO(101), the ZnO grains mainly form the pyramidal tip without the hexagonal cylinder.
Figure 2.9 AFM images (5 μm × 5 μm) of the (a) BZO0.4 (Rq 24.092nm) and (b) BZO0.6 (Rq 15.813nm) samples. Surface roughness of each sample, Rq, is shown in the parentheses.
Figure 2.9 AFM images (5 μm × 5 μm) of the (a) BZO0.4 (Rq 24.092nm) and (b) BZO0.6 (Rq 15.813nm) samples. Surface roughness of each sample, Rq, is shown in the parentheses.