Formation of plasmonic silver nanoparticles using rapid thermal annealing at low temperature and study in reflectance reduction of Si surface

In order to confirm the plasmonic behavior of AgNPs by RTP, Ag thin film was deposited on cleaned glass substrate by sputtering followed with RTP at 250 °C under the conditions mentioned in above experimental details. Plasmonic AgNPs layer generation using RTP at 250 °C temperature for different time intervals (5–30 min) onto glass is confirmed by absorbance spectra as shown in figure 1. AgNPs show plasmonic behavior with two peaks in absorbance spectrum. The primary peak at around 450 nm is responsible for surface plasmon resonance while secondary peak at around 360 nm is responsible for quadruple resonance. The plasmonic peak position of silver can be varied by tuning the size, shape and material of the nanoparticle as well as depends on the dielectric medium surrounding the particle [17, 18]. Figure 1 clearly shows plasmon peak for samples 1, 2, 3, 4 and 5 at 463.5 nm, 453.38 nm, 447.5 nm, 452.9 nm and 453.1 nm, respectively, with particle sizes 107.6 nm, 104.5 nm, 85.1 nm, 88.5 nm and 96.1 nm, respectively. A blue shift of plasmon resonance peak along with decrease in AgNPs sizes is observed from 5 min to 15 min RTP. A red shift is started from 20 min to 30 min RTP with increase in Ag NPs size with respect to 15 min RTP. The secondary plasmon peak in each absorbance spectrum observed around 360 nm can be attributed to quadruple resonance [18].

Zoom In Zoom Out Reset image size

The surface morphology of as deposited Ag thin film on c–Si (1 0 0) substrate is characterized by SEM and FESEM measurements and are shown in figure 2. The figure 2(a) shows nonvisibility of surface features in SEM image therefore FESEM measurement (figure 2(b)) was done at higher magnification and resolution which depicts non-continuous Ag thin film with flattened particles arrangement.

Zoom In Zoom Out Reset image size

Figures 3(a)–(e), figures 4(a)–(e) and figures 5(a)–(e) show SEM images depicting effect of RTP time duration (5, 10, 15, 20 and 30 min) and temperature (200 °C, 250 °C and 300 °C) for fabrication of AgNPs layer on polished c–Si substrate. Figures 3–5 show variation in surface morphology of formed AgNPs for different annealing durations of 5, 10, 15, 20 and 30 min and confirms that the average particles size of AgNPs is decreasing from 5 min to 20 min RTP durations and minimized for 20 min annealing in respective annealing temperatures as shown in figures 3(f), 4(f) and 5(f). Further rise in RTP duration to 30 min shows increase in average particle size at respective variable temperatures in all the figures 3–5. The shape of AgNP is found to be asymmetrical on increasing RTP temperature. Lowest average AgNPs with size of 60.42 nm is observed for 20 min RTP of Ag thin film at 200 °C.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

The average particle size variation with RTP duration at temperatures 200 °C, 250 °C and 300 °C are shown in figures 3(f), 4(f) and 5(f), respectively. The calculated average particle sizes from SEM images are summarized in table 1.

Thus SEM studies reveal that AgNPs sizes in irregular arrangement are decreasing with increase of RTP duration and minimized for 20 min RTP at temperatures 200 °C, 250 °C and 300 °C.

The changes in surface morphology of AgNPs due to RTP at different temperatures for various time durations can be explained by considering thin film growth mechanism. A typical thin film growth processes consists of several steps like adsorption, cluster formation, surface diffusion, desorption, nucleation, lateral growth leading to island formation, agglomeration of the island for coalescence and then filling of channels and holes in porous network of coalescence particles which can result into a continuous film [19–21].

As deposited silver thin film on Si surface by RF sputtering has intrinsic stress and found to be compressive in nature [22]. Zhao et al [23] and Liu et al [24] explored AgNPs array formation by annealing at temperature 400 °C and 300–800 °C, respectively, of Ag thin film deposited by thermal evaporation and dealt in detail the surface plasmonic behavior for solar cell applications. Though these research groups attributed various AgNPs formations in this methodology to supplied thermal energy, surface tension, thermal diffusion and recrystallization processes but reported literature lacks detailed explanation of systematic formation of AgNPs. Based on reported literature, a growth mechanism of AgNPs formation by RTP annealing of Ag thin film is discussed in detail in this report. In our case, rapid thermal heating provide thermal energy to the Ag particles onto Si surface. On increasing annealing time, atomistic mobility increases and so surface diffusion start increasing. The additional thermal energy imparted to Ag particles during annealing leads to recrystallization which occurs by nucleation and growth during solid state reaction [25]. On increasing RTP annealing time further coalescence amongst particles can be observed driven by higher surface diffusion. Upon cooling of this heat treated non-continuous Ag film on Si surface to room temperature, mismatching in thermal expansion coefficient of substrate (c–Si) (3.52  ×  10⁻⁶ K⁻¹ at 200 °C) and sputtered thin metallic Ag film (19  ×  10⁻⁶ K⁻¹) leads to shrinkage in thin film and so compressive stress exists at the interface of thin film and c–Si substrate due to mismatching in thermal expansion coefficient which results into the formation of plasmonic AgNPs on c–Si substrate. Here annealing temperature and annealing time also play important role in determining ultimate shape, size and interparticle distance amongst Ag NPs. Since c–Si substrate and Ag have 'fcc' structures and has no appreciable lattice mismatching [26, 27], therefore change in nature of stress due to chemical nature and surface topography of substrate can be ignored. If a thin film at temperature T_a is cooled to room temperature T_r, the existing thermal stress force (F_f) acting at the interface of thin film and substrate can be calculated by [28]

$$\begin{eqnarray}&&{{F}_{\text{f}}}=\frac{w\left({{\alpha}_{\text{s}}}-{{\alpha}_{\text{f}}}\right)\left({{T}_{\text{a}}}-{{T}_{\text{r}}}\right)}{\frac{1-{{\nu}_{\text{f}}}}{{{d}_{\text{f}}}{{E}_{\text{f}}}}+\frac{1-{{\nu}_{\text{s}}}}{{{d}_{\text{s}}}{{E}_{\text{s}}}}}\,,\end{eqnarray} \tag{ 1 }$$

where w is width of substrate, d_f is thickness of the film, d_s is thickness of substrate, α_s is thermal expansion coefficient of substrate, α_f is thermal expansion coefficient of film (Ag), T_a is annealing temperature, T_r is room temperature, ν_f is Poisson's ratio of film, ν_s is Poisson's ratio of substrate, E_f is Young's modulus of film, E_s is Young's modulus of substrate.

Since in our case ${{d}_{\text{f}}}{{E}_{\text{f}}}/\left(1-{{\nu}_{\text{f}}}\right)$ is very smaller than ${{d}_{\text{s}}}{{E}_{\text{s}}}/\left(1-{{\nu}_{\text{s}}}\right)$, therefore thermal stress in the film is

$$\begin{eqnarray}&&\sigma =\frac{{{F}_{\text{f}}}}{{{d}_{\text{f}}}w}=\frac{\left({{\alpha}_{\text{s}}}-{{\alpha}_{\text{f}}}\right) \Delta T{{E}_{\text{f}}}}{1-{{\nu}_{\text{f}}}}\centerdot \end{eqnarray} \tag{ 2 }$$

Based on available literature, the thermal expansion coefficient of silver is 19  ×  10⁻⁶ K⁻¹ [29] and that of Si substrate are 3.52  ×  10⁻⁶ K⁻¹, 3.72  ×  10⁻⁶ K⁻¹ and 3.84  ×  10⁻⁶ K⁻¹ at temperature 200 °C, 250 °C and 300 °C, respectively [30]. Young's modulus of elasticity and Poisson's ratio of silver film [31] are 82.5 GPa and 0.364, respectively. The calculated compressive stress (using equation (2)) has magnitudes 0.347 GPa K⁻¹, 0.442 GPa K⁻¹ and 0.537 GPa K⁻¹ at temperatures 200 °C, 250 °C and 300 °C, respectively. At 200 °C RTP temperature, Ag thin film results into AgNPs cluster formation for 5 min RTP. On further increase in RTP annealing time, residual stress relaxation takes place in thin film. New grain formation during recrystallization process usually occurs with decreases in average AgNPs size with the increase of annealing time [25]. The average AgNPs size reduction from 10 to 15 min RTP at each temperature during recrystallization can be attributed to degree of recrystallization and then minimum average AgNPs particle size is achieved for 20 min RTP at each temperature. Heat treatment during RTP results into residual stress relaxation with recrystallization amongst AgNPs and so energy minimization takes place for 20 min RTP which results into decrease in surface area of AgNPs. Further increase in RTP duration to 30 min results into enhanced surface diffusion which leads to particles agglomeration and results into island formation due to coalescence. Similar trends are observed for variable RTP duration at 250 °C and 300 °C as listed in table 1. At constant RTP duration of 5 min with the increase in RTP temperature, Ag NPs cluster size increases (table 1) due to increased strain energy. Similar trends are observed for each constant RTP duration with the variation of temperature from 200 °C to 300 °C as listed in table 1. Thus RTP time, RTP temperature and so generated thermal stress, surface diffusion play vital role in determining ultimate size of Ag NPs on Si substrate.

In order to understand effect of RTP temperature on AgNPs sizes, a graph between average size of Ag NPs and RTP temperature for constant RTP duration is summarized in figure 6. A larger increase in average size of Ag NPs is observed for 5 min RTP from 200 °C to 250 °C which becomes less significant for further increase in temperature from 250 °C to 300 °C. For 10 min RTP, the increase in average Ag NPs size is constant for the rise of temperature from 200 °C to 300 °C. This increase in average Ag NPs sizes decreases with temperature for 15 and 20 min RTP. The increase in average Ag NPs is minimum for 30 min RTP. Thus highest increase in average Ag NPs size with temperature in the range of 200 °C–300 °C is observed for 10 min RTP while it becomes minimum for 30 min RTP.

Zoom In Zoom Out Reset image size

Figure 7 shows summary of effect of RTP time durations on average AgNPs sizes at constant temperatures.

Zoom In Zoom Out Reset image size

The reflectance measurements of AgNPs on c–Si substrate in wavelength region 300–800 nm at constant RTP temperature for different RTP time duration are shown in figure 8 (at 200 °C), figure 9 (at 250 °C) and figure 10 (at 300 °C).The reflectance spectra clearly indicate highly reflective Ag thin film onto c–Si substrate due to highly packed flattened particles as shown in figure 2(b). After performing RTP of non-continuous silver thin films at constant temperature, average reflectance drops drastically due to creation of isolated AgNPs array as plasmonic layer. The appreciable change in reflectance of polished c–Si wafer due to surface plasmonic AgNPs may be attributed to the changes in size of AgNPs by RTP. It has been observed that in the typical samples, the average reflectance (for 5 min RTP at 200 °C) from around 19.38% is reduced up to 1.15% (for 20 minRTP at 200 °C). The tuning of AgNPs size at 200 °C RTP show negligible reflectance in long wavelength region (600–800 nm) and so can be efficiently utilize for significant photocurrent enhancement in Si based solar cell devices. This significant reflectance reduction for 20 min RTP can be attributed to the fact that the formation of NPs are more distinctly spaced and spherical in appearance to acquire energy minimization during RTP which favors appreciable forward scattering [32]. Moreover, dominant short wavelength (~350 nm) reflectance reduction with AgNPs size at 200 °C and 300 °C may be attributed to the quadruple resonance [32]. The effects of RTP duration with average reflectance are summarized in figure 11 at constant temperature. Figure 11 indicates maximum average reflectance reduction for 20 min RTP for all annealing temperatures due to minimum average size of AgNPs which may be attributed to relatively more spherical shape of particles. The effect of RTP temperature on average reflectance for constant RTP durations is summarized in figure 12. On increasing RTP temperature from 200 °C to 250 °C, reflectance decreases for 5 and 10 min RTP. Further rise in RTP temperature to 300 °C, reflectance increases for all RTP time durations.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Thus tuning of AgNPs size is observed due to RTP temperatures and durations which can be an effective way to design reflectance reduction of Si surface.