Portable 3D-Printed Paper Microfluidic System with a Smartphone Reader for Fast and Reliable Copper Ion Monitoring

Abstract

Copper ions (Cu²⁺) are the third most essential transition metal ions critical to human health. Rapid detection of Cu²⁺ in water and biological fluids is of significant importance. In this study, we develop a sensitive multi-channel paper microfluidic device integrated with a 3D-printed smartphone-based colorimetric reader for the rapid detection of Cu²⁺. A novel rhodamine derivative, 1-(N,N-dichloromethine) amino-4-rhodamine B hydrazine-benzimide (RBCl), exhibiting high selectivity and sensitivity to Cu²⁺, was synthesized and applied as the detection reagent. The interaction mechanism between RBCl and Cu²⁺ was investigated, revealing a structural transition from a colorless spirolactam (closed-ring) to an open-ring amide structure, resulting in a pink color upon Cu²⁺ binding. A multi-channel paper microfluidic device with eight detection zones was fabricated, enabling the simultaneous analysis of eight samples. To enhance portability and quantification, a 3D-printed smartphone colorimetric reader was integrated, providing a rapid and efficient detection platform. The system achieved highly specific Cu²⁺ detection within 2 min, with a detection limit as low as 1.51 ng/mL, meeting water monitoring standards in most countries. Excellent recoveries were demonstrated in real samples, including tap water, river water, blood serum, and urine diluent. This integrated paper microfluidic system is highly sensitive and specific, offering a promising solution for water quality monitoring and health assessment through its rapid sample-to-answer capability.

1. Introduction

Copper ions (Cu²⁺) are the third essential transition metal ions and are well known to control and adjust the fundamental physiological process in organisms, for instance, the deficiency of Cu²⁺ can inactivate some metalloenzymes such as cytochrome oxidase, tyrosinase, and super-oxide dismutase [1]. However, the excessive intake of Cu²⁺ can cause serious physiological damages, such as kidney- and liver-related diseases, neurodegenerative diseases, tachypnea, hypertension, and other diseases caused by the cellular dysfunction [2]. The concentration of Cu²⁺ in human blood normally ranges from 70 to 160 μg/dL for an adult and from 12 to 67 μg/dL for an infant [3,4]. The main sources of Cu²⁺ intake are water and food, thus many nations strictly regulate the maximum content of Cu²⁺. For example, the United States Environmental Protection Agency (USEPA) states that the Cu²⁺ in drinking water should be less than 1.30 mg/L [5]. In China, the concentration of Cu²⁺ needs to be less than 2.0 mg/L in water [6]. The European Food Safety Authority (EFSA) has issued a limit of Cu²⁺ intake per day of between 1.2 mg and 4.2 mg [7]. Obviously, it is important to monitor and measure the concentration of Cu²⁺ in a wide variety of fluids from groundwater, drinking water, and food to human body fluids. At present, some methods such as atomic absorption spectrometry (AAS) [8], inductively coupled plasma mass spectrometry (ICP-MS) [9], inductively coupled plasma atomic emission (ICP-AES) [10], voltammetry [11], and fluorescence spectrometry [12] have been commonly used to detect Cu²⁺. However, these methods usually need expensive equipment and professionally trained operators, both of which are far away from the rapid point of detection needed by end users.

Paper microfluidic devices are alternatives to the analytic system with many advantages, including convenience, easy operation, low cost, low sample volume, low pollution, and naked-eye readability [13,14,15,16]. To date, several methods have been presented to fabricate paper-based devices, including wax printing, inkjet printing, and screen printing [17,18,19]. Herein, wax printing is applied in this paper because of its easy and time-efficient manufacture.

Although paper microfluidics have many advantages, it is difficult to accurately quantify the results by naked-eye observation alone. Meanwhile, as widely used devices, smartphones have been applied for chemical and biological sensing [20]. Previously, smartphone-based microplate readers have been demonstrated to read the colorimetric signal of different assay results [21,22,23]. In this work, a smartphone-based colorimetric reader for paper microfluidic is developed to quantitatively detect analytes.

The recognition element of paper microfluidics is critically important for sensitivity and specificity. Rhodamine derivatives as Cu²⁺ and other metal ion probes have been reported [24,25,26,27]. These references showed that rhodamine derivatives transform from a spirolactam (closed) structure, which is colorless and non-fluorescent, to an open-ring amide, which has a pink color appearance with a strong fluorescence signal, after being induced by suitable metal ions. At present, most research focuses on the usage of rhodamine derivatives to develop liquid assays [28,29,30,31]. However, the rhodamine derivative also holds great potential as an attractive recognition element candidate for paper microfluidics development.

In this work, we develop a rapid detection system for Cu²⁺ by integrating a 3D-printed smartphone colorimetric reader with a multi-channel paper microfluidic device. A rhodamine derivative, 1-(N,N-dichloromethine) amino-4-rhodamine B hydrazine-benzimide (RBCl), highly selective and sensitive to Cu²⁺, is synthesized and applied in the paper microfluidic device for Cu²⁺ detection. The integrated system offers the advantages of portability, high sensitivity and specificity, low cost, and rapid sample-to-answer capabilities.

2. Experimental Section

2.1. Chemicals

Rhodamine B (99%) was purchased from Sigma-Aldrich (St. Louis, MO, USA). N,N-dihydroxyethyl -benzene (99%) was obtained from the China Medicine Group of Shanghai Chemicals (Shanghai, China). Different metal ions (Na⁺, Hg²⁺, Ni²⁺, Fe²⁺, Mg²⁺, Pb²⁺, Co²⁺, Cd²⁺, K⁺, Zn²⁺, Ca²⁺, Al³⁺, Fe³⁺, Cu²⁺) and nitrates were obtained from a commercial corporation. All chemicals were analytical grade and directly used without further processing. The silica gel was purchased from Qingdao Haiyang Chemical Co. Ltd. (Qingdao, China) with a particle size of 10 μm and a 120 Å mean pore size for column chromatography. A stock solution of RBCl (1 mg/mL) was prepared with ethanol (99.8%) and diluted to an appropriate concentration (such as 10⁻⁵ mol/L) before use. Whatman chromatography filter paper (Grade 1) was purchased from Sigma-Aldrich. Human serum and urine diluent samples were purchased from Sigma-Aldrich. Urine diluent is a non-biological solution that mimics human urine. Human serum is from human male AB plasma, USA origin, sterile-filtered. And its chemical safety instruction can be get form the official website of Sigma-Aldrich.

2.2. Instruments

Nuclear magnetic resonance spectra were measured on an INOVA400 spectrometer (Varian, Palo Alto, CA, USA) using tetramethylsilane (TMS) internal standard as a reference. Infrared spectra were recorded on a Thermo Fisher IS10 spectrometer (Thermo Scientific, Waltham, MA, USA), and the samples were blended with potassium bromide (KBr) to form a sheet before measuring. Elemental analyses were performed on a Perkin2400 (II) auto-analyzer (PerkinElmer, Waltham, MA, USA). The UV–visible–near-IR absorption spectra of dilute solutions were obtained on a SHIMADZU UV-2550 spectrophotometer (Shimadzu Corporation, Kyoto, Japan) using a quartz cuvette having a 1 cm path length. The wax patterns were printed with a Xerox Phaser 8560DN (Fuji Xerox, Tokyo, Japan) color printer. The color intensities were recorded with an ESEQuant lateral flow reader and a self-built smartphone colorimetric reader.

2.3. Synthesis of 1-(N,N-dichlormethine) Amino-4-rhodamine B Hydrazine-Benzamide (RBCl)

RBCl was prepared by a modified method according to the literature [32,33]. Firstly, rhodamine B hydrazine, a pink solid, was obtained with a yield of 85%. ¹H NMR (400 MHz, CDCl₃), 1.14 (t,12 H, CH₃, J = 7.8 Hz), 3.39 (q, 8 H, CH₂, J = 7.2 Hz), 3.60 (s, 2H, NH₂), 6.37 (dd, 2H, Xanthene-H, J₁ = 9.0 Hz, J₂ = 2.4 Hz), 6.42 (s, 2H, Xanthene-H), 6.50 (d, 2H, Xanthene-H, J = 9.0 Hz), 7.14 (dd, 1H, Ar-H, J₁ = 5.4 Hz, J₂ = 3.3 Hz), 7.49 (m, 2H, Ar-H), 7.94 (dd, 1H, Ar-H, J₁ = 5.6 Hz, J₂ = 3.4 Hz). Secondly, N,N-dichloroethyl-benzaldehyde, a pale yellow powder, was obtained with a yield of 92.5%. ¹H NMR (400 MHz, CDCl₃) δ (ppm): 9.87 (s, 1H, aldehyde), 7.63 (d, 2H, Ar-H, J = 2.31 Hz), 6.78 (d, 2H, Ar-H, J = 8.54 Hz), 3.68 (m, 2H, CH₂), 3.52 (m, 2H, CH₂). Thirdly, a mixture of rhodamine B hydrazine (0.228 g, 0.5 mmol) and N,N-dichloroethyl-benzaldehyde (0.124 g, 0.51 mmol) was dissolved in 30 mL of anhydrous ethanol and heated to stir at the reflux temperature. In order to ensure a sufficient reaction, a small amount of anhydrous magnesium sulfate and p-toluenesulfonic acid was added. After 2 h, the solution was cooled to room temperature, the yellow deposition was filtrated and rinsed with water and diethyl ether, then dried. The pale-yellow solid was obtained with a yield of 56%. MP, 224 °C. ¹H NMR (400 MHz, CDCl₃), δ = 1.26 (t, 12H, CH₃, J = 7.2 Hz), 3.33 (q, 8 H, CH₂, J = 6.81 Hz), 3.83 (m, 4H, CH₂), 4.61 (m, 4H, CH₂), 6.49 (s, 2H, Xanthene-H), 6.59 (d, 2 H, Xanthene-H, J = 9.0 Hz,), 7.48 (d, 2H, Ar-H, J = 4.00 Hz), 7.74 (d, 3H, Ar-H, J = 4.00 Hz), 8.01 (d, 1H, Ar-H, J = 2.00 Hz), 8.09 (d, 3H, Ar-H, J = 4.00 Hz,), 8.26 (s, 1H, Ar-H), 8.69 (s, 1H, imine-H). An element analysis calculated (%) for C₃₉H₄₃N₅O₂Cl₂: C 68.42, H 6.29, and N 10.23 found C 68.51, H 6.32, and N 10.18.

2.4. Fabrication of Paper Microfluidic

The dendritic multi-channel paper microfluidic device was fabricated using wax printing to create eight hydrophilic channels, each terminating in a detection zone, enabling the simultaneous measurement of eight samples. Each channel was designed with a main channel measuring 2 cm in length and 0.5 cm in width, ending in a 0.7 cm diameter reaction area. The patterns were created using the ChemDraw software (ChemDraw 22.0) and printed onto Whatman filter paper using wax printing. The printed wax formed hydrophobic walls to guide the flow of the samples through the channels. To evaluate the performance of each reaction zone, two types of paper microfluidic designs were tested in parallel. The first type, the single-dot paper microfluidic, consisted of circular Whatman paper with a 4 mm diameter, created using a hole puncher. The second type was the single-channel paper microfluidic, featuring a 2 cm long and 0.5 cm wide channel with a 0.7 cm diameter reaction area at the terminus.

The smartphone colorimetric reader was fabricated using 3D printing technology, following the approach outlined in our previous work [34,35,36,37]. The reader comprises three components: a smartphone adaptor, a paper microfluidic holder, and a convex lens holder, each of which was individually fabricated via 3D printing and subsequently assembled into a single unit. For image analysis, a MATLAB (MATLAB R2022a) (MathWorks, MA, USA) program was developed to automatically calculate the mean light intensity of each strip in the captured images. The program outputs the mean intensity values for the red, green, and blue (RGB) channels of each strip.

2.5. Detection of Cu²⁺ Ions

In a typical assay of multi-channel paper microfluidic, 100 μL of the RBCl solution (1 mg/mL) was added to the center zone. Then, 5 μL of samples was uniformly dispersed on each reaction zone of multi-channel paper microfluidic, respectively. After reacting for 2 min, the concerned reaction zone was cut off and the color intensity measurements were performed by the ESEQuant reader or the 3D-printed smartphone colorimetric reader.

In a typical assay of single-dot and single-channel paper microfluidics, 5 μL of the RBCl solution (1 mg/mL) and 5 μL of samples was uniformly dispersed on the reaction zone. After reacting for 2 min, the color intensity measurements of the reaction zone were performed by the ESEQuant reader or the 3D-printed smartphone colorimetric reader.

In a typical assay of a solution reaction, 50 μL of the RBCl solution (1 mg/mL) or 50 μL of samples was added to a tube. After reacting for 2 min, the UV absorption spectrum was measured by a multifunctional microplate reader.

2.6. Detection of Cu²⁺ Ions in Real Samples

The samples of tap water and river water were collected from the campus of the Washington State University and the Snake River near Wawawai Park in Pullman, WA, USA. The river water was simply filtered using filter membranes with a pore size of 2 mm. The concentrations of Cu²⁺ in tap water and river water used in the experiment were very low (<10⁻⁵ µg mL⁻¹), so the influence could be ignored. Human serum and urine diluent samples were purchased from Sigma-Aldrich without Cu²⁺. The pH of all the samples was adjusted to 4.5–5 with the MES (2-morpholino-ethanesulfonic acid) buffer solution. The operation procedure is consistent with the above experiments. All samples were analyzed with the RBCl by the UV absorption measurement in solution and color intensity measurements were conducted on the multi-channel paper microfluidic with the ESEQuant reader (Qiagen, Switzerland) and 3D-printed smartphone colorimetric reader.

3. Results and Discussion

3.1. Interaction Mechanism of RBCI-Cu²⁺

The synthetic route of RBCl and the interaction between RBCl and Cu²⁺ is summarized in Figure 1A. In order to investigate the complexation ratio between the ligands and Cu²⁺, a Job’s plot experiment was obtained [38]. According to the evolution curve of absorption by varying the concentration of RBCl and Cu²⁺ in ethanol solution, the total concentration was kept at a fixed value (0.1 μM and 0.05 μM). As shown in Figure 1B, the highest points of two Job’s plots showed that Cu accounts for 50% of the complex of RBCl-Cu²⁺, a fact which proved that a typical RBCl-Cu²⁺ complex with a molar ratio of 1:1 was obtained. In addition, the stability constant (K) between RBCl and Cu²⁺ was determined by two Job’s plots and the equation is shown as follows [39]:

$$K={\displaystyle \frac{x1}{(a1-x1)(b1-nx1{)}^n}}={\displaystyle \frac{x2}{(a2-x2)(b2-nx2{)}^n}}$$

(1)

where K is the stability constant, x1 and x2 represent the concentrations of the complex, and they are the same at the equilibrium state, and a1, a2 and b1, b2 are the initial concentrations of copper ions and the ligands, and their total concentrations were kept at 0.1 μM and 0.05 μM, respectively. n is the ligancy between Cu²⁺ and RBCl (n = 1). With n = 1, and with known a1, a2, b1, and b2, X can be calculated. In our work, a1, a2, b1, and b2 are 0.01 μM, 0.005 μM, 0.09 μM, and 0.0045 μM, respectively. Furthermore, K could be obtained with a value of 7.2 × 10⁸, which is higher than the reported values in the literature and indicates that the complex of RBCl-Cu²⁺ has better stability [40].

To study the structure of the RBCl-Cu²⁺ complex, the ¹H NMR spectra of RBCl and the RBCl-Cu²⁺ complex were measured and compared. The results are presented in Figure 1C, which shows that the peak a in RBCl almost disappeared and a new peak h’ (chemical shift δ ≈ 4.75) appeared in the RBCl-Cu²⁺ complex. The reason could be attributed to the rearrangement of amine and the formation of ethylamine after adding Cu²⁺. In addition, after the open-ring amide formation, a new peak g’ on xanthylium appeared at δ ≈ 6.35. From the above analysis, the possible interaction mechanism between RBCl and Cu²⁺ was the lactone ring-opening. Specifically, RBCl shifted from a spirolactam, which is colorless, to an open-ring amide, which has a pink color, after being induced by Cu²⁺. This change was also found in another report [41].

3.2. Principle of Integration Design

The design of the paper microfluidic device was based on the color change of RBCl before and after the interaction with Cu²⁺. Taking a single-channel paper microfluidic device as an example (Scheme 1A), the reaction zone turned from a white to a pink color in an obvious manner after adding the Cu²⁺ solution (1 mg/mL). In addition to naked-eye observation, the paper microfluidic device can be quantified by inserting it into a black strip cover (Scheme 1B) and assembled with a smartphone colorimetric reader. The design details of the smartphone colorimetric reader are described in Scheme 1C–E and mainly included three parts. The first part is a smartphone adaptor used to fit in a smartphone and connect the paper microfluidic holder. For different smartphones, different smartphone adaptors need to be replaced while the remaining components are the same. The second part is a paper microfluidic holder that allows a test strip to slide in. A smartphone camera is usually designed for taking pictures at a minimum distance due to the design for consumer applications. A compact smartphone reader that needs to read a colorimetric signal at a very close distance from the smartphone camera is not able to obtain a clear image by using the smartphone camera alone. An objective lens that can reduce the object distance and magnify the image is needed. In this device, a single convex lens with a 9 mm focal length was placed in the convex lens holder as the third part. After the integration of the black strip cover with the paper microfluidic and 3D-printed smartphone colorimetric reader (Scheme 1E), the well-designed system not only could realize naked-eye detection of Cu²⁺ but also provide a portable quantitative solution, thereby resulting in a remarkable application at the point-of-care.

3.3. Effect of pH

pH is the most important factor which affects the reaction between RBCl and Cu²⁺. In order to optimize it, the reaction zones were preliminarily treated by using 5 μL of Cu²⁺ solution (1 mg/mL) at different pH levels (2.5–11) by modulating it with 1 M HCl and 1 M NaOH. As shown in Figure 2, it was difficult to observe a color change when the pH was lower than 4.6 and greater than 9.1. This is due to the fact that the spirolactam ring structure in RBCI is affected by the pH value [42]. Moreover, the black precipitation was found after the pH was more than 6.7, which may be attributed to the formation of CuO [43]. Therefore, the pH was optimized in the range of 4.6–6 in the following measurements.

3.4. Sensitivity

The sensitivity of RBCl to Cu²⁺ was studied by the paper microfluidic and the solution system simultaneously. As shown in Figure 3A, the color change of RBCl occurred with the increasing concentrations of Cu²⁺, both on the paper microfluidic (upper) and in the solution system (below). After plotting the quantitative results, the absorption obviously increased with the addition of Cu²⁺ from 1 ng/mL to 800 μg/mL, and the absorption reached a plateau at a concentration of 200 μg/mL in solution, corresponding to the value of the logarithm (lg) [Cu²⁺] ≈ 5.3 in Figure 3B. Furthermore, the absorption was linear with Cu²⁺ concentrations from 1 ng/mL to 40 μg/mL at the lg scale (0~4.6). As shown in Figure 3C, the fitted core correlation is:

y_L = 0.507 + 0.140 lg [C] (R2 = 0.9729)

(2)

where y_L is the color intensity and C is the concentration of Cu²⁺.

For the sensitivity of the paper microfluidics, the linear correlation of the color intensity and the lg concentration of Cu²⁺ by the ESEQuant reader and the smartphone reader are shown in Equation (2) and Equation (3), respectively:

y_E = 7.72 × 10⁵ − 8289 lg [C] (R2 = 0.9675)

(3)

y_S = 5.88 × 10⁷ − 3.59 × 10⁶ lg [C] (R2 = 0.9638)

(4)

where y_E is the intensity obtained by using the ESEQuant reader, y_S is the intensity obtained by using the smartphone reader, and C is the concentration of Cu²⁺. The low limit of detection (LOD) of these three tests was calculated by using 3δ/K (K represents the slope and δ is the standard deviation). The LOD was 0.12 ng/mL for the liquid test, 3.46 ng/mL for the ESEQuant reader, and 1.51 ng/mL for the smartphone reader, respectively. Therefore, the superior sensitivity confirms the practical potential of this integrated system, which meets the water requirements in most countries.

3.5. Specificity and Anti-Interference Performance

In order to investigate the selectivity of RBCl to metals ions, different metal ions (Na⁺, Hg²⁺, Ni²⁺, Fe²⁺, Mg²⁺, Pb²⁺, Co²⁺, Cd²⁺, K⁺, Zn²⁺, Ca²⁺, Al³⁺, Fe³⁺, Cu²⁺) were added to the diluted solution of RBCl with a concentration of 10⁻⁵ mol/L in distilled deionized water. The concentration of the metal ions was 10 times that of RBCl. As shown in Figure 4A, an obvious pink color appeared after adding Cu²⁺, both in solution and on the single-dot paper microfluidic, while no obvious color change could be observed after adding other metal ions. The absorbance of the RBCI solution in the presence of Cu²⁺ at 550 nm was far over that in other metal ions (~5 times). Moreover, a multi-channel paper microfluidic with eight detection zones was applied to evaluate the interference from other metals. A mixture of different metal ions (i.e., Na⁺, Hg²⁺, Ni²⁺, Fe²⁺, Mg²⁺, Pb²⁺, Co²⁺, Cd²⁺, K⁺, Zn²⁺, Ca²⁺, Al³⁺, Fe³⁺, Cu²⁺) was prepared at a concentration of 50 μg/mL, respectively. After adding the Cu²⁺ solution and the mixture to the sample region, a similar pink color and a similar color intensity were observed in all reaction zones, as shown in Figure 4B. Therefore, these results prove that the proposed paper microfluidic is highly selective to Cu²⁺ and is strongly anti-interference from other metal ions.

3.6. Detection of Real Samples

In order to investigate the practical applicability of the integration device of a 3D-printed smartphone colorimetric reader with multi-channel paper microfluidics for Cu²⁺ detection, recovery experiments were conducted by adding appropriate amounts of Cu²⁺ to real samples including tap water, river water, human serum, and urine diluent samples. The results are summarized in Figure 5A for adding different concentrations of Cu²⁺, B for recovery, and C for the relative standard deviation (R.S.D.). The determination data were obtained with the average of the five separate determinations. In Figure 5A, one can see that the concentrations of Cu²⁺ was determined by the absorption method in the RBCl solution, ESE, and the smartphone method. In the three methods, the recovery rates in Figure 5B (from 99% to 101% by the absorption method; from 95% to 106% by the ESE method; from 96% to 100% by the smartphone reader) and the R.S.D. values (from 0.66 to 3.38 by the absorption method; from 0.67 to 3.5 by the ESE method; from 0.45 to 3.9 by the smartphone reader) in Figure 5C showed that the absorption method is slightly better than the other two methods. The accuracy and ereproducibility of the smartphone reader are superior to that of the ESEQuant reader. However, the solution absorption method and the ESEQuant method require a skilled technician to be operated, while the smartphone reader can be easily used by non-professional users for in-field tests.

4. Conclusions

In summary, we developed a portable system for Cu²⁺ detection by integrating a 3D-printed smartphone colorimetric reader with a multi-channel paper microfluidic device. A rhodamine derivative (RBCl) with high specificity for Cu²⁺ recognition was successfully synthesized. The interaction mechanism between RBCl and Cu²⁺ was investigated using a Job’s plot and ¹H NMR spectroscopy. The results revealed that an RBCl-Cu²⁺ complex with a 1:1 molar ratio was formed, with RBCl transitioning from a colorless spirolactam (closed) structure to a pink open-ring amide structure, exhibiting a high stability constant of 7.2 × 10⁸. Using RBCl as the detection reagent, a multi-channel paper microfluidic device was fabricated, enabling a highly sensitive and specific multi-sample detection of Cu²⁺. Additionally, a 3D-printed smartphone colorimetric reader was developed to interface seamlessly with the paper microfluidic device, offering a portable and user-friendly solution. The system requires only 2 min of reaction time per detection, demonstrating a rapid sample-to-answer capability. The experimental results show that the integrated system achieves a low LOD of 1.51 ng/mL for Cu²⁺ and exhibits excellent specificity in tap water, river water, blood serum, and urine diluent, highlighting its practical reliability for rapid Cu²⁺ detection by end users.