Quantifying Microplastic Leaching from Paper Cups: A Specklegram Image Analytical Approach

Abstract

The study proposes a novel speckle interferometric method for detecting and quantifying microplastic leaching from paper cups, addressing concerns raised by the World Health Organization regarding human health risks. Hot water at varying temperatures is placed in 36 paper cups from different manufacturers, and the specklegrams of the paper cups’ interior surface are recorded. The quantity of microplastics leached into water is estimated by the Neubauer chamber method, which increases with rising water temperature. Surface morphology analysis through atomic force microscopic images reveals thermal-induced melting and smearing of microplastics, decreasing roughness parameters. Co-occurrence matrix analysis of specklegrams correlates image parameters—inertia moment, homogeneity, energy, contrast, and entropy—with the microplastics count, showing surface modifications and altered pixel intensity distribution with increasing water temperature. Regression equations based on image parameters establish a strong correlation with the microplastics count, that are validated against the Neubauer chamber method. The study indicates contrast as the potential sensitive specklegram feature for microplastics detection and quantification.

1. Introduction

The growing concern for a healthy lifestyle and the inhibition of the spread of contagious diseases triggered the widespread use of paper cups and food packaging materials; their plastic coatings started posing health issues slowly over time. Today, the microplastics entering the food chain from the coatings of food containers and packaging materials are reported to result in serious health issues [1,2]. A recent study reports that humans ingest nearly 80 mg of microplastics daily [3]. Studies also report detection of microplastics in human excreta, liver, spleen, lung, colon, placenta, and breastmilk [4]. The toxic effects of microplastic exposure include organ dysfunction, oxidative stress, neurotoxicity, DNA damage, metabolic disorder, and immune response, along with developmental and reproductive toxicity [5,6,7,8]. Besides causing health issues, microplastics lead to environmental pollution, depletion of resources, and emission of greenhouse gases [9,10]. The current estimation of global plastic production is around 400 million metric tons per year, as per the 2021 statistics. If this trend continues, global plastic waste generation is forecasted to reach 1 billion metric tons by 2060 [11].

Plastic particles of size between 1 μm and 5 mm are usually referred to as microplastics [12]. Due to their small size and persistence in the ecosystem, they potentially impact marine life, terrestrial environments, and the food chain. The major microplastics detected include polyethylene, polyethylene terephthalates, polyvinyl chloride, polystyrene, polypropylene, and polycarbonate [13]. The hazardous effects of microplastics are exacerbated when they combine with a multitude of toxic pollutants like polychlorobiphenyls, dichlorodiphenyltrichloroethane, persistent organic pollutants, organo-halogenated pesticides, polycyclic aromatic hydrocarbons, nonylphenol, and dioxin [14]. This requires a comprehensive study of microplastic detection and mitigation. The literature reveals that researchers have adopted and developed several techniques for the detection of microplastics. Visual analysis and scanning electron microscopy have been used as techniques for the identification, quantification, and classification of microplastics in studies conducted by Karlsson et al. [15] and Shi et al. [16]. On the other hand, Huang et al. developed hyperspectral imaging as a potential online detection method for microplastics [17]. Majewsky et al. [18] and La Nasa et al. [19] adopted thermal analysis and thermoanalytical methods for the quantification and determination of microplastics. Furthermore, mass spectrometry and flow cytometry are employed for the analysis of microplastics in water bodies in independent investigations carried out by Fabbri et al. [20] and Li et al. [21]. Bianco et al. [22] investigated the fractal properties of holographic patterns from microplastics and microalgae, showing that these patterns can be used to distinguish between pollutants and bioindicators like diatoms. Their approach enables automated in situ detection of microplastics, using portable holographic systems to map pollutants and assess water quality. Valentino et al. [23] developed a polarization-sensitive holographic flow cytometer integrated into a Lab-on-Chip platform, which adds material specificity to the detection of microplastic fibers. This system can accurately distinguish synthetic fibers from natural ones, such as cotton and wool, without the need for sample pre-treatment or large concentrations. In both cases, quantification is achievable, but the techniques require additional steps, such as calibration and algorithm development, to translate the detected features into precise concentrations.

Of various spectroscopic techniques widely used for their sensitivity and accuracy, Araujo et al. [24] utilize Raman spectroscopy for the identification of microplastics, whereas Kappler et al. [25] and Peez and Imhof [26] employ Fourier Transform Infrared and Nuclear Magnetic Resonance Spectroscopy for the analysis, identification, and quantification of microplastics, respectively. Detection of microplastics in the environment is a tedious process due to their varied sizes, shapes, and properties like translucency and transparency. Among the optical methods, interferometric techniques are unique because of their ultra-sensitivity and accuracy. Electronic Speckle Pattern Interferometry (ESPI) is a proven optical interferometric technique as an emerging non-destructive evaluation method [27]. Speckle-based techniques utilize laser speckle patterns to correlate image parameters with microplastic counts, providing a non-invasive, high-throughput solution ideal for rapid and large-scale analysis. This technique removes the need for time-consuming manual counting or chemical treatment, making it both efficient and eco-friendly. While “laser speckle” refers to the interference patterns created by coherent laser light scattered from a rough surface, ESPI enhances this concept by integrating electronic image capture and computational analysis. ESPI enables real-time or high-precision measurement, extending the capabilities of traditional optical speckle techniques. Today, ESPI finds applications in various fields of science and technology for measuring strain, surface roughness, in-plane and out-of-plane displacement, and vibrations on film surfaces [28,29,30,31]. Apart from these, the ESPI has revealed its potential in biology, biomedical, and biomedical instrumentation [32,33]. Speckle refers to the granular patterns resulting from the interference of coherent and dephased wavelets, influencing intensity distribution over a surface. The object beam and the reference beam, upon interference, produce speckle patterns, which allow for the computation of minute differences in the deformation or surface modification of the object. By analyzing the specklegrams using image processing techniques, the material information can be obtained [34,35].

This study represents the first attempt at applying ESPI for investigating microplastic leaching from paper cups, proposing it as a surrogate technique for estimating leached microplastics using specklegram features. The primary objective is to estimate microplastic counts through regression analysis of specklegram-derived parameters. The study also explores how the regression-based method offers a faster, non-invasive alternative to conventional techniques.

2. Materials and Method

In the current study, 36 paper cups made by three manufacturers (i.e., 12 × 3 = 36) are employed. The paper cups of these manufacturers are labeled A, B, and C. To determine the degree of microplastic leaching from the paper cups, distilled water at 30 °C, 40 °C, 60 °C, 80 °C, and 95 °C is poured into the cups (Figure 1). They are left to cool down to room temperature (30 °C), and the respective water samples are labeled as WX30, WX40, WX60, WX80, and WX95, where X = A, B, and C. For the precise estimation of leached microplastics, 36 paper cups are utilized for each temperature separately. For recording the specklegram, a 2 × 2 cm² portion is cut from the base of all the paper cups. The sectioned paper cup samples corresponding to 30 °C, 40 °C, 60 °C, 80 °C, and 95 °C are labelled respectively as X30, X40, X60, X80, and X95.

To understand and estimate the leaching of microplastics from the paper cups, the water samples are placed in the Neubauer counting chamber beneath the coverslip and inspected using an optical microscope—Olympus CX31 (Olympus Medical Systems, Gurgaon, India) [36]. The Neubauer chamber of dimension 3 mm × 3 mm consists of nine squares, each of area 1 mm². The grid pattern on the Neubauer chamber is systematically examined, and microparticles within specific areas are counted. The quantity of particles per liter in the sample is estimated by calculating the average count per grid square or unit volume (1 mm² × 0.1 mm = 10⁻⁷ L). Microplastics per liter are computed by multiplying the average number of particles per square by 10⁷ [37].

The electronic speckle interferometer, consisting of a He–Ne laser (Holmarc Opto-Mechatronics, Kochi, India), (5 mW, 632.8 nm), beam splitter, spatial filter, glass window, diffuser plate, mirrors (M1, M2, M3, M4, M5), a CCD camera (Holmarc Opto-Mechatronics, Kochi, India) (1/2.5” 5 Multi-Purpose Camera (MPC) Metal Oxide Semiconductor (MOS) Color), and a computer is employed to acquire the specklegrams. The schematic setup of ESPI utilized in the present study is represented in Figure 2. To record the specklegram, the laser beam is initially split into the reference and object beam, which, on interference, produces the specklegram. The spatial filter configuration and the aperture help in recording noise-free specklegrams, which are captured by the CCD camera using the uEye V 3.90 software [34,38,39]. The specklegrams (400 × 400 pixels) thus recorded are processed and analyzed using Matlab R2022b software. From the specklegrams, the image features, energy, contrast, homogeneity, entropy, and inertia moment, are determined to understand the variations in surface morphology.

The extraction of texture information from an image is done by employing a second-order statistical tool, the Gray Level Co-occurrence Matrix (GLCM), which represents number of horizontal jumps from one gray level to another within its immediate neighborhood [40]. The co-occurrence matrix M(i,j) is where the matrix element represents the frequency of intensity value j, preceded by intensity value i, while moving horizontally across an image. The Matlab function “graycomatrix”, with an offset [0, 1] for 256 gray levels indicating the horizontal closeness of pixel components, is employed to get the GLCM [41]. From this, features like inertia moment, energy, contrast, homogeneity, and entropy are extracted and used to describe different characteristics of an image. Inertia Moment (IM) stands for the spread of the normalized co-occurrence matrix and is represented using Equation (1) [39]. The IM represents a measure of the distribution of pixel intensities relative to a reference point, often used to describe the shape or spatial characteristics of objects in an image. It helps in tasks like object recognition, segmentation, and feature extraction by quantifying how pixel values are spread around a centroid or axis.

$$\mathrm{IM}={\sum }_{i,j}{M(i-j)}^2$$

(1)

Energy (E) is another feature referring to the uniformity of an image, which is calculated as the sum of squared elements of M(i,j) and is shown in Equation (2). Also known as angular second moment, energy quantifies the level of repetition or regularity in an image’s texture. A high energy value indicates repetitive, structured patterns, while lower values suggest more variability or less regular patterns.

$$\mathrm{E}={\sum }_{i=1}^N{\sum }_{j=1}^N{M(i,j)}^2$$

(2)

Contrast (C) measures the sharpness in structural variations in an image. Larger values point to a higher degree of intensity variation, representing a more textured image [41], calculated using Equation (3). It measures the intensity difference between pixels, reflecting how much variation exists between light and dark areas of an image. Physically, higher contrast indicates sharper distinctions between objects or regions, making features more visible, while lower contrast suggests a more uniform or washed-out appearance.

$$\mathrm{C}={\sum }_{i=1}^N{\sum }_{j=1}^N{(i-j)}^{2}M(i,j)$$

(3)

Homogeneity (H) gives the closeness of pixel distribution, where greater values represent more uniformity in the gray levels, and is inversely correlated to contrast [41], as computed by Equation (4). This measures how similar or uniform the pixel intensities are within a region. A higher homogeneity value indicates that the pixel intensities are very close to each other, typically found in smooth regions with minimal texture.

$$\mathrm{H}={\sum }_{i=1}^N{\sum }_{j=1}^N{\displaystyle \frac{M(i,j)}{1 + |i - j|}}$$

(4)

The disorder or randomness in the pixel intensities of an image can be understood from the feature entropy (S). This measures the randomness or unpredictability in the pixel intensity distribution. It is derived from the probabilities of gray-level occurrences in the image or texture. Higher entropy corresponds to complex or chaotic patterns, such as noisy regions, while lower entropy is associated with more uniform or simple textures. Hence, the smaller the entropy, the more similar the images [42]; it is computed using Equation (5).

$$\mathrm{S}={\sum }_{i=0}^{N-1}{M(}_{i,j}\left){log}_2\right({M(}_{i,j}\left)\right)$$

(5)

3. Results and Discussion

The increased use of paper cups as containers for hot drinks has started posing environmental and human health issues because of microplastic leaching from the inner plastic coatings of the cups. Number of microplastics (N) leached into water samples WX30, WX40, WX60, WX80, and WX95 are estimated by the Neubauer chamber method. The changes in the average number of particles in the samples is depicted in Figure 3. The study points to the increase in the leaching of microplastics into water as the water temperature (T) rises in the paper cup.

The leaching of microplastics brings about surface modifications on the interior of the paper cup. The surface modifications can be visualized and quantified from the AFM images displayed in Figure 4. While Figure 4a,b gives the 2D representative images of paper cup surfaces A30 and A95, Figure 4a′,b′ give the respective 3D images in size compared to those on A95. The image analysis gives the parameters detailing the surface features in

Table 1. AFM image parameters detailing the surface features.

Parameters	Samples
	A30	A95
RMS roughness (nm)	6.3469	3.2417
Maximum Peak height (nm)	34.1735	15.8065
Average roughness (nm)	5.0207	2.7027
Average height (nm)	16.1875	8.5706
Surface skewness	0.2646	−0.0237
Surface kurtosis	2.8708	2.1704

. When A30 exhibits peaks of a maximum height of 34 nm, A95 exhibits a maximum peak height of 15 nm. Table 1 reveals that the RMS roughness and average height of A30 are superior to that of A95 because of the thermal-induced melting and smearing of microplastics due to surface tension, resulting in the smaller grain size displayed in Figure 4b. The positive value of skewness for A30 also suggests that it possesses a more asymmetric distribution than A95, as evidenced by the histogram of height distribution as seen in Figure 4a″,b″. Kurtosis is another measure of the image revealing the peakedness or flatness, indicating whether the distribution is centered around the mean. Here, the kurtosis value for A30 is superior to the kurtosis value of A95. The greater kurtosis value of 2.87 suggests that the image peaked around the mean more than the normal distribution. In other words, the particle height distribution in A30 has a heavier tail than A95. Hence, the distribution in A30 is called leptokurtic, while that of A95 is termed platykurtic.

The microplastic leaching-induced surface modification is analyzed using the ESPI technique, and the specklegram image features of the paper cup are correlated to number of microplastics leached. The specklegrams of the sectioned paper cup samples X40, X60, X80, and X95 are recorded and subtracted from X30, which is taken as the reference. The subtracted images are labeled SX40, SX60, SX80, and SX95. The representative specklegrams and the subtracted images for X = A are represented in Figure 5. The specklegrams SX40, SX60, SX80, and SX95 are analyzed to acquire the image features to draw insights into the leaching of microplastics from paper cups.

A co-occurrence matrix (CM) is a statistical method used in image processing and texture analysis to characterize the spatial relationships between pixel intensities in an image. It is computed using the frequency of pairs of pixel values occurring at specific spatial relationships within an image. The representative CM of the specklegrams of SA40, SA60, SA80, and SA95 with its 3D representation is displayed in Figure 6. The image features, inertia moment, energy, contrast, homogeneity, and entropy, are extracted from the CM. The spread of pixel intensities is more prominent for greater temperatures, owing to the roughening of the surface. The inertia moment measures the arrangement of pixel intensities in the CM around the mean intensity value. If the image is uniform, the CM will be diagonal, and any deviation from uniformity will appear as a spread in the CM. Hence, IM can serve as a sensitive measure of the image’s texture. The greater the spread in the CM, the more nonuniformity in the image. The IM values are computed for the 36 paper cups, and their boxplots corresponding to SX40, SX60, SX80, and SX95 are depicted in Figure 7. The average values for IM for the 36 paper cup samples, SX40 to SX95, are found to increase with water temperature and their variation is seen in Figure 8. The correlation of N vs. T (Figure 3) with IM vs. T gives the variation of N with IM (Figure 8b). Upon curve fitting the plot of N vs. IM, Figure 8b, we get a quadratic regression equation of the form as given in Equation (6) with a high R² value of 0.9965.

N (in million) = a IM² + b IM + c

(6)

where a = −1689.2, b = 5098, c = −3530.3, and IM in 10⁸.

The equation can be utilized for the estimation of number of microplastic particles (N) leached from the paper cup coating using the IM of the specklegram recorded for the paper cup.

The uniformity of the texture of images can be understood from the specklegram features, homogeneity (H), and energy (E). The mean H and E values are computed for the 36 paper cups from their corresponding boxplot (Figure 7). The variation of homogeneity with temperature (Figure 9a) indicates a transition from a homogeneous to a heterogeneous structure with an increase in temperature, as observed in the variation of CM and IM. The decrease in homogeneity can be attributed to the increasing number of particles leached from the paper cup coating, as evidenced by the plot of N vs. H (Figure 9b), which shows a second-order polynomial, Equation (7), with R² =0.95.

N (in million) = a H² + b H + c

(7)

where a = −2,107,632.98, b = 4,092,993.87, and c = −1,987,024.05.

Energy is another measure of image uniformity, and its variation with temperature is represented in Figure 9c. The decrease of energy with temperature is due to the creation of a significantly higher number of pits on the paper cup surface, which substantially influences the phase of the reflected waves, resulting in more regions of destructive interference in the specklegrams. As E reflects the surface modifications caused by the microplastic leaching, it could be correlated with the N vs. E plot (Figure 9d). Interestingly, similar to the variation of N with IM and H, N exhibits a second-order polynomial variation with E, having a high R² value of 0.92, as given in Equation (8).

N (in million) = a E² + b E + c

(8)

where a = −21,930, b = 35,841, and c = −14,523.

Image contrast (C) and entropy (S) are two critical features of an image that can throw light into its surface morphology. When C measures the brightness differences between the pixels of an image, the entropy quantifies the randomness or the complexity of the pixel values. The mean image contrast and entropy of the specklegrams A30 to A95 are computed by box plot analysis. Image contrast provides surface morphological information by distinguishing the regions of dark and bright areas in an image. The variation of C with T (Figure 10a) reveals that the contrast increases with increasing temperature. Increasing C, though slowly initially, exhibits a rapid rise above 80 °C, suggesting the possible surface modifications resulting in the clear distinction between the dark and bright regions. When the AFM image given in Figure 6 is analyzed using C, it becomes clear that with temperature, the specklegram contrast also increases from A40 to A95. The increased contrast level of A95 results from the emergence of a greater number of micropits on the paper cup surface. The leaching of microplastics and the thermal-induced fragmentation of the plastic coating results in the generation of a more significant number of distinct boundaries, as we see in the AFM image of A95, which suggests the possible correlation of C with N (Figure 10b), with a high R² value of 0.98. The regression equation connecting C and N is given by Equation (9).

N (in million) = a C² + b C + c

(9)

where a = −619,271, b = 70,009, and c = −1865.7.

Since entropy is seen as a measure of information generation in a dynamic system, in image analysis, it relates to the randomness or uncertainty in the pixel values. Thus, the greater the entropy, the greater the variability in the pixel values, indicating lesser surface uniformity. The variation of S with T (Figure 10c) shows the increasing S with T owing to variations in the surface. As explained earlier, the leaching of microplastics and the thermal-induced fragmentation of the plastic coating results in the generation of a greater number of microparticles, as revealed in the AFM image of A95, which accounts for the increasing value of S with T. The regression equation connecting S and N with good R² = 0.98 is given by Equation (10). Though Figure 10 reveals a good sensitivity for C and S above 60 °C, the microplastic number estimation using C and S (Equations (9) and (10)) is noted to be less sensitive compared to parameters IM, E, and H in estimating N (Equations (6)–(8)). Also, it is interesting to observe that the relations (Equations (6)–(10)) connecting N and the image features—IM, H, E, C, and S—follow a second-order polynomial variation.

N (in million) = a S² + b S + c

(10)

where a = −2242.9, b = 17,174, and c = −32,761.

The study has been validated by determining the number of microplastics leached from a paper cup with water at 95 °C using regression Equations (6)–(10). The microplastic counts estimated by the Neubauer chamber method and those calculated using specklegram features are compared in

Table 2. The number of particles leached from the paper cup with water at 95 °C estimated by the Neubauer chamber method and the value calculated using specklegram features.

Parameter	Parameter Values	Measured Value of Number of Particles by Neubauer Chamber Method (in Million)	Calculated Number of Particles Using Regression Equation (in Million)	Percentage Error
Contrast	0.053672	109	107	1.834
Homogeneity	0.973170	109	100	8.25
Energy	0.834450	109	114	4.58
Entropy	3.84390	109	114	4.58
Inertia Moment	1.2 × 108	109	154	41.28

. Among the parameters, contrast exhibits the highest accuracy with an error of only 1.83%, followed by energy and entropy, both showing moderate errors of 4.58%. Homogeneity displays a slightly higher error of 8.25%, while inertia moment demonstrates a significant error of 41.28%, indicating its limited reliability due to its sensitivity to noise or variability in speckle patterns. Contrast proves to be a robust predictor for microplastic quantification, as it directly measures variability in intensity without relying on complex spatial or statistical relationships. It is calculated as the ratio of the standard deviation of intensity values to the mean intensity, making it less susceptible to errors introduced by noise or artifacts. In contrast, parameters like entropy and inertia moment depend on probability distributions or spatial relationships, making them more prone to distortions caused by noise or uneven speckle patterns. Additionally, contrast captures immediate local intensity variations, avoiding dependencies on statistical uniformity or spatial arrangements, further enhancing its robustness. The results validate the specklegram-based method as a promising, non-invasive alternative for microplastic detection, with potential improvements in parameter selection and model refinement to enhance overall accuracy and reliability.

4. Conclusions

The issue of microplastics leaching from paper cups has undeniably become a threat to human health, as pointed out by the World Health Organization. This study proposes a speckle interferometric-based method for the detection and quantification of microplastics leaching from paper cups. For the study, hot water at various temperatures is taken in 36 paper cups (12 from 3 different manufacturers), and the specklegrams of the paper cups are recorded. An optical microscopic study was done to understand microplastic leaching. The quantification of microplastics in the samples was performed using the Neubauer chamber method, which revealed an increasing number of microplastics leached into the water with increasing water temperature. The surface morphology of the paper cups (X30, X40, X60, X80, and X95) is understood from the AFM images. The RMS roughness, peak height, average height, skewness, average roughness, and kurtosis are estimated from the AFM images and are observed to be decreasing with increasing water temperature, owing to the thermal-induced melting and smearing of microplastics due to surface tension. Constructing co-occurrence matrices of the specklegrams of the paper cups, the image parameters inertia moment, homogeneity, energy, contrast, and entropy are determined and correlated to the microplastic count obtained from the Neubauer method. The leaching of microplastic particles from the coating of paper cups results in surface modification. Reduction of particle size, the smearing of particles on the paper cup coatings, and increasing water temperature alter the pixel intensity distribution in the specklegram. The spread of the co-occurrence matrix of the specklegrams with T of water taken in the paper cup leads to increasing IM values. The reduction of particle size on the paper cup surface, as revealed through AFM analysis, explains the decrease in the parameters H and E with T. This indicates the transition of the pixel intensity distribution towards heterogeneity, resulting from an increased number of particles leached from the paper cup coating. The increasing value of C from A40 to A95 is attributed to the creation of a greater number of micropits on the paper cup surfaces owing to the leaching of microplastics from the paper cups. It is for the same reason that the image entropy is also seen to increase with T. The strong correlation of image parameters IM, E, H, S, and C with the number of microplastics (N) leached into the water enables the setting of regression equations for N based on IM, E, H, C, and S. The study was validated by quantifying the number of microplastics leached from a paper cup with water at 95 °C using regression equations. Among the features analyzed, C demonstrated the highest accuracy with minimal error, while calculations based on IM are found to be the least reliable.

Thus, the current work reveals the potential of speckle interferometric techniques in analyzing microplastic leaching from paper cups, thereby suggesting ESPI as a surrogate sensitive method for microplastic detection and quantification. The method effectively quantifies microplastic concentrations in water as the temperature rises, linking thermal effects to the extent of leaching. Surface analysis reveals that increased temperatures induce melting of microplastics, significantly altering their morphology. The main goal appears to be estimating microplastic counts using regression analysis on parameters derived from specklegrams, providing a strong correlation between these parameters and microplastic count for accurate quantification. This regression-based approach may serve as a potential alternative to traditional methods like the Neubauer counting chamber, offering a faster and non-invasive solution for microplastic analysis. This approach eliminates the need for labor-intensive manual counting or chemical processing, making it both time-efficient and environmentally friendly. Therefore, the present study proposes a speckle-based novel optical approach for environmental monitoring of microplastic leaching from paper cup.