Development of a Reliable Device for ‘Fluorokinetic’ Analysis Based on a Portable Diode Array MEMS Fluorimeter

Received: 3 March 2025 Revised: 25 March 2025 Accepted: 28 March 2025 Published: 3 April 2025 Citation: González-Arjona, D.; López-Pérez, G. Development of a Reliable Device for ‘Fluorokinetic’ Analysis Based on a Portable Diode Array MEMS Fluorimeter. Chemosensors 2025,13, 128. https://doi.org/ 10.3390/chemosensors13040128 Copyright: © 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/ licenses/by/4.0/). Article Development of a Reliable Device for ‘Fluorokinetic’ Analysis Based on a Portable Diode Array MEMS Fluorimeter Domingo González-Arjona * and Germán López-Pérez Department of Physical Chemistry, Faculty of Chemistry, University of Sevilla, C/Prof. García González 1, E-41012 Sevilla, Spain; [email protected] *Correspondence: [email protected]; Tel.: +34-954557175 Abstract: A device was developed to study the evolution of fluorescence spectra as a function of time. A previously designed fluorimeter based on the diode array mini-spectrometer CM12880MA was used. The control and measurement were carried out by programming a SAM21D microcontroller. Considerations regarding the optimization of acquisition speed, memory, and computer interface have been analyzed and optimized. As a result, a very versatile device with great adaptability, reduced dimensions, portability, and a low budget (under EUR 500) has been built. The sensitivity, controlled by the integration time of the photodiodes, can be adjusted between 10 µ s and 20 s, thus allowing sampling times ranging from 10 ms to more than 10 h. Under these conditions, chemical rate constants from 20 s −1 to 10 −8 s −1 can be experimentally determined. It has a very wide operating range for the kinetic rate constant determination, over six orders of magnitude. As proof of the system performance, the oxidation reaction of Thiamine in a basic medium to form fluorescent Thiochrome has been employed. The evolution of the emission spectrum has been followed, and the decomposition rate constant has been measured at 2.1 × 10 −3 s −1 , a value which matches those values reported in the literature for this system. A Thiochrome calibration curve has also been performed, obtaining a detection limit of 13 nM, consistent with literature data. Additionally, the stability of Thiochrome has been tested, being the photodecomposition rate constants 1.8 × 10 −4 s −1 and 3.0 × 10 −7 s −1 , in the presence and absence of UV light (365 nm), respectively. Finally, experiments have been designed to obtain, in a single measurement, the values of both rate constants: the formation of Thiochrome from Thiamine and its photo-decomposition under UV light to a non-fluorescent product. The rate constant values obtained are in good agreement with those previously obtained through independent experiments under the same experimental conditions. These results show that, under these conditions, Thiochrome can be considered an unstable intermediate in a chemical reaction with successive stages. Keywords: fluorescence; chemical kinetic; MEMS spectrometer; portable device; microcontroller ; Thiamine; Thiochrome; cost effective 1. Introduction This study focusses on the development of a system capable of performing fluorokinetics analysis by using a Micro-Electro-Mechanical System (MEMS)-based mini-spectrometer, adapted as a fluorimeter, controlled and operated by a microcontroller. The device is programmed under the open-source software declaration (FOSS) [ 1 ]. As an initial application, a kinetic study of the evolution of the fluorescent product produced in the oxidation of Thiamine (TA) has been carried out. Chemosensors 2025,13, 128 https://doi.org/10.3390/chemosensors13040128 Chemosensors 2025,13, 128 2 of 16 The application of the phenomenon of fluorescence continues to be a subject of current interest in scientific publications. There are numerous applications in the literature for the development of fluorescent markers, which have applications in diverse fields, ranging from the development of fluorescent protein-based biosensors that allow for the monitoring of physiological processes [ 2 – 4 ] to the monitoring and detection of contamination in environmental water, food samples, and pesticide residues [ 5 – 7 ]. Some of these applications use fluorokinetics methods combined with chemometrics for the detection of cancer toxins [ 8 ]. The worldwide economic impact relative to fluorescent detection devices is really important in many industrial applications. Thus, the food industry is projected to reach EUR 20 billion in 2025 [ 9 ], pesticide and water pollutant analyses are estimated to be worth about EUR 1.5 billion annually [ 10 ], and water management is expected to reach approximately EUR 4.5 billion in 2027 [11,12]. Kinetic fluorometry is also used in the study of photosynthesis using a technique called delayed fluorescence using far-red light for excitation. Response times typically range from milliseconds to several minutes [ 13 ]. Recently, strategies have been developed for imaging applications of fluorescent probes for ATP [ 14 ], which use longer excitation and emission wavelengths that allow for detection in deep biological samples. In this field, a very promising fluorescent probe (Bio-SiR) has been presented for the real-time imaging of ATP in cancer cells, which is a si-rhodamine-based fluorophore [ 15 ]. It is presented as the first tumor-targeting (near IR) molecule fluorescent probe for endogenous ATP imaging in real time for tumor-bearing mice. The basic instrument (hardware) and a proposed program (software) for kinetic measurements of fluorescence spectra are discussed in this work. The development is founded mainly on studies [ 16 , 17 ] and the references therein. Both papers constitute the fundamentals of the proposed system, the former focusing mainly on hardware requirements and the latter on software basics. The initial advantages are obviously the high reproducibility and sensitivity already demonstrated, the low budget, and portability. A new step has been taken to develop an inexpensive, portable, and simple diode array kinetic fluorimeter. Thus, it is possible to take advantage of the benefits and minimize the drawbacks pointed out in the bibliography (open source, stability, sensitivity, reliability, cost, etc.). More specifically, it is necessary to develop a system capable of recording fluorescence spectra with an acquisition frequency as wide as possible, minimizing difficulties derived from the slowness in obtaining and storing the measured spectra. The previous paper proposes the use of a development board based on the ATSAMD21G18 [ 18 ] microcontroller for the acquisition and control of the Hamamatsu CA12880MA [ 19 ] diode array spectrometer. The developed software was applied for stationary measurements of fluorescent samples, spectra, and static quantitative determinations. The time variable is only employed for the control of the spectrometer lines and the time that excitation light is on, allowing a very large sensitivity range. The second paper established the basis for both the external parameter verification (controlled by using a spreadsheet) and for the precise time control during the measurement. The add-on macro software embedded in the spreadsheet for the control of the system has been kept and is based on Parallax Data Acquisition Tools Release 2B [20] but using a newer, more versatile and user-friendly version [21]. To verify the usefulness and reliability of the proposed device, vitamin B1 or Thia-mine (TM) has been chosen as a starting compound acting as a fluorescent precursor. TM is an essential vitamin for human beings, the deficiency of which produces the well-known “beriberi” syndrome. It mainly affects the nervous and cardiovascular systems when blood levels are in the range of 9–14 nmol/L [ 22 ]. In the literature, many analytical methods can be employed for TM investigations, such as colorimetric, fluorescence, electrochemical, chemiluminescence, electrochemiluminescence, HPLC, optical detection by using nanoparticle methods, etc. [ 23 ]. The determination of TM and related compounds via fluorescence has been widely described in the literature for more than 80 years [ 24 ]. Its sensitivity and specificity have been used to develop methods for the determination of other compounds that react with Thiamine [ 25 ], and Chemosensors 2025,13, 128 3 of 16 even specific detectors have been developed for its determination [ 26 , 27 ]. Ryan et al. [ 28 , 29 ] developed and exhaustively described a kinetic method for the analytical determination of TM, which is the methodological basis for the present work. Their methodology is so well founded that it has been the subject of undergraduate teaching material [ 30 ]. Even more recently, there has been some interest in the study of the oxidation of TM to form fluorescent compounds, such as Thiochrome (TC) [ 23 , 31 , 32 ]. The kinetic behavior of these compounds provides a suitable framework for testing the versatility and limits of the fluorokinetic device presented in this study. It offers many advantages in this field, such as a high sensitivity for fluorescence measurements, fast acquisition times, low cost, great portability, and full spreadsheet integration for experimental data analysis. The results are compared with those published in the literature to evaluate the system’s overall performance. 2. Materials and Methods 2.1. Experimental Analytical-grade chemicals were purchased from Sigma Aldrich: Thiamine hydrochloride (CAS 67-03-8), Na 3 PO 4· 12H 2 O (CAS 10101-89-0), Na 2 HPO 4 (CAS 7558-79-4), sodium hydroxide (CAS 1310-73-2), hydrochloric acid (CAS 7647-01-0) and mercury (II) chloride (CAS 7487-94-7). Water was purified using a Millipore system (Milli-Q Reference model), and all measurements were performed at 22 ± 2 ◦ C. Hellma Suprasil precision quartz cells (type 101-QS) with a light path of 10 mm were used for experimental measurements. Figure 1shows a picture of the proposed device with the main components of the system. As can be noted, it is a portable device with small dimensions (120 mm × 80 mm × 60 mm), with an overall weight of 0.4 Kg. As excitation sources, it contains an LED (365 nm) and a laser diode (405 nm), which is included in the GroupGets board v1. The Hammatsu C12880MA has 15 nm of spectral resolution and a wavelength range from 340 to 850 nm, allowing one to acquire 288 spectrum points. The cell compartment has a standard size of 10 mm wide. Figure 1. Image of the proposed device showing the main components of the system: (a) fluorimeter box, (b) cell and mirrors compartment, (c) cell cap, (d) USB connection cable, (e) GroupGets breakout (v.1), Hamamatsu C12880MA spectrometer and (f) microcontroller ATSAMD21G18 board. The Euro coin is included for size reference. Chemosensors 2025,13, 128 4 of 16 2.2. ATSAMD21G18 Microcontroller Hardware The characteristics of the microcontroller selected in terms of processing speed, clock frequency (CLK), and internal 32 kB SRAM memory are interesting. However, two major difficulties were encountered at the beginning. Firstly, the process to obtain the internally stored data, using the common read mode, was particularly slow. Secondly, the available microcontroller memory is limited to storing the different spectra for a kinetic process. Furthermore, in the measurement process, there was no possibility to reduce the noise level by averaging without sacrificing its speed. Strategies for accelerating data acquisition, averaging, and memory optimization will be explained in the following sections. More detailed information on these aspects can be found in the Supplementary Materials. 2.3. Speeding up the Software Code for “analogRead” Despite all the advantages mentioned in relation to the selected microcontroller, there was an added difficulty with the speed of processing the “analogRead” instruction, when the default compilation of the high-level language is employed. This instruction uses the microcontroller analog-to-digital converter (ADC) and is essential for reading/storing the light intensity values for each wavelength. A comparative speed study of the three microcontrollers (UNO, 2560, and ZERO) using the compiled instruction was performed to obtain a complete spectrum of 288 wavelength (WL) data. The time required for the first two boards was about 37 ms, but, incredibly, for the ZERO board, it was about 250 ms. Undoubtedly, the compiler used to translate the instruction for this board was inefficient and, with this timing, the use of the direct compiled instruction for ZERO makes it useless for kinetics purposes. Fortunately, this speed issue was pointed out on different internet forums. On the Arduino forum, an example code for “speeding-up” the analog-to-digital converter (ADC) compilation was presented [ 33 ]. This code involved specific port control for the SAM21D analog-todigital converter (more information is available in the Supplementary Materials). The time required for retrieving a whole spectrum varies from 8 ms for an average of two samples to 89 ms for an average of 64 samples. A study of the error associated with the number of averages shows that the standard deviation is lower, and the absolute error changes its sign for an average between 32 and 64 samples. 2.4. Timing Control One of the most important advantages of microcontrollers is the precise and fast timer control. Internal timers can be set up to obtain the number of microseconds or milliseconds elapsed since they were triggered. The time required to perform a measurement process involves different tasks, and each one has its own time consumption. To perform one single measurement and control the lines of the C12880MA, this process is fully described in the previous paper. The main difference is the faster procedure described in Section 2.3. Next, the selected number of data points must be stored in the data matrix. To read a new measurement, the program checks whether the selected sample time is over. If the gap time is large enough, the program writes into Excel and plots one WL selected value to check the measurement evolution. Once the gap time is over, the sequence is repeated until it reaches the number of measurements selected. For checking purposes, those timer values can be exported to Excel by checking the ‘Debug mode’ box in the PLX-DAQ controller add-on. A detailed description of the timing process can be found in the Supplementary Materials. Chemosensors 2025,13, 128 5 of 16 2.5. Memory Optimization Flash EPROM and SRAM memory types are implemented in microcontrollers, each one of them depending on its characteristics, and are usually assigned a specific task. Their organization and uses are widely discussed in the Arduino Memory Guide [ 34 ]. Different alternatives for the storage of kinetic data have been considered, especially taking into account two fundamental issues: the time required for their storage and the amount of data to be stored. There are also other collateral considerations such as minimizing the cost and complexity of implementation. Although it might be thought that the amount of memory available in the microcontroller is scarce, with some optimization, it is possible to store a complete spectrum every 10 ms, using just 0.5 kB. In this way, about 50 complete spectra could be stored, without consuming the entire available SRAM memory. Nonetheless, in fluorescence measurements, the area of the spectrum corresponding to the excitation source can be neglected. This fact allows a larger number of spectra to be stored in memory. On the other hand, it is possible to store more than 10,000 points if a single WL is selected (at maximum emission). With these numbers, the decision was made to use the best option offered by ATSAMD21G18, which implies an easier, cheaper, and faster development, since it has a solid base in its hardware, regarding the spectrophotometer and the development of its software. In addition, the optimization of the available SRAM provides a higher storage speed with a reasonable software modification effort to trim and organize the measured data spectra. Thus, a single array variable is used to store the succession of clipped spectra during a kinetic experiment, once the number of wavelengths and the total number of wavelengths to be stored are known. This array of clipped spectra for different acquisition times can be exported as a two-dimensional table with two output modes: columns containing the selected WL and rows for the selected time, or the transposed mode with time in cols/rows and WL. More detailed information about memory optimization and organization can be found in the Supplementary Materials. 2.6. Acquisition Software PLX_DAQ As described in previous papers, the main advantage of using PLX-DAQ software Release 2B with Excel is to read and write specific parameters via USB without having to reload the code into the microcontroller every time a parameter changes. It also allows for the storage of the results and their graphing during the measurement. Moreover, using the different checkboxes included in the PLX-DAQ, the program flow and the parameter verification and validation can be performed easily. One possible drawback is that the use of an Excel spreadsheet is mandatory, because this spreadsheet program has fundamental RS232 input/output control procedures, compatible with USB. However, its use has been developed by PLX-DAQ with open source. Version 2.11 [ 21 ] has some additional improvements, such as the selection of the spreadsheet that interacts with the USB interface and a debug output that allows for scrutinizing possible communication problems. However, the main drawback is the strict language required to build the message syntax. Special characters, usually employed for control in the Serial communication interface, can interfere with Excel: comma, spaces, line feed, and carriage return characters. Thus, it is highly recommended to carefully read the user guide. Figure 2shows an Excel screenshot of the control datasheet and the PLX-DAQ add-on (orange framed) for a dummy experiment. Chemosensors 2025,13, 128 6 of 16 Figure 2. Spreadsheet screenshot for the fluorokinetics control, showing input parameters (red frame), debugging information about the timing (green frame), PLX-DAQ add-on for Excel (orange frame), estimated parameters (blue frame) and actual kinetic data (cyan frame). See text for a detailed description. 2.7. Input Parameters The variables which control the fluorokinetics experiments are related to the spectroscopic and kinetic times. In the red frame shown in Figure 2, the first column contains a brief description of the measured parameter and employed units, if applicable; the second column contains information about their valid range. The third column is used to write the desired actual values, and the fourth column contains their estimation from the actual values. The timer value (in µ s) for debugging is displayed in the green frame. Their values are useful for additional testing and validation of the kinetic measurement process. There are some interrelationships between some input parameters, and efforts have been made to maintain the maximum performance of the device (memory, time, WL). Thus, the number of kinetics points is limited by the number of WLs stored up to a maximum of 25kB. The minimum kinetic time interval is limited by the time required to operate the CA12880MA. Therefore, once the input parameters are typed in the Excel cells, it is necessary to recheck their values to verify that all device limits are matched (blue frame in Figure 2). Thus, the number of WLs to be stored is estimated from the available memory. The minimum sampling time (in milliseconds) is estimated from the internal timer values (in microseconds) with a 5% increment for safety. The exact time at which the first spectrum is recorded corresponds to half the integration time of the first measurement (see a detailed description in Supplementary Materials). The time between measurements must be greater than the minimum sampling time. If the difference between the sampling time and the integration time is large enough (about 100 ms), it is possible to export and plot the actual kinetic data during the measurement process for a single selected WL (cyan frame in Figure 2). In this case, it is highly recommended not to operate the add-on and/or Excel during the measurement process (see Supplementary Materials for more information). Once the input parameters are set, they are checked and, if necessary, recalculated. If no errors are detected, they can be validated by marking the ‘Check_parameter’ checkbox on the add-on to proceed. As the complete data matrix for a kinetic experiment could be very large, PLX-DAQ v2.11 allows one to select more than one datasheet to export data. Thus, one datasheet Chemosensors 2025,13, 128 7 of 16 contains the control parameters, debugging, and the ADC data corresponding to the WLs selected for plotting. Other datasheets can store the entire output matrix of selected time/WL data for different experiments. This procedure can be a bit tricky initially, and special care has to be taken to avoid writing the output matrix on the control sheet or trying to read the input parameters from the data matrix sheet. Another checkbox, ‘Export_Data_matrix’, is used to minimize the possible error of writing into the wrong datasheet. 2.8. Measurement Routines In addition to the usual code for microcontroller programming, setup and loop, several routines have been developed to efficiently perform each experiment. The most important ones are listed here (more detailed information can be found in the Supplementary Materials). •Boosting analogRead procedure: AdcBooster(Avr_level ). •Generate nCLK pulses: F_pulse_clk3MHz(n). •Management of the CA12880MA control lines for a single spectrum: F_readSpectrometer(Ex_L ). •Estimation of the wavelength associated from the manufacturer data: Wavelength(order ). •Computes the associated indexes according to the WL calibration data: WL_Index_Data(&w_H, &w_L, &index_H, &index_L ). •Output matrix spectra generation by row: create_selected_matrix_by_row(&Row_Index_L, &Row_Index_H, &actual_col, *matrix). •Trimmed selection of the whole data matrix, including acquisition time: printDataMatrixExcel(mode, &Row_Index_L, &Row_Index_H, &N_points, &Delta_ms, off_t_ms, *Data_Matrix). •Clean up in Excel datasheet of the parameter estimation: Clean_estimated_param(). 2.9. Software Flowchart A very simplified flowchart of the measurement process is described in Figure 3(more detailed information can be found in the Supplementary Materials). Figure 3. Simplified scheme of the measurement process. The sequential procedure is operated by pressing the button on the fluorimeter, following the text messages from the add-on. Chemosensors 2025,13, 128 8 of 16 The device is connected via a USB port to a known COM port number. The Excel spreadsheet containing the enabled PLX_DAQ macro must use the same port value. The INIT and SETUP processes are executed only once, by writing the information in the Excel control sheet. The procedure is then controlled by pressing the device button briefly and following the instructions in the PLX-DAQ text messages. Once the measurement has been completed, special care should be taken to select the Excel sheet for data export. After exporting, select the control sheet again to be able to modify the parameters for a new measurement. This procedure is similar to the one already explained in the previous document [ 16 , 17 ]. The complete fluorokinetics source program, with multiple commentaries, can be obtained from the authors, upon request. 3. Results and Discussion As a test of the versatility, capability, and accuracy of the designed fluorokinetics device, the rate of TC generation from the oxidation of TM in a basic medium has been studied. 3.1. TC Formation from TM Evolution As mentioned in the Introduction, the reaction kinetics are very well known in the literature and are even used as an example at the higher educational level [ 30 ]. The authors recommend using Hg(II) as the oxidant agent in basic media, despite other possible indications [ 28 , 29 ]. TC has an emission maximum at 450 nm when excited with UV light (365 nm). Figure 4shows the TC clipped spectrum obtained with the fluorimeter after complete oxidation of a 5 ppm TM solution in phosphate buffer (pH 12). Figure 4. (A) Screenshot of the parameters employed to obtain the emission spectrum of TC. (B) Emission spectrum obtained by Hg(II) oxidation of a 5 ppm TM solution at pH 12. Ordinate corresponds to the ADC values. No blank correction is performed. Figure 5shows the progress of the TM oxidation with the evolution of the TC emission spectrum. These plots can be easily obtained from the exported matrix of the radiation intensity data at the selected WLs and times during the course of the TC formation reaction. The 3D view is a striking example of the visualization of fluorokinetics measurements obtained by the designed device. Chemosensors 2025,13, 128 9 of 16 Figure 5. Three-dimensional evolution of TC formation during Thiamine oxidation. Readings of the ADC converter of the microcontroller at different wavelengths. Experimental conditions are analogous to those described in Figure 4. When the kinetics are relatively slow (sampling time above 2.0 s), measurements can be made by switching off the excitation light between points. However, if it is desired to obtain the rate constant at zero time, measurements have to be made at the mixing reactant early stages with short sampling times to obtain a good resolution. Figure 6shows the evolution of the logarithm value of the ADC reading at the wavelength of the emission maximum (450 nm). The acquisition was carried out for about half an hour. The oxidation reaction tends to be a limiting value corresponding to the amount of initial TM. Figure 6. Evolution of the natural logarithm of the ADC at the wavelength of the emission maximum (450 nm), until TM depletion (blue dots). Pseudo-first-order analysis for initial reaction times (red dots and line) and the experimental conditions are the same as Figure 4. Chemosensors 2025,13, 128 16 of 16 25. Holzbecher, J.; Ryan, D.E. The fluorimetric determination of phosphate with thiamine. Anal. Chim. Acta 1973,64, 147–150. [CrossRef] 26. Kusube, K.; Abe, K.; Hiroshima, O.; Ishiguro, Y.; Ishikawa, S.; Hoshida, H. Electrochemical Derivatization of Thiamine in Flow Injection System. Application to Thiamine Analysis. Chem. Pharm. Bull. 1983,31, 3589–3594. [CrossRef] 27. Kusube, K.; Abe, K.; Hiroshima, O.; Ishiguro, Y.; Ishikawa, S.; Hoshida, H. Determination of Vitamin K Analogues by High Performance Liquid Chromatography with Electrochemical Derivatization. Chem. Pharm. Bull. 1984,32, 179–184. [CrossRef] 28. Ryan, M.A.; Miller, R.J.; Ingle, J.D., Jr. Intensified Diode Array Detector for Molecular Fluorescence and Chemiluminescence Measurements. Anal. Chem. 1978,50, 1772–1777. [CrossRef] 29. Ryan, M.A.; Ingle, J.D., Jr. Fluorometric Reaction Rate Method for the Determination of Thiamine. Anal. Chem. 1980,52, 2177–2184. [CrossRef] [PubMed] 30. Bower, N.W. Kinetic Fluorescence Experiment for the Determination of Thiamine. J. Chem. Edu. 1982,59, 975–977. [CrossRef] 31. Stepuro, I.I. Thiamine and vasculopathies. Prostaglandins. Leukot. Essent. Fatty. Acids. 2005,72, 115–127. [CrossRef] 32. Bubeshko, N.N.; Stsiapura, V.I.; Stepuro, I.I. Fluorescent Properties of Thiochrome in Solvents of Different Polarity. J. Appl. Spectrosc. 2011,78, 337–343. [CrossRef] 33. Speeding up Analogread() at the Arduino Zero. Available online: https://forum.arduino.cc/t/speeding-up-analogread-at-thearduino-zero/426099/1 (accessed on 27 January 2025). 34. Arduino Memory Guide. Available online: https://docs.arduino.cc/learn/programming/memory-guide (accessed on 26 January 2025). 35. Miller, J.; Miller, J. Statistics and Chemometrics for Analytical Chemistry, 6th ed.; Prentice Hall: London, UK, 2010; ISBN 978-0-273-73042-2. 36. Anwar, Z.; Sheraz, M.A.; Ahmed, S.; Mustaan, N.; Khurshid, A.; Gul, W.; Khattak, S.-U.-R.; Ahmad, I. Photolysis of Thiochrome in Aqueous Solution: A kinetic study. J. Photochem. Photobiol. B Biol. 2020,203, 111766–111775. [CrossRef] 37. Ahmad, I.; Abbas, S.H.; Anwar, Z.; Sheraz, M.A.; Ahmed, S.; Mobeen, M.F.; Mustaan, N.; Gul, W. Stability-Indicating Photochemical Method for the Assay of Thiamine by Spectrophotometry. J. Spectrosc. 2018,2018, 3178518. [CrossRef] 38. Define and Solve a Problem by Using Solver. Available online: https://support.microsoft.com/en-us/office/define-and-solve-aproblem-by-using-solver-5d1a388f-079d-43ac-a7eb-f63e45925040 (accessed on 26 January 2025). Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.