A Self-Color-Changing Film with Periodic Nanostructure for Anti-Counterfeit Application

Abstract

A self-color-changing film aimed at enhanced security and anti-counterfeit packaging is presented. Its function is to change color automatically when flipped under visible light. It is low-cost, takes a few seconds to check by the naked eye, and does not need any special tools to evaluate. The design of the color-changing, anti-counterfeiting film is based on a frequency-selective surface (FSS). The film is designed with aluminum nanocubes. They are laid out as an array in a plane with equal distance from one another. This arrangement allows us to select certain wavelengths of light to pass through by the size of the cubes and the separation distance between them. The performance is evaluated by a finite element analysis (FEA) method. The results show that the intersection of transmittance and the reflectance curves cause the film to change its color automatically when flipped. We also propose a method to predict the color of the film based on the transmittance values. The accuracy of this method is verified by actual colors from experiments with an error of no more than 12.8%, analyzed by the CIE chromaticity diagram.

1. Introduction

Anti-counterfeiting technology (ACT) is a popular topic that has garnered a lot of attention in recent times [1,2] Counterfeiting is responsible for global economic losses reaching USD 1.82T per year [3,4,5,6]. The current trend of new anti-counterfeiting techniques gravitates toward those that are low-cost, easy-to-recognize, but challenging to imitate [6]. Techniques that can easily be recognized by the naked eye in a few seconds and do not require any special tool to analyze are in high demand [6,7,8,9]. Among the reported ACTs [10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32], only the security film has all of these features.

The security film consists of a substrate and a coating agent. The substrate comprises a polymer, paper, glass, or fabric, and the coating agent can be fluorescent, graphene or metal nanoparticles. In 2021, Saad et al. presented a new technique to prepare polyester fabric as a new anti-counterfeiting [33] material. Fluorescence was applied to polyester fabric by a microwave irradiation method. The anti-counterfeiting polyester fabric worked by absorbing the radiating UV to cause discoloration under UV. The efficiency was obtained from a reflectance and transmittance measurement. The results showed spontaneous discoloration behavior under UV that was attractive and easy to detect visually. However, the invention still relied on UV as a catalyst; therefore, it was not convenient to use in places where the UV was not readily available.

Graphene, on the other hand, has good optical properties and can be designed to possess an unusual form of optical response by using light to induce free-electrons on the surface of nanoparticles with an electric field for oscillation [34,35,36,37]. The oscillation gives the transmittance a convex characteristic, resembling a notch filter. The point where the lowest transmittance occurs is called the resonant frequency. This phenomenon is called localized surface plasmon resonance (LSPR) [38].

In addition to graphene, metal nanoparticles can also induce the LSPR. In 2020, Kuroiwa et al. [38] introduced silver nanoparticles on a glass substrate as a color film using a laser printing technique. By sizing the appropriate silver particle dimension via adjustment to the laser intensity, Kuroiwa et al. were able to create nanoparticle-based film in any color from the LSPR effect. Although graphene and nanoparticles have good light-absorbing properties, the behavior for displaying different colors can also be easily imitated by ink or other chemical absorbers that absorb light at the same wavelength. A more sensible alternative is to use light transmission and reflection instead of absorption. This mechanism is readily observable in a frequency-selective surface (FSS) structure.

An FSS occurs when an array of metallic nanoparticles are laid together in a plane on an insulator substrate to create a resonant effect. An isolated nanoparticle forms only localized surface plasmon resonance, which is suitable for making light-absorbing materials. However, if a nanoparticle is placed in proximity to other nanoparticles in a proper arrangement, a mutual induction occurs between particles that results in the frequency-selective surface behavior of the film. FSS can transform from a convex transmittance curve (or a notch) to a low-pass or a high-pass filter. The intersection of transmittance and reflectance curves causes the color-changing phenomenon. Such a phenomenon would be difficult to imitate, which is the goal of our work.

Our research focuses on designing an anti-counterfeiting film that changes color under visible light. It differs from that proposed by Saad, since ours does not require a UV excitation. Our technique also goes further than Kuroiwa et al. or other graphene research. We not only make use of the LSPR, but also take advantage of the FSS structure in our design. Our proposed film composite uses low-cost materials such as aluminum (Al) and polyethylene terephthalate (PET) as the main components. PET is a commonly known polymer that possesses many desirable features such as waterproof, flexible, durable, transparent and portable.

This paper is organized as follows. Section 1 introduces the background of anti-counterfeit technology. Section 2 describes the design principle of the proposed self-color-changing film and the simulation domain setup. Section 3 discusses results from the parametric studies of the nanocube when surrounded by PET. It also investigates the effects of the nanocube size and substrate components on transmittance and reflectance. Finally, the conclusion is provided in Section 4.

2. Concept and Structure Design

An unconventional luminous phenomenon that is difficult to imitate but easily observable must be considered when designing a high security anti-counterfeiting method. This paper focuses on applying the physics of light to create new anti-counterfeiting technologies.

2.1. Principle of Self-Color-Changing Film

The concept is described from a side-view perspective of the packaging to which the film is affixed. “A” and “B” refer to the film’s front and back sides, respectively. When a light source is placed in front of the film side (A), the observer sees color from the light that is reflected off the film, i.e., blue. The color of the light that is transmitted through the film to the back side (B) is a complementary color to the one on the other side, which is yellow in this case. The complimentary color of a color is defined as the color that is opposite to the color of interest on a standard color wheel. On the other hand, when a white light source is placed on the backside of the film side (B), the light that is reflected from the film is blue, as if the film can swap colors. Hence, the color of the light that is transmitted through to side (A) shows up as yellow.

The film’s ability to change colors when swapping the light source location or flipping the film cannot be recreated via any chemical process. Hence, this new security feature achieves its goal of quick and easy inspection by the naked eye.

For the proposed film to display different colors on both sides, the transmittance and reflectance curves must intersect. The intersection in this manner splits the incident wave diagram into two regimes: high transmittance and high reflectance. The wavelengths with a transmittance that is greater than −10 dB penetrate through the film, and the wavelengths with a reflectance that is greater than −10 dB reflect off the film. In practice, when the transmittance is lower than −10 dB or 10%, we assume that the film does not transmit wave in that regime. Similarly, when the reflectance is lower than −10 dB or 10%, we assume that the film does not reflect wave in that regime [39]. The intersection that separates the light characteristics into two regimes ensures two different colors appearing on the two sides of the film. For the case of VIS, when the intersection point occurs at the center of the diagram (λ = 550 nm), the color that is observed on the front and the back sides is blue and yellow, respectively.

2.2. Description of the Proposed Film

The structure of a proposed film in Figure 1 consists primarily of a conductor (metal) and an insulator (polymer). Metal is made into nanocubes. The size of each cube within the polymer is less than the wavelength (λ) of light, in order to induce the color change due to the plasmonic phenomenon [40]. In 2014, Schade et al. [41], and in 2020, Kuroiwa et al. [38], presented plasmonic transmittance spectra of nanometal particles. They both found that the plasmonic behavior can be observed by the presence of a convex curve (or a notch) in the transmittance spectra, indicating a resonant frequency. The wavelength at the lowest point of the convex curve is called the resonant wavelength [41]. If a resonance wavelength occurs, it indicates the existence of a plasmonic phenomenon in metal, which causes the change in the color of the particles. In our research, we study the plasmonic behavior of metals from transmittance. The size of the metallic nanoparticles can be varied to obtain a desired resonant frequency. The array arrangement of such particles can form a frequency-selective surface (FSS).

The proposed film is a thin PET film of a few microns. It has three layers. The first is a flexible and highly transparent PET layer [42]. This insulating substrate layer is approximately 1 micron thick. The second layer is an array of metallic nanocubes with an initial size of 110 nm, uniformly spreading over the entire surface of the substrate. The spacing between the nanocube is filled with PET. The nanocube is made of aluminum, which has good plasmonic properties and is inexpensive [43]. The last layer is a PET coating on top of the nanocube. It is approximately 1 micron thick to prevent air from oxidizing the aluminum. Light is incident upon the film surface either from the top or the bottom and it transmits through to the other side. The total film thickness is around 2.2 microns and the structure is symmetrical with respect to the XY plane.

2.3. The Domain Setup

This section describes the domain setup for the parametric studies of the proposed film. The two setup principles to achieve the new anti-counterfeiting properties are as follows.

First, the setup method must be able to model objects that are smaller than the minimum wavelength of VIS (<400 nm) to allow us to control the optical properties of nanoparticles, such as color. The finite element analysis (FEA) method is suitable for this study [44]. FEA is a well-known technique with high potential for analyzing complex nano-scale structures [45,46]. FEA has two essential setup processes: defining boundary conditions and mesh sizes.

Second, the boundary conditions to be used must be characterized by a frequency-selective surface (FSS) behavior that allows the transmission or the rejection of wave to cause the self-color-changing of the film. The setup in this paper uses periodic boundary conditions on the side to create an array with infinite extent. In this study, we use COMSOL Multiphysics as a simulation tool.

Our security film relies on FSS behavior for achieving spontaneous color change. Such behavior is dependent on the optical properties of the material. Therefore, the material properties are crucial to our studies. The boundary condition of the metal in FEA is usually defined with a perfect electric conductor (PEC) [45]. However, this paper will use the actual optical property of aluminum and PET obtained from [47,48], respectively, instead of PEC, for the most realistic results.

In Figure 1, the illumination of VIS starts from the upper perfectly matched layer (PML), traveling through the air layer down to the film, with an input power of 1 watt. Aluminum nanocubes are surrounded by PET and enclosed by an air layer. The air layer is sealed on the top and bottom with a PML boundary condition for absorbing all the reflected waves, assuming that the top and bottom regions of the film are infinite in extent [45]. The PML layers have a height of 19.25 µm. All the domains are meshed by tetrahedral meshes with a maximum size of no more than 0.1 times the shortest VIS wavelength (or in this case 0.1 × 400 nm). In our study, the largest mesh size is no more than 40 nm.

In this research, parametric studies are as follows:

• Width (W) of the nanocube in the range of 10 to 200 nm;
• Length (L) of the nanocube in the range of 10 to 200 nm;
• Height (H) of the nanocube in the range of 10 to 200 nm.

The parametric studies aim to find the best size of the nanocube to create an LPF.

3. Results and Discussion

In this section, the results that are obtained from the simulation are presented in the form of transmittance over various wavelengths. In addition, adjusting the size of the nanocube allows us to control the frequency-selective behavior of the film and the placement of the intersection of the reflectance and transmittance curves, in order to achieve the self-color-changing characteristic.

3.1. The Parametric Studies

In the study process, the initial W, L and H values are 110 nm. The parametric study starts out by adjusting the parameter W. When the optimal size of W is obtained, we can use it for the next parametric study. Finally, when we obtain the most satisfactory size of L, we use the best size of W and L to initiate the adjustment of the final parameter, H. The final result is the combined effect of all three parameters. Figure 2 demonstrates that at a wavelength of more than 500 nm, varying W does not affect the transmittance too significantly. The transmittance curves appear to overlap and do not differ much. At a wavelength of less than 500 nm, it is found that the transmittance decreases substantially. From the results, at W = 110 nm, the transmittance in wave in the 400 to 550 nm regime differs greatly from that in the 550 to 700 nm regime, and the convex curve on the graph indicates that the nanoparticles are plasmonic and have a resonant wavelength between 400 nm and 450 nm. In the study cases of L and H, the transmittance gradually changes from a convex curve to a low-pass characteristic when varying W. At L = 140 nm and H = 130 nm, the proposed film exhibits a low-pass filter behavior, as shown in Figure 3 and Figure 4. Such a low-pass characteristic is an essential feature that divides the transmittance diagram into two regions.

Figure 5 demonstrates that when W is more than 110 nm, the transmittance decreases substantially. The transmittance in the 400 to 550 nm regime is close to that in the 550 to 700 nm regime, which does not have an intersection point. Therefore, W = 110 nm is chosen as the value of choice for the subsequent studies. In the study cases of L and H, the transmittance changes from a low-pass filter to a convex curve when L and H are more than 140 nm and 130 nm, as shown in Figure 6 and Figure 7, respectively. Therefore, we conclude that the best size of a nanocube to exhibit a low-pass filter is W = 110 nm, L = 140 nm, and H = 130 nm. The low-pass filter behavior leads to an intersection of reflectance and transmittance curves, as shown in Figure 8 for the 1.4-microns film thickness. As the transmittance decreases, the reflectance increases. This causes both sides of the film to show two different colors and change color automatically when the film is flipped.

From the parametric studies, the transmittance can be controlled by varying the width (W), length (L), and height (H) of the nanocube. Adjusting the width (W) of the nanocube can induce the plasmonic effect, as observed by the emergence of a convex on the transmittance curve, consistent with the research by Schade et al. [41] and Kuroiwa et al. [38]. However, varying the width (W) alone to cause the convex is not sufficient to create a color-changing security film. Such behavior occurs when the transmittance curve exhibits a low-pass characteristic that divides the transmittance diagram into two regimes: high transmittance and high reflectance. Therefore, we need to vary the length (L) to adjust the nanocube size, in order to induce a frequency-selective surface behavior. The frequency-selective surface behavior rejects some wavelengths to allow the proposed film to show a low-pass filter behavior. The color of the light passing through or reflected from the film can be obtained [30].

The incident light may not always strike the film’s surface at a normal incidence in actual use; therefore, we simulate how the light incidence at an oblique angle with respect to the film’s x-axis affects the transmittance and the reflectance, as in Figure 9. From the simulation, the film still works well. The transmittance does not vary much when the incident angle is between 0 and 30 degrees. In the range of 60–70 degrees, the color of the proposed film changes from blue to yellow without the need to reverse the film, because it falls in the critical angle regime. This phenomenon is a characteristic that is specific to our film and is not available in other security techniques.

3.2. Validation of the Results

To confirm the results, we compare our results with the experimental results of some related studies. From the results, we discover that the variables affecting the frequency selectivity of the film are W and L of the nanocube. The horizontal area of a cube obtained by the product of W and L is, therefore, a significant parameter that can influence the film’s optical properties. In 2014, Schade et al. [41] presented a prototype of nanoparticles that were made of aluminum. The results of Schade et al.’s experiment also showed a significant relationship between the resonant frequency and the horizontal cross-sectional area of nanoparticles, as shown in Figure 10.

Finding the resonant frequency is challenging when a low-pass filter characterizes our film. Because of each adjustment, the resonant frequency changes with an accompanying low-pass filter behavior. Furthermore, it is difficult to locate the resonant frequency because the characteristic does not have a prominent convex character. However, our nanocube size is approximately the same as that which was previously studied by Schade et al., and they found that at resonant frequency, the transmittance was in the range of 10% to 20%. In such a range, we find that the resonant wavelength of our best nanocube size is 500 nm, which is consistent with the trend of the relationship between the resonant frequency (or wavelength (λ) [nm]) and the horizontal area [nm²] as:

$$Area=0.3493{\lambda }^2-258.8\lambda +57890$$

(1)

Equation (1) is derived from Schade et al.’s experiment, which showed that particles of different sizes had different resonant frequencies. In our case, the horizontal area is W × L = 110 nm × 140 nm = 15,400 nm² and the resonant wavelength is 500 nm. Such findings show good consistency when plotted against Equation (1), shown as the red dot in Figure 10. However, Schade et al.’s prototype was not immediately suitable for security film, as their nanoparticle size was not the optimum dimension. Nevertheless, the optimum values of W and H can be obtained from our studies.

The next study is to transform the transmittance to color, in accordance with the Commission International d’Eclairage: CIE standard, which defines three primary colors standardized at the wavelengths of 700 nm (red), 546.1 nm (green), and 435.8 nm (blue), respectively [49]. The CIE chromaticity diagram is a diagram that shows all the colors that humans can perceive. It helps to pinpoint the location of color that is seen by the human eye.

3.3. The Color Analysis Process

The color analysis process begins with sampling the transmittance for three wavelengths of RGB color and normalizing the three transmittance values. Let the sum be 100 percent; the normalized transmittance values are converted to RGB values by multiplying the transmittance percentage by the ratio of the maximum RGB value (255) to the maximum value of transmittance in percentages (100). The color can be identified from the CIE (x,y) coordinates on the CIE chromaticity diagram.

The diagram consists of X and Y axes, representing the amount of red and the amount of green, respectively. The amount of blue (Z), determined by Z = 1 − (X + Y), is collectively known as the tristimulus coefficient.

We create the CIE chromaticity diagram on a MATLAB R2020b. The RGB values are converted to CIE (x,y) values via a transformation matrix from [50]. The overall process is shown in Figure 11.

3.3.1. Sampling and Normalization of RGB Values

From the parametric studies, the transmittance values of the optimized nanocubes at wavelengths of 700 nm, 546.1 nm, and 435.8 nm, are 43.29%, 23%, and 1.4%, respectively. To obtain 100% as the sum of the transmittance percentages of the three colors, we normalize the percentage of each color by the ratio of 100% to the summation of the transmittance percentages, for example: 100/(43.29 + 23 + 1.4) = 1.477 and obtain the results as 63.94%, 33.97%, and 2.09% of RGB color, respectively.

3.3.2. Calculation of the XYZ Tristimulus Values

The RGB values are scaled to 0–255 for general image’s pixels (8 bits). The percentages of transmittance are mapped to the RGB values, as shown in

Table 1. Mapping of the transmittance to RGB values.

Colors	Wavelength (nm)	% Transmittance (Normalized)	RGB Values
Red (R)	700.0	63.94	163
Green (G)	546.1	33.97	87
Blue (B)	435.8	2.09	5

, by multiplying them with a factor of 255/100 = 2.55 and rounding the decimal numbers to the nearest integers. (0% transmittance would correspond to 0 RGB value, while 100% would be 255 in RGB value.)

We can transform the normalized RGB values to XYZ tristimulus values by using the transformation matrix [50] as follows:

$$\left[\begin{array}{c}\mathrm{X}\\ \mathrm{Y}\\ \mathrm{Z}\end{array}\right]=\left[\begin{array}{ccc}0.4124564& 0.3575761& 0.1804375\\ 0.2126729& 0.7151522& 0.0721750\\ 0.0193339& 0.1191920& 0.9503041\end{array}\right]\left[\begin{array}{c}\mathrm{R}/255\\ \mathrm{G}/255\\ \mathrm{B}/255\end{array}\right]$$

(2)

where X, Y, and Z are tristimulus values, and R, G, and B are red, green, and blue values, respectively.

3.3.3. Calculation of CIE (x,y) Values

After the XYZ tristimulus values are obtained, we can find the CIE (x,y) values for the color identification of the film from:

$$\mathrm{x}=\frac{\mathrm{X}}{\mathrm{X}+\mathrm{Y}+\mathrm{Z}}$$

(3)

$$\mathrm{y}=\frac{\mathrm{Y}}{\mathrm{X}+\mathrm{Y}+\mathrm{Z}}$$

(4)

From Equations (2)–(4) and the data in Table 1, the pixel locations of the colors in the CIE chromaticity diagram are plotted, as in Figure 12.

3.4. Validation of the Color

This section confirms the accuracy of the proposed color analysis method by comparing with the experimental results. In 2020, Kuroiwa et al. [38] created a security film with nanoparticles coated on top by shooting a laser onto them. The particle size was changed to the desired size by adjusting the laser intensity, which resulted in different colored particles. In Kuroiwa’s experiments, the nanoparticles were similar in size to ours. Therefore, it is appropriate to compare our results with those of Kuroiwa et al.’s. The transmittance and the color of the nanoparticles are measured by using an ultraviolet−visible-near-infrared. Our analysis begins by implementing the measured transmittance, following the steps in Figure 11. The results are summarized in

Table 2. Results comparison.

Sample No.	% Transmittance [38]			% Transmittance (Normalized)			Colors from RGB Analysis (Proposed)	Colors from Measurement [38]
	R	G	B	R	G	B
1	46.29	32.96	51.11	35.51	25.28	39.21
2	85.92	72.60	30.00	46.22	37.65	16.13
3	73.33	30.00	33.33	53.66	21.95	24.39
4	88.52	76.30	28.15	45.87	39.54	14.59

. The experimental colors by Kuroiwa et al. [38], compared to our results on the CIE chromaticity diagram, are shown in Figure 13 and Figure 14. Since the true colors seen by the human eye are not those derived from RGB values, but rather those derived from CIE (x,y) values, we present a comparative analysis of color error that is obtained from the proposed color analysis method with the experimental colors from the CIE chromaticity diagram in

Table 3. Color error analysis.

Sample No.	CIE (x,y) Values (Proposed)		CIE (x,y) Values [38]		% Error
	x	y	x	y	x	y
1	0.3089	0.2856	0.3208	0.3000	3.7	4.8
2	0.3755	0.4071	0.4083	0.4083	8.0	0.3
3	0.4397	0.3856	0.4083	0.3417	7.7	12.8
4	0.3789	0.4153	0.3670	0.4000	3.2	3.8

.

The CIE chromaticity diagrams in Figure 12, Figure 13 and Figure 14 show colors as they would appear when viewed by the naked eye. The diagram divides color groups into regions on the diagram. Similar colors are close to each other in the same region. The dots in the figures show a comparison between the colors from the experiment and our proposed technique. The locations of these points are in the same color shade. In other words, the results show that the dots from our proposed method lie in close proximity to the ones from Kuroiwa’s technique. This implies that our method can accurately convert transmittance to actual colors. We also learn that sampling only specific wavelengths (RGB) when analyzed with our technique can predict the overall characteristics of the transmittance curve. This technique can be applied to complex tasks that require a lot of effort to calculate or measure the transmittance characteristic. By calculating or measuring just three wavelengths (RGB), transmittance across the visible light regime can be mapped to true colors, without the need for all the transmittance values across the entire VIS regime.

4. Conclusions and Future Work

This paper proposes a design of an anti-counterfeiting security film based on a frequency-selective surface (FSS). Our film can display different colors on each side of the film and the colors are swapped when the film is reversed, due to the film’s FSS behavior.

In this design, aluminum nanoparticles are arranged in an array on a two-dimensional plane that is embedded in PET. This arrangement allows us to adjust the wavelengths based on the separation distance, as well as the dimension of the cubes. The performance of the film is evaluated by an FEA method.

The parametric studies show that the nanocube’s width (W) and length (L) are the critical parameters in the film design. By adjusting the size of the nanocube appropriately, it is possible to obtain the transmittance characteristic that changes the color of the light when passing through or reflecting off the film. The condition requires that the size of the nanocube must be smaller than the wavelength of the incident light. This technology has a low cost when it is mass produced. It has a simple structure and the color-changing phenomenon can be observed clearly by the naked eye. Additionally, to bring it to industrial use, the authors plan to create the best-sized prototypes as presented in this paper by roll-to-roll nanoimprint lithography or other technology [38,41].

The average efficiency of light transmittance in the 550 nm to 700 nm region exceeds 50% and the average reflectance is less than 10%. The average light transmittance efficiency in the 400 nm to 550 nm region is less than 10% and the average reflectance is more than 50%, enabling our film to perform well as an LPF.

We also present a method for analyzing the colors of nanofilm by converting the transmittance to color. This technique yields colors that are close to the actual colors obtained experimentally. The CIE chromaticity diagram confirms the correctness of our analyzed colors with the colors that are seen by the human eye. The result shows less than 13% of color error and reaffirms the accuracy and efficiency of our method.