Forensic discrimination of blue fountain pen inks based on dielectric constant property

Abstract

The present work aims to show the analysis and comparison of fountain pen inks that are crucial in forensic questioned document examination, particularly in developing nations where the likelihood of fraud is higher in situations involving checks, marriage papers, birth and death records, and similar documents via using dielectric properties of material. Dielectric constant measurement is a new and deep method for discriminating fountain pen inks. To achieve this goal the dielectric constant of the identification of fountain pen ink has been studied to differentiate commercially used blue colour fountain pen inks in Turkey. The data was obtained by designing and setting up an alternating current (AC), Function Generator, and Oscilloscope to measure the resistance of each fountain pen ink sample. The measurements were performed in the frequency range between 1 Hz and 3 MHz at room temperature. Then, resistance measurements were used to calculate the dielectric constant. To support the result, as a conventional method, TLC-IA (image analysis) was applied to the same samples to discriminate these samples based on the intensity profile of red, green, and blue (RGB) by using the software PyCharm Community 2024.1.1. The results from each method supported each other. The distinction between samples can be made based on their dielectric constants in the frequency range of 2–2.5 MHz, and their loss factors in the range of 0–1 MHz have also been found to be distinguishing variables. Also, for the result of TLC-IA similarity ratios, the mean was calculated as 51.72% while the minimum value was 9.66%. For example, Sample 6 was distinguished from other samples with these two methods. So, dielectric properties, the new method, allowed us to identify the different fountain pen inks with the obtained results.

Article highlights

A novel alternating current (AC) circuit method, distinct from existing forensic techniques, examines dielectric properties.

The study compares dielectric properties for ink separation with traditional forensic approaches.

Results show dielectric properties can distinguish ink types, suggesting improvements in forensic accuracy.

Introduction

Ink analysis is a crucial and fundamental aspect of forensic document examination (FDE). It is conducted to compare, identify, characterize, and differentiate various types of inks found on a questioned document. For numerous years, a variety of methods have been employed to examine different types of ink [1]. For example, Thin Layer Chromatography (TLC) is widely used as a standard technique for distinguishing inks, providing a dependable and easily accessible tool for analysis. Fourier Transform Infrared (FTIR) Spectroscopy is commonly used to obtain precise information about the organic composition of ink. This technique operates within the frequency range of 1.2 × 10¹³ Hz to 1.2 × 10¹⁴ Hz and provides excellent insights into the chemical structure of ink components. In addition, UV/Visible Spectroscopy is commonly employed to differentiate fountain pen inks depending on their dye composition and colour. This technique typically operates within the frequency range of 7.5 × 10¹⁴ Hz to 1.5 × 10¹⁵ Hz. Although Inductively Coupled Plasma Atomic Emission Spectroscopy (ICP-AES) and Inductively Coupled Plasma Mass Spectrometry (ICP-MS) provide accurate analysis, they are not widely employed for routine tests due to their expensive maintenance and expensive [1, 2–3]. In recent years, new methods such as impedance and dielectric spectroscopy techniques have been used in forensic science. The applicability of impedance spectroscopy in forensic science has been discussed in a book published in September 2024 [4]. Furthermore, this approach has been supported [5, 6]. As a result, studying at low frequencies and analysing the electrical properties of ink samples can be a new way to discriminate ink samples. For this aim, dielectric properties have been studied.

The dielectric property is a key measure that describes the electrical insulating ability of materials and reflects the quantity of energy stored and released inside them. The dielectric constant quantifies the capacity of a material to retain electric potential energy when subjected to an electric field, using induced polarisation. Dielectric materials exhibit polarisation when exposed to an electric field, which can be categorised into three types: ionic, dipolar, and electronic [7, 8–9]. Ionic polarisation arises in materials characterised by ionic bond structures, such as NaCl. Applying an electric field causes the ions to experience a force, which in turn causes them to move and change the length of the bond. This ultimately results in the creation of a dipole moment. Dipolar polarisation is a phenomenon observed in materials that possess permanent dipole moments, such as H₂O. In the presence of an electric field, these dipoles undergo oscillations, resulting in the generation of a dipole moment. Electronic polarisation, a phenomenon observed in all materials, refers to the displacement of electron clouds within atoms or molecules when subjected to an electric field, leading to the creation of a dipole moment [7]. Electronic polarisation is a phenomenon that happens in all materials. It refers to the separation of electron clouds within atoms or molecules when an electric field is applied, leading to the creation of a dipole moment. The polarization-induced dipole moment plays a crucial role in determining the polarizability and dielectric properties of materials [7, 8–9].

Dielectric permittivity is defined as the ratio of the permittivity of a substance to the permittivity of dry air. Relative permittivity is a quantitative measure of how easily a material can become polarised when exposed to an electric field. It arises from multiple causes, such as electronic, atomic, dipolar, space charge, and ionic effects [10, 11–12]. Structural disruptions, usually produced by local variations in polarization, often lead to significant changes in the dielectric constant and other physical parameters. External electric fields induce polarisation in the electron distributions of molecules or atoms, and dielectric materials are commonly employed to store electrical energy by separating charges [13].

An examination of the existing body of research indicates that dielectric constant, often referred to as relative permittivity (ε_r) and conductivity (σ)[14] are fundamental variables frequently employed to ascertain the way electromagnetic (EM) radiation is propagated, absorbed, and reflected in biological tissues [15]. These properties have been quantified to comprehend the interaction between electromagnetic fields and the human body [16]. Additionally, they have been utilised for the identification of cancer tissue [17], biopsied tissues [18], the detection of blood cells in urine [8, 17, 18, 19, 20–21] applications in the food industry [7, 14, 20, 21, 22, 23–24] and addressing concerns about seawater [25].

According to the traditional methods in forensic sciences, classic high-frequency techniques are commonly employed for wide-ranging examination or general screening purposes. On the other hand, the dielectric measurements performed at lower frequencies (in the kHz or MHz range) offer more intricate insights into the interior composition of materials. At lower frequencies, the effects of polarisation and other intrinsic features of materials become more prominent, leading to more accurate and detailed outcomes. Hence, dielectric measurements in forensic science facilitate more intricate examinations, namely in areas like forgery detection and identity verification, in contrast to conventional techniques. The dielectric constant is barely used in forensic science. In this field, the importance of dielectric constant cannot be found out yet.

This research examines dielectric properties through many methodologies and introduces a novel approach utilising an alternating current circuit. The merits and demerits of conventional methods are also analysed comprehensively. After this introduction, the remainder of the work is structured as follows: The subsequent section provides a detailed explanation of the experimental part, followed by the study's materials and methods. The results and discussion section delineates the results derived from the experiments and their subsequent analysis and discusses the findings, incorporating comparisons with alternative methodologies. The conclusion is the last section and ultimately presents overarching findings and recommendations for subsequent research.

Experimental admittance

The step of sample preparation and the dielectric constant measurement set-up using an alternating current (AC) were shown in Fig. 1 and Fig. 2., respectively. Electrical measurements tested electric conductivity by putting probes into the prepared ink solutions in the Eppendorf tubes. The electric conductivity of fountain pen inks (FPI) solutions and resistance (Z) were measured using an AC by keeping the frequency between 1 Hz and 3 MHz. The variable resistance icon shows the value for the FPI solutions obtained resistance values (R2). 1 kΩ resistance was added to the system to find out the circuit's current. Therefore, the voltage on 1 kΩ resistance was divided by this value (Eq. (7)). The current value is the same for all the circuits. The voltage on the FPI was calculated by subtraction between the source voltage and 1 kΩ resistance voltage (Eq. (8)). To find out the R₂ value, V_R2 has been divided by founded I (A) value Eq. (9).

TLC is regarded as an environmentally friendly analytical technique and is the preferred method in laboratories with financial constraints. The key advantages of this technology over other chromatographic procedures such as HPLC and Gas Chromatography (GC) are its simple setup, low cost, and lack of maintenance requirements. Previously, TLC was exclusively employed for qualitative and semi-quantitative analysis. Recent developments in digital image technology allow for a more precise quantitative analysis of the TLC plate based on the image and several studies have been conducted on quantitative determination employing Thin Layer Chromatography- Image analysis (TLC-IA), where a flatbed scanner is utilized to capture the image of the coloured analyte spots [26]. Due to being a cost-effective and straightforward analysis method TLC is frequently employed for the qualitative identification of sample content [27]. In this study, TLC was effectively used to differentiate between the coloured components of inks and RGB (stands for red, green, and blue) [28] analysis.

Materials and methods

Samples collection

A total of 10 ink bottles/refill cartridges of FPI, as shown in Table 1, are commercially used in Türkiye and were obtained from the local market and online shopping websites.

Table 1. The list of studied blue fountain pen inks in the present study

Sample No	Brand	Model	Sample Name
1	Monopol	Bottle	1 M
2	Lankongque	Bottle	2 M
3	Parker-Quink	Bottle	3 M
4	Lamy	Refill	4 M
5	Kaweco	Refill	5 M
6	Faber Castell	Refill	6 M
7	Parker-Quink	Refill	7 M
8	Mikro	Refill	8 M
9	AIHAO	Roller	9 M
10	Waterman Paris	Bottle	10 M

Sample preparation

Samples for dielectric constant analysis were prepared by transferring 10 µL fountain pen ink to separate test tubes and diluting it until 3.5 ml with deionized water as seen in Fig. 1. Then 1.5 ml of each sample was taken to be studied via SFG-1003-Gw Instek Function Generator (Direct Digital Synthesis (DDS), 1 Channel, 3 MHz) and AATech ADS-3072B Digital Oscilloscope to test the electrical conductivity of samples [15]. While all the sample preparation steps are illustrated in Fig. 1, the experiment circuit and set-up are shown in Fig. 2 and Fig. 3, respectively [23, 25, 29].

Fig. 1 [Images not available. See PDF.]

Illustrated sample preparation and set-up

Fig. 2 [Images not available. See PDF.]

AC circuit for Dielectric Constant Set-Up

Fig. 3 [Images not available. See PDF.]

Dielectric Constant Set-Up and Sample Analysis

The dielectric constant has been shown by the symbol ${\epsilon }_r\prime$, and ${\epsilon }_r^{\prime \prime }$ is the dielectric loss factor. Conductivity is indicated by the symbol σ and the relation between σ and ${\epsilon }_r^{\prime \prime }$ is given by;

1

$${\epsilon }_r={\mathrm{\epsilon }}_r\prime -j{\mathrm{\epsilon }}_r^{\prime \prime }$$

2

$$Z=\sqrt{\frac{{\mathrm{\mu }}_0}{{\mathrm{\epsilon }}_0{\mathrm{\epsilon }}_r}}$$

3

$${\mathrm{\epsilon }}_r^{\prime \prime }=\frac{1}{Z\mathrm{\omega }}$$

4

$${\epsilon }_{\text{r}}^{\prime }={\epsilon }_0{\epsilon }_r$$

5

$${\mathrm{ϵ}}_0\approx 8.854\times {10}^{-12}\text{F/m}$$

6

$${\mathrm{\mu }}_0\approx 4\mathrm{\pi }\times {10}^{-7}\text{H/m}$$

The following equations, Eq. (7) and Eq. (8), are used to find the circuit's current and the voltage on the variable resistor, respectively.

7

$$I(A)=V_{R1}/R1$$

8

$$V_{R2}=V_{\mathit{Source}}-V_{R1}$$

9

$$R2=\frac{V_{R2}}{I(A)}$$

R2 is assumed as Z (impedance), so 1/R₂ gives us the conductivity value in S/cm (shape of sample tube ignored) and the generated model for the ${\epsilon }_r\prime$ and $\epsilon _r^{_{}^{″}}$ are shown below, respectively [13].

10

$${\mathrm{\epsilon }}_r\prime =15+\frac{25}{1+{\left(\frac{f}{20·{10}^9}\right)}^2}+\frac{2·{10}^3}{1+{\left(\frac{f}{1.5·{10}^6}\right)}^2}+\frac{2·{10}^5}{1+{\left(\frac{f}{{10}^3}\right)}^2}+\frac{2.5·{10}^7}{1+{\left(\frac{f}{20}\right)}^2}$$

11

$${\mathrm{\epsilon }}_r^{\prime \prime }=\frac{15·{10}^9}{f}+\frac{2·\frac{f}{{10}^9}}{1+{\left(\frac{f}{20·{10}^9}\right)}^2}+\frac{2·{10}^3·\frac{f}{{10}^6}}{1+{\left(\frac{f}{1.5·{10}^6}\right)}^2}+\frac{2·{10}^5·\frac{f}{{10}^3}}{1+{\left(\frac{f}{{10}^3}\right)}^2}+\frac{2·{10}^6·f}{1+{\left(\frac{f}{20}\right)}^2}$$

Thin layer chromatography

Each extracted ink sample was spotted with a volume of 1 µL on a TLC plate (Silica gel 60 F254(20 × 20 cm), Merck, Germany) using a micropipette. According to the literature optimum solvent system was chosen as Ethyl Acetate: Ethanol: Distilled Water (85:10:5) [30]. The retention factor (R_f) and colour tones of the separated bands were determined. The procedure was repeated three times for each sample to obtain better reproducibility [2, 3, 31].

Hardware and software

Properly evaluating and spotting the spots in the TLC images, a website, ASPOSE [32], was used to split the TLC images into the same sizes. Then the software PyCharm Community 2024.1.1 [33] was used to calculate the RGB values and create the heat maps of correlations.

Optimization of TLC conditions

The activity of sources is contingent upon the specific material and instrumental system employed [34]. Even in High-Pressure Liquid Chromatography (HPLC), the dielectric constant of the solvent has a notable impact [35]The mobile phase had the greatest impact on the separation of the samples. Given the diverse range of ink samples, it was advantageous to discover a singular mobile phase system capable of effectively separating most ink samples, yielding reliable and consistent data that can be used for comparison. The list of the mobile phases is shown in Table 2, and the chosen solvent system is shown in bold.

Table 2. The list of mobile phases in the present study

Mobile Phase	Relative Ratio
Ethyl acetate: Methanol: Ammonium Hydroxide	70:35:30
Ethyl acetate: Ethanol: Water	85:10:5
n-Hexane: Ethyl acetate	50:50
Ethyl acetate: Butanol: NH3	60:35:5
Toluene: acetone: ethanol: NH3	30:60:7:2
Butanol: Acetic Acid: Water	60: 15: 25
Ethyl acetate: ethanol: H2O	70:35:30

Optimization of AC set-up conditions

Initially, analyses were conducted using a fixed resistor with a resistance of 0.1 Ω. After discovering that the fixed resistance at this specific value did not produce the expected response to the applied frequency (as shown in Fig. 4), more resistance experiments were conducted until a complete wave was recorded on the oscilloscope. Resistors with values of 50Ω, 100Ω, and 1 kΩ were tested, and it was determined that the most suitable resistance was 1 kΩ Fig. 5. The purple wave represents the wave originating from the source, whereas the yellow wave corresponds to the wave that has passed through the fixed resistance sample.

Fig. 4 [Images not available. See PDF.]

0.1 Ω fixed resistance and wave (yellow)

Fig. 5 [Images not available. See PDF.]

1 kΩ fixed-resistance and wave (yellow)

Image processing software

For the discrimination of blue fountain pen inks, the components of each sample were first separated by TLC. All the images were split into the same sizes to be normalized. The normalized-RGB color components (RGB) were calculated by using the equations as follows [19, 22]:

12

$$r_R=\frac{R}{R+G+B}$$

13

$$r_G=\frac{G}{R+G+B}$$

14

$$r_B=\frac{B}{R+G+B}$$

Computing similarity

Normalization is crucial for enabling the comparison of RGB profiles from various images by adjusting the values to a standardized range. This guarantees that each channel has an equal impact on the study. Ensuring that the values are scaled to the range of 0–1 ensures that the color channels are equally adjusted, which is essential for precise comparison and analysis. This prevents any one channel from unduly affecting the results. Moreover, in statistical analyses such as correlation, it is crucial that the data possess identical scale and distribution. To normalize RGB profiles, each color channel's value is divided by the sum of the color values. This technique calculates the proportion of each channel in relation to the total intensity, guaranteeing that the channel values are confined to the range of 0–1. After the process of normalization, the Pearson correlation coefficient is employed to quantify the extent of the linear association between two sets of data. The Pearson correlation coefficient is computed using the following formula [2, 26].

15

$$r=\frac{{\sum }_{i=1}^n\left(X_i,-,\overline{X}\right)\left(Y_i,-,\overline{Y}\right)}{\sqrt{{\sum }_{i=1}^n{\left(X_i,-,\overline{X}\right)}^2{\sum }_{i=1}^n{\left(Y_i,-,\overline{Y}\right)}^2}}$$

The similarity score between RGB profiles is calculated by taking the average of the Pearson correlation coefficients for each color channel (R, G, B). To represent the similarity score as a percentage, the resulting score is multiplied by 100. Similarity (percentage);

16

$$\left(\frac{r_R+r_G+r_B}{3}\right)\times 100$$

Result and discussion

The dielectric properties results dielectric constant is a crucial factor in defining the electrical characteristics of a substance [7]. Measuring the dielectric constants of materials including plastics, polymers, paints, and coatings can aid in identifying the chemical compositions and origins of these materials [6, 8, 36, 37].

The measurements between the frequency range of 1–100 Hz, 1–100 kHz, 1 MHz, 1.5 MHz, 2 MHz, and 3 MHz showed us dielectric constant measurement is suitable for discrimination of FPI. It has been determined that samples can be distinguished using their dielectric constants within the 2–2.5 MHz frequency range in Fig. 6 and by their loss factors within the 0–1 MHz range in Fig. 7. Also, the heatmap of the dielectric constant for all studied frequency ranges is shown in Fig. 8. In contrast, Fig. 9 shows the specific frequency range (2–3 MHz) for the discrimination of most of the inks that are differentiated properly, as a supporting figure of Fig. 6.

Fig. 6 [Images not available. See PDF.]

Dielectric Constant (real part) of samples

Fig. 7 [Images not available. See PDF.]

Dielectric Loss Factor (imaginary part) of samples

Fig. 8 [Images not available. See PDF.]

Heatmap of dielectric constant for samples

Fig. 9 [Images not available. See PDF.]

Heatmap of dielectric constant for the frequency range of 2–3 MHz

The TLC-IA results the result of TLC-IA was used to accurately evaluate the similarity between the normalized RGB profiles of two pictures. Figure 10 shows that samples 3 M and 10 M have a similarity percentage of 28.52% while sample 10 M has been correlated with itself in Fig. 11 and has a 100% similarity percentage. The obtained similarity percentages and normalized RGB results for all the samples are given in Fig. 12 and Fig. 13, respectively. For the similarity ratios, the mean was calculated as 51.72% while the minimum value was 9.66% in Fig. 12. These score values were described as a threshold of similarity but need to be improved. Also, the similarity of normalized RGB profiles has been shown in Fig. 13 and it can clearly be said that all the results of TLC-IA support each other. It is so clear to compare each ink.

Fig. 10 [Images not available. See PDF.]

The similarity of 3 M and 10 M fountain pen inks

Fig. 11 [Images not available. See PDF.]

The similarity of 10 M fountain pen ink with itself

Fig. 12 [Images not available. See PDF.]

1M_10M Similarity Percentage Heatmap

Fig. 13 [Images not available. See PDF.]

Similarity of Normalized RGB Profiles, TLC-IA

Conclusions

In this study fountain pen ink samples were differentiated based on qualitative examination of TLC-IA and Dielectric Constant Set-Up. TLC is widely employed for routine examination because of its straightforward methodology and affordable price. Dielectric Constant is also used extensively in the field of electrical studies due to its interaction between the material and electromagnetic field.

The Dielectric Constant System, in conjunction with TLC-IA, may effectively distinguish between fountain pen inks. This study introduces a valuable methodology for examining the composition of fountain pen inks, which is advantageous for forensic document analysts. It is important to acknowledge that this study is preliminary and just briefly explores the analysis of fountain pen inks in this aspect. Hence, comparable investigations should encompass a substantial quantity of fountain pen inks procured from a diverse array of sources or manufacturers. Because of the possibility of inconsistent contact at the connection points, in the dielectric constant system, the obtained results, while not as conclusive as those from TLC-IA, demonstrate potential for future advancement. Sample 7 M and 9 M have a direct correlation with the ratio 1 at the heatmap of the dielectric constant and these the same samples have 96.77% similarity percentages in the similarity percentage heatmap and a correlation of 0.98 in the Heatmap of the dielectric constant for the frequency range of 2–3 MHz. The same samples are indistinguishable with both techniques, but the remaining samples were effectively separated using both approaches.

The same experiment setup is used for all the ink samples. Then polarization does not change, and consequently, dielectric constant and tangent loss remain unchanged. Therefore, the study of the effect of polarization is out of the scope of this study.

Dielectric properties can be used for;

• understanding ink formulation, and its composition, as it indicates the distinct dielectric properties of its solvents, pigments, dyes, and additions

• determine the resemblance or dissimilarity between two ink samples, which may vary in their formulations.

Acknowledgements

The authors thank the Assistant Professor Javad Jangi Golezani (ORCID:0000-0002-1217-0862) and Research Assistant Enis Kranda (ORCID: 0000-0003-4390-0669) for their help and support. 

Author contributions

O. Simsek did the experiments and wrote the main manuscript. S. S. Seker did the theoretical part.

Funding

This research did not receive a specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability 

Data Generation and Collection: "This dataset was generated entirely by the authors through the capture and processing of original images. The images were created specifically for this research project and followed a standardized procedure to maintain consistency and reliability across the dataset." Repository and Accessibility: "The dataset, including the images and any derived data files, has been deposited in an accessible data repository, facilitating transparency and supporting compliance with data-sharing mandates. This setup ensures that the data are readily available for reusability and reproducibility by other researchers." Data File Formats: The primary dataset consists of image files in JPEG format. Data Identifiers: Each image file is uniquely named to reflect its place in the dataset sequence, ranging from Blue_01 to Blue_10. This naming convention allows each image to be easily identified and referenced for research purposes.

Declarations

Ethics approval and consent to participate

Not applicable as this article does not contain any biological studies with human participants or animals.

Consent for publication

All the authors have given consent for publication.

Competing interests

The authors declare no competing interests.

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

1. Purba, MK; Garg, M. A review on spectroscopic techniques for examination of ink. Bull Environ Pharmacol Life Sci; 2022; 5, pp. 299-303.

2. Sharif, M; Batool, M; Chand, S; Farooqi, ZH; Tirmazi, SAAS; Athar, M. Forensic discrimination potential of blue, black, green, and red colored fountain pen inks commercially used in pakistan, by UV/visible spectroscopy, thin layer chromatography, and fourier transform infrared spectroscopy. Int J Anal Chem; 2019; [DOI: https://dx.doi.org/10.1155/2019/5980967]

3. Sharif, M; Jalees, MI; Ali Shah Tirmazi, SA; Athar, MM; Durrani, AI; Batool, M. Discrimination of pakistani fountain pen inks by gas chromatography-mass spectrometry (GC-MS). Int J Anal Chem; 2022; [DOI: https://dx.doi.org/10.1155/2022/7186625]

4. Tejashwini, DM; Naik, R; Sunagar, V; Nagaswarupa, HP; Naik, YV. An Introduction to forensic electrochemistry. ACS Symp Ser; 2024; [DOI: https://dx.doi.org/10.1021/BK-2024-1481.CH001]

5. Lazanas, AC; Prodromidis, MI. Electrochemical impedance spectroscopy─a tutorial. ACS Measure Sci Au; 2023; 3, 3 pp. 162-193. [DOI: https://dx.doi.org/10.1021/ACSMEASURESCIAU.2C00070/ASSET/IMAGES/LARGE/TG2C00070_0032.JPEG]

6. Lee, A; Watanabe, H; Matsumiya, Y; Choi, KH; Ahn, KH; Lee, SJ. Dielectric characterization of pigment inks for electrohydrodynamic jet printing. Ind Eng Chem Res; 2014; 53, 44 pp. 17445-17453. [DOI: https://dx.doi.org/10.1021/IE5031437]

7. Wasusathien, W; Thongsopa, C; Santalunai, S; Thosdeekoraphat, T; Santalunai, N. Measurement of dielectric properties in mixtures of various rice cultivars for purpose of detecting contamination in industry. Electrotech Rev; 2024; [DOI: https://dx.doi.org/10.15199/48.2024.02.37]

8. Zhu, CZ; Ting, HN; Ng, KH. Detection of pregnancy using dielectric properties of urine. J Microw Power Electromagn Energy; 2018; 52, 3 pp. 182-197. [DOI: https://dx.doi.org/10.1080/08327823.2018.1497425]

9. Al Faruq, A; Zhang, M; Bhandari, B; Azam, SMR; Khatun, MHA. New understandings of how dielectric properties of fruits and vegetables are affected by heat-induced dehydration: a review. Drying Technol; 2019; [DOI: https://dx.doi.org/10.1080/07373937.2018.1538157]

10. “Dielectric Constant - an overview | ScienceDirect Topics.” [Online]. Available: https://www.sciencedirect.com/topics/chemistry/dielectric-constant. Accessed 18 Jun 2024

11. Newnham, RE; Giniewicz, JR. Nonmechanical properties of composites. Comprehensive Comp Mater; 2000; [DOI: https://dx.doi.org/10.1016/B0-08-042993-9/00052-8]

12. Ilic, J. “Wood: electrical properties”, encyclopedia of materials. Sci Technol; 2001; [DOI: https://dx.doi.org/10.1016/B0-08-043152-6/01744-7]

13. Seker, S; Abatay, H. New frequency-dependent parametric modeling of dielectric materials. AEU-Int J Electron C; 2006; 60, 4 pp. 320-327. [DOI: https://dx.doi.org/10.1016/J.AEUE.2005.04.004]

14. Silva, VES; Costa, DD; Sinfrônio, FSM; Barros, AK. Application of dielectric constant for identification of dilution in Raw Milk. J Adv Agricul Technol; 2021; [DOI: https://dx.doi.org/10.18178/joaat.8.1.25-29]

15. Porter, E et al. Minimum information for dielectric measurements of biological tissues (MINDER): a framework for repeatable and reusable data. Int J RF Microwave Comput-Aided Eng; 2018; [DOI: https://dx.doi.org/10.1002/mmce.21201]

16. Porter, E; La Gioia, A; Santorelli, A; O’Halloran, M. Modeling of the dielectric properties of biological tissues within the histology region. IEEE Trans Dielectr Electr Insul; 2017; 24, 5 pp. 3290-3301. [DOI: https://dx.doi.org/10.1109/TDEI.2017.006690]

17. Martellosio, A et al. Dielectric properties characterization from 0.5 to 50 GHz of breast cancer tissues. IEEE Trans Microw Theory Tech; 2017; 65, 3 pp. 998-1011. [DOI: https://dx.doi.org/10.1109/TMTT.2016.2631162]

18. Canicattì, E et al. Dielectric characterization of breast biopsied tissues as pre-pathological aid in early cancer detection: a blinded feasibility study. Diagnostics; 2023; [DOI: https://dx.doi.org/10.3390/DIAGNOSTICS13183015]

19. Nasir, N et al. Electrical detection of blood cells in urine. Heliyon; 2020; [DOI: https://dx.doi.org/10.1016/j.heliyon.2019.e03102]

20. Abdelkafi, A; Büschel, P; Fendri, A; Kanoun, O. Impedance investigation of milk dilution. AMA Service GmbH; 2020; [DOI: https://dx.doi.org/10.5162/sensor2015/a7.2]

21. Abdullah, SN; You, KY; Hisham Khamis, N; Chong, CY. Modelling the dielectric properties of cow’s raw milk under vat pasteurization. Progr Electromag Res M; 2019; [DOI: https://dx.doi.org/10.2528/PIERM19052202]

22. Lonappan, A; Thomas, V; Bindu, G; Hamsakutty, V; Mathew, KT; Rajasekharan, C. Analysis of human breast milk at microwave frequencies. Prog Electromagn Res; 2006; 60, pp. 179-185. [DOI: https://dx.doi.org/10.2528/PIER05080102]

23. Mabrook, MF; Petty, MC. Effect of composition on the electrical conductance of milk. J Food Eng; 2003; 60, 3 pp. 321-325. [DOI: https://dx.doi.org/10.1016/S0260-8774(03)00054-2]

24. Phimphisan, S; Sa-ngiamvibool, W. Determination of water added in raw milk using interdigital capacitor sensor. Przeglad Elektrotechniczny; 2015; 91, 9 pp. 265-267. [DOI: https://dx.doi.org/10.15199/48.2015.09.65]

25. S. Guo, X. Han, Y. Peng, and J. Li, “Frequency Dependence of Conductivity Characteristics of Seawater Ionic Solution under Magnetic Field”, https://doi.org/10.1051/matecconf/20166702021.

26. Kerr, E; West, C; Kradtap Hartwell, S. Quantitative TLC-image analysis of urinary creatinine using iodine staining and RGB values. J Chromatogr Sci; 2016; [DOI: https://dx.doi.org/10.1093/CHROMSCI/BMV183]

27. MuniZ-Márquez, DB; Martínez-Cervantes, MA; Martínez-Pérez, A; AguiLar, CN; AguiLar-Zarate, P; Wong-Paz, JE. A simple quantitative method using TLC-image analysis to determine fructooligosaccharides (FOS) in food samples. Turk J Chem; 2022; [DOI: https://dx.doi.org/10.55730/1300-0527.3436]

28. Djozan, D; Baheri, T; Karimian, G; Shahidi, M. Forensic discrimination of blue ballpoint pen inks based on thin layer chromatography and image analysis. Forensic Sci Int; 2008; 179, 2–3 pp. 199-205. [DOI: https://dx.doi.org/10.1016/j.forsciint.2008.05.013]

29. Angkawisittpan, N; Manasri, T. Determination of sugar content in sugar solutions using interdigital capacitor sensor. Measure Sci Rev; 2012; [DOI: https://dx.doi.org/10.2478/v10048-012-0002-0]

30. Quinteros ED. Analysis of Blue Fountain Pen Ink by UV-Vis Spectroscopy, ATR-FTIR Spectroscopy, and Thin Layer Chromatography Enhanced with RGB Spot Evaluation to Construct a Database for Comparison and Identification. In: CUNY Academic Works. 2022. https://academicworks.cuny.edu/cc_etds_theses/1000.

31. Sharif, M; Chand, S; Tirmazi, SAAS; Farooq, U; Athar, MM; Batool, M. Investigation and discrimination of ballpoint pen inks by analytical techniques and chemometrics. Int J Anal Chem; 2022; [DOI: https://dx.doi.org/10.1155/2022/7450539]

32. “ASPOSE.” [Online]. Available: https://products.aspose.app/words/tr/splitter. Accessed 09 Jun 2024

33. “PyCharm: the Python IDE for data science and web development.” [Online]. Accessed: 09 Jun 2024

34. Dorey, R. Ceramic Thick Films for MEMS and Microdevices; 2012; Oxford, William Andrew Publishing: pp. 85-112. [DOI: https://dx.doi.org/10.1016/B978-1-4377-7817-5.00004-3]

35. Moldoveanu, SC; David, V. Essentials in Modern HPLC Separations; 2013; Amsterdam, Elsevier: pp. 363-447. [DOI: https://dx.doi.org/10.1016/B978-0-12-385013-3.00007-0]

36. Hoekstra, P; Delaney, A. Dielectric properties of soils at UHF and microwave frequencies. J Geophys Res; 1974; 79, 11 pp. 1699-1708. [DOI: https://dx.doi.org/10.1029/JB079I011P01699]

37. Pringle, JK; Cassella, JP; Jervis, JR; Williams, A; Cross, P; Cassidy, NJ. Soilwater conductivity analysis to date and locate clandestine graves of homicide victims. J Forensic Sci; 2015; 60, 4 pp. 1052-1060. [DOI: https://dx.doi.org/10.1111/1556-4029.12802]