1. Introduction

Information privacy protection is increasingly valued in modern society. Nowadays, simple storing process can no longer meet people's needs for information security. Encryption of special or critical data has become particularly important. Color information has the advantages of being directly observable to the naked eye and having high discrimination, and has been used in fields such as watermarking and realistic scene reproduction. Metals such as gold, silver and copper are important materials to generate color [1–3]. The metal particles with nanostructures have unique absorption characteristics based on localized surface plasmon resonance (LSPR) [4,5]. The optical absorption and scattering at a certain frequency result from the resonation of the free electrons in metal particles with the electric field oscillation of incident light. The resonance frequency is closely related with the size and shape of metal nanoparticles (NPs) [6–8]. Furthermore, regulating interparticle distance and dielectric environment result in different reflective colors in a rather wide range. It is well-known that Lycurgus Cup of ancient Rome was made 1600 years ago just utilizing the LSPR property.

In recent years, laser color-printing has attached much attention with increasing regulation methods of noble-metal NPs. Silver nano-island and nano-cap were formed by pulsed laser deposition [9] and dewetting at high temperature [10], respectively. The utilized methods also include electron beam lithography (EBL) [8,11,12], drop-casting of colloidal solution [13], and laser processing [14–16]. EBL finely manipulates the nanostructure of NPs and prints high-resolution images, which was used to print images with a pixel-spacing of 250 nm, reaching the diffraction limit [17]. Besides, the patterns with nearly 3000 different colors and shades were developed [18]. EBL-fabricated silicon nanostructure and surface-relief aluminum metasurface were also applied in reversible switching and encryption of high-density optical data [19,20]. Although such a high-performance printing method has been proposed, a low-cost and feasible printing mode is still required. Accordingly, drop-casting of colloidal solution was applied to print a color image with an area of 10 × 10 cm² [21], but lacked the ability to print precise and complex patterns. Different from the two printing methods mentioned above, laser processing is based on the deformation of metals by the laser-generated thermal energy which can print patterns accurately in a short time. The patterns with the dot per inch (DPI) of 25,000 were obtained from green to red regions by using a femto-second pulsed laser [22]. In addition, the pattern of 4,600 DPI was drawn by an infrared fiber laser scanner [23]. Moreover, the deformation process of NPs is also dependent on the surface profile of oxide substrate. TiO₂/Ti₂O₃ transition during laser processing results in deformation and coloring of titania [24]. Computer-generated holograms (CGHs) has been used as image-based optical physical unclonable functions (PUFs) by laser fabrication. Using industrial nanosecond infrared fiber lasers to create holograms on silver surfaces, the error effect of the nanostructure processing provides the necessary randomness for PUF [25]. Commonly, laser processing relies on the transient thermal energy of laser beam which has the limited ability of finely tuning morphology of metal NPs. Instead, chemical solution reaction is a relatively slow and moderate process, and its regulation ability in particle size and spacing is closely related to reaction time and solution concentration. It was found that chemical etching selectively dissolves metal or semiconductor, and can be combined with laser interference to be applied in high-efficiency poly-Si solar cells [26] and in controlling the wettability characteristics [27]. However, it still remains a challenge to combine laser printing with chemical etching for information encryption and steganography.

Recently, we proved the surface treatment of KCl solution accelerates the speed and efficiency of plasmonic holographic data storage [28]. The optimization process comes from the direct oxidation at the outer layer of nano-Ag to AgCl through the Cl^- ions solution adhesion [29]. This process is accompanied by the mutual conversion between Ag and Ag ⁺. Electron backflow may be suppressed by contacting with a wide bandgap semiconductor to improve the stability of information storage. It was also reported that the reactivity of Au NPs depends on both the pH value of halide ion solution and the wavelength of excitation light [30]. However, the chemical reaction period in halide ion solution is rarely explored, and the combination effect of laser processing and chemical solution is not paid any attention. In this work, a fine color modulation strategy for the Ag/Ta₂O₅ bilayer system is proposed based on the bi-path of photo-thermal reaction and liquid-phase-chemical reaction. The size and spacing of nano-Ag particles are both modulated. The printed color is precisely modulated in a certain range of visible region. The colored information with micro-nano hybrid structures is hidden by setting the exposure number and the energy density during laser processing in the background region and target region, which can be decrypted after the sample immersion in KCl solution in a certain liquid reaction period. This work opens up new avenues for color-printing and steganography.

2. Experimental

Ta₂O₅ solution was prepared by sol-gel method. Tantalum pentachloride powder (0.60 g, Shanghai Macklin Biochemical Co. Ltd.) was dissolved in ethanol (12.65 mL) with stirring for 2 h. Meanwhile, F127 (0.30 g, Sigma, USA), as a kind of pore forming agent, was added into ethanol solution (12.65 mL) and then stirred at 65 °C for 20 min until the polymer was completely dissolved. These two kinds of solutions were finally mixed with each other and stirred for 2 h after adding deionized water (2.0 mL). Tantalum pentoxide precursor was obtained followed by standing still at 300 K for 15 h. Ta₂O₅ nano-porous films were prepared by immersing glass substrates in tantalum pentoxide precursor with a rate of 0.45 cm/s, and annealing at 600 °C to remove the polymer. The thickness of the prepared Ta₂O₅ nano-porous films were 35 ± 5 nm, measured by a profilometer (Alpha-Step D-120, KLA-Tencor). A silver layer (17 ± 2 nm) was magnetron-sputtered onto the Ta₂O₅ substrate to construct Ag/Ta₂O₅ bilayer systems (Fig. S1 in Supplement 1). A fiber laser marking device (Daheng Optics, F6S20) was used to apply laser processing. The sample was scanned with a 1064 nm infrared pulse laser beam, with a beam diameter of 0.04 mm, a spectral pulse width of 100 ns, and the total power of 20 W. In addition, the pulsed laser beam was modulated in the following parameters: the spacing of line-scanning from 0.01 mm to 0.50 mm, the scanning speed from 100 mm/s to 1000 mm/s, and the pulsed repetition frequency (PRF) from 20 KHz to 70 KHz. Alternately, the Ag/Ta₂O₅ bilayer film was dissolved in the KCl solution of 0.1 mol/L for the immersion period from 1 s to 120 min, followed by the drying process with an air-gun. The whole preparation process is shown in Fig. 1.

 

Fig. 1. Fabrication of Ag/Ta₂O₅ nanocomposite films by laser processing and chemical solution dissolution.

Download Full Size | PDF

3. Results and discussion

3.1 Photothermal reaction

The infrared pulsed laser beam is incident on Ag/Ta₂O₅ films with the modulation parameters of PRF f, energy E, spacing ${L_s}$, and scanning speed $\textrm{}v$ (Fig. 2(a)). Various colors are obtained under one-directional scanning (ODS) by changing v, ${L_s}$, E and f (Fig. 2(b)). Here, we chose a white-light-emitting LED as illumination source, and its emission spectrum is shown in the Fig. S2 in Supplement 1. Samples were placed on white paper and then we used the CMOS camera of cell phone to capture the color. Specific parameters for each coordinate position of Fig. 2(b) are listed in Table 1. The default modulation parameters in the Fig. 2(b) are $f$ = 20 KHz, $E$ = 50%, ${L_s}$ = 0.06 mm, and $v$ = 800 mm/s. Each column represents the color variation by changing one of the parameters, where Column i represents the change of f from 20 KHz to 60 KHz, Column j the change in E from 20% to 100%, Column k represents the change in ${L_s}$ from 0.3 mm to 0.03 mm, and Column l represents the change of v from 1000 mm/s to 100 mm/s. For the sample ${S_{2l}}$ with the color of light-pink, the modulation parameters of $f$ = 20 KHz, $E$ = 50%, ${L_s}$ = 0.06 mm, and $v$ = 800 mm/s are adopted. A series of concave and circular areas are formed by the laser beam irradiation, which is observed by scanning electron microscope (SEM, FEI Czech Republic S.R.O) (Fig. 2(c)). From the magnified image inserted in Fig. 2(c), it can be confirmed that the original continuous silver film is converted into Ag NPs by the thermal effect of the pulsed laser [31,32]. The size distribution of Ag NPs is characterized by the Nano measurement software which can measure the size of every particle in SEM, and carried out statistics of the calculated results. The size distribution of Ag NPs varies from 7.2 nm to 30.2 nm and the average particle size is 15.5 nm. When the modulation parameters are set as $f$ = 20 KHz, $E$ = 50% ${L_s}$ = 0.06 mm, and $v$ = 100 mm/s, the color of the sample ${S_{5l}}$ changes to deep-pink. Meanwhile, a strip-shaped concave region is observed (Fig. 2(d)), in which the size distribution of Ag NPs varies from 3.1 nm to 20.3 nm and the average particle size is 9.3 nm. In addition, after laser processing, a decrease in film thickness of about 17 nm was observed at the directly ablated position (Fig. S1 in Supplement 1). Increasing f and E has a similar effect of color-deepening to that of decreasing v, and also results in a series of strip-shaped concaves, as shown in columns i and j, respectively, in Fig. 2(b).

 

Fig. 2. Laser processing induced plasmonic coloring in Ag/Ta₂O₅ composite films. (a) Modulation parameters for laser processing. (b) Plasmonic color modulated by $f$, E, ${L_s}$, and v. SEM images for(c) ${S_{2l}}$, (d) ${S_{5l}}$, (e) ${S_{5i}}$, and (f) ${S_{5j}}$.

Download Full Size | PDF

Table 1. Specific Parameters for Each Coordinate Position in Fig. 2(b) ^a

View Table | View all tables in this article

From the above investigation, it is found that the spatial arrangement of plasmonic particles plays a role in regulating the film color [33–35]. Hence, a method of orthogonal secondary scanning (OSS) is proposed to construct more color gamut compared with that by the single exposure. The second scanning direction is perpendicular to the original one, as shown in Fig. 3(a). On the substrate of Ta₂O₅/glass, the reflection spectrum of the sample is hard to be detected. A spectrophotometer (UV-2600, Macy Instrument, China) is used to measure the transmission spectrum of the sample. Then we use the plug-in of Chromaticity Diagram for Origin 2021 to calculate the color of the sample in the CIE 1931 diagram (Fig. 3(b)). Sample colors under the specific parameters are shown in Fig. 3(c), and the details for each coordinate position of are listed in Table 2. When ${L_s}$ is changed from 0.1 mm to 0.03 mm while the other parameters are set constant as $v$ = 800 mm/s, $f$ = 20 KHz and $E$ = 50%, the maximum absorption value in visible region is decreased towards zero, which means the sample color has become lighter. The corresponding absorption curves are shown in Fig. S3 in Supplement 1. SEM measurements indicate that micro-nano hybrid blocks are formed in the overlapping position of exposure. The block size also becomes smaller and tend to disappear (Fig. 3(d)). At the gap position of the blocks, the circular boundary is still observed clearly for the case of ${L_s}$ = 0.1 mm. Due to the fact that only the single variable of scanning spacing causes a decrease in block size, it is inferred that these blocks are not the remnants by direct laser ablation, but rather particle aggregate formed by thermal radiation shaping outside of laser line scanning. After DOS processing, the color gamut range is expanded from the black dots part to the red one, as shown in Fig. 3(b).

 

Fig. 3. Laser processed plasmonic coloring in Ag/Ta₂O₅ composite films using DOE. (a) Modulation parameters for laser processing using “second exposure” method. (b) Comparison of chromaticity diagram between ODS (black dots) and OSS (red dots) methods. A white triangle shows the gamut of the sRGB color space, representing colors achievable by common computer monitors. (c) Plasmonic color modulated by v, ${L_s}$, E and f. (d) Absorption peak value and the corresponding SEM images for the sample of different parameters of ${L_s}$.

Download Full Size | PDF

Table 2. Specific Parameters for Each Coordinate Position in Fig. 3(b)^a

View Table | View all tables in this article

Furthermore, the block is investigated in detail by SEM, which is divided into the regions I, II and III under the laser processing condition of $v$ = 500 mm/s, ${L_s}$ = 0.1 mm, $f$ = 20 KHz and $E$ = 50%, as shown in Fig. 4(a). As the thermal radiant intensity from the pulsed laser decreases from Region III to Region I, the small-sized and near-spherical Ag NPs is only formed in Region I which is the center position of the block (Fig. 4(b)). As extending to Region II, porous structures and small-sized particles appear (Fig. 4(c)). It is observed that the silver layer is melted thermally and recombined into large-sized and tightly-arranged particles at the edge of the block (Region III, Fig. 4(d)) [36,37]. Statistical analysis was conducted on the particle size and particle spacing in the three regions. Then the histograms of distribution statistics for the particle size and spacing of Ag NPs are obtained, as inserted in Figs. 4(b-d). The average particle size is 81.13 nm, 90.27 nm and 141.50 nm, and the average particle spacing is 93.9 nm, 181.17 nm and 32.74 nm in Regions I, II and III, respectively. Obviously, the size distribution of Ag NPs in Region I of these blocks plays a key role in the color adjustment of the sample.

 

Fig. 4. SEM image of the Ag block. (a) Top-view of the sample under the condition of $f$ = 20 KHz, $E$ = 50%, ${L_s}$ = 0.1 mm, and $v$ = 500 mm/s. Ag particle size distribution in the center (b), middle (c), and edge (d) areas of the block.

Download Full Size | PDF

Alternately, the sample can be repeatedly irradiated by ODS method, that is the sample experiences multiple one direction scanning processes (MODS). The target area is scanned for 1, 2, and 3 times under the high energy density ($f$ = 60 KHz, $E$ = 100%, ${L_s}$=0.06 mm, and $v$ = 100 mm/s), abbreviated as H-MODS (1), H-MODS (2) and H-MODS (3), respectively, as shown in Fig. 5(a). The color saturation degree is enhanced with increasing the scanning times, resulting in the color coordinates extending towards the edge on the chromaticity diagram. The absorption spectra of samples modulated by the ODS method with different scan times are shown in Fig. S4 in Supplement 1. From the inserted images of Figs. 5 (c-e), the average particle size is calculated as 9.3 nm, 12.4 nm and 19.0 nm, and the population density of Ag NPs as 4.6 ${\times} $ ${10^{11}}$ ${/}c{m^2}$, 3.0 ${\times} $ ${10^{11}}/c{m^2}$, and 1.0 ${\times} $ ${10^{11}}/c{m^2}$ under the ODS treatment for one, two, and three times, respectively. As the number of scans increases, the average size is increased while the particle density is decreased. It indicates that the thermal radiation energy received from the pulsed laser has an accumulation effect on shaping the plasmonic particles [2,38]. Due to the lack of enough thermal energy, a single ODS process results in densely distributed and small-sized particles in the center position of the scan area, while the repeated treatment causes the formation of larger-sized Ag-particles. It should be pointed out that, different from this process, the repeated printing ability becomes weak if reducing scan energy density. The patterns are even unrecognizable, which will be illustrated in detail in the following text.

 

Fig. 5. (a) Plasmonic color modulated by the ODS method with different scan times. (b) The corresponding coordinate position in chromaticity diagram. SEM images for the sample of (c) H-MODS (1), (d) H-MODS (2) and (e)H-MODS (3). A white triangle shows the gamut of the sRGB color space, representing colors achievable by common computer monitors.

Download Full Size | PDF

3.2 Chemical reaction

Laser processing is a drastic process for the Ag/Ta₂O₅ two-layer film, while chemical solution immersion may be a relatively mild one, and the reaction is as followed:

The morphology of silver nanoparticles will undergo slow shape-evolution versus immersion time [30]. Therefore, using halide solution immersion (HSI) to adjust the plasmonic color will be very precise. The untreated Ag/Ta₂O₅ film appears gray-white, corresponding to the deposition of a large number of Ag particles, followed by their aggregation to form a continuous silver film. From the absorption spectrum, it can be seen that the initial film exhibits strong signal throughout the visible light region (Fig. 6(a)). Afterwards, the sample is immersed in KCl solution of 0.1 mol/L, converting to brown within the first 5 minutes. However, the color gradually turns back to light-yellow, with a slower color change from 5 minutes to 120 minutes (Fig. 6(b)). Meanwhile, absorption at the wavelength longer than ∼550 nm decreases significantly, while the absorption intensity below the green wavelength increases. A distinct LSPR band appears and tends to be narrowed. Besides, with the prolongation of immersion period in KCl solution, the absorption peak position undergoes a blue shift. The color position in the chromaticity diagram is shown in Fig. 6(c), almost moving linearly versus the immersion period. In addition to color changes, the sample also exhibits a significant transition from reflectivity to transmittance. In Fig. 6(d), the virtual image of a metal badge is reflected by the initial film, while the graph behind the film can be observed clearly for the sample with the immersion of KCl solution for 45 min. The decrease in reflectance and increase in transparency can be attributed to a decrease in silver film thickness (Fig. S1 in Supplement 1). We also carried out SEM measurement of the films at different immersion periods. The continuous silver film on Ta₂O₅ substrate with small gaps is observed at initial stage (as shown in Fig. S5 in Supplement 1). After HSI for 5 min, the densely arranged Ag NPs became dispersed, with an average particle size of 42.43 nm and an average gap of 30.93 nm (Fig. 6(f)). As treatment time prolonging, particle size and spacing can be adjusted continuously. For the immersion time of 15 min, the average particle size is 41.74 nm and the average spacing between the adjacent particles is 38.85 nm (Fig. 6(g)), while the average particle size of 42.81 nm and the average spacing of 41.48 nm is calculated for the case of 45 min (Fig. 6(h)). Different from the results by laser processing that particle size and spacing are modulated simultaneously, HSI period mainly controls the particle spacing while has little effect on particle size. The increase in particle spacing also contributes to the increase in transmittance. Using such a characteristic, the HSI is arranged after the laser processing to precisely control the color of printed images.

 

Fig. 6. (a) Absorption spectra and (b) sample colors after HSI of different periods. (c) Chromaticity diagram of the film with different KCl solution immersion times. A white triangle shows the gamut of the sRGB color space, representing colors achievable by common computer monitors. (d) Reflectance (up) and transmittance (down) of the Ag/Ta₂O₅ film, before and after the HSI, respectively. Top-view of SEM for the samples with the HSI periods of (e) 0 min, (f) 5 min, (g) 15 min, and (h) 45 min.

Download Full Size | PDF

3.3 Encryption scheme

Due to the fact that both laser processing and chemical solution immersion can adjust the color of the plasmonic film, combining these two methods may yield unexpected results. After pulsed laser treatment of Ag/Ta₂O₅ films, further immersion with KCl solution can cause the sample color continue to be varied, as shown in Fig. S6 in Supplement 1. The idea of encryption strategy for printed information begins with the modulation of the thermal energy density in the photothermal reaction process. Only part of the Ag particles undergoes the variation in size and spacing simultaneously for secondary scanning under low energy density (such as $f$ = 20 KHz, $E$ = 50%, ${L_s}$ = 0.1 mm, and $v$ = 1000 mm/s), i.e., L-MODS (2), which is different from the case under high energy density (H-MODS, such as $f$ = 60 KHz, $E$ = 100%, ${L_s}$ = 0.06 mm, and $v$ = 100 mm/s). It is found that small-sized Ag particles appear in the crescent shaped exposure position (Fig. 7(b)). And comparing the results of Fig. S7 in Supplement 1 with that of Fig. 7(b), it is concluded that the particle size and spacing for the non-crescent shaped region change little whether the secondary laser scanning is applied or not. This indicates that the process of Ag morphology transformation can be divided into two stages. Firstly, silver films are gradually melted into small-sized particles. Subsequently, the small-sized Ag particles begin to aggregate with each other and enter the growth stage after further thermal radiation, as demonstrated in the previous section. The second process results in a gradual increase in particle size. Therefore, we can use the intermediate state during formation of Ag particles by laser processing, i.e., the first stage, to realize information steganography. The detail operation steps are as followed. A background color is produced by the low-energy-density pulsed laser treatment as $f$ = 20 KHz, $E$ = 50% ${L_s}$=0.1 mm, and $v$ = 1000 mm/s. The region where is not exposed forms a pattern of the letter “N” with the color of silver-gray. Then we use the same scanning parameters to carry out the secondary printing of the letter patterns of E, N, and U in the just exposed region. Only a small population of silver particles are converted into the small-sized ones, which plays little role in variation of optical absorption band. Thus, colors of the three letters printed later are almost indistinguishable from the background color by visual observation (Fig. 7(c), left part). However, the small-sized plasmonic particles are more sensitive to KCl solution than the larger ones [6]. After the long-term chemical reaction in KCl solution, the crescent shaped position still exists (as shown in Fig. S8 in Supplement 1), and the small-sized particles in this position tend to aggregate into disordered silver block, accompanying with the weakened LSPR absorption. Then the secondary printed letters emerge gradually (Fig. 7(c), right part). The absorption bands of the background and encryption regions of Ag/Ta₂O₅ thin films are correspondingly measured, as shown in Fig. 7(d). The original absorption curves for the two kinds of regions are very similar, exhibiting equivalent LSPR intensity and peak position. After HSI for 90 min, the LSPR intensity for the encrypted region significantly decreases but the peak position remains almost unchanged, while the absorption peak position for the background region shows a significant blue-shift but the LSPR intensity only slightly decreases. Such a quite different kinetics of chemical reaction provides a possibility for redout of the hidden information.

 

Fig. 7. (a) Schematic diagram of bi-path encryption strategy. (b) Top-view of SEM images for Ag/Ta₂O₅ films after L-MODS and HSI. (c) Actual optical photos of hidden (left) and decrypted (right) patterns (d) Absorption spectra of the sample at encrypted and background regions.

Download Full Size | PDF

The different sensitivity of Ag NPs with different sizes to halide solution can be attributed to the size dependence of its redox potential (or Fermi energy level) [39,40], which simultaneously affects the thermodynamics and kinetics of Ag NPs dissolution. In the viewpoint of thermodynamics, the smaller the size of Ag NPs, the easier it is to be oxidized. According to Plieth's research, the oxidation-reduction potential $E_p^0$ of Ag NPs can be calculated by the following equation [41,42]:

4. Conclusion

Laser processing and chemical solution immersion methods can both achieve color printing of Ag/Ta₂O₅ bilayer films. By adjusting the pulsed laser parameters and chemical solution reaction time, the color and reflection-transmission characteristics of the films are effectively controlled. When the laser pulse energy is relatively high, repeated printing can obtain richer color gamut information, but the repeated printing with reducing the pulsed laser energy hardly changes the color of the sample. By further immersion of the film in chemical solutions, the color contrast is significantly enhanced, which can be used to achieve encrypted printing of various color information. This work provides a new path for the permanent storage of high-density data and the secure display of color information.