Spectrophotometric and smartphone-based colorimetric methods utilizing polyvinylpyrrolidone-capped silver nanoparticles for determining doxorubicin in human plasma samples

Abstract

Doxorubicin (DOX) or adriamycin is a common anticancer drug with a narrow therapeutic index. Therefore, sensitive and reliable quantification of DOX is crucial for therapeutic drug monitoring purposes. In this study, both a spectrophotometric and a smartphone-based colorimetric method were fabricated to detect DOX in plasma samples. Both methods utilize polyvinylpyrrolidone (PVP)-capped silver nanoplates, which undergo color with varying DOX concentrations. The colorimetric method offers significant beneficial features of fast detection time, simplicity, and the ability to be easily observed by the naked eye without any need for expensive instruments. The linear dynamic ranges are 0.25–5.0 µg/mL and 0.5–5.0 µg/mL, with the lower limit of quantification (LLOQ) of 0.25 and 0.5 µg/mL for spectrophotometric and smartphone image analysis, respectively. The smartphone-based image analysis was performed using a smartphone application (PhotoMetrix), which relies on univariate calibration using the histograms of the RGB image. Using the smartphone camera, the image histograms were automatically generated and processed. The proposed probe can potentially be utilized to detect DOX in clinical samples with a mean accuracy and precision of 88.7% and 3.2%, respectively. The results demonstrated that these methods can accurately determine DOX concentrations in plasma samples, highlighting the potential of integrating digital imagery and smartphone applications with chemometric tools.

Introduction

Doxorubicin (DOX) is a water-soluble anticancer drug, which is used for treating various types of cancer including ovarian, lung, leukemia, and breast¹. Its mechanism of action involves disrupting DNA based pairs, leading to helix damage and inhibiting DNA synthesis. Additionally, DOX inhibits the enzyme topoisomerase II, resulting in cell death. Furthermore, DOX generates reactive oxygen species (ROS) in a complex form with iron, causing further DNA damage². However, the use of DOX is associated with a variety of concentration (dose)-dependent side effects, including cardiotoxicity, severe gastrointestinal reactions, and bone marrow suppression. As a result, real-time and fast measuring DOX levels in blood is crucial to optimize therapeutic dosages and minimize adverse effects³. Various studies have reported that the therapeutic concentration of DOX in plasma ranges from 1.10 to 27.00 ng/mL⁴. DOX was detected in different human samples, such as hair⁵, serum⁶, saliva⁷, urine⁸, and exhaled breath condensate (EBC)⁹, which have been utilized as biofluids for the monitoring of DOX. In particular, EBC, with its less complex matrix composition compared to other biological samples, emerges as a promising alternative for drug monitoring in early detection and follow-up procedures¹⁰.

Several analytical techniques such as high-performance liquid chromatography (HPLC), electroanalytical methods, Raman spectroscopy, and chemiluminescence have been investigated for detecting DOX, but they often have limitations in terms of high cost, complexity, limited selectivity, sensitivity, and response times^11–19. Thus, the necessity for the fabrication of a simple, cost-effective, and user-friendly detection approach with significant implications for medicine and pharmacology is of great importance³.

The advancement of nanomaterials with enhanced mechanical, chemical, and physical properties, particularly colorimetric nanoprobes based on noble metal nanoparticles, has opened up new opportunities for their use as sensors for drug monitoring^20,21. Smartphones have become increasingly popular tools in analyte concentration determination due to their widespread availability, advanced computational capabilities, high-resolution cameras, and other features, surpassing traditional instruments in terms of affordability, portability, and user-friendliness, with the added benefit of easy transmitting data to other devices²². Compared to UV–Vis spectroscopy, the utilization of a smartphone as a detection tool offers a multitude of advantages of low cost, improved accessibility, the ability to directly analyze solid and liquid materials without extensive sample preparation, and the capacity to conduct detailed analyses within the user application without the necessity of data exportation. In addition to the advances in smartphones’ ability to capture high-definition images, the introduction of smartphone-based applications has further enabled opportunities for converting the images into digital data and fast analysis. Recently, Böck et al. introduced a smartphone-based application for image analysis that could be utilized in the digital interpretation of the color change in various samples. They introduced the PhotoMetrix as a free tool to be used in the analysis of different analytes in diverse samples^23,24.

A variety of materials has been employed for colorimetric sensing in both environmental and biological samples. These materials include metal nanoparticles such as Au NPs²⁵, Ag NPs^26,27, and CuNPs^28,29, carbon dots^3,30, as well as semiconductor quantum dots (semi-QDs)³¹. For example, Saenchoopa et al. employed a wet-chemical reduction procedure to functionalize Ag particles with γ-aminobutyric acid (GABA) and citrate as stabilizer. The resulting GABA-citrate@Ag NPs were utilized for colorimetric quantification of Hg²⁺ in environmental samples, resulting in a linear dynamic range of 5–35 µM and limit of detection (LOD) of 2.37 µM³². Also, Fangfang et al. produced Bi,N-doped CDs in which a discernible alteration in color from yellow to orange-red was observed upon the addition of DOX. The dynamic range of the developed probe was reported as 0.05–30 μM with a LOD of 169 nM³³. Also, 11-mercaptoundecanoic acid (MUA)-CDs was synthesized using the hydrothermal method and utilized for the colorimetric detecting DOX in human spiked samples. In the presence of DOX, the MUA-CDs color was changed from colorless to dark yellow. Regarding that, a ratiometric colorimetric approach was developed to accurately measure the DOX concentration in 2.50–29.80 μM range with a LOD of 0.75 μM in human and live cell samples³⁴.

Nanoparticles, including Ag NPs, have been widely employed across multiple biomedical and environmental applications^35–39. These particles possess numerous advantages, such as diverse shapes, size-tunable optical features, crystalline structure, ease of synthesis⁴⁰, high electron densities, exceptional catalytic performance, and acceptable biocompatibility⁴¹. Ag NPs have been frequently employed in identifying various drugs, including ampicillin, cyclosporine, methylcobalamin, tramadol, naratriptan, rizatriptan, sumatriptan, and zolmitriptan^42,43.

To quickly and effectively monitor drug concentrations, simple methods are often required that allow for determination without extensive analysis. Colorimetric methods are straightforward techniques that can accurately measure the concentration of substances in situ. By integrating these methods with smartphone technology, we can enhance both the efficiency and sensitivity of the measurements. In this research, we have employed a novel smartphone-based colorimetric method to measure DOX concentrations in plasma samples utilizing PVP-capped Ag nanoplates as the main material. This report presents the first colorimetric and smartphone-based probe for the detection of DOX in real samples. In the presence of the DOX molecules, the color of the Ag nanoplates changed from blue to yellow or green-yellow color, caused by the etching effect of DOX, which converted Ag nanoplates (blue) to Ag NPs (yellow). A smartphone-based application, specifically the PhotoMetrix application, was utilized for the capturing images, converting them to digital data, and analyzing the images. Also, the absorbances were recorded using a spectrophotometer to develop a spectrophotometric method and validate the accuracy of the developed smartphone-based approach. The primary objective of this work is to create a dependable tool for future studies that use Ag nanoplates as an etching material. This tool will enable on-site identification of clinical specimens and biomarkers through portable colorimetric or naked-eye detection devices.

Experimental section

Materials

DOX has been obtained a pharmaceutical company (Tehran, Iran). The DOX-free plasma samples were obtained from the Tabriz city branch of Iranian Blood Transfusion Research Center (Tabriz, Iran). Sodium borohydride (NaBH₄, 96%), sodium chloride (NaCl), zinc sulfate (ZnSO₄), acetonitrile (ACN), hydrogen peroxide (H₂O₂, 30 wt%), polyvinyl pyrrolidone (PVP K-30), tri-sodium citrate (Na₃C₆H₅O₇), and silver nitrate (AgNO₃) were obtained from Merck (Germany). Doubly distilled water was purchased from Ghazi Pharmaceutical Co. (Tabriz, Iran). Acetate buffer (7.5 mM, pH 6.0) was prepared from acetic acid (Merck) and sodium acetate (C₂H₃NaO₂, Merck). The pHs were adjusted at different pHs by sodium hydroxide (NaOH) and hydrochloric acid (HCl), which was purchased from Merck (Germany).

Instrumentation

Zetasizer Ver. 7.11 (Malvern Instruments Ltd, MAL1032660, England) was used to evaluate the zeta potential and size distribution of produced silver nanoparticles for the dynamic light scattering (DLS) investigation. The surface morphology of the nanoparticle was investigated using field-emission scanning electron microscopy (FE-SEM) with a working voltage of 3 kV (Hitachi-Su8020, Czech), and energy-dispersive spectroscopy (EDS) was used to examine the chemical components of the nanoparticle. Nanosurf (AG Gräubernstrasse 124,410 Liestal, Switzerland) was used the atomic force microscope (AFM) in a tapping mode to ascertain the topography of the synthesized materials. The absorption spectra were measured at room temperature using a molecular absorption instrument of U-3010 spectrophotometer (Hitachi, Japan). The infrared spectra were obtained using Nicolet FT–IR 6700 spectrometers. The pH meter (model 744, Metrohm Ltd., Switzerland). A handmade rectangular photography box with 8 × 15 × 8 cm (width, length, and height) measurements was created to capture the color changes. White-black paper sheets were taped to the box’s six sides to create a comparable environment and lighting. A hole (1.5 × 2 cm) was created in the front section of the box to adjust the distance between the smartphone camera and the glass cell. To give adequate light to take the picture, it is set up on top of the box. The flashlight was off when shooting the pictures to avoid reflection from the surface of glass cell. All of the images were captured in the same approach after the fixing of the glass cell. A single digital image each sample was recorded after putting them within a 1.5 mL glass cell. An 8 × 8 pixel area of interest was chosen for the PhotoMetrix application. Digital images were captured using Samsung mobile phone (Galaxy A32) with a camera of 48-megapixel and a 6.5-inch touchscreen with a screen resolution of 720 × 1600 pixels.

Synthesis of Ag nanoplates

About 60 mg of PVP and 4 mL of AgNO₃ solution (10 mM) were added to a 10-mL volumetric flask and agitated to prepare a solution. Then, about 8 mL of tri-sodium citrate solution (75 mM) was added, after vigorous stirring, about 0.96 mL of H₂O₂ was added to the mixture. Under continuous stirring conditions, 3.2 mL of NaBH₄ (100 mM) was added to the solution, causing the color of the mixture to deep yellow, indicating the production of Ag NPs. Further changes in the color of the mixture occurred by turning to pale yellow. After 30 min of stirring at room temperature, the color of the mixture eventually became blue, signaling the endpoint of the synthesis and the production of Ag nanoplates. The obtained nanoparticle mixture was stored at 4 °C for further use.

Sample preparation

500 μL of plasma was mixed with 300 μL of ACN and 200 μL of ZnSO₄ (1.2 M) to precipitate proteins⁴⁴. The solutions were vortexed and centrifuged for 15 min and the supernatant liquid was collected and analyzed by the method. The drug-free plasma was obtained from the Iran blood transfusion organization (Tabriz, Iran). The real sample donors have agreed to participate in the study and signed a consent form that was approved by the Tabriz University of Medical Sciences ethics committee with a license code of IR.TBZMED.REC.1400.425.

General procedure

Plasma samples with DOX concentrations ranging from 0.25 to 5 µg/mL were prepared by adding an appropriate volume of DOX standard solution. Then 100 μL of acetate buffer (7.5 mM, pH = 6), 300 μL of Ag nanoplates, and 20 μL of NaCl (600 mM) were added to the mixture. After incubating for 5 min at room temperature, the solutions were ready for visual and absorbance analysis.

Smartphone-based colorimetric detection

Images (64 × 64 pixels), captured in a laboratory studio, were acquired and analyzed using the PhotoMetrix application (version 1.1.23). The partial least squares (PLS) model was selected using RGB histogram data. The graphical user interface of the PhotoMetrix application, which facilitates multivariate analysis through PLS calibration model, is depicted in Fig. S1. The digital images of the analyte were consistently captured under controlled conditions to ensure accurate results and minimize errors caused by different mobile phones and external light sources. The RGB color data was automatically converted to a logarithmic scale on the phone to measure the color intensities, and a calibration curve was created based on the captured images from the analytes. This calibration curve are used to calculate the DOX levels in unknown samples (Scheme 1).

Scheme 1.

Smartphone-based DOX detection.

Results and discussions

Characterization of the nanoparticles

Ag noble metal nanoparticles exhibit unique physicochemical properties and with size tunable optical characteristics. Figure 1a demonstrates the FESEM of Ag nanoplates, showing their structural morphology. Ag nanoplates exhibit an uneven, plate-like shape, which is their distinctive features. The FESEM results indicates that the main parts of the nanoparticles have a relatively uniform size distribution, with an average size of ~ 30 nm. To further confirm the FESEM results, TEM was employed. The TEM results approve the finding from FESEM, confirming the plate-like shape and the average size of Ag nanoplates around 30 nm (Fig. 1b). The TEM images provides a more detailed view of nanoparticles, highlighting their flat structure, and the edges that give to their unique features. EDX analysis results showed that Ag nanoparticles were synthesized correctly and with high accuracy. The high percentage of oxygen indicates that PVP is present as a capping agent in the composition of Ag nanoparticles. Also, the low percentage of Ag indicates that this metal is properly dispersed in the 3D structure of PVP. Boron as an impurity comes from NaBH₄ used during the synthesis of nanoparticles. And phosphorus is present as an element (Fig. S2). DLS results showed that the majority of the nanoparticles have a size value of 40.55 nm with the surface charge of −29.1 mV (Fig. S3a and b). Negative zeta potential indicates the attachment and development of negatively charged groups on the surface of the Ag nanoplates. The formation of a thin layer of PVP on the surface of Ag nanoplates effectively enhances the negative charge on the nanoplates surface. The transfer of n-electrons of ketone groups of PVP molecules (a nucleophilic) to Ag nanoplates (an electrophilic) can increase the surface charge density. As a result, PVP-capped Ag nanoplates with negatively charged surface prevents aggregation, thereby enhancing the colloidal stability of the nanoparticles by electrostatic repulsion forces⁴⁵. AFM was also used to assess the surface topology of the nanoparticles. AFM results revealed that the Ag nanoplates are polydispersed with plate-like shape, having the height variations range from −15 nm to 12.3 nm with minor stacking detected between the nanoparticles (Fig. 1c). These results are fairly supported by the results obtained by FESEM and DLS techniques. FT-IR spectrum of Ag nanoplates showed some peaks at 3415 cm⁻¹, 2919 cm⁻¹, 2851 cm⁻¹, 1619 cm⁻¹, 1149 cm⁻¹, 1090 cm⁻¹, and 778 cm⁻¹ corresponding to polyphenols, aromatic C–H stretching, C–H stretching, N–H bending, C–O stretching (alcohols), C–N stretching of aliphatic amines, and aromatic C-H bending, respectively (Fig. S3c). The FT–IR spectrum confirms the proper attachment and stabilization of Ag nanoplates by PVP molecules.

Fig. 1.

FESEM images of (a) Ag nanoplates and (b) before, (c) after the addition of the DOX, and (d) topographic AFM images of Ag nanoplates.

UV–Vis spectroscopic measurements

The UV–Vis spectra of the synthesized Ag nanoplates were recorded with one main peak at 487 nm (Fig. S4). This peak corresponds to in-plane quadrupole Plasmon resonances, arising from the impact of phase retardation in the collective electron oscillations during the interactions of light-matter in an electrical field⁴⁶.The absorbance originates from the collective oscillation of the free electrons in the metal nanoparticles and are known as the Mie resonance or surface Plasmon under a specific frequency of electromagnetic field. The peak position is highly sensitive to the shape and geometrical dimensions of the particles including triangular nanoplates morphology (edge length and height), surrounding media, and metal anisotropy as well as the metal dielectric constant. Upon the addition of DOX and its etching effect, which changes the morphology of the Ag nanoplates, the color of the Ag nanoplates changes from dark blue to yellow or green–yellow color is occurred. This color change is linearly related to the concentration of DOX.

Mechanism of action

In general, the etching effect of DOX can be achieved by changing the structure and morphology of the Ag nanoplates, which causes the color of the Ag nanoparticles to change from blue to yellow. The mechanism of this phenomenon is as follows. DOX can act as an oxidizing agent in the presence of oxygen and convert the Ag into Ag⁺ ions (Ag → Ag⁺ + e⁻). After the oxidation of Ag, DOX is reduced by the following reaction and the quinone group is converted to hydroquinone⁴⁰.

Oxidation of Ag atoms and production of Ag ions (Ag(s) → Ag⁺ + e⁻) occur mostly at the edges and corners of the nanoplates due to their high reactivity. Repeated oxidation of metallic Ag to ions causes the sharp edges and plate-like corners of the Ag nanoplates to gradually lose their morphology and become spherical or round. This process, also known as the etching effect, is directly related to the concentration of doxorubicin. High concentrations of DOX increase the etching effect, which also changes the color of the Ag nanoparticles from the plate state to spherical Ag nanoparticles. In general, spherical nanoparticles are thermodynamically more stable than the plate-like form. The etching process also causes clear changes in the optical properties of the nanoparticles.

Analytical section

Optimization

Optimizing parameters is a critical part of developing an analytical method for detecting analytes in various matrices. One of the key factors to consider is pH, particularly when the analyte contains acidic or basic groups. Selecting an appropriate buffer with the right concentration and pH is essential to achieve the maximum signal relative to the analyte. In this work, two types of buffers, namely sodium phosphate/phosphoric acid and sodium acetate/acetic acid, were investigated in different concentrations and pHs. First, the effect of various pH values was evaluated using phosphate buffer (10 mM) from 3.0 to 8.0. Results revealed that the maximum absorption was obtained at pH = 6, caused by the enhanced etching effect of DOX in its ionized form (Fig. 2a). Also, the concentration and type of buffer were optimized and sodium acetate/acetic acid buffer at 7.5 mM was selected as the optimum buffer and concentration (Fig. 2b). Furthermore, the type and concentration of salt have a significant effect on the absorbance by regulating the ionic strength of the solution. To study the effect of different types of salt, NaCl and KCl were tested at various concentrations. It is clear that with a 0.6 M of KCl, the highest absorption was achieved (Fig. 2c). To investigate the effect of Ag nanoplates, different volumes (100, 150, 200, 250, 300 and 350 µL) were tested, as shown in Fig. 2d. It was confirmed that the interactions between Ag nanoplates and DOX, specifically the etching effect, reached their highest level at 300 µL of Ag nanoplates.

Fig. 2.

(a) Effect pH of buffer (DOX 1 µg/mL, PBS (10 mM), (b) buffer concentration (DOX 1 µg/mL, PBS, 10 mM, pH = 6), (c) concentration of salt (DOX 5 µg/mL, acetate/acetic acid buffer 7.5 mM, pH = 6), and (d) concentration of Ag nanoplates on the final absorbance (DOX 5 µg/mL, acetate/acetic acid buffer 7.5 mM, pH = 6.0, KCl 0.6M).

Analytical figures of merit of the developed colorimetric method

The ability of the Ag nanoplates-based colorimetric method to measure various DOX concentrations was also assessed under the optimized conditions. Figure 3 presents the calibration curve, along with by smartphone images taken at DOX concentrations ranging from 0.25 to 5.0 µg/mL. A linear correlation was observed between the signal and DOX concentrations in plasma, as measured by the spectrophotometry method (Fig. 3a and b). The regression equation was Y = − 0.08416 [DOX] + 1.15 with a good correlation of 0.9977.

Fig. 3.

(a) and (b) Calibration curve and corresponded absorption spectra for DOX with increasing concentrations (0.25–5 µg/mL), (c) and (d) PhotoMetrix application-based calibration curve. (Ag nanoplates 300 µL, acetate/acetic acid buffer (7.5 mM, pH = 6), KCl solution (0.6 M).

The PhotoMertix application on a smartphone was used to provide digital images (Fig. 3c and d). Two color models, namely RGB (red, green, blue) and HSV (Hue, Saturation, and Value), were utilized to measure the image-based digital colors. The application supports multichannel for image analysis. The calibration equations using RGB model are as the following equations.

It is noteworthy that there is a strong correlation in the G channel of the image data with a correlation coefficient of 0.942 in the 0.5 to 5.0 5.0 µg/mL concentration range. Using the calibration equation of G channel i.e. Y = 1.418 [DOX] + 119.375, the mean relative error of the concentrations was calculated as 9.0%. The analysis of DOX solutions indicated a decrease in the color intensity. HSV model was also utilized for the analysis of images. The resulting equation can be represented as the follows.

A comparison of the regression equation and correlation coefficients of both RGB and HSV analysis showed that the HSV channel model provides more reliable results than the RGB channel. Also, the back-calculated concentration of HSV model had a mean relative error of 7.4%. It is noteworthy that the image histograms were automatically calculated and processed using a smartphone application, eliminating the need for data conversion or transfer. The HSV model minimizes the interfering influence of ambient lighting, enhancing the reliability of the model for detecting DOX concentration.

Table 1 outlines the performance analysis, materials used, and methods employed to measure DOX in biological samples. While the detection capability of the developed probe does not match that of previously reported methods, it is significant that this method represents the first colorimetric method for detecting DOX in plasma samples. Therefore, this technique can be recommended as a versatile probe for detecting DOX in clinical trials.

Table 1.

The reported analytical methods for the detection of DOX.

Composite	Method	Linear range	LOD/LLOQ	Sample	References
Nafion composite	Electrochemical	5–2000 nM	1 nM	Plasma	48
GQD	Electrochemical	0.018–3.60 µM	0.016 µM	Plasma	49
MWNTs	Electrochemical	0.0025–0.25 µM	1.0 nM	Whole blood	50
MWCNTs	Electrochemical	0.04–90 µM	0.0094 µM	Serum, urine	51
CoFe2O4/MNPs	Electrochemical	0.05–1150 nM	10 pM	Serum, urine	52
MWCNTs/Pt	Electrochemical	0.05–4.0 µM	0.002 µg/mL	Plasma	53
Fe3O4-GO-SO3H	Electrochemical	0.043–3.5 µM 0.026–3.5 µM 0.86–13.0 µM	0.0049 µM 0.0043 µM 0.014 µM	Plasma Urine CF	11
Poly arginine	Electrochemical	0.069–1.08	LLOQ: 0.069 nM	Plasma	11
Polytoluidine blue	Electrochemical	340–8.6	LLOQ: 0.34 nM	Plasma	14
Ag nanoplates	Spectrophotometric smartphone	0.25–5.0 µg/mL 0.5–5.0 µg/mL	0.5 µg/mL 0.25 µg/mL	Plasma	This work

Validation of the developed method

Validation is one of the key aspects of an analytical method that confirms the performance of the developed probes in real samples. Repeatability and accuracy are two primary parameters in the method validation. Food and drug administration (FDA) guidelines requires two types of repeatability, namely intraday and interday, to confirm a method for clinical use⁴⁷. Two or three concentrations from the calibration dynamic range are selected, and their concentrations are calculated using the developed method to calculate the repeatability and accuracy. As summerized in Table 2, the developed probe exhibits a relative standard deviation (RSD%) between 1.1 and 6.4% and relative error percentage (RE%) ranges from − 9.4 to 15.0%, confirming accuracy and precision of the probe for DOX determination. According to the FDA guidelines, the maximum allowable RSD% and RE% values for repeatability and accuracy are ± 15%.

Table 2.

Repeatability and accuracy of the fabricated Ag nanoplates-based method for the DOX detection.

Concentration (µg/mL)	Intraday		Interday
	Precision (RSD%)	Accuracy (RE%)	Precision (RSD%)	Accuracy (RE%)
0.25	2.0	–	6.4	–
1.0	2.6	− 9.4	1.1	12.0
3.0	3.8	15.0	3.2	− 3.8

The specificity of the Ag nanoplates nanoprobe was also investigated in the presence of various interfering substances that could potentially affect the absorbance of the probe. The tested substances include the ions (Na⁺, Zn²⁺, K⁺, Co²⁺, Fe²⁺, Al³⁺, Mg²⁺, Ca²⁺), biomolecules (glycine, cysteine, tryptophan), and some co-administrated medications (carvedilol, metoprolol, losartan) which their influences were studied under the optimized analytical condition with the DOX concentration of 1.0 µg/mL (Table 3). In this study, the effect of interfering agents was assessed as the percentage change in the final signal of the probe when these agents were present compared to when they were absent. The results demonstrate that the Ag nanoplates-based colorimetric probe specifically detects DOX in plasma samples.

Table 3.

Selectivity of the proposed method in the presence of various interfering agents.

Agent	Analyte concentration (µg/mL)	Interfering concentration (µg/mL)	Signal in the absence of interfering agent	Signal in the presence of interfering agent	RE (%)
Na+	DOX (1 µg/mL)	0.03	1.0525	1.0064	4.38
Zn2+	DOX (1 µg/mL)	0.03	1.0722	1.431	2.71
K+	DOX (1 µg/mL)	0.03	1.099	1.0862	1.16
Co2+	DOX (1 µg/mL)	0.03	1.1201	1.092	2.50
Fe2+	DOX (1 µg/mL)	0.03	1.1155	1.0901	2.28
Al3+	DOX (1 µg/mL)	0.03	1.1011	1.0792	1.99
Mg2+	DOX (1 µg/mL)	0.03	1.1384	1.1041	3.01
Ca2+	DOX (1 µg/mL)	0.03	1.1129	1.0944	1.66
Carvadiaol	DOX (1 µg/mL)	0.03	0.8809	0.9331	5.42
Matoprolol	DOX (1 µg/mL)	1	1.0386	1.0184	1.94
Losartan	DOX (1 µg/mL)	1	1.075	1.0426	3.01
Glysine	DOX (1 µg/mL)	1	0.8854	0.8527	3.69
Cysteine	DOX (1 µg/mL)	1	1.0717	1.0312	3.78

The stability of the Ag nanoplates was also investigated under two storage conditions at different temperatures (refrigerator and ambient) and times (up to 30 day). The data obtained showed that the produced nanomaterials were stable for long periods at 4–8 °C temperatures, but lower temperatures can cause changes in the properties and morphology of the produced nanomaterials (Fig. S5). It seems that ambient temperature can cause instability of nanoparticles by increasing the leaching process of PVP molecules as stabilizing agent. Also, low temperatures can force the aggregation of nanomaterials and cause its instability. The results of the pH effect on nanoparticle stability indicate that these materials exhibit better stability at neutral pH levels. The results suggest that the stability of silver nanoparticles is dependent on pH levels, with high stability observed at pH 6. At lower pH values, the nanoparticles tend to aggregate or dissolve, which diminishes the SPR effect. Conversely, at higher pH levels, the Ag nanoplates may oxidize and convert to silver oxide (Ag₂O) or change their surface chemistry.

Application

The application of the developed colorimetric probe was evaluated to quantify the amount of DOX in human plasma samples (Table 4). An appropriate concentration of DOX was added to the samples and then stirred to provide a homogeneous mixture. Subsequently, the amounts of DOX were determined using the calibration equation. The results indicated that the mean recoveries were 100.7% and 104.5% for spectrophotometric and PhotoMetrix based detections, respectively, confirming the accuracy of the developed probe for detecting DOX in biological samples.

Table 4.

Application of the developed colorimetric and PhotoMetrix application-based method for the detection of DOX in plasma samples.

Concentration (µg/mL)	Founded (µg/mL)		Recovery (%)
	Spectrophotometric	PhotoMetrix	Spectrophotometric	PhotoMetrix
1.0	1.1	0.93	110.0	93.0
2.0	1.8	2.21	90.0	110.5
5.0	5.1	5.50	102.0	110.0

Conclusions

In summary, a smartphone-based colorimetric probe was fabricated to detect DOX in plasma samples using Ag nanoplates as the main reagent. Under optimized conditions, this probe is able to detect DOX in the concentration range from 0.25 to 5.0 µg/mL in human plasma samples. The developed probe has been validated, and the results confirm its accuracy and selectivity, making it a reliable tool for the interference-free detection of DOX in therapeutic drug monitoring. The probe, when used with the PhotoMetrix application for image analysis, can detect DOX concentrations as low as 0.5 µg/mL. It offers rapid analysis times, low costs, user-friendliness, and portability.

Author contributions

E.B. and S.K.D. carried out all experiments and prepared the first draft, M.P. and A.J. investigated, conceptualized and reviewed the final version, J.S. conceptualized the sensing approach, probe fabrication, and validation section, and reviewed the final version.

Data availability

Data will be available upon request. Elmira Behboudi and Saeedeh Khadivi-Derakhshan are responsible for this information.

Declarations

Competing interests

The authors declare no competing interests.

Ethical approval

This project was approved by the Research Ethics committees of Tabriz University of Medical Sciences ethics committee with the confirmation code IR.TBZMED.REC.1400.425. Also, this study complies with all regulations. Informed consent was obtained from all subjects and/or their legal guardian(s).

Consent for publication

All authors approved the publication of the current study.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-98460-8.