Activity Analysis of Iron in Water Using a Simple LED Spectrophotometer

Abstract

Access to clean water is a vitally important part of the lives of all people, yet there is often a concern about the contamination of water by heavy metals. The detection and measurement of heavy metals in water can be performed using colorimetric reagents paired with ultraviolet−visible light (UV−vis) spectrophotometry. The presented activity includes construction of a simple, inexpensive single-wavelength spectrophotometer using LEDs as the light source and detector. Using a colorimetric reagent, 2,4,6-tripyridyl-s-triazine, the concentration of iron can be measured by the LED spectrophotometer. Multiple natural water samples were collected and analyzed by this technique, and the performance of the LED spectrophotometer was compared to that of a conventional UV−vis spectrophotometer. The LED spectrophotometer had comparable results to those of the conventional UV−vis spectrophotometer while teaching students about simple circuit design, spectroscopy, and water quality measurements.

Graphical Abstract

INTRODUCTION

Water is a vital part of everyday lives, and the access to clean water can be a concern as the result of changes to municipal water treatment or natural disasters.^1,2 According to the United States Geological Survey, the total withdrawal of water by the U.S. is approximately 1.3 trillion liters of water per day (including irrigation and mining), with the total public- supplied water use at 159 billion liters of water per day.³ With such a large dependency of the U.S. on water, it is important that the water has low levels of toxic chemical contaminants. Under the authority of the Safe Drinking Water Act 40 CFR 141, the U.S. Environmental Protection Agency (USEPA) sets regulatory standards for drinking water; the USEPA requires maximum concentration limits on toxic metal concentrations in drinking water, including arsenic, lead, chromium, and mercury.^4,5

The compound of interest for this activity is iron in natural waters. There is no regulatory limit to the concentration of iron in drinking water,^4,5 but there can be nontoxic effects from water with high concentrations of iron, such as damage to plumbing fixtures (ref 6, pp 170−224). In addition, there is a wide range of iron concentrations in natural waters,⁶ which allows the activity to provide varying results with different water sources without the need for spiked water samples. The purpose of this activity is to teach students about water quality via spectrophotometry and environmental analytical chemistry. Through the creation of a simple and inexpensive spectrophotometer and a fast, observable reaction with an iron indicator, students can measure the concentration of iron in natural waters, directly linking them with a natural, relevant phenomenon.

The construction of the light-emitting diode (LED) spectrophotometer introduces students to the concept of analytical instrumentation, including circuit design and spectroscopy. Spectrophotometers with a single multiwavelength light source require a diffraction grating to isolate single wavelengths.^7,8 Student users of these commercial spectrophotometers may see them as “black boxes” with complicated circuitry, many mechanical parts, and software that returns interpreted data. In contrast, the LED spectrophotometer is a simple design that gives students an intimate look at the function of a spectrophotometer and may provide a better understanding of the interaction of light with a solution. LEDs can be manufactured to emit light over a narrow range of wavelengths, and LEDs are commercially available and relatively inexpensive. A unique feature of LEDs is their function as both a light source and detector at similar wavelengths, as demonstrated in other published LED spectrophotometer activities.^9,10

HAZARDS

The organic chemicals and nitric acid should be handled and disposed according to their respective safety data sheets. Goggles and gloves should be worn at all times while handling solutions. While the current and voltage from the battery pack with 2 AA batteries are low, care should be taken to make sure the wires are placed into the breadboard before the battery pack is turned on.

EXPERIMENTAL SECTION

LED Spectrophotometer Construction

The spectrophotometer consists of a miniature breadboard, two LEDs (bright yellow; 585−590 nm maximum emission wavelength), four wire connectors with alligator clips, a AA battery pack for two AA batteries in series (with on/off switch), one 100 Ω resistor, and a multimeter capable of measuring 0−2 V, which were acquired from Spark Fun Electronics (Niwot, CO). Polystyrene cuvettes were purchased from VWR International, LLC (Radnor, PA). The construction of the LED spectrophotometer is detailed in the Supporting Information. Briefly, the cuvette is glued to the center of the breadboard, and the LEDs are placed on opposite sides of the cuvette and bent to 90° (parallel to the breadboard) so that the top of the LEDs are aimed directly at each other through the cuvette. The wires with alligator clips are placed to create a circuit between the detector LED and the multimeter, and between the battery pack, the resistor, and the source LED. A graphical display and circuit diagram are shown in Figure 1.

Figure 1.

Graphical display (top) and electrical circuit diagram (bottom) for the LED spectrophotometer.

The typical range for the voltage response is between 0 and 200 mV, but this is dependent on the position of the LEDs and the brightness of the source LED (which can vary by manufacturer). With distilled water in the cuvette, the LEDs can be carefully moved until the output voltage is as high as possible, which indicates the optimal position (highest transmittance of light).

The approximate cost for a single spectrophotometer was $9.00, not including batteries or the multimeter.

Calibration Solutions

An iron(II) [Fe²⁺] stock solution (nominally 100 mg/L) was made in the lab using laboratory distilled water, and nitric acid was added for a nominal concentration of 1.5% (mass percent) nitric acid. Solutions of 100 mg/L Fe²⁺ are commercially available. Calibration solutions were prepared by volumetric dilutions in laboratory distilled water. The procedures are described in the Supporting Information; the calibration solutions concentrations were 1, 0.5, 0.04, and 0 mg/L Fe²⁺ in water.

Collecting the Water Samples

For the present study, water was collected from five different locations in Maryland: a river, a pond, water from a household well (the sample was taken prior to household water treatment), and Atlantic Ocean coastal water. The water samples were collected in polypropylene bottles and stored in a refrigerator until analysis. In addition to the natural water samples, a laboratory-spiked water sample was prepared by adding the stock solution to distilled water to create a 0.2 mg/L solution and labeled as an unknown spiked solution.

Measuring Iron in the Solution

The iron indicator reagent (2,4,6-tripyridyl-s-triazine; TPTZ) was purchased from the Hach Company (Loveland, CO) in reagent powder pillows for 25 mL water samples. In addition to the iron indicator, the reagent powder includes a reducing agent that converts all precipitated or suspended iron to Fe²⁺.¹¹ In most aquatic systems, the dominant form of ionic iron is in the Fe²⁺ state, although in highly oxidizing and acidic conditions, Fe³⁺ can be present (ref 6, pp 100−103). The structure and reaction of TPTZ with Fe²⁺ are shown in Figure 2. The maximum applicable Fe²⁺ concentration for the reagent is 1.8 mg/L.¹¹ The absorption spectrum for the iron-TPTZ coordination compound is shown in Figure 3, with the maximum absorption between 580 and 600 nm.

Figure 2.

Observed reaction between Fe²⁺ and the TPTZ iron indicator reagent. The resulting coordination compound is an observable violet color.

Figure 3.

Absorption spectra of 1 mg/L Fe²⁺ solution with TPTZ (orange solid line) and distilled water (blue dotted line).

To prepare samples, one packet of the iron indicator reagent was added to 25 mL of each calibration solution and water sample. The solutions were shaken briefly and allowed to sit for at least 3 min before analysis. Sequentially, the calibration solutions and samples were added to the cuvette (with distilled water rinses between each measurement), and the measured voltage was recorded. Each solution was measured three times to determine the variability of the detector response. In addition, the measured voltage of distilled water (no reagent) was recorded to determine V₀ (see equation below). As described by the Beer−Lambert Law, the light intensity through a solution at a specific wavelength is

$${\text{log}}_{10}\frac{I_0}{I}=\epsilon \times b\times c$$ (1)

where I is the intensity of the light through the test solution, I₀ is the intensity of the light through the solvent without the compound of interest, ε is the absorptivity of the compound of interest at the specified wavelength, b is the path length of the cell (often 1 cm), and c is the compound concentration of the test solution. The value on the left side of the equation is equivalent to the absorbance (A), which is the typical response reported by a conventional UV−vis spectrophotometer. As the voltage produced by the LED detector is proportional to the intensity of the light detected, the measured voltage can be substituted for the intensities, therefore

$${\text{log}}_{10}\frac{V_0}{V}=\epsilon \times b\times c$$ (2)

where V₀ is the measured voltage response of the LED detector from the light through the solvent (distilled water) without the compound of interest, and V is the measured voltage response of the LED detector from light through the solution with the compound of interest.

Conventional UV−Vis Spectrophotometer

To evaluate the performance of the LED spectrophotometer, the calibration and sample solutions described above were analyzed using a Cary 100 UV−vis spectrophotometer (Agilent Technologies, Santa Clara, CA). The absorbance of the solutions was measured at 585 nm with spectral bandwidth of 1.5 nm and collection time of 0.1 s.

RESULTS

The results from the calibration solutions, using both the LED and conventional spectrophotometers, are shown with linear regression statistics (y = mx + b; with the slope m and the intercept b) and calibration curves in Table 1 and Figure 4, respectively. As can be noted in Table 1 by the correlation coefficient for the linear regression, there is a good-fitting linear response (R² = 0.998, based on 4 points) for the LED spectrophotometer, which is comparable to the conventional spectrophotometer. As noted in Kvittingen et al., LED spectrophotometers do have a nonlinear response over a large range of concentrations;¹⁰ for the purposes of this experiment and within the range of the calibrants of this activity, we treated the response as linear for ease of calculation.

Table 1.

Comparison of Linear Regression Results of the Calibration Curve of the Two Spectrophotometers

	Spectrophotometer
Statistic	LEDb	UV-Visc
Slopea	0.36	0.46
Intercept	0.013	0.0061
Correlationd	0.99822	0.99992

^am within y = mx + b regression.

^bUsing log₁₀(V₀/V) for y.

^cUsing measured absorbance for y.

^dThe correlation value represents the Pearson product−moment correlation coefficient (based on 4 points).

Figure 4.

Calibration curve of the mean responses for the individual calibration solutions using the LED spectrophotometer (top) and the conventional UV−vis spectrophotometer (bottom).

Using the standard deviation (σ; n = 3) of the signal of the 0 mg/L solution and the slope of the calibration curve, the limit of detection (LOD) and limit of quantitation (LOQ) can be determined for both the LED and conventional spectrophotometer using the following equations:

$$\text{LOD\hspace{0.17em}}=\frac{3.3\times \sigma }{\text{ slope }}$$ (3)

$$\text{LOQ\hspace{0.17em}}=\frac{10\times \sigma }{\text{ slope }}$$ (4)

The LOD and LOQ for the LED spectrophotometer were 0.022 and 0.065 mg/L, respectively, and the LOD and LOQ for the conventional spectrophotometer were 0.0069 and 0.0221 mg/L, respectively.

Using the linear regression of the measured voltage response (and absorbance), the concentrations of iron in the five water samples were calculated and are reported in Table 2. In the spiked “unknown” water sample (0.2 mg/L), the calculated concentration of iron was 0.22 and 0.20 mg/L for the LED and conventional UV−vis spectrophotometers, respectively. This result suggests that the accuracy of the LED spectrophotometer based on the reagent-based iron indicator is similar to the conventional UV−vis spectrophotometer using the same wavelength. For the water samples, the precision of the LED spectrophotometer ranged from 0.92% to 5.6% relative standard deviation (RSD) for the two detected concentrations, while the conventional UV−vis spectrophotometer precision ranged from 0.25% to 11% RSD for the three detected concentrations. For both spectrophotometers, the high variability was observed at the lowest iron concentrations (the river, pond, and ocean water samples).

Table 2.

Calculated Results for Iron Concentration in the Water Samples

	Mean Fe2+ Concentration Values, mg/L (SD) by Spectrophotometer
Sample Source	LED	UV-Vis
Spiked unknown	0.215 (0.012)	0.204 (0.0005)
River	<LODa	<LODa
Pond	<LOQb	0.0276 (0.00316)
Well	0.808 (0.0074)	0.805 (0.0035)
Ocean	<LODa	<LODa

^aSamples below the limit of detection, LOD.

^bSamples below the limit of quantitation, LOQ.

As can be observed with the measurement of iron in the pond sample (and their respective LOD/LOQ values), the sensitivity of the LED spectrophotometer is poorer than the conventional UV−vis spectrophotometer. The sensitivity of the LED spectrophotometer could be improved by reducing the impact of stray light (using an opaque cover for the spectrophotometer and light slits on the cuvette) and changing the resistance between the battery and the source LED. The experiments above were conducted in laboratories with fluorescent lighting and no windows allowing natural sunlight; therefore, the impact of stray natural light was not determined.

CONCLUSIONS

The activity presented provides a modular tool for teaching electrical circuits, analytical chemistry, and water quality to a wide range of student audiences. For middle school and early high school students, the LED spectrophotometer can be preassembled and the focus of the activity can be on water sampling and the analysis of iron in various water bodies in their local area. As the spectrophotometer is portable, the tool can be used outside (with shielding from natural light) to demonstrate field laboratory measurements. Upon introduction of the assembly of the spectrophotometer, high school students would learn the general structure of an electrical circuit diagram, as well as general analytical chemistry related to calibration and quantitation using Excel spreadsheet calculations.

For undergraduate students, the activity can demonstrate the previously noted educational goals, while also introducing the discussion of spectroscopy, which includes the identification of the λ_max for the greatest sensitivity and optimal design of a spectrophotometer. Design parameters such as the presence of a light shield, orientation of the LEDs, resistance of the resistor, and battery voltage (using a AAA, AA, or 9 V battery) could be safely evaluated (the LEDs do have a maximum input voltage). The use of TPTZ introduces the concept of a coordination compound that results in a visible color change. In addition, concepts such as linear regression and regression statistics can be introduced when examining a calibration curve plot.

Using drinking water sourced from an underground well (as demonstrated in the presented activity), students can evaluate the concentration of iron in drinking water. Iron in drinking water can be an excellent tool to discuss the differences between well-sourced water and municipality-supplied water; while iron in drinking water is not toxic, it has qualitative aspects that can be undesirable, such as color or taste. In addition, a household that has an individual water-treatment system for the removal of iron can be used to measure the iron concentration before and after treatment to evaluate the performance of the treatment system. As noted in the demonstrated activity, waters such as ocean water or surface waters may not have a high enough concentration of iron to detect using the LED spectrophotometer; for a greater number of water samples with detectable water, samples should be taken from groundwater sources or surface water near point sources of iron, such as mining facilities or industrial runoff.

In conclusion, the LED spectrophotometer performs well and is comparable to the conventional UV−vis spectrophotometer for measuring moderate to high levels of iron in water. Given the ease of construction and the inexpensive nature of the materials used to build the LED spectrophotometer, the instrument provides a unique tool to teach students about general spectroscopy and water quality.