Oxidation State-Differentiated Measurement of ... - ACS Publications

Oct 30, 2011 - marvels brought by modern semiconductor innovations, from light-emitting diodes to night vision equipment, would not be possible. On th...
0 downloads 0 Views 3MB Size
ARTICLE pubs.acs.org/ac

Oxidation State-Differentiated Measurement of Aqueous Inorganic Arsenic by Continuous Flow Electrochemical Arsine Generation Coupled to Gas-Phase Chemiluminescence Detection Mrinal K. Sengupta† and Purnendu K. Dasgupta* Department of Chemistry and Biochemistry, University of Texas at Arlington, 700 Planetarium Place, Arlington, Texas 76019-0065, United States

bS Supporting Information ABSTRACT: The electrochemical reduction of inorganic As on a graphite cathode depends on the current density. We observed that while only inorganic As(III) is reduced to AsH3 at low current densities, at high current densities both forms of inorganic As are reduced. We describe a unique electrochemical reactor in which the cylindrical anode compartment is isolated from the outer concentric cathode compartment by a Nafion tube in which a hole is deliberately made and the entire anode compartment is inside the cylindrical cavity of a small volume (∼115 μL) cathode chamber. The evolved arsine is then quantitated by gas-phase chemiluminescence (GPCL) reaction with ozone; the latter is generated from oxygen formed during electrolysis. For the dimensions used, inorganic As(III) can be selectively determined at a current of 0.1 A while total inorganic As (both As(III) and As(V)) respond equally at an applied electrolysis current at 0.85 A, without any sample treatment. For a 1-mL sample, the system provides a limit of detection (LOD, S/N = 3) of 0.09 μg/L for total As (i = 0.85 A) and an LOD of 0.76 μg/L for As(III) (i = 0.10 A); As(V) is obtained by difference. Comparison of ICPMS results for total As in groundwater samples that span a large range of concentration and total inorganic As determined by the present method showed a high correlation (r2 = 0.9975) and a near unity slope. The basic electrochemical arsine generation technique and current-differentiated oxidation state speciation should be applicable as the front end to many other arsenic measurements techniques, including atomic spectrometry.

A

rsenic, the 20th most abundant element in the earth’s crust, has had a long and enigmatic relationship with mankind. On one hand, without the extensive use of arsenic, many of the marvels brought by modern semiconductor innovations, from light-emitting diodes to night vision equipment, would not be possible. On the other hand, the toxicity of arsenic is legendary. This is immortalized even in the title of one of the best known plays of the last century.1 Arsenic pollution problems are common worldwide. The last two decades have brought to light that South Asia (Bangladesh and the Gangetic delta in particular) has been so afflicted with natural groundwater arsenic poisoning (>600 000 with diagnosed arsenicosis, >20 million at risk) that the World Health Organization has labeled it the greatest environmental calamity in recorded history.2 4 Naturally occurring high concentrations of inorganic arsenic in drinking water is causing chronic poisoning, leading both to cancer and noncancerous effects.5 Physiological and toxicological effects of As are best understood from its chemical form.6 In drinking water and groundwater, arsenic exists almost solely in the inorganic form; however, arsenite (As(III)) and arsenate (As(V)) greatly differ in their toxicity.7 Many methods have been reported for the speciation and measurement of As(III) and As(V) in drinking water, but few are sufficiently sensitive, inexpensive, and portable. Some form of initial separation, chromatographic or otherwise, is commonly used in conjunction with and ahead of a mass-selective or element-specific detector.8 10 For the simple r 2011 American Chemical Society

speciation of inorganic As(III) vs As(V) in drinking water, chromatographic separation is somewhat of an overkill, and it detracts from an eventual portable instrument. However, nonchromatographic front-end techniques that have been advocated, such as solvent extraction,11 solid-phase extraction,12,13 coprecipitation,14 capillary microextraction,15 cloud point extraction,16 etc., all represent added costs of consumables and disposal that are significant in developing countries and do not particularly promote the ease of carrying out analysis in the field. Electrochemical methods such as stripping voltammetry and cathodic hydride generation-based techniques have been widely explored for arsenic speciation. Matrix isolation provided by hydride generation has been investigated because of the manifold advantages of selectivity, small sample requirement, reasonable analysis time, and low capital/operational cost.17 Except under extreme conditions (e.g., g4.5 M HCl),18 generally As(V) cannot be easily reduced for analytical purposes; at best, it responds poorly. Consequently, a two-step measurement is typically carried out: A direct measurement of the sample gives the results for As(III) (with little or no response from As(V) depending on the cathode material and conditions). This is followed by a second Received: July 29, 2011 Accepted: October 29, 2011 Published: October 30, 2011 9378

dx.doi.org/10.1021/ac201972m | Anal. Chem. 2011, 83, 9378–9383

Analytical Chemistry

ARTICLE

Figure 1. (a) Electrochemical reactor (ECR) schematic. (b) System schematic, peristaltic pump (PP1, PP2), sample injection valve V), gas liquid separator (GLS) consisting of microporous tube (μPT) in outer jacket (OJ), liquid outlet (LO), restriction tube (RT), gas outlet (GO), oxygen reservoir (OR), waste outlet (WO), ozone generator (OZG), chemiluminescence chamber (CC), photomultiplier tube (PMT), activated Mn oxide catalyst (MC).

analysis of the sample which has undergone a chemical prereduction step.10,19 21 Such methods do work but at the expense of a chemical reduction step that is generally too slow, especially for use in the field. We have previously reported on borohydride-based arsine generation coupled to a gas-phase chemiluminescence (GPCL) detector that can differentiate between As(III) and As(V) based on reaction pH and thus can sensitively measure total As or just As(III), followed by a sequential measurement of As(V).22 Several comparative and validation studies of the GPCL detection scheme have been described.23 25 This GPCL detector is sufficiently economical, sensitive, and portable; it is currently being used in South Asia.26 Sodium borohydride is the primary cost of the analysis; this moisture-sensitive reagent also has a limited lifetime in tropical climate. Finally, we recently reported a graphite cathode electrochemical arsine generator that operates in the batch mode.27 Response to As(V) was significantly lower than As(III) but was sufficient for total As to be measured after oxidation to As(V) (the oxidation step, e.g., with a drop of NaOCl, is simple and rapid, compared to having to reduce As(V) to As(III)). Nevertheless, chemical pretreatment is needed and unequal response of As(III) and As(V) poses intrinsic limitations for accurate speciation. Continued experience with the behavior of graphite electrodes for arsenic reduction and the design of electrochemical reactors have led us to a unique continuous flow electrochemical arsine generator design where a portion of the catholyte flow is deliberately and controllably leaked into the anode chamber to constitute the anolyte flow. More importantly, exactly what is reduced depends on the current density at the cathode. For our cell dimensions, at a “low” applied current of 0.1 A, only As(III) responds. At a “high” applied current of 0.85 A, As(III) and

As(V) both respond equally. The approach thus facilitates speciated determination of inorganic As without having to carry out prior redox manipulations. While other analytical techniques such as atomic spectroscopy can be used, we describe measurement by coupling with our favorite detection technique, GPCL, to form an inexpensive, speciation capable inorganic arsenic analyzer capable of measuring inorganic As down to sub-μg/L levels.

’ EXPERIMENTAL SECTION Instrumentation. The arrangement is based on continuous flow of the carrier electrolyte and continuous aspiration of the effluents. A two-channel (PP1) and a four-channel (PP2) peristaltic pump (www.rainin.com) are used. The sample is introduced by a 1 mL volume loop injector into a water carrier stream (for matrix matching), and this is merged with an equal flow rate of 0.1 M H2SO4 and enters the cathode chamber of the electrochemical reactor (ECR). The ECR is shown in detail in Figure 1a. Construction of the ECR and other relevant details are given in the Supporting Information. Briefly, the ECR is constructed of a 1/4 in. (6.3 mm) diameter electrical discharge machining (EDM) grade graphite rod. A 2.4 mm diameter hole is drilled axially through the entire length. Both the top and bottom termini are then provided with 1/4 28 threads and a flat seat to facilitate connections. The active length of the interior of the reactor is 25 mm. The graphite functions as the cathode. Stainless steel tubes are epoxied through holes in the walls of the graphite rod to provide fluidic access to the cathode interior cavity; one of these also serves as the anode connection. Concentric within the cathode chamber is a Nafion tube (dry dimensions: 1.27 mm o.d., 1.05 mm i.d.) that separates the cathode and anode compartment. A 0.8 mm 9379

dx.doi.org/10.1021/ac201972m |Anal. Chem. 2011, 83, 9378–9383

Analytical Chemistry diameter hole is made in the Nafion tube that allows some of the influent catholyte to leak into the anode chamber, and this constitutes the anolyte flow. Because both the catholyte and anolyte exits are actively aspirated peristaltically and electrolytic gas is formed in both chambers, the aspirated flow rates represent combined gas and liquid flow rates. A 0.25 mm diameter Pt wire runs through the center of the Nafion tube and functions as the anode. The bottom end is sealed with a hot-melt adhesive plug, and the top end exits through a miniature barbed tee where it is also similarly sealed. EDM graphite is porous; the entire exterior of the graphite ECR is coated with graphite-epoxy to preclude liquid seepage through the walls. A power supply PS supplies power to the graphite cathode and the Pt anode and operates in the constant current mode. The gas liquid mixture aspirated from the cathode is pumped into the central channel of a jacketed microporous polypropylene tube (μPT). The μPT is composed of a 15 cm long microporous polyvinyldene fluoride tube (1 mm i.d., 2.2 mm o.d., www. membrana.com) that is put in a 2.7 mm i.d. PTFE tube (10SW, www.zeusinc.com) jacket and serves as the gas liquid separator (GLS). A restriction tube RT (0.56 mm i.d.  400 mm) provides resistance to the liquid flow out of the GLS before the liquid is delivered to waste. The back pressure forces the cathode gas (H2 and AsH3) into the annular jacket space of the GLS from where it is actively pumped into the chemiluminescence detection chamber CC. The anolyte containing liquid and electrolytic oxygen gas is aspirated through the tee arm at the top of the ECR and is dumped into a bottle that functions as the oxygen reservoir, OR. The pressure built up in OR automatically drives the head space oxygen though a miniature ozone generator (OZG, www.enaly. com) into CC, a reflective enclosure placed directly atop a miniature photomultiplier tube (PMT) serving as the chemiluminescence detector. The head space in OR is kept at a constant level by pumping out liquid from OR to waste at a constant rate. Operating Sequence. PP1 starts pumping the electrolyte to the ECR while spent anolyte and O2 are aspirated from the anolyte outlet and spent catholyte and cathode gases are aspirated from the catholyte outlet by PP2. Electrolysis is conducted either at 0.85 or 0.1 A depending upon whether total As or As(III) is being measured; this requires respectively ∼12 14 or 4 6 V. The baseline stabilizes after sufficient oxygen has accumulated in OR and sample injection (1 mL) is now begun. The electrogenerated AsH3 in the catholyte effluent is separated from the liquid by the GLS and enters CC where it reacts with the ozone formed from the anodically generated oxygen, producing chemiluminescence. The PMT (H5784, www.hamamatsu.com) was operated throughout using a control voltage of 0.85 V; this provides about one-third of the maximum sensitivity of which this detector is capable. New samples can be injected about every 4 min for total As and about 7 min for As(III). Because of lower current, in the latter case, gas evolution occurs at a slower rate and reactant gas passage through CC becomes correspondingly slower. Except as stated, results reported herein are based on peak height and reported as average ( SD (n = 3). CAUTION: Arsenic is a class A carcinogen, and appropriate precautions must be taken while handling arsenic-containing standards and sample solutions. Note that arsine gas is extremely toxic, and the instrument must be thoroughly tested for potential leaks before it is used. Because an excess of ozone is used, the arsine is completely oxidized to oxides of arsenic. The arsenic oxide particles are trapped, and excess ozone is destroyed by

ARTICLE

Figure 2. Response of As(III) and As(V) at different electrolyte concentrations. Input flow rate 3 mL/min, low volume ECR (0.1 cm3), electrolysis current 0.85 A.

passage through a MnOOH catalyst cartridge MC (Carulite 200, www.caruscorporation.com). It should be borne in mind when disposing of this cartridge that it may contain a significant amount of arsenic.

’ RESULTS AND DISCUSSION System Optimization. On the basis of previous work with 22 different cathode materials,27 we continued with a graphite cathode and chose Pt as anode. To study the efficiency of arsine generation as a function of cathode area, we varied both the length and the inner bore of the graphite ECR to make four devices A D. The overall volumes of the cylindrical chamber were 0.1, 0.3, 0.5, and 1.1 cm3, respectively (A D active i.d.  length in cm: 0.2  3.8; 0.3  4.3; 0.4  4; 0.5  5.8). Note that these volumes represent the total volume of the cavity. The actual cathode chamber volume is significantly lower, especially for the lower volume devices. These early experiments were conducted without electrolyte composition optimization (this strongly influences relative response of As(III) vs As(V)), at a constant current of 0.85 A and independent (no leakage between chambers) electrolyte flow of 5 mL/min 0.5 M H2SO4 through each electrode chamber. For devices A D, 200 μg/L As(III) produced signal heights of 5.7, 5.2, 4.7, and 4.2 V while 200 μg/L As(V) produced signals of 3.5, 2.9, 1.8, and 0.8 V, respectively. These data are also shown as a plot in Figure S1 in the Supporting Information. In all cases, the smallest volume ECR performed the best and was adopted henceforth with slightly altered dimensions, as stated. Both As species are reduced more efficiently at higher current densities; the difference is larger than the data prima facie indicate, considering that the residence time is also prolonged with larger ECR volumes. This dependence on current is particularly steep for As(V). Electrolyte Concentration and Flow Rate. Sulfuric acid is an inexpensive electrolyte and generally performs better than alternative inorganic acids for electrochemical reduction of As.27,28 Figure 2 shows the results for H2SO4 as the electrolyte ranging in 9380

dx.doi.org/10.1021/ac201972m |Anal. Chem. 2011, 83, 9378–9383

Analytical Chemistry

Figure 3. Response of As(III) and As(V) at different electrolyte flow rates. Electrolyte used is 0.05 M H2SO4. Low volume ECR (0.1 cm3), electrolysis current 0.85 A.

concentration from 0.025 M to 4 M (after dilution with the merging water stream). It should be noted that aside from the obvious effects of the acidity, at the low concentration end, the resistance of the cell increases appreciably. At constant current, this increases the total applied power and the steady-state temperature increases, promoting better removal of dissolved arsine. While the response of As(III) initially increased some from 0.025 M to 0.05 M H2SO4 and then decreased gradually with increasing concentration, that for As(V) remained constant at [H 2 SO4 ] e 0.05 M. At higher acid concentrations, the As(V) response decreased near-exponentially (note logarithmic abscissa of Figure 2). Fortuitously, with a 0.05 M H2 SO4 electrolyte, both As species responded equally and this was then chosen as the electrolyte concentration. For the present setup, inlet liquid flow rates can only be varied if the aspiration flow rates are varied proportionally. With such an arrangement, the results for an inlet electrolyte (and total aspiration) flow rate range of 1 7 mL/min are shown in Figure 3. Note that both responses increase slightly but remain identical with each other for flow rates up to 3 mL/min. As we are looking at peak heights rather than area, this increase can simply be attributed to temporal compression of the signal. It would be intuitively understood from the data that in terms of peak area measured in photon counts, the signal decreases monotonically with increasing flow rate. Even with peak height, the signal for As(V) decreases past 3 mL/min. The faster reacting As(III) does not show a decrease until a flow rate of 6 mL/min. The only reagent presently used in this work is 0.1 M H2SO4; this is inexpensive and no significant savings are realized by lowering the flow rate. On the other hand lowering the flow rate adversely affects the sample throughput rate. We chose therefore an inlet flow rate of 3 mL/min. Effect of Applied Current. We have observed previously that an increase in either the applied current and the current density affects the arsine generation efficiency.27 The present cathode has an active area of ∼>1 cm2, and the volume of the cathode chamber itself is