An environmental friendly procedure for photometric determination of hypochlorite in tap water employing a miniaturized multicommuted flow analysis setup.

A photometric procedure for the determination of ClO(-) in tap water employing a miniaturized multicommuted flow analysis setup and an LED-based photometer is described. The analytical procedure was implemented using leucocrystal violet (LCV; 4,4',4''-methylidynetris (N,N-dimethylaniline), C(25)H(31)N(3)) as a chromogenic reagent. Solenoid micropumps employed for solutions propelling were assembled together with the photometer in order to compose a compact unit of small dimensions. After control variables optimization, the system was applied for the determination of ClO(-) in samples of tap water, and aiming accuracy assessment samples were also analyzed using an independent method. Applying the paired t-test between results obtained using both methods, no significant difference at the 95% confidence level was observed. Other useful features include low reagent consumption, 2.4 μg of LCV per determination, a linear response ranging from 0.02 up to 2.0 mg L(-1)  ClO(-), a relative standard deviation of 1.0% (n = 11) for samples containing 0.2 mg L(-1)  ClO(-), a detection limit of 6.0 μg L(-1)  ClO(-), a sampling throughput of 84 determinations per hour, and a waste generation of 432 μL per determination.

1. Introduction

Since the use of chlorine for tap water treatment was adopted in England in the 1880s [1], water chlorination has been one of the most common disinfectant methods used by waters suppliers [2]. Because chlorine is an efficient agent for inactivating several types of microorganisms, it has been preferred as a disinfecting agent to assure the bacteriological quality of the drinking water [3]. The water bacterial contamination can also occur in the distribution network; therefore, to prevent this occurrence, a free chlorine residual in excess of 0.2 mg L−1 must be maintained throughout the distribution lines [3, 4]. Therefore, the availability of sensitive and reliable analytical procedure for chlorine determination can be considered as an essential condition to assure the quality of tap water. Determination of free chloride in tap water has been carried out by employing as detection techniques amperometry [5], spectrophotometry [6], ion selective electrode [7], chemiluminescence [8], and so forth.

In this work, we intend to develop a photometric procedure for the determination of residual chlorine in tap water using leuco crystal violet (LCV; 4,4′,4′′-methylidynetris (N,N-dimethylaniline), C25H31N3) as chromogenic reagent. LCV is colorless in aqueous solution, but when oxidized at pH 4, it forms crystal violet dye (CV+), which absorbs electromagnetic radiation with a maximum at 592 nm [9, 10]. LCV has been employed in procedures for the determination of hydrogen peroxide [11], iodine, iodide and hypoiodous acid [9, 12, 13], iridium [14], and oxidized manganese [10]. Although, LCV has been widely used in analytical procedures, it has not yet been used for hypochlorite determination, which will be attempted in this investigation.

Nowadays, efforts have been made to environmentally friendly develop analytical procedures, and thus accomplishing the Green Analytical Chemistry (GAC) recommendation [15-17]. Reduction of reagent consumption and waste generation are among the requisites necessary to attain the GAC recommendation [18, 19]. Multicommuted flow analysis (MCFA) [20-22] and multisyringe flow injection analysis (MSFIA) [18, 23] are efficient tools for handling low volumes of sample and reagent solutions, thus providing facilities to save them. Solenoid minipumps have been employed to propel solution in flow analysis systems replacing the peristaltic pump, presenting as an advantage small dimension [24-26], a feature that has been exploited in order to downscale the flow system [27, 28]. This downscaling approach seems to be a powerful tool to develop analytical procedure focused to the GAC recommendation. In the research reported here, these features are exploited to design a miniaturized flow system setup based on the MCFA process for photometric determination of hypochlorite in tap water.

2. Experimental

2.1. Reagents Solutions

All chemicals were of analytical-grade reagent. Purified water with electric conductivity lower than 0.1 μS cm−1 was used throughout.

An acetate buffer solution (0.2 mol L−1) was prepared by dissolving 1.6408 g of sodium acetate in 20 mL of water. After dissolution, pH was adjusted to 4.0 using a 2.0 mol L−1 HCl solution. Afterwards, the volume was made up to 100 mL with water. A 1.0 g L−1 leuco crystal violet (LCV; 4,4′,4′′-methylidynetris (N,N-dimethylaniline), C25H31N3) stock solution was prepared by dissolving 0.1 g of solid material (Eastman Kodak) in 1 mL of phosphoric acid (Merck) 85% (v/v). After dissolution, the volume was made up to 100 mL with water. A 0.1 g L−1 LCV working solution was prepared daily by adequate dilution of the stock solution with water.

A 1000 mg L−1 hypochlorite (ClO−) stock solution was prepared by adequate dilution of a 10% (w/v) reagent solution (Fluka) in 100 mL of a 0.01 mol L−1 NaOH solution. This solution was standardized using the iodometric titration method [29]. Hypochlorite working solutions ranging from 0.02 up to 2.0 mg L−1 were prepared daily by adequate dilution of the ClO− stock solution using a 10−4 mol L−1 NaOH solution.

A set of tap water samples was collected from several points of Piracicaba City. Prior to analysis, samples were alkalized by adding 1.0 mL of a 10−1 mol L−1 NaOH solution to a volumetric flask, and the volume was made up to 100 mL with sample.

2.2. Equipment

The apparatus included a microcomputer running a software written in QuicH BASIC 4.5, a digital multimeter with a serial interface (Minipa, ET-2231), three solenoid minipumps (Bio-Chem Valve Inc., 090SP115-8), a homemade interface to drive the solenoid pumps [30], which was coupled to the microcomputer through the parallel output port, a homemade regulated power supply (−12 V, +12 V) to feed the photometer, an orange high-bright LED (λ = 590 nm), and a photodiode OPT301 (Burr Brawn). A homemade flow cell machined in acrylic, with an optical path length of 20 mm and an inner diameter of 1.0 mm, and a homemade bubble-removing microdevice machined in acrylic, with an inner volume of 10 μL, was similar to that used elsewhere [28]. All flow lines were of Tygon tubing with 0.56 mm inner diameters. The system control and data acquisition were performed by the microcomputer, running software written in Quick BASIC 4.5.

2.3. Analytical Procedure

The diagram of the proposed setup is shown in Figure 1. The network constituted by a transistor (Te), and resistors were used to control the intensity of the light beam emitted by the LED. The glass cylinders (g) conducted the light beams  I1 and I2 from the LED up to the flow cell channel, and from there up to the photodetector (Det), respectively. When the software was run, the solenoid minipumps P1, P2, and P3 were switched in an alternating On/Off sequence, as shown in Table 1, so as to fill each flow line with the respective solution. Afterwards, mini-pump P2 was switched On/Off 40 times to wash the flow cell with carrier solution. The calibration step included the reading of both the dark signal (D) and the reference signal (S). In the first case, the reading was done with LED switched Off while, in the second one, the LED emission was enabled by turning forward the variable resistor (10 kΩ) wired to the base of the transistor (Tr). The emission intensity was adjusted to obtain an electric potential difference (S) of 2000 mV generated by the photodetector (Det). The measurements D and S were converted to digital signal by the multimeter and sent to the microcomputer through the serial interface. These measurements were used for the absorbance calculation.

Figure 1

Diagram of the flow analysis module. S = sample or ClO− standard solution; R = 0.1 g L−1 LCV solution; Bf = acetate buffer, pH 4.0; Int = homemade power driver interface; Mic = microcomputer; Dm = digital multimeter; P1, P2, and P3 = solenoid minipumps; B1, B2, and B3 = flow lines, Tygon tubing 40, 15, and 25 cm long, respectively, and 0.56 mm inner diameter; bd = bubble removing device; Ap = acrylic plates, length, width and thickness of 40, 30, and 10 mm, respectively; h = holes, 2.0 mm diameter; Tm = Teflon membrane; fc = cutaway of the flow cell; W = waste; g = glass cylinders, 30 mm length and 1.0 mm diameter; LED = λmax of 590 nm, high bright (10,000 mcd); Te = transistor BC547; Det = photodetector OPT301; I1 and I2 = electromagnetic radiation beam from LED and to detector, respectively; S0 = signal output (mV).

Table 1

Minipumps switching pattern.

Step	Event	P1	P2	P3	Cycles	Volume (μL)
(a)	Filling flow lines	On	Off	Off	30	360
(b)	Washing flow cell	Off	On	Off	40	240
(c)	Photometer calibration	Off	Off	Off	0	—
(d)	DPD and acetate buffer insertion	On	On	Off	4	48
(e)	Sample insertion signal reading	Off	On	On	12	144
(f)	Signal reading	Off	On	Off	40	240

The number 1 and 0 indicated that respective mini-pump is switched ON or OFF, respectively. Cycles indicate the selected number of times that each valve was switched On/Off.

A micropumping On/Off switching event is named as a sampling cycle, thus the volumes of sample and reagents solutions necessary to carry out the analysis were found by varying the number of sampling cycles. As shown in Table 1 (step d), micro-pumps P1 and P2 were switched simultaneously, so that the coil (B1) was loaded with a mixture of reagent and buffer solutions. Afterwards (step e), micro-pumps P2 and P3 were switched On/Off also simultaneously, so that slugs of the sample and a mixture of buffer and reagent solutions merged into the coil (B2). The reaction for forming the compound to be monitored proceeded while the sample zone was displaced towards the photodetector (Det). While steps (d) and (f) were performed, the microcomputer carried out data acquisition and saved this data as an ASCII file, to allow further processing. The absorbance was calculated using the following equation: Absorb = −log [(S0 − D)/(S − D)], where S0 was the sample signal, with Ss and Dk defined as in the previous paragraph.

While sample processing was occurring, a plot of the analytical signal was displayed on the screen, in order to allow its visualization in real time. After the last event (step f), the software returned to step (d), so as to start another analytical run.

To find the adequate volume of reagent solution, the assays were carried out using 20 sampling cycles for sample and varying from 1 up to 5 the On/Off switching event applied to minipumps P1 and P2 (step e).

The assays described above were carried out using a chromogenic reagent solution with a concentration of 0.1 g L−1 LCV and a 1.0 mg L−1 ClO− standard solution. To find the appropriate reagent concentration, experiments were performed applying 12 and 8 sampling cycles for sample and reagent solution, respectively. Maintaining the hypochlorite concentration at 1.0 mg L−1, the LCV concentration was varied from 0.02 up to 0.1 g L−1.

Aiming to prove the effectiveness of the proposed setup, a set of tap water samples was analyzed. To allow accuracy assessment, samples were also analyzed using a reference procedure [31]. In this procedure, hypochlorite in a pH 6.3 medium reacts with the chromogenic reagent N,N′-diethyl-p-phenylenediamine (DPD), which is oxidized to a red-colored semiquinonoid cationic radical (DPD*) that was monitored at 520 nm.

3. Results and Discussion

3.1. General Comments

According to Hatch [9] and Lambert et al. [13], the stoichiometric relationship of LCV oxidation by hypoiodous acid or iodine in acid medium results in the generation of two electrons for each oxidized molecule, as indicated by the equations shown below:

where LCV is leuco crystal violet and CV+ is the colored crystal violet specie.

Since LCV could be oxidized by hypochlorite, following a reaction mechanism similar to that described for HIO, one would expect that a relationship according to the equation shown below occurs:

Preliminary assays carried out using hypochlorite working solutions showed that absorbance at 592 nm was directly proportional to hypochlorite concentration, thus this feature was employed to develop the analytical procedure.

Reduction of reagent consumption and waste generation are among the requisites necessary to attain the GAC recommendation [18, 19], and the volumes of solutions delivered by the solenoid micropumps per stroke have been used to calculate reagent consumption and waste generation [28]. Nominally, the solenoid micropumps used in this work should deliver a solution volume of 8 μL per stroke, nevertheless, in an earlier work, it was verified that volume delivered per stroke was lower than the expected value [28]. Thus, considering this fact, the volume of solution delivered by stroke under working condition was determined, and was done by applying for each micropump 10 On/Off pulses. The water volume was collected into a vial and weighted to determine the actual volume, which was found to be 6.0 μL per stroke.

3.2. Effect of the Reagents Concentration

The effect concerning the LCV concentration was investigated using a 1.0 mg L−1 ClO− standard solution and applying 12 and 8 sampling cycles for sample and reagent solution, respectively. The LCV concentration varied from 0.02 up to 0.10 g L−1. The absorbance increased up to a concentration of 0.08 g L−1 and showed a tendency to a constant value for higher concentration. Thus, the 0.10 g L−1 solution was selected, so as to assure an excess of reagent throughout the sample zone.

3.3. Effect of the LCV Solution Volume

The assays commented on in previous sections were carried out by inserting aliquots of both LCV and buffer solution, each one with a volume of 48 μL and an aliquot of sample solution with a volume of 72 μL. As is shown in Table 1 (step e), micropumps P2 and P3 were also switched On/Off 12 times, therefore, the volume of the sample zone was 144 μL. Aiming to minimize the reagent consumption, a set of experiments was carried out by varying the volume of the solution aliquot from 6 up to 30 μL, which was done by varying the number of sampling cycles (Table 1, step d) from 1 to 5. The curve in Figure 2 shows that for volumes higher than 18 μL, the variation of the analytical signal was not significant. Thus, the volume of 24 μL was selected as a compromise, in order to save reagent solution while, on the other hand, assuring a sufficient amount for allowing the reaction development to occur.

Figure 2

Effect of chromogenic reagent volume on the analytical signal.

3.4. Effect of Sample Volume

With the objective of verify the effect caused by the sample volume on the analytical signal, assays were carried out by varying the volume of sample aliquot from 6 up to 102 μL, yielding the results shown in Figure 3. Analyzing this curve, we observe that for volumes higher than 66 μL, the signal increment decreased step by step. Thus, we would expect that signal magnitudes tend toward a constant. When sample volume varied within the range of 18 to 60 μL, a linear relationship (R2 = 0.998) was achieved. This behavior is quite different from that observed in the usual FIA system employing injection by loop. In order to achieve a compromise between sensitivity and throughput, a sample volume of 72 μL was selected.

Figure 3

Effect of sample volume on the analytical signal. Standard solution 1.0 mg L−1 ClO−; 24 μL of reagent and buffer solution.

3.5. Performance of the Proposed System

Once the operational conditions of the system were established, a set of assays was carried out, in order to verify the overall performance of the system. The results achieved are compared with other procedures and are summarized in Table 2. As we can see, the proposed procedure presented useful features: a wide linear response range, a low limit of detection, which was estimated as three times standard deviation of blank divided by the slope of the analytical curve, high throughput, low volume of waste generated, and low reagent consumption. Comparing these data with those presented in the referenced papers [32-34], we observe that, excepting throughput (see [33]), the results are highly favorable for the proposed system.

Table 2

System performance comparison.

Parameter	Proposed procedure	Reference [32]	Reference [33]	Reference [34]
Linear equation	A = 0.069 + 0.281C	A = 0.011 + 0.170 C	—	—
Working range (mg L−1)	0.02–2.0	0.6–4.8	0.05–10.0	2–14
Linear coefficient (R2)	0.999	0.998	0.999	0.998
RSD (%), n = 11	1.0	<2.0	0.9	1.4
LOD, 3σ (μg L−1)	6.0	600	30	510
Determination per hour	84	15	110	45
Reagent consumption* (μg)	2.4	40	150	1300
Effluent generation* (μL)	432	3430	1360	2200

*Per determination.

An additional assay was carried out in order to verify whether LCV also reacted with combined chlorine. The assay was done using two sets of standard solutions. The first one containing 1.0 mg L−1 ClO− plus 1.0 mg L−1 NH4+ and 2.0 mg L−1 ClO− plus 2.0 mg L−1 NH4+; the second one, prepared to contain equal ClO− concentration without ammonium. In the first case, absorbances generated processing both solutions were ≈97% lower than those observed when standard solutions were prepared without ammonium. Therefore, we could consider these results as an indication that the LCV does not reacted with combined chloride.

3.6. Sample Analysis

A set of tap water samples was analyzed, employing the operational conditions shown in Table 1, yielding the results shown in Table 3. In order to make accurate assessment, the samples were also analyzed using a reference method [30]. Student's paired t-test calculated for the values obtained by both procedures for the 95% confidence level was found to be 1.360. Since the theoretical value is 2.447, there is no significant difference at the 95% confidence level.

Table 3

Comparison of results obtained by proposed procedure and the reference method.

Sample	Hypochlorite (mg L−1)
	Proposed	DPD [31]
A	0.471 ± 0.008	0.462 ± 0.005
B	0.026 ± 0.001	0.026 ± 0.001
C	0.026 ± 0.00	0.027 ± 0.001
D	0.471 ± 0.004	0.428 ± 0.005
E	0.274 ± 0.006	0.270 ± 0.005
F	0.027 ± 0.001	0.026 ± 0.000

Results are average of the four consecutive sample analyses. No significant difference at 95% confidence level, ttabled = 2.447, tcalculated = 1.360.

4. Conclusions

The performance of the proposed setup proved that multicommuted flow injection analysis (MCFA), implemented employing solenoid minipumps to propel solutions, afforded facilities for downscaling the flow setup. Its reduced dimension allowed a drastic reduction of reagent consumption and waste generation, without any sacrificing of precision, accuracy, and throughput.

Although the LCV had been employed for many analytical purposes, it had not yet been used for hypochlorite determination. In this work, we proved that LCV can be used as a chromogenic reagent for the photometric determination of ClO− in tap water.