1 Introduction

Microfluidic paper-based analytical devices (µPADs) have been used in many applications to analyze biological, environmental, and food samples at the microscale because they are portable, low-cost, efficient, and environmentally friendly [1,2,3,4]. Moreover, µPADs are easy to fabricate with affordable tools, and can be modified and customized the device structure to meet the typical requirements of an analytical application [5, 6]. The device is basically composed of microchannels that enable the handling and control of sample delivery to particular reaction zones for chemical reactions to produce detectable signals of an analyte. Commonly, the µPAD is made of filter paper or chromatography paper designed with hydrophobic barriers to generate microchannels for liquid sample transportation, mixing, and separating, and a detection zone where the reagents are immobilized and the detecting signal is developed [7].

Recently, the development of µPADs has focused on designing devices for simple signal readouts, controlling fluid transportation, and enhancing sensitivity by using functional materials [8,9,10]. A counting-based measurement on paper analytical devices (cPADs) and distance-based µPADs were designed for simple result readouts without the need for any manipulation [10,11,12]. Controlling fluid flow in microfluidic devices could improve detection sensitivity and achieve multi-step protocol for sophisticated assays [9, 13]. Regulating fluidic flow in microfluidic devices can be achieved by varying the dimensions of microfluidic channels, and adding hydrophilic or hydrophobic materials into the channels [5]. Enhancing sensitivity of detection was achieved by immobilizing nanomaterials such as gold nanoparticles, silver nanoparticles, and multiwalled carbon nanotubes [14, 15].

A hybrid microfluidic device is created when a pristine microfluidic device encounters limitations in multiple detections, reagent immobilization, and controlling fluidic transportation for the detection of target analytes [16,17,18,19]. The hybridization in a microfluidic device is typically achieved by integrating it with components such as an electrochemical sensor, micro pump, nitrocellulose membrane, and μPAD [17, 20,21,22]. For example, a hybrid paper-based microfluidic was made by combining a digital microfluidic (DMF) and a μPAD for colorimetric detection of nucleic acid, and glucose [23]. The μPAD’s function was to store reagents for glucose detection while the DMF could not be immobilized due to its hydrophobic surfaces. Recently, capillary flow-driven microfluidics were integrated with a μPAD for the detection of heavy metals [22]. The capillary-driven flow microfluidics were used for sample delivery while the μPAD was used for reagent storage and detection signal development. These works provide great examples of using hybrid analytical devices for novel analytical methods.

While µPAD has been well developed, a relatively new paper-based analytical device, a photo paper-based microfluidic device (PPMD), is an alternative type of device fabricated using photo paper and adhesive film using a lamination method [24]. The fabrication of PPMD can be done simply by cutting and laminating with affordable tools such as a desktop cutter and a desktop laminator. The fluidic flow in the PPMD is driven by capillary force, and the flow distance is linearly proportional to time. The PPMD has been used for several applications including the colorimetric detection of methyl paraoxon, thioredoxin-1, and the electrochemical detection of bisphenol A [21, 24, 25]. The PPMD’s channel is photo paper, which is a robust substrate that can be patterned with functional materials for a wide range of fluidic flow control for a multi-step assay protocol [24]. In addition, it was also possible to pattern electrodes for electrochemical sensors in the PPMD’s channel [25]. Despite a number of advantages, there have been a limited number of studies conducted to explore more characteristics of the PPMD, tailoring device functionality through hybridization, and usage for analysis of a wide range of samples including biomarkers, water, and food.

Table salt is a daily food that is commonly iodized between 15 and 40 ppm as recommended by the World Health Organization (2007). An essential microelement, iodine, is required by the human body and is used by the thyroid gland to synthesize triiodothyronine (T3) and thyroxine (T4). Various methods have been used for iodine content analysis including iodometric titration [26], spectrophotometry [27], and high-performance liquid chromatography (HPLC) [28]. For more sensitive and accurate measurements, inductively coupled plasma mass spectrometry (ICP-MS) has been used [29]. In addition to the use of heavy instruments for the detection of iodine, simple iodine testing devices have been developed for onsite measurements [30, 31]. Because the devices are cost-effective and simple to use without the need for complicated operation, these point-of-need devices are essential for the control of iodine deficiency disorders, especially in areas where household coverage with adequately iodized salt is significant. In Cambodia, for example, reports have resulted in concerning findings pertaining to iodization efforts [32,33,34]. A staggering 60% of collected commercial table salt samples were found to be non-iodized. Additionally, a significant 92% of samples with added iodine were deemed unsatisfactory, suggesting the problem of a prevalence iodine deficiency in the population.

In this work, we expand our exploration of a photo paper-based microfluidic device (PPMD) by conducting studies on fluidic velocity through using various commercially available photo papers, increasing the height of the channels, and integrating with a μPAD. By combining PPMD with a conventional µPAD, we made a hybrid photo paper-based microfluidic device (hPPMD). This hPPMD is designed as an alternative field-testing device and can be used for colorimetric detection of iodine in salt, as an example of application. This iodine sensing device is intended as a tool to assist in combating iodine deficiency by providing a rapid method for determining iodine content of salt for consumption. Freeware for computers and smartphones was used for quantitative analysis of the color signal of the detection. The accuracy of the method was analyzed using the spiking method and confirmed by spectrophotometric results.

2 Materials and method

2.1 Chemicals and apparatus

All chemicals used in this experiment were of analytical grade. Starch soluble (9005-84-9) and potassium iodide (KI, 7681-11-0) were obtained from Merck (Darmstadt, Germany). Potassium iodate (KIO3, 7758-05-6) was purchased from Fisher Scientific (Canada). Citric acid monohydrate (5949-29-1) was obtained from Scharlau (Spain). All the standard and reagent solutions were prepared using absolute pure water (18.2 MΩ.cm, Labaqua). Single-side glossy photo papers (HL (260 gsm), Epson (S042187), Epson (S041140), Formtec (IH-3810), HP (HP C7891A), and Color inkjet (HC4020)) were purchased from local stores. Adhesive plastic film (G2140, Printec, Korea), Whatman filter paper grades 1, 2, 4, 5, and 6 (Whatman) and filter paper 5C (Advantec, Korea) were materials used for device fabrication and optimization. A Silhouette cutting machine was used to cut each layer before the lamination. A Dino-Lite microscope was used to capture contact angle images on the photo paper and plastic film. The UV/Vis spectrophotometer (X4PC, Munroscientific) was used as the reference method to compare the results.

2.2 Fabrication of microfluidic channels

The fabrication process initially started by designing an electronic file using the Silhouette Studio program (Version 3.0). After the channels design of device was done, it was uploaded and USB-connected to a digital cutter (Silhouette Curio) for cutting with its blade. Materials for the fabrication of the microfluidics including photo paper, filter paper, and adhesive film were cut with speed and pressure of 2 cm/s and 6, respectively. After the components of the PPMD were prepared, they were laminated manually. The order of the layers was photo paper, cut-adhesive film, and cut-adhesive film as the last layer. To prepare a hPPMD, a circular shape of cut-filter paper was inserted between the first and second layers of the adhesive films of the PPMD. We used photo paper (HL (260 gsm)) for the fabrication of hPPMD for fluidic flow observation and iodine detection because the color of the photo paper significantly contrasts with the color signal (purple) from the detection.

2.3 Reagent and standard solution preparation

A standard stock solution of iodine (1000 ppm) was prepared by dissolving 1.6850 g of KIO3 in 1000 mL of deionized water. The stock solution was further diluted to prepare the calibration solution. A 3.0% KI solution was prepared by dissolving 3.0 g of KI in 100 mL of deionized (DI) water. A 2.0% starch indicator solution was prepared by dissolving 2.0 g of soluble starch in 100 mL of DI water. The solution was gently heated on a hot plate until it dissolved. A 5.0% citric acid solution was prepared by weighing 5.0 g of citric acid in a 100 mL volumetric flask. The flask was then filled with DI water.

2.4 Colorimetric detection procedure for iodine

In the analysis step, the reagents were first immobilized on filter paper located in the detection zones. DI water was added to the control filter paper instead of reagents. Two microliters of 2.0% starch, 5.0% citric acid, and 3.0% KI were added to each filter. The reagents were dried for a few minutes before being assembled into the hPPMD. Subsequently, 40 µL of the standard or sample was added to the sample zone. The liquid was then split and allowed to passively flowed to each detection zone via capillary action. The color reaction in the presence of iodine was captured using a smartphone camera (17 MP, Sony XZ2). The color intensity developed was then analyzed using ImageJ software and RGB Color Detector (a smartphone application) for quantification of gray and red intensity values, respectively. The value was normalized by subtracting it from 255. Higher concentrations of iodine in the test solution produced a more intense color with a high normalized gray value (ranging from 0 to 255).

2.5 Method validation procedure

For verification, a standard calibration curve was constructed using different iodine standard concentrations (1–100 ppm). The accuracy of the device was studied by calculating the recovery of the iodine-spiked salt samples at 30 ppm with a 20% salt matrix solution. The obtained results were compared with spectrophotometric results as a reference method. The limit of detection (LOD) and limit of quantification (LOQ) were calculated as 3.3 × SD/m and 10 × SD/m, respectively, where SD and m are the standard deviations of the response and slope of a linear calibration curve, respectively [35].

3 Results and discussion

3.1 Fabrication and characterization of the photo paper-based microfluidic device

The width of a photo paper-based channel is defined by the width of a guided channel that is cut by the cutting tool. To achieve a finely guided channel using the digital cutter, the cutting speed should be slow; the cutting blade should be sharp enough; and the cutting mat should be sticky enough to fix the movement of the cutting material. In our cutting setup, we achieved a width of the channel as small as 0.25 mm while the larger channel width was easily made (Fig. S1). The guided channel was created by laminating a cut-adhesive film onto the surface of the photo paper. Then it was laminated by a second layer of adhesive film to form a photo paper-based fluidic channel that allows fluidic to flow.

After the photo paper-based channel was formed, we loaded color solution into the channel to observe fluidic flow characteristics. The color solution flowed along the guiding channel with a certain velocity without much penetration outside the channel, indicating that the channel wall was properly created (Fig. 1a). The flow of the solution was induced by capillary wicking and capillary-driven flow created by the hydrophilic surface of the photo paper (Contact angle: 48.0°) and the hydrophobic surface of the top cover (Contact angle: 119.0°). The fluid flow in the photo paper-based channel was not only along the guided channel but also had wicking down into the matrix of photo paper. To observe the absorption of the fluid on the photo paper surface, the contact angle was measured as a function of time after dropping 5 µL of the water droplet. It took approximately 11 min for the water droplet to be completely absorbed (Fig. S2).

Fig. 1
figure 1

a A photo paper-based microfluidic channel with a color solution added to visualize the channel and its higher magnification. b Average fluid velocity and contact angle of various types of photo paper. c Average fluid flow velocity versus the fluidic channel height

Full size image

The surface energy of photo paper induces a capillary-driven flow of fluid in the PPMD. Different commercially available photo papers can have different surface energy properties. We used six brands of commercial photo papers to observe their surface contact angle. Those photo papers were used for the fabrication of photo paper-based microfluidic channels and for observing their fluidic transport characteristics. Figure 1b shows the measured average velocity and contact angle of different types of photo paper. It is clear that the photo paper-based channel, fabricated with the photo paper having a higher surface contact angle, induced slower fluidic velocity. Conversely, the use of photo paper with a lower surface contact angle resulted in faster fluidic transport. The velocity made by the photo papers ranged from about 0.33–4.56 mm/s. This finding suggests that the selection of photo papers to be used for the fabrication of PPMD is important when a certain flow velocity is required.

To extend our investigation on PPMD, we studied the velocity of the fluid in the device when the channel heights were varied. In this case, we selected the HL photo paper for the fabrication of photo paper-based channels for the flow observation. The height of the fluidic channel could be increased by adding more channel layers to the device. One layer was equivalent to 100 µm. Figure 1c shows the average velocity of the fluid at different heights of channels from 0.1 to 0.6 mm. It should be noted that we could achieve good lamination of the adhesive films only up to 6 layers, due to the difficulty of alignment of each layer. The average fluid velocity varied with channel height. As the height of the channel increased, the fluid tended to flow more rapidly. At a 0.6 mm (6 channel layers) channel height, the velocity increased by approximately four times (1.31 mm/s) compared to that of the single channel-layered device (0.33 mm/s). This can be explained by the relative importance of the channel wall, which is a hydrophilic surface (contact angle: 75.0°). When the channel height increases, the hydrophilic area also increases and contributes to an increase in the capillary-driven flow, resulting in an increase in fluidic velocity.

The average fluid velocity was studied with respect to various widths of the channel. With an increase in the channel width, a higher fluid flow velocity was obtained, which decreased slightly over wider channels (Fig. S3). This phenomenon is because of the relative importance of the channel wall. When the channel is smaller, higher fluidic resistance is created. When the fluidic channel is greater than 3 mm in width, the flow slightly decreases and slightly changes in the larger channel. A similar characteristic was observed and explained in previous studies [24]. The maximum fluidic velocity is obtained when the channel width is 3.0 mm, while the fluidic velocity decreases in wider channels. This decrease in velocity is attributed to the reduced driving pressure in wider channels.

3.2 Fluidic flow in the hybrid photo paper-based microfluidic device

Fluidic flow in both μPAD and PPMD is a type of capillary flow that is continuous from the inlet to the outlet when there is sufficient sample volume. This continuous flow was and was not observed in the hPPMD, depending on the orientation of the flow direction. When the liquid solution flowed from a photo paper-based channel to a filter paper-based channel of the μPAD, the solution was transferred continuously from one channel to another (Fig. 2a). The filter paper-based channel with high wettability, therefore, creates capillary action and a quick intake of the solution into its porous matrix to create flow. On the other hand, when the solution flowed from the filter paper-based channel, the liquid solution could not continue into the photo paper-based channel (Fig. 2b). The liquid solution was not transferred from the filter paper-based channel to the photo paper-based channel because the photo paper-based channel had lower surface wettability compared to the filter paper. Having a surface contact angle of 48.0°, photo paper could absorb water approximately in the amount of 0.12 mg/mm2 when immersed in water for 200 s, while filter paper absorbs it immediately upon immersion. Additionally, the filter paper has a higher surface absorption capacity than the photo paper-based channel, thus, the liquid solution prefers to stay on the surface of the filter paper. The fluidic flow characteristic observed in the hPPMD, which is an integration of a filter and a photo paper-based channel, can be the basis for developing a microfluidic valve in paper-based microfluidic devices.

Fig. 2
figure 2

a Images of the photo-filter paper hybrid channel with a color solution and b images of the filter-photo paper hybrid channel with a color solution indicated that the fluidic flow was unable to pass to the photo paper channel from the filter paper channel. c The fluid flow distance versus time in the 2 mm width of photo paper, filter paper, and hybrid photo paper-based channel

Full size image

We created four microfluidic channels made of filter, filter-photo, photo-filter, and photo paper with a channel dimension of 2 × 70 mm for the evaluation of fluid flow characteristics. The liquid solution was loaded into each microfluidic channel and then we measured the position of the convex meniscus of the solution as a function of time (Fig. 2c). As expected, the lateral flow of liquid solution in a filter paper-based channel had the same behavior as Washburn flow, while the lateral flow of liquid in a photo paper-based channel was linearly proportional to time [24, 36]. In the filter-photo paper-based channel, the velocity of the liquid solution was initially similar to the flow behavior in the conventional paper-based microfluidic channel, but at the junction between the filter and photo paper channels, the flow was unable to continue to the photo paper-based channel. In the case of a photo-filter paper-based channel, the flow was continuous from the loading zone to the end of the channel. The flow in the filter paper-based channel, after the junction between the photo and filter paper-based channels, had a faster speed than in the photo paper-based channel.

The fluid transport length through the photo paper-based channel was measured and compared with the length of the flow through the filter paper-based channel. It was observed that the fluid traveled longer distances through the filter paper than through the photo paper (Fig. S4). In the photo paper-based channel, most liquids undergo capillary-driven flow rather than capillary wicking. Hence, most fluid fills in the opening between the plastic cover and the substrate surface, rather than in the paper matrix. In the filter paper, the liquid is wicked into the porous network of the cellulose fiber, enabling longer distances through the channel.

3.3 Fabrication of a hybrid photo paper-based microfluidic device for iodine detection

Figure 3a shows the structure of the hPPMD which consists of three layers. Photo paper was used as the substrate in the first layer to allow the liquid sample to flow along the channels. The second and third (upper) layers were single-adhesive plastic films. The second layer was cropped to create microfluidic channels. The cover was also cropped with a small hole at the edge of each detection zone to provide an air outlet for a liquid sample to flow. At the detection zones where there was integration of the µPAD, the filter paper was cut into circular shapes and inserted. This µPAD was used for ease of reagent immobilization and complex development. In this hPPMD, three detection zones were created: one for the control zone and two for the detection zones, where color signals would be developed in the presence of iodine. Each zone was connected by a channel to the sample splitting zone located at the center of the hPPMD. A cross-sectional view of the hPPMD is shown in Fig. 3b. The thickness of the paper substrate measured by using a micrometer was about 327 µm. The fluidic channel height was fixed at 100 µm according to the manufacturer’s information. The channel height can be increased by adding more layers of plastic film. Each added layer increases the channel height by 100 µm. At the detection zones, filter paper with a 5.5 mm diameter and immobilized reagents were inserted. The images of the hPPMD, both with and without a loaded color solution are shown in Fig. 3c.

Fig. 3
figure 3

Schematic design of the hPPMD showing: a The patterns of each layer of the hPPMD consisted of a photo paper (substrate), two adhesive films (guided channel and cover) and three circular-shaped filter papers (reagent storage) inserted at detection zones. b Cross-sectional view of the hPPMD. c The fabricated hPPMD images before and after the color solution have been added

Full size image

3.4 Optimization for colorimetric detection of iodine

Obtaining an optimal method for analytical application is necessary for achieving higher detection sensitivity. As stated earlier, a PPMD is a relatively new type of microfluidic device. The scope of utilization of the device platform is still small, unlike the µPADs that have been used for the detection of many analytes. The colorimetric detection of iodine in salt follows a reaction: the reduction reaction of iodate to iodine. The iodine is subsequently formed into a complex with the starch indicator to produce a purple-colored compound as shown in Eqs. (1) & (2).

$${\text{IO}}_{{3}}^{ - } + {\text{5I}}^{ - } + {\text{6H}}^{ + } \to {\text{3I}}_{{2}} + {\text{3H}}_{{2}} {\text{O}}$$
(1)
$${\text{I}}_{{2}} + {\text{Starch}} \to {\text{starch}} - {\text{iodine complex}}$$
(2)

We first conducted experiments to confirm whether using photo paper or filter paper for the detection zone provides higher color signal. Three microfluidic devices were made for comparative study: PPMD, µPAD, and hPPMD. After loading a salt solution containing iodine at 40 ppm into each device, a color signal was developed in the detection zone (Fig. 4a). The lowest color intensity, which was obtained in the detection zone made of photo paper indicates low detection sensitivity. On the other hand, the color intensity was much higher in the detection zones of the µPAD and hPPMD was much higher. This confirmed that using filter paper for detection provided a stronger signal of detection for the iodine assay. The higher sensitivity can be attributed to the better stability of the reagent stored on the filter paper, in comparison to the photo paper. Even though the detection signal of the µPAD was clear enough, the detection signal was not evident only in the detection zone but also in the channel of the µPAD, creating a neck. This phenomenon was not observed in the hPPMD. The flawless circular color signal in hPPMD was achieved because fluidic flow into the filter paper matrix was unfavorable to flow back to the photo paper-based channel. There was no a fibrous network connecting filter paper-based to photo paper-based channels, so back and forth flow back and forth was not likely to happen in the hPPMD. Acceptable homogeneity of the color signal was obtained even without surface treatment at the detection zone.

Fig. 4
figure 4

a Colorimetric detection of iodine using paper analytical devices: PPMD, µPAD, and hPPMD. Optimization of b reaction time, c filter paper type, d KI concentration, e citric acid concentration, and f starch concentration for best achieving the detection signal

Full size image

The effect of the reaction time and various grades of filter paper on the analytical response was evaluated with a sample volume of 40 µL. The sample volume was also pretested in the device to ensure sufficient volumes of liquid flow in all channels. The effect of reaction time is shown in Fig. 4b. The color developed immediately as the sample reached the detection zone and was captured by a camera. The color intensity of the iodine starch complex reached a maximum intensity at approximately 1 min and continued to fluctuate slightly. However, after 5.5 min, the intensity began to drop. This reduction can be explained by the reformation and sublimation of elemental iodine. As the solution is absorbed into the photo paper matrix and exposed to air, elemental iodine reforms from the triiodide complex. Subsequently, this elemental iodine slowly sublimates. This sublimation process leads to the gradual loss of iodine from the detection zone, resulting in a decrease in color intensity over time. We selected a reaction time of 1 min as the optimal time to capture the analytical signal.

The selection of filter papers for use at the detection zone included six filter paper substrates with different pore sizes namely Whatman No. 1, Whatman No. 2, Whatman No. 4, Whatman No. 5, Whatman No. 6, and Advantech 5c. The experimental conditions, except for variations in the type of filter paper, were kept fixed. Among the filter paper substrates tested, Whatman No. 4 filter paper provided the highest mean color intensity (Fig. 4c). This can be explained by the low particle retention (20–25 µm typical porosity) of this filter paper which meant the substrate could easily infiltrate the complex form. The resulting color was more uniform throughout the detection zone. As a result, Whatman filter paper No. 4 was selected for further study.

The effect of the reagent concentration on the mean color intensity in the detection zones was investigated. The effects of various concentrations of KI, citric acid, and starch on the analytical responses are shown in Fig. 4d, e, and f. KI concentrations from 0.1 to 10.0% were studied. The mean color intensity increased as the KI concentration increased to 3.0%. At higher concentrations, a plateau was observed. Therefore, 3.0% KI was selected as the optimum concentration. For citric acid, the response reached its highest intensity at a 5.0% acid concentration and stopped increasing at higher concentrations. Therefore, 3.0% citric acid was selected as the optimal acid concentration for hPPMD. Figure 4f shows a graph of the increased concentrations of starch from 0.1 to 3.0%. The resulting mean color intensity was highest at 2.0% starch. Thus, 3.0% KI, 5.0% citric acid and 2.0% starch were selected as the optimal conditions for calibration curve preparation.

3.5 Colorimetric detection of iodine in salts

After the selection of filter paper (Whatman No. 4), incubation time (1 min), and reagent concentrations (KI: 3.0%, citric acid: 5.0%, and starch: 2.0%) were completed, we performed the colorimetric detection of iodine. It is important to highlight that incorporating filter paper into the detection zone of the hPPMD offers the distinct advantage of producing a flawless circular color signal. This configuration ensures that the color signal is confined solely to the detection zone, preventing its dispersion into other parts of the channel. This not only facilitates straightforward measurement but also enhances the overall accuracy of the measurement process.

The images of the response in the detection zone at various concentrations (0–100 ppm) of iodine and their analysis are presented in Fig. 5. The hPPMDs were captured by smartphone camera for further color intensity analysis using a computer and smartphone (Fig. 5a). After the iodine standard solution was loaded into the hPPMD, the color developed at two detection zones (T1 and T2), and the control zone remained unchanged (Fig. 5b). By using ImageJ software on a computer, we could construct a calibration curve plotting the mean color intensity versus iodine concentrations (Fig. 5c). To obtain a linear plot for the entire range, the iodine concentrations were fitted to logarithmic values. The linearity coefficient of the calibration curve (R2) in the range of 1–100 ppm was 0.9977 with the equation y = 44.758x + 139.92. The LOD and LOQ in the test solution were determined to be 1.62 ppm and 4.32 ppm, respectively, as iodine, which corresponded to 2.23 ppm and 5.95 ppm as iodate (IO3-), respectively. For another method of color intensity analysis, we used a smartphone with the RGB Color Detector application installed. This method is straightforward because the images can be taken and analyzed with the same tool without needing to be transferred to another device. The calibration curve was obtained by plotting the mean color intensity versus iodine concentrations (Fig. 5d). The linearity coefficient of the calibration curve (R2) was good and suitable for the analysis.

Fig. 5
figure 5

a Tools for quantitative analysis of the color intensity from the detection: i computer and ii smartphone. b hPPMD used for colorimetric detection of iodine in salt solution containing various concentrations of iodine from 1 to 100 ppm. c Calibration curve of iodine concentration versus mean intensity analyzed by computer software, ImageJ. d Calibration curve of iodine concentration versus mean intensity analyzed by a smartphone application, RGB Color Detector

Full size image

The results indicated that the hPPMD was sufficiently sensitive and suitable for detecting the iodine present in the salt sample. The results of the real sample analysis are presented in Table 1. The accuracy of the developed hPPMD was also applied to real salt samples using a spiking technique. Non-iodized fine salt and coarse salt samples were used in this study. The 20% salt solution was spiked with an iodine standard at 30 ppm in the final solution and analyzed using hPPMD. No iodine was detected in the non-spiked samples (< LOD). After spiking, the iodine recoveries in fine salt and coarse salt were 100.53 ± 6.03% and 91.44 ± 4.12%, respectively. These percentage recoveries indicate that hPPMD can be used for iodine assays in real salt samples. The salt samples analyzed with this hPPMD were also detected by a UV/Vis spectrophotometric method using the same colorimetric reagent at 583 nm as the optimum wavelength. The result from the hPPMD was very similar to that from the spectrophotometric method.

Table 1 Results of the hPPMD and spectrophotometry analysis of the iodine-spiked salt solution
Full size table

A comparison of iodine detection using our hPPMD with other paper-based devices for iodine detection is presented in Table 2. The testing time in this hPPMD was comparable to the reported devices [37, 38] and notably shorter than a device employing fluorometric detection [39]. With this proposed device, the sample volume needed per test was similar to most paper-based methods, except for the card titration method, which required at least 125µL [40]. In terms of sensitivity, the LOD and LOQ of this hPPMD were lower than those of the ink stamp method [37] and were not much higher than another reported method [38,39,40]. However, the device demonstrates sufficient sensitivity for the detection of iodine in iodized salt, typically in the range of 15–40 ppm.

Table 2 Comparison of iodate detection in hPPMD with other paper-based devices
Full size table

4 Conclusions

The hybrid photo-paper-based microfluidic device (hPPMD) was developed by combining a photo paper-based microfluidic device (PPDM) and a microfluidic paper-based analytical device (µPAD). The PPMD was designed and used for liquid sample delivery, while µPAD was used for storing reagents in the detection zone. We have characterized the fabricated device, including its channel surface properties and fluidic flow. By adding more layers to increase the height of the guided channel, the fluidic flow was increased. We were able to achieve approximately four times the speed when six layers of guided channel were added. The fluidic flow characteristics observed in the hPPMD proposed an idea for the development of a microfluidic valve in paper-based microfluidic devices. Flawless reagent immobilization in the hPPMD allowed us to achieve a flawless color signal developed at the detection zone. Under the optimal conditions such as paper type, reagent concentration, and time to readout of results, the fabricated hPPMD was successfully applied for iodine detection in salt samples at various concentrations ranging from 1 to 100 ppm. The accuracy of the method was confirmed by a fortified method and a spectrophotometric result comparison. The recoveries were in the acceptable range. Under optimal conditions, this hPPMD can detect iodine in salt solutions down to 1.62 ppm, which is sufficiently sensitive for salt iodine determination. The developed hPPMD for iodine detection should be a potential analytical device that opens a route to aid in the prevention and control of iodine deficiency disorders, especially in the developing world. The hPPMD may be a new technological means for such efforts, and have applications in food, water, and biomarker analysis.