An Environmentally Friendly Compact Microfluidic Hydrodynamic Sequential Injection System Using Curcuma putii Maknoi & Jenjitt. Extract as a Natural Reagent for Colorimetric Determination of Total Iron in Water Samples

The miniaturization of analytical systems and the utilization of nontoxic natural extract from plants play significant roles for green analytical chemistry methodology. In this work, the microfluidic hydrodynamic sequential injection (HSI) with the LED-phototransistor colorimetric detection system has been proposed to create an ecofriendly and low-cost miniaturized analytical system for online determination of iron in water samples using Curcuma putii Maknoi & Jenjitt. extracts as high stability and good selectivity of a natural reagent. The proposed method was designed for online solution mixing and colorimetric detection on a microfluidic platform. The Curcuma putii Maknoi & Jenjitt. extracts and standard/samples were sequentially aspirated to fill the channel before entering the built-in flow cell. The intensity of iron-Curcuma putii Maknoi & Jenjitt. extract complex was monitored under the optimum conditions of flow rate, sample volume, mixing zone length, and aspiration sequences, by altering the gain control of the colorimetric detector to achieve good sensitivity. The results demonstrated a good performance of the green analytical systems. A linear calibration graph in the range of 0.5–6.0 mg L−1 was obtained with a limit of detection at an adequate level of 0.11 mg L−1 for water samples with a sample throughput of 30 h−1. The precise and accurate measurement results were achieved with relative standard deviations in the range of 1.61–1.72%, and percent recoveries were found in the range of 90.6–113.4. The proposed method offers cost-effective, easy operation over an appropriate analysis time (2 min/injection) with good sensitivity and is environmentally friendly with low consumption of solutions and the use of high stability and good selectivity of nontoxic reagents. The achieved method was demonstrated to be a good choice for routine analysis.


Introduction
Iron is one of the most important metals in the environmental geochemical, industrial, and biological processes. It can be chemically combined with many other elements to form iron ores. Te natural deposits and refning of iron ores and industrial wastes are essential parameters for the existence of iron in natural water such as rivers, lakes, and groundwater. Te iron level in groundwater can be increased by the dissolution of iron compounds in soils or ferrous boreholes and handpump components through leaching from home water drinking, cooking, washing, cleaning, and agricultural purposes [1,2]. Generally, iron exists in natural fresh waters in the range of 0.5 to 50 mg L −1 . Te presence of iron above the certain level produces a reddish color of water, metallic taste, a distinct odor, turbidity, and a tendency to stain clothes. In submerged paddy felds, a ferric ion is reduced to the more soluble ferrous ion under the conditions of low pH and anaerobic environment. Te excess ferrous form can lead to cytotoxicity in rice as it is absorbed by the roots and accumulated in plant tissues, leading to tissue destruction and yield loss [3,4]. Terefore, monitoring of the iron levels in water is extremely essential for water quality evaluation and water pollution control to determine the suitability of water for consumption and other applications. Te recommended limit of iron in groundwater and drinking water by the World Health Organization (WHO) is 0.3 mg L −1 . Water for irrigation the iron levels above 0.1 mg L −1 may cause drip emitter clogging irrigation [5,6].
Various analytical techniques have been developed for the determination of iron in water samples. Tese methods include ultraviolet-visible spectrophotometry (UV-Vis) [7,8], atomic absorption spectroscopy (AAS) [9], liquid chromatography [10], potentiometry [11], and voltammetry [12]. Among these techniques, UV-Vis spectrophotometry is the analysis instrument with easy operation and simple colorimetric detection in the visible region with several reagents to form color complexes and can be encountered in almost every laboratory. However, this technique involves high cost and large sizes that consume high amounts of chemicals. Moreover, complete chemical reactions are important for the detection of the products. Tus, microfuidic technology is currently of interest for the development of a detection system with portability advantages, low-cost integrated miniaturized devices, providing a low solution consumption. To obtain automatic operation and sufcient sensitivity of the detection device, fow-based analytical systems such as fow injection (FI), sequential injection (SI), and hydrodynamic sequential injection (HSI) have gained more attention recently for good alternative choices. Among these fow-based techniques, HSI presents an excellent system in terms of cost, chemical consumption, and automatic operation. However, its drawbacks revealed to human health and environmental concerns because most of the reagents are toxic such as Tiron, 1,10 phenanthroline, 2-(5bromo-2-pyridylazo)-5-[N-n-propyl-N-(3-sulfopropyl) amino] aniline, ferrozine, nitroso-R, 4,7-diphenyl-1,10phenanthroline, bathophenanthroline, 2,2′-bipyridine, deferiprone, thiocyanate, and 3-hydroxy-1,2-dimethyl-4(1H)-pyridinone [2,[13][14][15]. To solve this problem, a natural reagent along with a microfuidic HSI system for the determination of iron is considered as a remarkable technique. Te use of natural extract ofers safer and greener colorimetric reagents for humans and the ecosystem. Some plant extracts were utilized as natural reagents in conjunction with digital images by a mobile phone and FI system for iron determination such as green tea [16], guava leaves [17], and Phyllanthus emblica Linn. [18] that showed successful quantitative analysis in real samples. In all of these plant extracts, phenolic compounds were found as common active compounds. Tey are comprised of a hydroxyl group (-OH) bonded directly to an aromatic ring and are capable of binding or chelating with metal ions. Tus, plant extracts containing phenolic compounds can be used as a natural complexing agent for metal ions.
Curcuma putii Maknoi & Jenjitt. is a new species discovered in central Tailand [19]. Te leaves of this species are rich in phenolic compounds that have antioxidant activity. Tese compounds contain a large number of OH groups and a strong ability to complex with metals. Consequently, it is possible that Curcuma putii Maknoi & Jenjitt. extract could be applied as a great complexing agent with high stability, good selectivity, and sensitivity for iron quantifcation by the microfuidic HSI system. Terefore, in this work, the combination of a compact and cost-efective microfuidic HSI and Curcuma putii Maknoi & Jenjitt. extract was proposed for the determination of total iron in water samples based on a green chemical analysis process. Te proposed method exhibited higher stability than the other plant extracts that have previously been reported for iron determination and ofered good sensitivity with a homemade colorimetric detection device. Te special feature of the detector is its ability to adjust its sensitivity using a gain-adjustable signal. Typically, there is no distinction in the narrow concentration range since the generated signals all have the same shade. In this work, amplifcation of the signal with adjustable gain can assist to distinguish a small concentration range. Te satisfactorily selectivity and reproducibility results were achieved with a portable device and an environmentally friendly approach which is suitable for water quality monitoring.

Chemicals and Materials.
All chemicals were of analytical reagent grade, and deionized water was used for the preparation of all solutions throughout the experiment. A stock standard solution of iron (II) at 1000 mg L −1 was prepared by dissolving 0.7022 g of ammonium ferrous sulfate hexahydrate (Sigma-Aldrich, England) in water containing 1.0% (v/v) concentrated sulfuric acid (BDH, England); then, the volume was adjusted to 100 mL. Te stock standard solution of iron (III) at 1000 mg L −1 was prepared in the same way with 0.1 g of ferric chloride (BDH, England). A series of working standard solutions were prepared daily by diluting the stock standard solution to the desired concentrations. Acetate bufer (0.2 M, 500 mL) pH 4.8 was prepared by dissolving 7.21 g of sodium acetate trihydrate (Carlo Erba, Italy) in water containing 2.72 mL of acetic acid (Carlo Erba, Italy). Maknoi & Jenjitt. Extraction. Dry Curcuma putii Maknoi & Jenjitt. leaves of 1.0 g were extracted with 40 mL of 70% (v/v) ethanol using ultrasonic for 1 hour and subsequently fltered through a flter paper (Whatman No. 4). For ongoing usage during the day, the fltrate was maintained at room temperature. Te extract was prepared from fresh Curcuma putii Maknoi & Jenjitt. leaves daily, and the concentration of the total phenolic component was also determined before use. Te use of the same extraction methods with the same batch of Curcuma putii Maknoi & Jenjitt. leaves was one of these criteria. Te total phenolic contents of Curcuma putii Maknoi & Jenjitt. extracts were quantifed using the Folin-Ciocalteau assay [20] and were found to be 2,021 mg L −1 (80.33 ± 2.12 mg/g dry weight), which was employed in the experiment.

Curcuma putii
Te most active compounds of phenolic compounds, including favonoids and tannins, have the ability to chelate metal ions. Tis research believed that favonoids or tannins were the major compounds in Curcuma putii Maknoi & Jenjitt. extract for chelating with metal. Terefore, the ferric test and Shinoda's test procedures were used for tannins and favonoids determination, respectively [21,22]. Moreover, tannins are generally classifed into hydrolysable (gallotannins and ellagitannins), condensed, and complex tannins. Terefore, the analysis of tannins in Curcuma putii Maknoi & Jenjitt. was preliminarily studied using the acid butanol test, nitrous acid test [21], and potassium iodate test [23]. Te acid butanol test is specifc for condensed tannins, forming a red-orange to red-crimson product. Hydrolysable tannins ellagitannins can be determined with a nitrous acid test, which forming a red or pink color and slowly changing color to purple or blue while potassium iodate can also be used for the detection of gallotannins, forming pink color.
To confrm the major active compound in Curcuma putii Maknoi & Jenjitt. extract, liquid chromatography-mass spectrometry (LC-MS) was used for the identifcation of phenolic compounds. Tis method mainly aims to identify the compounds present in the extracts. LC-MS was used under the following instrumental operating conditions: column-Hypersil GOLDTM PFP (100 × 2.1 mm i.d., 1.9 μm), mobile phase: (A) 0.1% formic acid in water and (B) methanol, gradient elution: 0% B in A for 2 min, linear increasing from 0% B in A to 100% B in A for 8 min, and 100% B for 2 min, fow rate 1 mL/min, and detection: UV 254 nm.

Study of the Curcuma putii Maknoi & Jenjitt. Extract-Iron
Complex Formation. In this study, 10 and 100 mg L −1 of both iron (II) and iron (III) complexes with Curcuma putii Maknoi & Jenjitt. extracts were extracted by using various solvents including deionized water, acetate bufer (pH 4.8), and ethanol in order to study the extraction efciency. Te procedure is briefy described as follows: Both 2.5 mL standard solutions of iron (II) and iron (III) of 1,000 mg L −1 were mixed with 10 mL of Curcuma putii Maknoi & Jenjitt. extract in three extraction solvents (water, acetate bufer pH 4.8, and ethanol) and adjusted to the fnal volume of 25 mL with acetate bufer. After 30 min, the mixtures of each extraction solvent were scanned for absorption spectra measurement at the visible range of 500-800 nm using a UV-Vis spectrophotometer (Termo Fisher Scientifc, USA). Te Curcuma putii Maknoi & Jenjitt. extracted without iron was used as a blank solution. Te complex formation of both 10 mg L −1 iron (II) and iron (III) with Curcuma putii Maknoi & Jenjitt. extracts were prepared similarly but with 100 mg L −1 standard solutions of iron (II) and iron (III).

Stability Study of the Curcuma putii Maknoi & Jenjitt.
Extract. Te Curcuma putii Maknoi & Jenjitt. extract was prepared in the suitable extract solvent from the previous experiment and let to stand in the brown bottle at room temperature for 0-7 hours and up to 30 days before mixing with 10 mg L −1 of iron (III). Te mixtures were prepared as previously described. Finally, the absorption spectra were measured to investigate absorption characteristics.

Microfuidic HSI Design and Assembly.
Te microfuidic HSI colorimetric system ( Figure 1) was designed for the determination of total iron using Curcuma putii Maknoi & Jenjitt. extract as a natural reagent. It consisted of a 5.0 mm thickness black polymethyl methacrylate (PMMA) platform (80 mm × 140 mm) that contained a microfuidic channel (0.5 mm width × 0.5 mm deep) and a built-in fow cell with a laser cutting machine (MK2230, Mass Business Co. Ltd., Tailand) for online sampling and reagents mixing and product detection, 1.0 mm thickness clear PMMA (80 mm × 140 mm) as a top cover on the platform, and peristaltic pump (Ismatec, Model ISM796B-230 V, Switzerland) ftted with Tygon pump tubings (1.14 mm i.d.) which were connected to PTFE tubing in each port of the microfuidic channel on the platform. All ports were connected with 2-way solenoid valves (Biochem Valve, 075T2NC12-32, 20 PSI 12 VDC, USA) or a 3-way solenoid valve (NResearch, HP161T031, 12 VDC 100 PSI, USA) which was controlled by Control Solenoid Valve 3.0 software of a homemade controller connected to the personal computer to detect the fow of solutions by turning on/of the valves, and a simple homemade light-emitting diode phototransistor (LED-PT)-based colorimetric detector using a red LED (maximum emission: 660 nm) as a light source which was assembled in a built-in fow cell in the microfuidic platform, and a data acquisition unit (eDAQ Australia).

Detection of Total Iron.
First of all, the microfuidic HSI colorimetric system ( Figure 1) was cleaned thoroughly with deionized water by switching solenoid valves (SV1-SV5) to the fow cell position for 2 minutes and propelling the carrier solution (acetate bufer) for 2 minutes. Ten the Curcuma putii Maknoi & Jenjitt. extract was aspirated into the microfuidic channel on the platform for 6 seconds by a peristaltic pump, 2-way solenoid valve 2 (SV2), and 3-way solenoid valve 5 (SV5, position "out") while 2-way solenoid valves 1, 3, and 4 (SV1, SV3, SV4) were closed. In this step, the solution was flled into the zone A (50 μL, R1), and the excess volume was discarded into the reservoir for waste collection. Te aspiration of the standard solution of iron (III)/sample solution (75 μL zone B, S) and Curcuma putii Maknoi & Jenjitt. extract (50 μL zone C, R2) were similarly operated as previously described, but 2-way solenoid valve 3 (SV3) and 2-way solenoid valve 4 (SV4) were opened for 8 and 6 seconds to fll in the zone B and zone C, respectively. Tus, the sequence of solution zones was R1 S R2, respectively, for efcient mixing between the natural reagent and standard/sample solution. Finally, injection step, all solution zones were pushed by acetate bufer carrier stream through the mixing zone with 2-way solenoid valve 1 (SV1) into the fow cell with 10 mm path length to monitor the color of complex formation and record the signal profles continuously on a personal computer. Each new cycle was continuously operated with the control solenoid valve 3.0 software of a homemade controller. In addition, the solution aspiration was controlled by opening only one 2-way solenoid valve per step and switching 3-way solenoid valve 5 at the position "out" while switching 3-way solenoid valve 5 at the position "in" for colorimetric detection of Curcuma putii Maknoi & Jenjitt. extract-iron complex formation.

Sample Preparation.
Water samples were collected of diferent sources including well water samples and tap water from Sakae Krang Sub-District, Mueang District, Uthai Tani, Tailand. Samples were fltered through a Whatman flter paper No. 6 and kept in polyethylene bottles at 4°C. All bottles were rinsed with 10% (v/v) of nitric acid or sulfuric acid and triple-rinsed with distilled water before further usage to reduce metal ion sorption from the sample storage bottle. Te samples were analyzed within 24 hours after sampling.
To investigate the efcient detection of the proposed method for the determination of total iron, the standard samples were prepared using the known concentration of the iron standard solution to give three standard samples that contained iron (II), iron (III), and iron (II) plus iron (III). Each sample has the fnal concentration of 2.0 mg L −1 as a reference value.  ferric test procedure was tested with 2.0% (w/v) of ferric chloride, resulting in a greenish-brown precipitates appearing in the extract. It is possible that the phenolics detected were tannins. Some research reported that the ferric test could be used to distinguish hydrolysable tannins from condensed tannins. Condensed tannins give greenish-brown precipitates while hydrolysable tannins form bluish-black color and precipitates [21]. Terefore, it is possible that the extracts contain condensed tannins. To confrm that Curcuma putii Maknoi & Jenjitt. extract contains condensed tannins (e.g., catechin and gallocatechin), the extract was tested using acid butanol, nitrous acid, and potassium iodate. Te formation of a crimson product with the acid butanol test indicates the presence of condensed tannins in Curcuma putii Maknoi & Jenjitt. while hydrolysable tannins (ellagitannins and gallotannins) with nitrous acid and potassium iodate tests, gave negative results. Terefore, the preliminary conclusion is that the phenolic compounds in Quercetin is a strong chelating agent with both iron (II) and iron (III). Iron attached to the 4-carbonyl group of C ring and 5-hydroxyl group of A ring, resulting in the break of double bond and deprotonation to generate the two Fe-O bonds. Te next possible site is the 3′,4′-dihydroxyl site, either one or both H atoms can be removed from hydroxyl groups and bonded with iron. Moreover, iron (II) and iron (III) ions are most likely chelated with two quercetin molecules in 1 : 2 and 1 : 1 ratios, respectively. However, the formation of ironquercetin or metal-ligand complex depends on pH, solvent, oxidizing agent, and reactant forms [24,26].

Results and Discussion
Several studies have reported that quercetin has a stronger afnity for iron (III) than for iron (II) [27][28][29]. Furthermore, iron (III) at pH 5.0 could oxidize sufciently to be reduced to iron (II) by quercetin (Figure 4(b)) [29]. Bijlsma et al. reported the reduction of iron (III) to iron (II) through the transfer of electrons from the favonoid (catechol group) to iron and the favonoid is oxidized to a semiquinone radical, and fnally to a quinone [25].

Curcuma putii Maknoi & Jenjitt. Extract-Iron Complex
Formation. Te absorption spectra of iron (III), iron (II), and Curcuma putii Maknoi & Jenjitt. extract in water, acetate bufer, and ethanol were recorded using UV-Vis spectrophotometer (Termo Fisher Scientifc, USA) in the range of 500-800 nm. Figures 4(a)-4(c) represent the absorption spectra of Curcuma putii Maknoi & Jenjitt. extract-iron complexes in water, acetate bufer, and ethanol, respectively. Te results indicated that the use of water and acetate bufer as the extraction solvents (Figures 4(a) and 4(b)) might not be able to extract the active species from the Curcuma putii Maknoi & Jenjitt., which was the main active component for chelating with iron. It could also be possible that the formation of iron and the active site of the active species did not appear, resulting in no signifcant detection of iron complex absorption in the 650-700 nm region, and their absorption spectra were similar to those of iron standard solution prepared in water and acetate bufer. In ethanol, both iron (II) and iron (III) could bind to the active site of the active species of Curcuma putii Maknoi & Jenjitt. a blue-green or brown color appears and exhibits the maximum absorption at 667 nm as shown in Figure 4(c). However, the absorbances values of iron (III) complexes were higher than those of iron (II) complexes. Tis is probably due to the fact that the iron (III)-Curcuma putii Maknoi & Jenjitt. complex has a higher molar absorptivity and stronger afnity. Tus, iron (III) standard solution was chosen for the construction of calibration curves for further studies and determination of total iron by the microfuidic HSI colorimetric system in order to obtain the best sensitivity for water sample monitoring by using the microfuidic system.
Some research studies reported that some chemicals (e.g., favonoids and tannins) could reduce iron (III) to iron (II). Terefore, to test for the reducing efect of the sample, preliminary studies were conducted to investigate some chemicals (e.g., favonoid, tannin, steroid, and terpenoids.) found in the Curcuma putii Maknoi & Jenjitt. extract. From the report based on the phenanthroline method, iron (III) did not form a reddish complex with phenanthroline but only with iron (II) as it appeared to show the absorption spectra at about 510 nm [30] or 500 ± 20 nm [17]. Tus, if the active species in Curcuma putii Maknoi & Jenjitt. extract could reduce iron (III) to iron (II), a reddish complex of iron (II) and phenanthroline would be formed, leading to the absorption spectra at about 500 ± 20 nm. In the experiment, iron (III) was mixed with Curcuma putii Maknoi & Jenjitt. extract and 3.0% (w/v) phenanthroline. Te mixing procedure is similar to that in Section 2.3 but phenanthroline was added before adjusting the fnal volume. In addition, the desired fnal concentrations were 1.      to the extract signal after 20 days of storage was 3.93% Te percentage diference of signal between the initial signal (frst measurement point) and the extract storage signal for 20 days was 9.55%, which is calculated by dividing the diference between the target signal (storage 20 days) and the initial signal by the initial signal and multiplying the result by 100, both did not exceed 10%. For more than 20 days, the results provided 5.23% for the deviation between the initial signal to the fnal signal (stored for 30 days) and 13.84% for the percentage diference between the initial signal and the fnal signal (stored for 30 days), respectively. Te results indicated that Curcuma putii Maknoi & Jenjitt. extract has a high stability of the determination of total iron, while other plant extracts whose stability has been reported at least within 4 and 72 hours (3 days) [16,17].

Optimization.
Te microfuidic HSI colorimetry was used for determination of total iron by varying important parameters of experimental conditions, such as fow rate, sample volume, mixing zone length, and aspiration sequences. Te investigated parameter was varied while the  Journal of Analytical Methods in Chemistry others were fxed. Te iron (III) standard solutions of 0.0-6.0 mg L −1 were employed during optimization. Ten, the optimum conditions were selected to obtain high sensitivity and sufciency for total iron determination.

Efect of Flow Rate.
Flow rate was the one of infuencing parameters on the sensitivity of Curcuma putii Maknoi & Jenjitt. extract-iron complex detection because it afected solution dispersion in a microfuidic platform. Te results shown in Figure 7(a) indicate that a low fow rate led to a longer time for both solution aspiration and the movement into the detector, resulting in a large dispersion which could reduce the sensitivity and the sample throughput. Te outcome of increasing fow rate from 0.1 to 1.0 mL min −1 was a decrease in dispersion thus causing higher sensitivity. However, at fow rate above 1.0 mL min −1 , a lower sensitivity was obtained due to the limited reaction time. So, a fow rate of 1.0 mL min −1 was selected.

Efect of Sample Volume.
Te sample volumes were controlled with the microfuidic channel on the platform which was designed in accordance with calculating length, width, and depth of the S zone position to obtain volume in the range of 25-150 μL, as shown in Figure 7(b). Te sensitivity increased with the increase in sample volume up to 75 μL, then remained constant. Tus, to achieve high sensitivity and sampling rate, a sample volume of 75 μL was selected for further studies.

Efect of Mixing Zone Length.
In order to achieve an efcient mixing of sample and reagent, the carrier was required for the fow cell throughput, leading to maximum sensitivity for product detection. Te mixing zone lengths of 50-200 cm were studied. Te results shown in Figure 7  zone led to extremely large dispersion of the detection zone, resulting in lower sensitivity. Tus, a 75 cm mixing zone length was chosen because it provided the highest sensitivity.

Efect of Aspiration Sequence.
Te aspiration sequences of the solution are studied by using four patterns with diferent carrier solutions as shown in Table 3. Te results revealed that the pattern of reagent (zone R), sample (zone S), and reagent (zone R), using a natural reagent as a carrier provided the highest sensitivity. However, some drawbacks often occurred during the experiment operation such as the color of the natural reagent easily and quickly adhered to the microfuidic channel. Tis caused small colloids that were retained in the fow cell, resulting in an unstable signal, unreliable, and faulty results. To compromise between sensitivity and reduce these drawbacks, the sequence of reagent (R), sample (S), and reagent (R) with acetate bufer as a carrier (No. 2) was selected.  Table 4, indicating that the percentage of recovery was in the range of 90.6-113.4 with satisfactory accuracy and correspondence with the AOAC acceptable range.

Interference Study.
Te efects of some interference species on iron determination were studied, especially cations and anions such as Zn 2+ , Cu 2+ , Mn 2+ , Cd 2+ , Ni 2+ , Cr 3+ , Co 2+ , Pb 2+ , Na + , Ca 2+ , Mg 2+ , Al 3+ , NH + 4 , Cl − , NO − 2 , NO − 3 , and PO 3− 4 . Tese species of known concentration were added to iron (III) standard solution containing a fxed concentration of 1.0 mg L −1 to evaluate the tolerance limit of interference species on iron detection. Te tolerance limit was defned as the maximum of interference species concentration that caused the iron concentration deviation of less than ±5% of the iron concentration without interference. It was found that there was no interference in excess of 500 mg L −1 of Na + , Ca 2+ , and  2+ , and Pb 2+ , indicating that some species (Mn 2+ , Zn 2+ , Ni 2+ , Cd 2+ , and Pb 2+ ) had a positive or negative efect that interfered the determination of total iron. However, these species were absent in the studied samples.
Tus, it could be considered to have no interference efect in this case.    determine iron ion in water samples. Comparing the spectrophotometric method that used 1,10 phenanthroline as a chromogenic reagent and hydroxylamine as a reducing agent, the results as shown in Table 5 indicate that the proposed microfuidic HSI-Curcuma putii Maknoi & Jenjitt. extract system had good correlation with the spectrophotometric method. Considering the linear regression equations (y � 0.9779x + 0.0824) of the correlation graph between both methods, where X and Y represented the results obtained from spectrophotometric and the microfuidic HSI methods, the slope was close to 1 and r 2 value was 0.9918. Moreover, both methods were also compared using the t-test at 95% confdent level [31]. It was found that there was no signifcant diference among all water samples (t critical � 2.36, t calculated � 0.16, DOF � 7). Moreover, when comparing the amount of total iron obtained from standard samples (sample numbers [6][7][8] with the amount of total iron (2.0 mg L −1 ) obtained from iron (II) and iron (III) standard solutions, it was found that there was no signifcant diference. All results indicated that the proposed microfuidic HSI-Curcuma putii Maknoi & Jenjitt. extract system gave a comparable performance to the spectrophotometric method. Moreover, the proposed method could reduce some major drawbacks of the standard spectrometric method. Te frst point, the proposed method successfully determined total iron in water samples with microlitervolumes of nontoxic reagents and samples, while the standard spectrometric method required volume of milliliter of highly toxic reagents. It indicated that the proposed method produced less toxic waste and was more environmentally friendly. Te second avail, the detection device of the proposed method is much cheaper than a standard spectrometric instrument and could be easily fabricated. Besides, the automatic operation of the detection system could shorten the analysis time, which is convenient for routine work. However, the proposed method has a limitation from the air bubble within the solutions that could interfere with the detection signal of the microfuidic HSI system. Tis drawback could be solved by removing the air bubble in the microfuidic HSI with the carrier solution during the operation and degassing the solutions before use.

Conclusions
A compact microfuidic HSI system for homemade colorimetric detection using Curcuma putii Maknoi & Jenjitt. extract was proposed as a green analytical methodology for iron quantifcation. Te extract was nontoxic, low in cost, had high stability (stable for at least 20 days) with good selectivity, and was an easily available reagent that could be applied to the microfuidic HSI system. Te developed method ofered high accuracy and precision as well as sufcient sensitivity with the adjustable gain control detection device and optimum conditions without the need to purify natural reagents prior to use. Te proposed method provided cost-efective, easy operation with an appropriate analysis time (2 min/injection) and environmental protection with low consumption of solutions and chemical usage reduction. Terefore, it could be considered as an alternative analytical method for routine analysis of total iron in water samples.

Data Availability
Te data used to support the fndings of this study are included within the article.

Conflicts of Interest
Te authors declare that they have no conficts of interest.