Seasonal variation of redox species and redox potentials in shallow groundwater: A comparison of measured and calculated redox potentials

Journal of Hydrology 444–445 (2012) 187–198

Contents lists available at SciVerse ScienceDirect

Journal of Hydrology journal homepage: www.elsevier.com/locate/jhydrol

Seasonal variation of redox species and redox potentials in shallow groundwater: A comparison of measured and calculated redox potentials A. Ramesh Kumar a, P. Riyazuddin b,⇑ a b

Chemical Laboratory, Central Groundwater Board, South Eastern Coastal Region, E1, Rajaji Bhavan, Besant Nager, Chennai 600 090, India Department of Analytical Chemistry, University of Madras, Guindy Campus, Chennai 600 025, India

a r t i c l e

i n f o

Article history: Received 14 July 2011 Received in revised form 7 March 2012 Accepted 9 April 2012 Available online 17 April 2012 This manuscript was handled by Laurent Charlet, Editor-in-Chief, with the assistance of Prosun, Associate Editor Keywords: Redox potential Redox couples Redox equilibrium Mixed potential Groundwater Platinum electrode

s u m m a r y The seasonal variation of redox potential (Eh) and redox species such as As(V)/As(III), Cr(VI)/Cr(III), Fe(III)/  Fe(II), NO 3 =NO2 , and Se(VI)/Se(IV) were studied in a shallow groundwater for a period of three years (May, 2004–January, 2007). The study area was Chrompet area of Chennai city, India. Groundwater samples from 65 wells were monitored for pH, electrical conductivity, dissolved oxygen (DO), and major ions during pre-(May) and post-monsoon (January) seasons. The objective of the study was to gain insight into the temporal variation of the redox species due to groundwater recharge and to identify the redox reactions controlling the measured Eh of the groundwater. The study revealed that the shallow groundwater was ‘‘oxic’’ with DO ranging between 0.25 and 5.00 mg L1, and between 0.38 and 5.05 mg L1 during pre-(May, 2004) and post-monsoon (January, 2005) seasons, respectively. The measured Eh (with respect to standard hydrogen electrode, SHE) ranged between 65 and 322 mV, and between 110 and 330 mV during pre- and post-monsoon seasons, respectively. During post-monsoon seasons, DO and Eh increased in most of the wells due to groundwater recharge. The calculated Eh using the redox couples As(V)/As(III),  NO 3 =NO2 , O2/H2O and Se(VI)/Se(IV) neither agreed among themselves nor with the measured Eh during all the seasons. It shows that in the shallow groundwater, the various redox couples are in disequilibrium among themselves and with the Pt electrode. However, 41% (n = 122) of the Eh values calculated from Fe(III)/Fe(II) couple agreed with the measured Eh within ±30 mV, the uncertainty of Pt-electrode measurement. The post-monsoon seasons showed higher values of As(V)/As(III) and Se(VI)/Se(IV) compared to the pre-monsoon seasons, whereas Fe(III)/Fe(II) behaved in the opposite manner. This pattern of variation is consistent with the increased oxidizing nature, as shown by the higher DO and Eh values observed during post-monsoon seasons. The results showed that the Fe(III)/Fe(II) is the dominant redox couple to equilibrate with Pt electrode. However, the measured Eh can only be used in a semi-qualitative way and can be interpreted with other redox indicating parameters. The measured Eh though represent ‘mixed potential’, is a useful indicator for characterizing the speciation and temporal variation of redox sensitive species. Ó 2012 Elsevier B.V. All rights reserved.

1. Introduction Oxidation-reduction (redox) reactions involve the transfer of electrons from one chemical species to another. Redox reactions play an important role in the chemistry of natural waters and affect the mobility and availability of many inorganic and organic species (Appelo and Postma, 2005). For example, As(III) and As(V) have different adsorption characteristics; Cr(III) and Cr(VI) have different solubilities. Redox processes can alternately mobilize or immobilize potentially toxic metals associated with naturally occurring aquifer materials (Cao et al., 2001; Smedley and Kinniburgh, 2002; Sracek et al., 2004; Wang et al., 2007), contribute to the degradation or preservation of anthropogenic contaminants (Bradley, 2003; ⇑ Corresponding author. Tel.: +91 44 22351269; fax: +91 44 22353309. E-mail address: [email protected] (P. Riyazuddin). 0022-1694/$ - see front matter Ó 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jhydrol.2012.04.018

McMahon and Chapelle, 2007), and generate undesirable byproducts such as dissolved ferrous iron (Fe2+), hydrogen sulfide (H2S), and methane (CH4) (Barcelona et al., 1989; Ryu et al., 2004). Redox reactions in natural aqueous systems have been the focus of many geochemical studies including groundwater contamination (Nicholson et al., 1983), acid mine drainage (AMD) (Bachmann et al., 2001; Gezahegne et al., 2007; Mok et al., 1988; Nordstrom et al., 1979), and waters associated with marine sediments (Berner, 1963; Vershinin and Rozanov, 1983). Redox reactions and redox sensitive aqueous species differ in nature from most other reactions in aquifer systems. For many reasons, therefore, it is desirable to have an indicator of redox status of the system. The potential of a platinum electrode (with respect to standard hydrogen electrode, SHE) has been found to roughly obey the Nernst equation in certain natural aquatic systems, including groundwater, AMD and marine sediments. As an alternative to

188

A. Ramesh Kumar, P. Riyazuddin / Journal of Hydrology 444–445 (2012) 187–198

electrometric measurement of redox potential (Eh), redox couples could be used to calculate Eh values. The common redox couples useful for this purpose are Fe(III)/Fe(II) (Grenthe et al., 1992; Grundl and Macalady, 1989; Holm and Curtiss, 1989; Matia et al., 1991), As(V)/As(III) (Cherry et al., 1979; Mok et al., 1988) and S2 =SO2 4 (Ioka et al., 2011). Being an intensity factor, Eh characterizes as a summary parameter, the tendency, but not the capacity, of a system for oxidation or reduction. It is a direct potentiometric measurement of the equilibrium established between all oxidized and reduced species in solution and is based on the Nernst equation, which relates potential to the activities of the electroactive species. In practice, Eh is determined by measuring the potential difference between an inert metal electrode (Pt) in contact with a solution and a reference electrode (Ag/AgCl). Electron exchange with the reduced and oxidized species takes place at the inert metal electrode. The measurement is done at zero net exchange current across metal–solution interface i.e. when electrochemical reversibility is maintained. This reversibility is attained when the individual exchange currents of electroactive redox ions exceed 106 A cm2, which corresponds to concentration of >105 mol L1 of the redox pair (Nordstrom et al., 1979). One of the fundamental requirements for a meaningful Eh measurement in a given time is the attainment of equilibrium between the different redox couples in a solution and the electrode surface, thus resulting in a stable value (Stumm and Morgan, 1996; Whitfield, 1972). In natural waters several redox couples co-exist; therefore all the redox couples are expected to contribute to equilibrium with defined potential. When at least one redox couple is irreversible, the system cannot arrive at equilibrium. Redox couples involving multielectron transfers and structural reorganization between the reduced and oxidized forms are kinetically slow; hence they do not attain equilibrium even though their concentration exceeds 105 mol L1 (Grenthe et al., 1992). Consequently, kinetics apart from equilibrium thermodynamics plays a significant role (Frevert, 1984; Grenthe et al., 1992). The measured Eh, as summary parameter, can thus differ significantly from the calculated potentials for each redox couple. Thus, consistent values of measured and calculated potentials serve a convenient way to check the attainment of redox equilibrium. Lindberg and Runnells (1984) concluded from their studies that neither measured Eh, nor any single calculated Eh, represents ‘‘a master redox value for natural waters’’. For many natural systems, low exchange currents over the electrode interface render measurements unreliable. At disequilibrium, individual redox couples exchange current independently over the solution–electrode interface and the overall potential (measured Eh) is obtained when the sum of all anode exchange currents equal the sum of all cathode currents. This potential is usually referred to as ‘mixed potential’ (Emix) (Grenthe et al., 1992; Grundl, 1994; Stefansson et al., 2005; Stumm, 1984). In dilute solutions containing redox sensitive species that give rise to an exchange current below 106 A cm2 (concentration of electroactive species <105 mol L1), oxygen begins to be sensed by the Pt electrode and undergo surface redox reactions (Stefansson et al., 2005; Whitfield, 1974). Mixed potentials can disturb redox measurements very much, because they are not reversible and mostly are not constant. Even though Pt-electrode Eh measurements may lack a strict thermodynamic basis, when the major redox couples present in a system have been identified, such measurements may permit the generalization of our understanding of geochemical reaction processes and a system’s approach to equilibrium. Indeed Eh measurements were used to interpret several geochemical processes such as mobilization of trace elements (Ahmed et al., 2004; Cao et al., 2001; Hermann and NeumannMahlkau, 1985; Wang et al., 2007), speciation of toxic elements (Kim et al., 2002; Zheng et al., 2004), characterization of contami-

nant plume migration (Chapelle et al., 1996; Christensen et al., 2000; Kehew and Passero, 1990), groundwater pollution from landfill (Nicholson et al., 1983), recharge-discharge process in confined aquifers (Barcelona et al., 1989; Champ et al., 1979; Jackson and Patterson, 1982), redox zonation of aquifers (Chen and Liu, 2003; Edmunds et al., 1984; Hsu et al., 2010; Smedley and Edmunds, 2002; Wang et al., 2007), and temporal variation of redox sensitive elements (Groffman and Crossey, 1999; Gruau et al., 2004; Zheng et al., 2004). The present communication is a part of a study regarding the speciation of trace elements in shallow groundwater of a tannery area of Chennai city, India. The aim of the present study was to gain insight into the temporal variation of the species As(V)/As(III), Fe(III)/Fe(II), Se(VI)/Se(IV), and Cr(VI)/ Cr(III) due to groundwater recharge and to identify the redox reactions controlling the measured Eh of the groundwater. 2. Materials and methods 2.1. Study area The study area is a suburban area (Chrompet) of Chennai city, the capital of Tamil Nadu state, India. It lies between N latitude 12°560 1200 to 12°590 5600 and E Longitude from 80°060 1400 to 80°110 3000 and has an aerial extent of 45 km2. The area forms part of Adyar basin and contains about 15 tanks. Most of the tanks do not possess their command area, because a major part of the area has already been converted to residential/industrial plots over the years. Presently most of the tanks are used as waste disposal sites. The area is underlain by charnockite rocks of archean age. These rocks are exposed over a wide area and are massive and well foliated. These rocks are absolutely devoid of primary porosity and groundwater occurrence is chiefly confined to the secondary porosity developed by weathering, fracturing, jointing and faulting. Groundwater occurs under phreatic conditions in the shallow weathered and fissured zones and under semi-confined conditions in the fractured, faulted and sheared zones (CGWB, 2003). Groundwater is developed through mostly dug wells as the thickness of weathered zone is sufficiently high (15 m). The depth of sampled wells ranged between 3.32 and 14.15 m below ground level (bgl). The depth to water table (DTW) ranged between 0.70 and 13.77 m, and between 0.38 and 8.56 m bgl during pre-(May 2004) and post–monsoon seasons (January 2005), respectively. The wide variation in the water table is due to the topography of the area. Leather tanning is the major industrial activity in the area. The area was away from residential areas when the tanneries came up nearly a century ago, but now it is densely populated. Before the establishment of common effluent treatment plant (CETP) during 1995, the effluents were discharged untreated into open sewers and low-lying areas. Also, the area lacks adequate domestic sewerage system. Over the years, these factors led to severe groundwater contamination. More details about the study area have been described previously (Kumar and Riyazuddin, 2011a). 2.2. Sample collection and preservation Groundwater samples were collected from 65 open dug wells representing the phreatic aquifer during the months of May (premonsoon) and January (post-monsoon) for a period of three years (i.e. 2004–2007). The sampling wells were selected to represent the industrial and domestic sewage contamination. All the wells are regularly pumped for domestic or industrial purposes. Wells were purged before sample collection and groundwater pH, electrical conductivity (EC), dissolved oxygen (DO) and Eh were continuously monitored. Samples were collected after the stabilization of

189

A. Ramesh Kumar, P. Riyazuddin / Journal of Hydrology 444–445 (2012) 187–198

these parameters. Ground water levels were measured during all sampling seasons in order to assess the effect of rainfall on groundwater table. Water samples were collected separately for major ions, total dissolved metals, and for speciation analysis of arsenic, selenium and chromium. Samples for major ions were collected in 1 L precleaned polypropylene bottles. Samples for total dissolved metals were filtered (0.45 lm, Millipore) and acidified with HNO3 (Aristar grade, BDH) to pH < 2. Samples for arsenic and selenium speciation analysis were filtered (0.45 lm, Millipore), collected in 500 mL polypropylene bottles and acidified with HCl (ultra-pure grade, BDH) to pH < 2. This preservation procedure stabilized arsenic and selenium species at least for 30 days. Samples for the analysis of Cr(III) were collected in 500 mL containers after filtration and stored at 4 °C under dark conditions (Sacher et al., 1999; Whalley et al., 1999). For Cr(VI) analysis, samples were collected separately in 500 mL containers and the sample pH increased to >8 by adding 1 mol L1 NaOH (Clesceri et al., 1998; Parks et al., 2004). The samples were kept in ice boxes and transported to the laboratory and stored at 4 °C until analysis. Field blanks and field-spiked samples were prepared and treated exactly as samples. 2.3. Analytical methods All reagents used were of analytical grade or higher and were prepared in de-ionized water of resistivity 18 MX. cm obtained from a Milli-Q (Millipore) water purification system. The sample pH and EC were measured onsite. In situ (downhole) measurement of Eh and DO were made by DO and ORP probes (YSI 200) equipped with lengthy cables. The probes were inserted 30 cm below the water table. Stable readings of Eh were achieved  within 20– 30 min. The measured Eh values were normalized to 25 °C and referenced to the SHE using standardized Zobell solution (Nordstrom, 1977). The Eh probe was kept in a vial of deoxygenated water during fieldwork to limit its exposure to air. Before each measurement, the Pt electrode was cleaned with dil. HCl in the field. Dissolved Fe(II) and Fe(III) concentrations were determined onsite by the modified ferrozine procedure (Viollier et al., 2000) using a portable spectrophotometer (HACH DR 2010). Major ionic constituents were determined following APHA procedures (Clesceri et al., 1998). NO3 was determined by cadmium reduction spectrophotometric method and NO2 was determined by spectrophotometric method using diazotized sulfanilamide with N-(1-napthyl)-ethylenediamine dihydrochloride. Total dissolved metals were determined using ET-AAS (GBC Avanta) after HNO3 digestion. As(V)/ As(III) and Se(VI)/Se(IV) concentrations were determined by the HGAAS methods reported earlier (Kumar and Riyazuddin, 2006; Kumar and Riyazuddin, 2008). Cr(III)/Cr(VI) were determined by the modified ammonium pyrolidinedithiocarbamate (APDC)– methylisobutylketone (MIBK) procedure (Kumar and Riyazuddin, 2011a; Subramanian, 1988). Eh values of the samples were calculated via the Nernst equation using the analytical values for redoxactive species and other constituents as inputs to the equilibrium speciation program PHREEQC (Parkhurst and Appelo, 1999) using wateq4f.dat database. The redox reactions considered for the present study are listed in Table 1. In cases, where either the oxidized or reduced chemical species’ concentrations were at or below analytical detection limits, they were not used in the calculation of Eh values. 2.4. Quality control Analysis of five field blanks showed no evidence of contamination. The recoveries of Se(IV) and Se(VI) from 15 field spikes ranged between 96.0% and 104.5%, and between 94.0% and 103.0%, respectively. The accuracy of the analytical results was verified by an

Table 1 Redox reactions and equilibrium constants used for the calculation of Eh.a. Reaction

Log k

DH (kcal)

H3 AsO4 þ H2 H3 AsO3 þ H2 O

22.5 13.02

117.480 9.680

86.08 30.26

134.790 –

28.57

43.760

Fe2þ Fe3þ þ e O2 þ 4Hþ þ 4e 2H2 O 2 þ  SeO2 4 þ 2H þ 2e SeO3 þ H2 O  þ  þ 2H þ 2e

NO þ H2 O NO 3 2 a

(Parkhurst and Appelo, 1999).

independent method of analysis using ICP-MS. Regression analysis of total selenium values obtained from HGAAS and ICP-MS methods showed that the calculated slope (1.0035) and intercept (0.0017) values were not significantly different from the ideal values of 1 and 0, respectively. Recoveries of As and Se from the analysis of certified reference material SRM 1640 was satisfactory (As: certified value: 26.7 ± 0.4 lg L1; analytical result: 1 27.5 ± 1.2 lg L recovery: 103%; Se: certified value: 22.0 ± 0.5 lg L1; analytical result: 24.1 ± 0.9 lg L1 recovery: 109%). Recoveries of added Cr(III) and Cr(VI) spikes ranged between 90.0% and 95.5%, and between 90.0 and 92.5%, respectively. 3. Results and discussion 3.1. Redox status of the groundwater The range, mean and standard deviation of the hydrochemical parameters monitored are given in Table 2. All the groundwater samples had measurable DO (range 0.25–5.00 mg L1; average 2.74 mg L1). Only two samples had 0.25 mg L1 of DO and five samples had DO between 0.50 and 1.00 mg L1. The remaining 58 samples had DO > 1.00 mg L1 with the maximum of 5.00 mg L1. As per the geochemical classification, waters with DO > 0.5 mg L1 are considered ‘oxic’ and with DO < 0.50 mg L1 are considered ‘suboxic’ (Canfield and Thamdrup, 2009; Berner, 1981). The Eh of the two samples with DO 0.25 mg L1 were 120 and 65 mV, respectively. The Eh of the five samples that had DO < 1.00 mg L1 ranged between 90 and 242 mV, respectively. For samples with DO > 1 mg L1, the Eh varied between 85 and 322 mV. These indicate that with the exception of a few, most of the samples are ‘oxic’ in nature. In general, ‘oxic’ groundwaters should be devoid of dissolved iron, but it is not uncommon to encounter measurable concentrations of dissolved Fe in the presence of DO in natural waters (Applin and Zhao, 1989; Chen and Sung, 2009; Groffman and Crossey, 1999; Matia et al., 1991; White et al., 1990). The Eh values observed in this study were comparable with that of other similar ‘oxic’ groundwaters (Chen and Liu, 2003; Edmunds et al., 1984; Smedley and Edmunds, 2002). For example, the Eh values of groundwater of Sherwood sandstone aquifer of UK ranged between 200 and 400 mV. Chen and Liu (2003) reported a somewhat lower value (range 0–200 mV) of Eh for the oxic groundwater of Choshui delta, Taiwan. The lower values were attributed to the higher dissolved organic matter (DOC) content of the samples. In the principal aquifers of US, shallow water table witnessed lower DO than the deeper ones (<150 m) due to the higher DOC content of shallow waters (McMahon and Chapelle, 2007). In the present study, we did not observe such vertical variations, as the maximum depth of sampled wells was 15 m bgl. Though we have not measured DOC of our samples, the wide spatial variation observed in the DO content could be due to the varying amount of soil organic matter contributed by disposal of tannery and domestic wastes. Some authors observed positive correlation between DO and Eh (Chen and Liu, 2003; Edmunds et al., 1984; Smedley and Edmunds,

A. Ramesh Kumar, P. Riyazuddin / Journal of Hydrology 444–445 (2012) 187–198

Mean

7.24 209 3.32 25.5 2149 141 53 230 13 354 383 178 91 0.05 0.28 77 173 23 0.78

3.2. Temporal variations in the redox status of the groundwater

0.21 48 1.26 0.4 1687 116 50 275 29 141 462 203 126 0.07 0.28 698 344 42 1.00

6.90–7.84 78–320 0.81–4.21 32.6–35.0 355–11,950 14–592 12–321 26–2530 ND–50 67–884 39–3035 34–1850 3–436 ND–0.69 0.03–0.68 ND–856 ND–2650 ND–311 1.80–10.52

7.42 169 2.44 33.9 2633 172 87 265 6 414 511 242 98 0.54 0.18 104 239 57 5.18

0.17 53 0.97 0.5 1715 107 61 314 11 169 470 263 85 0.20 0.14 185 382 49 2.12

6.81–7.85 118–351 0.85–5.12 23.8–26.8 320–8500 24–420 2–206 30–1350 ND-170 73–531 71–2057 10–800 ND–644 ND–0.56 0.02–1.19 ND–1000 ND–1450 ND–200 0.02–5.98

0.22 51 1.06 0.7 1288 79 39 178 25 103 310 159 101 0.07 0.28 162 314 31 0.86

2002; Walton-Day et al., 1990), but in our study the correlation was poor (r = 0.17). This could be due to the mixed redox process and the general lack of attainment of equilibrium (Smedley and Edmunds, 2002). Such a poor correlation between these redox indicating parameters was observed in DO saturated groundwaters of Tuscan basin aquifer (Rose and Long, 1988b).

mV mg L1 °C lS cm1 mg L1 mg L1 mg L1 mg L1 mg L1 mg L1 mg L1 mg L1 mg L1 mg L1 lg L1 lg L1 lg L1 m bgl

ND: not detected; Eh: redox potential; EC: electrical conductivity; DO: dissolved oxygen, WL: water level; bgl: below ground level.

7.17 197 3.00 26.3 2276 155 66 235 15 346 428 220 107 0.05 0.28 337 189 37 0.86 6.67–7.57 105–305 0.60–6.10 25.1–27.6 250–10,735 20–563 6–263 30–2150 ND–170 65–707 40–2970 10–1160 ND–644 ND–0.56 ND–1.19 ND–4230 ND–1560 ND–190 0.02–5.98 0.20 68 1.32 0.7 1646 121 64 271 14 166 90 302 116 0.07 0.18 75 366 76 2.46 7.43 185 2.45 33.7 3011 195 102 289 7 390 612 282 109 0.02 0.20 299 262 87 5.75 6.95–7.94 95–395 0.22–6.16 25.0–36.2 745–10,540 28–688 10–318 30–2100 ND–70 73–1061 60–2837 34–1487 0.86–769 ND–0.50 0.06–0.99 3–3052 ND–2000 ND–561 1.40–11.96 0.32 58 1.11 0.5 1585 105 53 265 20 164 449 196 92 1.40 0.09 172 310 153 1.71 7.38 204 2.96 27.9 2677 177 86 274 10 397 551 201 105 0.27 0.06 95 143 76 2.34 6.5–8.20 110–330 0.38–5.05 26.7–29.7 600–10,260 6–563 13–263 31–2050 ND–150 27–902 57–2921 28–1090 0.26–381 ND–8.52 ND–0.41 ND–850 ND–1462 ND–958 0.38–8.56 0.26 49 1.29 0.6 1851 130 74 334 25 126 579 294 116 1.98 0.05 383 232 58 2.66 7.59 171 2.74 31.8 3171 193 117 343 12 376 693 306 133 0.39 0.07 199 173 64 5.53 pH Eh DO Temp. EC Ca Mg Na K HCO3 Cl SO4 NO3 NO2 PO4-P Mn Fe(II) Fe(III) WL

6.96–8.18 65–322 0.25–5.00 30.0–33.5 605–11,660 18–779 22–399 41–2600 ND–170 73–650 71–3205 46–1606 2–452 ND–11.75 ND–0.30 ND–2050 ND–1543 ND–377 0.85–13.77

January 2007

Mean May 2006

Range SD Mean

January 2006

Range SD Mean

May 2005

Range SD Mean

January 2005

Range Mean

SD May 2004

Range

Unit Parameters

Table 2 Range, mean and standard deviation (SD) values of hydrochemical parameters of groundwater samples collected from the study area during Pre-(May) and post-monsoon seasons (January).

SD

Range

SD

190

Rainfall is the chief source of groundwater recharge and hence all the wells showed an increase of water table after rainfall. The difference in the DTW between post-(January 2005) and pre-monsoon (May 2004) seasons ranged between 0.05 and 7.45 m bgl. The DO of groundwater increased after monsoon in 45 samples in the range of 0.50 and 2.25 mg L1. This has been reflected in the redox status as indicated by the rise in Eh values, but quantitative comparisons between the two redox indicators are poor. Also, it is important to note that though the seasonal variation of DO and Eh is due to groundwater recharge, neither measured nor calculated Eh values were correlated with water level fluctuations. The DO showed a marginal decrease in 20 samples, in the range of 0.10–1.05 mg L1. Interestingly, of these 20 samples, Eh values increased in the range of 10 to 69 mV in 15 samples. These observations show the ‘mixed potential effect’ as observed by several authors (McMahon and Chapelle, 2007; Rose and Long, 1988a). Notwithstanding, these limitations on the validity of ‘mixed potential’, useful qualitative interpretations can be inferred. Fig. 1 shows the comparison of the average values of DO and Eh of pre- and post-monsoon seasons. As can be seen, the groundwater recharge results in the increase of the ‘oxic’ nature of water. This in turn has a significant influence on the mobilization of oxyanionic form of elements as will be discussed later. 3.3. Temporal variation of redox sensitive species The measured concentrations of redox species of As(V)/As(III),  Se(VI)/Se(IV), NO 3 =NO2 and Fe(III)/Fe(II) during the pre- and post-monsoon seasons for some representative samples are given in Table 3. In addition to these, Cr(VI)/Cr(III) were monitored in 15 samples that showed total chromium >50 lg L1, the limit for human consumption. Table 4 shows some representative results of Cr(VI)/Cr(III) speciation analysis. Of these redox sensitive elements, the concentration of total arsenic and total selenium were less than 10 lg L1 throughout the study period, except in a few cases. Fe(II) was in the range between a few tens and a few hundred lg L1 levels, whereas Fe(III) was less than Fe(II) in most of 1 the samples. The level of NO 3 was quite high (range 2–452 mg L  1 as NO3) compared to NO2 (range 0.01–11.75 mg L as NO2), and the latter was detected in a few samples especially during premonsoon seasons. During the three years study period, it was observed that the post-monsoon groundwater was more ‘oxic’ than that of pre-monsoon seasons. Consequently, though As(V) and Se(VI) were the dominant species during pre-monsoon, their concentration increased significantly during post-monsoon seasons (Fig. 2a and b). Unlike arsenic and selenium, of the two species of chromium, only Cr(VI) exists as oxyanion, whereas Cr(III) exists in the cationic form, except under strongly alkaline conditions i.e. pH > 11. Therefore, during post-monsoon seasons, Cr(III) concentration decreased. The concentration of Cr(VI) did not increase appreciably, but the ratio of Cr(VI)/Cr(III)+Cr(VI) species showed a significant increase as seen from Fig. 2c (Kumar and Riyazuddin, 2011a). In the case of arsenic and selenium the number of samples having detectable As(V) and Se(VI) concentrations have also increased during post-monsoon seasons. For example, during May 2004, As(V) and Se(VI) were detected in 38 and 55 samples, respectively,

A. Ramesh Kumar, P. Riyazuddin / Journal of Hydrology 444–445 (2012) 187–198

200

3.5

a

b

3.0 2.5

DO, mg L-1

150

Eh, mV

191

100

2.0 1.5 1.0

50 0.5 0.0

0 May 04 Jan 05 May 05 Jan 06 May 06 Jan 07

May 04 Jan 05 May 05 Jan 06 May 06 Jan 07

Season

Season

Fig. 1. Average values (n = 65) of (a) Eh (mV) and (b) DO (mg L1) of groundwater during pre-(May) and post-monsoon (January) seasons.

but, during January 2005 As(V) and Se(VI) were detected in 59 and 65 samples, respectively. Similar trend was noticed during the other pre- and post-monsoon seasons, as well (Kumar and Riyazuddin, 2011b). The reason was alkaline pH which favors the desorption of oxyanions from sediments and soil, resulting in their mobilization (Masscheleyn et al., 1990; Smedley and Edmunds, 2002; Sracek et al., 2004). Also, oxyanion forms of elements are quite stable in oxic groundwaters (Johannesson and Tang, 2009). Therefore, oxyanion forms of elements such as As, Cr, Se and Sb are common in aerobic groundwaters than in anaerobic groundwaters (McMahon and Chapelle, 2007; White et al., 1990). The trend in the behavior of iron was apparently opposite to that of arsenic and selenium. This is due to the fact that under the observed pH range, As(V) and Se(VI) oxyanions are relatively mobile, whereas Fe(III) precipitates out as hydroxide. The premonsoon seasons are characterized by higher Fe(II) and lower Fe(III) concentrations in most of the samples (Fig. 2d). During post-monsoon, the total Fe concentrations decreased considerably, together with the decrease of Fe(II) and Fe(III) due to the increased oxidizing nature of groundwater. The trend observed is in line with the theoretical prediction and measured Eh. However, the concentration of Fe(III) increased in two wells (samples 18 and 6) during the post-monsoon periods of January 1995 and 1997. This could be due to the higher DO content and/or due to the artifacts of the sample filtration (Horowitz et al., 1996). As Fe(III) is insoluble in the pH range of groundwaters of this study, the measured Fe(III) (mostly colloidal) were not of much use for comparisons.  NO 3 =NO2 species did not show any temporal variation trend as that of the other redox couples. Unlike other redox species, dissolved nitrogen originates mainly from anthropogenic activities i.e. domestic and tannery wastes. Its concentration is controlled by the local sources, hence no spatial or temporal variations were observed. Concentration of NO 2 was comparatively less than that of NO 3 , as the latter is the dominant form of nitrogen in ‘oxic’ waters. 3.4. Aqueous speciation of redox couples We have modeled the aqueous speciation of redox couples As(V)/As(III), Fe(III)/Fe(II) and Se(VI)/Se(IV) using the analytically determined redox species concentrations and other hydrochemical parameters as inputs to PHREEQC program. The species distribution of As(V)/As(III), Fe(III)/Fe(II) and Se(VI)/Se(IV) is almost similar for all the samples; hence the simulated speciation of a few samples are given here. Table 5 shows the aqueous speciation of As(III) and As(V). The predominant form of As(III) is the undissociated H3AsO3 species, because pH of all the samples were lower than the pK1 of H3AsO3 (i.e. 9.2). On the other hand, As(V) predomi-

nantly exists as the HAsO2 and H2 AsO 4 4 species as the pK1 of 2 H3AsO4 is 2.3. Se(IV) exists as HSeO 3 and SeO3 , and Se(VI) exists as the completely dissociated form of SeO2 (Table 6). Fe(II) and 4 Fe(III) exist as a variety of inorganic complexes (Table 7). In addition to the inorganic complexation, Fe(II) forms complexes with organic ligands (Theis and Singer, 1974) but, we have not considered it in the present work.

3.5. Comparison of measured vs. calculated potentials The Eh calculated from all the five redox couples are plotted against the measured Eh values in Fig. 3. As can be seen, the calcu lated Eh values of O2/H2O, Se(VI)/Se(IV) and NO 3 =NO2 are higher than the measured values during all the seasons as the Eh calculated from these couples lie above the 1:1 diagonal line. On the other hand, the calculated Eh values from As(V)/As(III) couple lie below the 1:1 line. Most of the calculated Eh values based on arsenic couple are negative values. The Eh calculated from Fe(III)/ Fe(II) couple clustered around the 1:1 line, yet the deviations are high. The calculated Eh values of all the redox couples clustered around their E0 values and hence appear as horizontal bands. Essentially, Fig. 3 is of similar in nature as that of the plot of calculated vs. measured Eh values of Lindberg and Runnells (1984). In the latter, the scatter of values of redox couples appears as vertical bands, as the calculated Eh values were plotted in x-axis. The number of data points of the redox couples varies, because Eh cannot be calculated if one of the redox pair is below the detection limit (BDL) of the analytical method. Quite clearly, there is no indication of good agreement between the measured and calculated Eh values from different redox couples. Of the 390 samples collected during the study period, only 8 samples had all the five calculated Eh values. It is certainly less compared to the total number of measurements; therefore, we have considered samples with four calculated Eh values (36 samples). The calculated and measured Eh values of these 44 samples are plotted against pH values in Fig. 4, in order to see the negative relationship between pH and Eh, and its influence on the redox equilibrium because redox reactions are often sensitive to pH. The pH of the 44 samples ranged between 6.86 and 8.12. The wide range of difference between the calculated Eh values from different redox couples present in the same samples indicate the disequilibrium between various redox couples. With increasing pH, calculated Eh values tend to decrease, and it is more pronounced at pH values >7.50. Figs. 3 and 4 fully support Stumm and Morgan’s (1996) contention that equilibrium between redox couples in natural waters is not to be expected, and that the best one can hope for is a dominant redox couple to equilibrate with the Pt electrode.

Season

1

May 04 January May 05 January May 06 January May 04 January May 05 January May 06 January May 04 January May 05 January May 06 January May 04 January May 05 January May 06 January May 04 January May 05 January May 06 January May 04 January May 05 January May 06 January May 04 January May 05 January May 06 January May 04 January May 05 January May 06 January May 04 January May 05 January May 06 January May 04

2

3

5

10

17

21

22

31

35

05 06 07 05 06 07 05 06 07 05 06 07 05 06 07 05 06 07 05 06 07 05 06 07 05 06 07

pH

EC (lS cm1)

DO (mg L1)

Eh (mV)

Fe(II) (lg L1)

Fe(III) (lg L1)

As(III) (lg L1)

As(V) (lg L1)

Se(IV) (lg L1)

Se(VI) (lg L1)

NO2 (mg L1)

NO3 (mg L1)

7.33 7.37 7.47 7.31 7.60 7.31 7.51 7.58 7.47 7.41 7.43 7.31 7.38 7.05 7.54 7.17 7.55 7.17 7.40 6.99 7.26 7.09 7.17 7.09 7.83 7.00 7.25 6.67 7.23 6.98 7.39 7.20 7.25 6.87 7.40 6.87 8.18 7.32 7.57 6.90 7.42 6.90 7.82 7.53 7.73 7.57 7.44 7.25 7.85 6.50 7.21 7.19 7.39 7.19 7.75

1975 1460 2120 1955 1800 1700 2850 1590 3300 1650 2700 1300 3520 2540 2990 3330 2640 2330 6480 5960 5950 5620 6050 5610 4830 4440 5440 4830 4880 4220 2410 2740 2340 2660 2250 2350 1240 1185 1270 1240 1250 1040 2350 1210 2550 985 2910 975 2720 2170 2340 1225 1790 1125 3260

2.35 2.85 1.55 3.00 1.50 2.56 1.95 1.52 2.95 3.15 2.00 1.98 2.05 2.30 2.9 3.56 1.89 2.56 0.25 0.38 1.75 3.10 1.36 0.85 1.82 2.15 1.05 1.90 2.01 2.56 1.75 2.00 2.15 2.30 1.99 2.56 4.58 4.68 2.60 2.50 4.08 3.89 4.60 4.50 4.89 1.20 3.87 3.82 0.53 1.25 1.45 1.10 0.95 2.36 0.85

150 175 120 195 120 195 186 200 210 210 195 185 180 185 185 193 172 200 120 150 110 175 146 175 180 285 250 215 199 215 229 250 250 210 205 210 215 225 180 220 197 223 265 245 350 235 285 351 242 307 360 305 285 287 145

251 122 193 40 280 101 ND ND 12 ND 20 9 94 ND 162 ND 320 12 121 52 183 60 245 62 122 11 258 10 200 38 ND 1154 12 850 20 1450 31 ND 50 ND 30 4 ND ND ND ND ND ND ND 22 23 ND 40 37 52

112 40 160 12 10 20 25 12 41 ND 9 1 112 ND 90 ND 65 7 15 ND 69 22 65 9 102 19 100 ND 79 2 19 97 7 136 5 200 73 5 103 ND 85 1 ND 25 ND ND ND ND 24 5 39 ND 12 4 16

0.45 ND 0.69 ND 0.15 0.25 1.75 0.25 0.16 ND 0.25 0.15 1.05 ND 0.25 ND 0.18 1.25 ND ND ND ND ND ND 0.54 ND 0.42 ND 0.35 0.21 0.25 0.20 ND ND ND ND 0.73 1.85 2.56 ND 1.25 3.85 1.00 0.25 ND ND 0.85 0.25 0.25 0.24 0.26 1.25 ND ND 0.95

2.60 2.75 1.36 ND 0.95 2.28 2.81 2.87 1.90 ND 3.29 1.70 1.53 1.93 1.04 ND 2.71 2.31 0.25 ND ND ND ND ND 3.04 3.45 2.10 ND 2.23 4.04 1.60 1.38 ND ND 0.62 1.85 9.48 10.40 3.09 ND 5.56 6.67 1.85 1.90 ND ND 1.55 1.29 0.61 1.28 1.49 3.77 ND 2.25 1.55

ND ND ND ND ND ND ND ND ND ND ND ND ND 0.23 ND 0.25 ND ND ND ND ND ND 0.20 ND ND ND ND ND ND ND ND ND ND 0.31 0.26 0.35 ND ND 0.15 0.23 ND 0.35 0.15 0.40 0.25 ND ND 0.38 ND ND ND ND ND ND ND

ND 3.31 0.95 1.95 0.75 2.46 3.84 3.13 1.10 1.85 1.38 1.36 3.06 3.91 2.95 2.25 1.01 2.05 ND 2.37 2.08 1.52 2.14 2.00 0.53 1.27 0.31 0.95 0.20 0.82 2.27 9.63 ND 4.94 1.00 5.94 0.84 5.12 0.60 3.39 0.25 2.23 0.84 3.85‘ 1.05 1.44 0.20 1.46 1.56 2.59 0.33 2.51 0.34 2.04 0.65

0.08 0.04 ND 0.07 ND ND 0.09 0.08 ND 0.07 ND 0.07 0.08 ND ND 0.03 ND 0.03 11.03 7.50 0.50 0.08 0.40 0.08 ND ND ND 0.06 ND 0.06 ND ND ND 0.04 ND 0.04 ND ND ND 0.04 ND 0.04 ND ND ND ND ND ND 0.06 ND ND 0.03 ND 0.03 ND

112 8 25 107 14 32 108 87 217 104 150 80 295 133 133 150 140 101 345 207 173 483 302 283 172 150 385 150 214 100 212 264 175 7 215 55 40 69 76 23 57 20 44 36 55 33 56 30 346 367 143 45 96 40 140

A. Ramesh Kumar, P. Riyazuddin / Journal of Hydrology 444–445 (2012) 187–198

Well No.

192

Table 3 Representative analytical results (mean, n = 3) of redox sensitive species found in groundwater during pre-(May) and post-monsoon (January) seasons.

56

55

54

07 ND: not detected.

06

05

07

06

05

07

06

05

07

06

January May 05 January May 06 January May 04 January May 05 January May 06 January May 04 January May 05 January May 06 January May 04 January May 05 January May 06 January

05

7.14 7.43 7.11 7.55 7.25 7.42 7.50 7.59 7.36 7.84 7.50 7.86 7.54 7.72 7.29 7.69 7.25 7.62 7.50 7.45 6.98 7.33 7.09

2840 2560 2420 2640 2410 2560 2060 3080 1310 2190 1950 1765 1825 2000 1030 900 1580 3330 2570 3660 2720 1965 2720

1.25 2.85 3.50 1.08 1.87 3.00 2.52 4.32 1.65 2.89 3.00 1.80 2.05 1.95 1.53 1.70 3.04 2.05 3.05 1.25 2.38 2.01 3.07

150 154 127 93 163 135 145 154 148 143 197 145 185 124 192 139 205 153 210 168 200 141 221

12 72 ND 40 15 162 ND 322 ND 251 131 205 322 744 250 341 426 322 ND 352 ND 140 ND

15 50 ND 55 5 98 ND 30 ND 51 19 125 108 108 36 64 24 108 ND 99 ND 42 ND

0.52 ND ND 0.26 0.45 0.25 0.35 0.27 2.33 0.15 2.54 0.28 1.12 1.25 2.56 ND 3.12 ND 1.24 2.28 0.52 ND ND

5.02 0.95 ND 1.19 2.53 1.01 5.29 0.79 3.88 0.94 4.00 4.90 7.42 4.96 7.45 1.58 4.46 0.20 8.18 7.73 1.86 ND 2.05

ND ND ND ND 0.18 0.23 ND ND ND ND 0.15 ND 0.18 ND ND ND 0.15 ND ND 0.15 ND ND 0.15

2.21 0.95 1.05 0.37 0.50 0.49 1.60 ND 0.45 ND 0.63 0.16 1.37 ND 0.95 ND 0.84 ND 2.92 0.70 1.84 ND 1.85

ND ND ND ND 0.02 ND ND ND ND ND ND ND ND ND ND ND ND 0.03 0.03 ND 0.06 ND 0.06

96 12 214 68 187 2 85 16 6 14 60 2 12 23 10 16 12 258 186 107 212 144 180

A. Ramesh Kumar, P. Riyazuddin / Journal of Hydrology 444–445 (2012) 187–198

193

The O2/H2O is the only redox couple for which all the 390 calculated Eh values are available. As all the groundwater samples are from open dug wells tapping phereatic aquifer, significant DO was observed in all the samples (range 0.25–5.0 mg L1 May 2004). Serious criticisms have been expressed on the validity of Eh measurements in oxic waters (Grenthe et al., 1992; Stefansson et al., 2005; Whitfield, 1974). In well aerated waters the Pt electrode is susceptible to the formation of Pt-oxide coating, which then becomes insensitive to changes in the oxygen partial pressure (RiosMendoza et al., 2003; Whitfield, 1974). The air-water exchange reactions are fast compared to other geochemical redox reactions. Therefore, in a shallow unconfined aquifer, it is very unlikely that the redox species will ever reach overall redox equilibrium. On the other hand, for groundwaters which are under confined conditions, DO become depleted along the flow path and hence the Eh decreases (Edmunds et al., 1984; Jackson and Patterson, 1982). The calculated Eh of O2/H2O couple is the highest of all the five couples and the measured values, due to the high E0 value of the couple. Such high calculated Eh values were reported by other authors, as well (Chen and Sung, 2009; Levy, 2007; Lindberg and Runnells, 1984; Stefansson et al., 2005; White et al., 1990). The difference between calculated and measured Eh ranged between 358 and 699 mV. Sato (1960) noted in his pioneering work that measured Eh values were generally much lower than those which would be expected if reversible equilibrium were established between oxygen and water. Sato (1960) suggested that the mechanism of oxygen reduction in natural water involves H2O2 as an intermediate in oxic environments. Subsequent works reported fairly good agreement between Eh calculated from H2O2/O2 couple and measured Eh (Barcelona et al., 1989; Rios-Mendoza et al., 2003). Since we have not determined H2O2 concentration, it may be inappropriate to compare theoretical ‘‘equilibrium’’ Eh values (i.e., calculated on the basis of the O2/H2O couple and standard potentials) with platinum electrode measurements in oxic groundwaters. The calculated Eh based on As(V)/As(III) couple were consistently lower than the measured Eh during all the seasons. Most of the calculated Eh values are negatively biased (<0). Differences of up to 435 mV have been observed between calculated and measured Eh values. Mok et al. (1988) reported consistency between measured and calculated Eh values based on As(V)/As(III) couple within the experimental uncertainty of ±30 mV (Whitfield, 1969). However, the groundwater samples of these authors were from a mine area, having a pH of 2.7–2.8, which is far from the normal groundwater pH range. Also, the potentials calculated from arsenic speciation were higher (428–454 mV) than the present study, because of the acidic pH range. Holm and Curtiss (1989) reported consistency between calculated and measured potentials (±30 mV) for 10 out of 36 groundwater samples. However, the concentration of arsenic species was more than one order of magnitude compared to that of the present study. In the study of Nicholson et al. (1983), Eh values based on As(V)/As(III) couple were consistently lower than the measured Eh, similar to the results of our findings. Also, the concentration range of As species are comparable to the samples of this study. The inconsistency between the measured and calculated Eh values of Nicholson et al. (1983) has been attributed to the artifacts of sampling, analytical error and the redox reactions due to DOM. But, we do not see such effects could be the reasons for the inconsistency observed in our study. Nicholson et al. (1983) study is based on data from the Bordon landfill site and slow reactions with DOM are probably responsible for the lack of equilibrium. In the present study the low concentration of arsenic species and the slow kinetics of the equilibrium are the reasons for the lack of equilibrium (Stefansson et al., 2005). Cherry et al. (1979) proposed the concept of ‘‘redox window’’ within the Eh–pH plot in order to find the predominant arsenic species based on the measured Eh, pH and total arsenic concentra-

194

A. Ramesh Kumar, P. Riyazuddin / Journal of Hydrology 444–445 (2012) 187–198

Table 4 Representative analytical results of chromium species found in groundwater during pre– (May) and post-monsoon (January) seasons. Season

pH

EC (lS cm1)

DO (mg L1)

Eh (mV)

Cr(III) (lg L1)

Cr(VI) (lg L1)

Cr(III)/CrT

Cr(VI)/CrT

5

May-04 January-05 May-05 January-06 May-06 January-07 May-04 January-05 May-05 January-06 May-06 January-07 May-04 January-05 May-05 January-06 May-06 January-07 May-04 January-05 May-05 January-06 May-06 January-07 May-04 January-05 May-05 January-06 May-06 January-07 May-04 January-05 May-05 January-06 May-06 January-07

7.40 6.99 7.26 7.09 7.17 7.09 7.03 7.94 7.43 7.38 7.52 7.45 7.39 7.20 7.25 6.87 7.40 6.87 7.54 7.50 7.63 7.57 7.44 7.35 7.75 7.14 7.43 7.11 7.55 7.25 7.43 7.50 7.60 7.45 7.67 7.35

6480 5960 5950 5620 6050 5610 4660 3200 3330 3250 2560 2500 2410 2740 2340 2660 2250 2350 11,660 10,260 10,540 10,735 11,950 8500 3260 2840 2560 2420 2640 2410 4290 4200 4060 2280 2490 2180

0.25 0.38 1.75 3.10 1.36 0.85 2.35 3.10 1.65 3.19 2.00 3.50 1.75 2.00 2.15 2.30 1.99 2.56 2.65 3.15 2.05 4.20 2.85 3.56 0.85 1.25 2.85 3.50 1.08 1.87 0.25 2.50 1.60 3.76 2.04 2.96

120 150 110 175 146 175 180 195 150 195 148 220 229 250 250 210 205 210 215 200 285 210 228 220 145 150 154 127 93 163 65 140 195 152 208 149

15.4 8.3 6.7 5.5 6.7 3.3 10.9 10.9 14.3 3.1 10.3 1.2 13.6 5.0 15.6 9.1 8.7 3.6 2.0 14.7 20.1 7.7 22.1 6.9 13.0 4.5 10.6 2.7 15.8 3.8 90.2 16.2 26.0 7.9 54.2 8.0

41.3 42.9 55.8 30.1 35.8 35.6 32.8 37.8 46.9 27.4 50.4 18.9 42.5 55.7 46.9 40.6 42.7 42.1 90.7 73.8 85.7 52.8 78.4 54.2 39.8 45.7 40.2 12.6 24.9 38.7 39.6 80.5 89.7 42.8 95.6 64.9

0.27 0.16 0.11 0.15 0.16 0.08 0.25 0.22 0.23 0.10 0.17 0.06 0.24 0.08 0.25 0.18 0.17 0.08 0.02 0.16 0.19 0.13 0.22 0.11 0.25 0.09 0.21 0.18 0.39 0.09 0.69 0.17 0.22 0.15 0.36 0.11

0.73 0.84 0.89 0.85 0.84 0.92 0.75 0.78 0.77 0.90 0.87 0.94 0.76 0.92 0.75 0.82 0.83 0.92 0.98 0.84 0.79 0.87 0.78 0.89 0.75 0.91 0.79 0.82 0.61 0.91 0.31 0.83 0.78 0.85 0.64 0.89

17

18

35

As(V)/As(III) concentration, µg L -1

38

2.0

a

As(III) As(V)

1.5

1.0

0.5

0.0

1.0

b

Se(IV) Se(VI)

3

2

1

0 May 04 Jan 05 May 05 Jan 06 May 06 Jan 07

Season

Season

c

Cr(III)/TCr Cr(VI)/TCr

0.8

Cr species ratio

4

May 04 Jan 05 May 05 Jan 06 May 06 Jan 07

0.6

0.4

0.2

0.0 May 04 Jan 05 May 05 Jan 06 May 06 Jan 07

Season

Fe(III)/Fe(II) concentration, µg L -1

7

Se(VI/Se(IV) concentration, µg L -1

Well No.

250

Fe(II) Fe(III)

d

200 150 100 50 0 May 04 Jan 05 May 05 Jan 06 May 06 Jan 07

Season

Fig. 2. Average (n = 65; for Cr species n = 15) concentration (lg L1) of (a) As(V)/As(III), (b) Se(VI)/Se(IV),(c) relative distribution of Cr(III)/[Cr(III) + Cr(VI)] and Cr(VI)/ [Cr(III) + Cr(VI)], and (d) Fe(II)/Fe(III) of groundwater during pre-(May) and post-monsoon (January) seasons.

195

A. Ramesh Kumar, P. Riyazuddin / Journal of Hydrology 444–445 (2012) 187–198 Table 5 PHREEQC simulated aqueous speciation of As(III) and As(V) in some groundwater samples. Well No./season

As(III)

As(V)

Measured concentration (mol L

1

)

Species 1 May-04

21 May-04

6.014  10

9

3.420  108

2.471  108

Concentration (mol L

1

)

9

H3AsO3

5.871  10

H2AsO3 H4AsO3+

1.434  1010 1.614  1016

HAsO3 2

1.429  1017

AsO3 3 H3AsO3

1.837  1025

H2AsO3

4.789  109 3.011  1015

HAsO3 2 H4AsO3+ AsO3 3 21 January-05

Measured concentration (mol L1)

Simulated

3.475  10

8

HAsO2 4 H2 AsO 4 AsO3 4 H3AsO4

2.941  108

6.707  108

HAsO2 4 H2 AsO 4

1.110  1016 2.336  1022

H3AsO3

2.427  108

H2AsO3 H4AsO3+

4.480  1010 6.621  1016

HAsO3 2

3.085  1017

AsO3 3

2.734  1025

Simulated Species

1.389  107

Concentration (mol L1) 2.484  108 9.905  109 3.309  1012 8.433  1014 6.315  108

AsO3 4

3.861  109 5.118  1011

H3AsO4

4.738  1015

HAsO2 4 H2 AsO 4

9.542  108

AsO3 4

4.350  108 9.292  1012

H3AsO4

3.686  1013

Table 6 PHREEQC simulated aqueous speciation of Se(IV) and Se(VI) in some groundwater samples. Well No./season

Se(IV)

Se(VI)

Measured concentration (mol L1)

Species 17 January-06

22 January-05

38 January-05

3.806  10

HSeO 3

9

Measured concentration (mol L1)

Simulated Concentration (mol L1) 3.663  10

9

6.267  10

8

10

SeO2 3

1.420  10

5.070  109

H2 SeO3 HSeO 3

2.333  1013 4.374  109 6.969  1010

2.032  109

SeO2 3 H2SeO3 HSeO 3

6.333  1014 1.715  109

SeO2 3

3.171  1010

H2SeO3

2.453  1014

Simulated Species

4.880  108

8.991  108

Concentration (mol L1) 6.267  108

SeO2 4 HSeO 4

2.476  1013

SeO2 4 HSeO 4

4.808  1014

SeO2 4 HSeO 4

7.551  1014

4.880  108

8.991  108

Table 7 PHREEQC simulated aqueous speciation of Fe(II) and Fe(III) in some groundwater samples. Well No./season

1 May-04

56 May-04

Fe(II)

Fe(III)

Measured concentration (mol L1)

Simulated

4.500  106

Fe2+ FeHCOþ 3 FeCO3 FeSO4 FeCl+ FeOH+ Fe(OH)2 FeHSOþ 4 FeðOHÞ 3

5.778  106

Species

Measured concentration (mol L1) Concentration (mol L1) 3.064  106 1.017  106

Simulated Species

2.008  106

Concentration (mol L1)

Fe(OH)3 Fe(OH)2+

1.507  106 4.529  107

2.403  107 1.316  107 2.404  108 2.347  108 7.382  1012 4.999  1014

FeðOHÞ 4 FeOH2+ FeSO4+ Fe3+ FeCl2+ FeðSO4 Þ 2

4.854  108 7.654  1011 1.106  1015 7.744  1016 1.244  1016 1.007  1017

7.539  1015

FeCl2+

2.431  1018 4.161  1019

Fe2 ðOHÞ4þ 2

Fe2+ FeHCOþ 3 FeSO4 FeCO3 FeCl+

3.995  106 9.290  107 3.663  107 3.970  107 4.455  108

FeOH+ Fe(OH)2 FeHSOþ 4 FeðOHÞ 3

1.938  106

1.569  106 2.736  107 9.573  108 2.853  1011 5.214  1016

4.639  108 2.326  1011 7.376  1014

Fe(OH)3 Fe(OH)2+ FeðOHÞ 4 FeOH2+ þ FeSO4 Fe3+ FeCl2+ FeðSO4 Þ 2

4.752  1014

FeCl2+

1.174  1018 8.496  1020

Fe2 ðOHÞ4þ 2

2.000  1016 3.991  1017 1.221  1017

196

A. Ramesh Kumar, P. Riyazuddin / Journal of Hydrology 444–445 (2012) 187–198 Table 8 Comparison of Ehc, Ehp and Ehm values (mV) of shallow groundwater.

Eh (mV) Calculated

700 600 500 400 300 As(V)/As(III) Fe(III)/Fe(II) O2 /H2O Se(VI)/Se(IV) NO 3- /NO 2-

200 100 0 -100 -100

0

100 200 300 400 500 600 700 800

Eh (mV) Measured Fig. 3. Comparison of measured vs. calculated redox potential (Eh) of shallow groundwater.

800 700 600

Eh, mV

500 400 300 200

Μ

Μ Μ Μ Μ ΜΜ ΜΜ Μ ΜΜ Μ Μ Μ Μ ΜΜ Μ Μ Μ Μ Μ Μ Μ ΜΜ Μ Μ Μ Μ Μ

Μ ΜΜ ΜΜ

100

Μ Μ Μ Μ

Μ

0 -100 6.8

7.0

7.2

7.4

7.6

7.8

8.0

8.2

pH Fig. 4. Measured (shown as M) and calculated redox potential (Eh) of shallow groundwater as a function of pH. The vertical length of points indicates the spread of Eh of a single sample. Refer Fig. 3 for legend.

tion. From our literature survey it appears that it has not attracted much interest by geochemists. The predicted redox potential (Ehp) based on ‘‘redox window’’ corresponding to the total arsenic concentration and pH of a few samples are given in Table 8. For comparison, the corresponding measured and calculated Eh values from As(V)/As(III) are also given in Table 8. It appears that the difference between Ehp values based on ‘‘redox window’’ and the calculated Eh values were small compared to that of measured and calculated Eh values. This may be considered as an evidence that the As(V)/As(III) was not the dominant redox couple to equilibrate with Pt electrode in the present study.  The Eh calculated from Se(VI)/Se(IV) and NO 3 =NO2 couples were higher than the measured values during all seasons. Though  the concentrations of NO 3 =NO2 species were sufficiently high (i.e. 5 1 >10 mol L ), equilibrium was not attained. Runnels and Lindberg (1990) have shown that in synthetic solutions, Se(VI)/Se(IV) do not attain equilibrium with Pt electrode even at concentrations as high as 790 mg L1. This disagreement is not surprising considering that these redox couples characteristically exhibit non-equilibrium with respect to other more electroactive redox couples such as Fe(II) (Fe(III). The kinetics of redox reactions often depend strongly on the number of electrons that must be transferred to form stable reaction products. This is because multiple electron transfers are generally associated with complex changes in molecular configuration, while single electron transfers often involve

Well No.

Season

pH

As(III)

As(V)

Ehc

Ehp

Ehm

21 21 21 21 35 48 48 48 54 55 55 55 55 56 56

May-04 January-05 May-06 January-07 January-05 January-05 May-06 January-07 January-05 May-04 January-05 May-05 January-06 May-05 January-05

8.18 7.32 7.42 6.90 7.14 6.93 7.48 7.58 7.50 7.42 7.54 7.72 7.29 7.45 7.50

0.73 1.85 1.25 3.85 0.52 0.75 2.56 2.58 0.35 0.28 1.12 1.25 2.56 2.28 1.24

9.48 10.40 5.56 6.67 5.02 9.50 5.43 8.26 5.29 4.90 7.42 4.96 7.45 7.73 8.18

131 8 37 24 33 39 56 44 18 87 33 72 10 46 29

50 30 35 100 75 25 22 20 25 40 28 20 40 25 25

215 225 197 223 150 178 198 198 185 145 185 124 240 168 210

As(III) and As(V) are in lg L1; Ehc calculated Eh based on As(V)/As(III) concentrations; Ehp predicted Eh from ‘redox window’ (Cherry et al., 1979); Ehm measured Eh using Pt electrode.

much less change in molecular structure. Both the oxidized (Fe(III)) and reduced (Fe(II)) forms of Fe-aqua complexes have an octahedral structure. The activation energy to transform Fe(II) ! Fe(III) is relatively small and the reaction can proceed rapidly at groundwater temperature and pressure (Christensen et al., 2000). Of the 296 Eh values calculated from Fe(III)/Fe(II) couple, the measured and calculated Eh values differed by ±30 mV or less for 122 samples (41%). Holm and Curtiss (1989) reported consistency between measured and calculated Eh in 55% (n = 46) of the groundwater samples. The difference between calculated and measured Eh was plotted as a function of total dissolved iron concentrations in Fig. 5. As can be seen, deviation as well as agreement between measured and calculated Eh values are observed for the entire dissolved iron concentration. This reflects the effect of mixed potential, and similar observations have also been reported even in anoxic iron-rich groundwaters (Barcelona et al., 1989). Grundl and Macalady (1989) showed that ionic strength (I) of >0.005 mol L1 is necessary to obtain redox equilibrium of Fe(III)/ Fe(II). In the present study, the I of the groundwater ranged between 0.008 and 0.152 mol L1during pre-monsoon (May 2004), and between 0.009 and 0.134 mol L1 during post-monsoon seasons (January 2005). It appears that the I of the samples contributed towards the attainment of stable Eh, albeit the waters are oxic. We do not claim redox equilibrium in these 41% of samples just because of the agreement between measured and calculated Eh values within ±30 mV, because it can result due to mixed

Eh(measured)-Eh(calculated), mV

800

250 200 150 100 50

0 -50 -100 -150 -200 -7.5

-7.0

-6.5

-6.0

-5.5

-5.0

-4.5

-4.0

Log Total dissolved Fe, molality Fig. 5. Plot of difference between measured and calculated Eh from Fe(III)/Fe(II) vs. log of total dissolved Fe concentration.

A. Ramesh Kumar, P. Riyazuddin / Journal of Hydrology 444–445 (2012) 187–198

potential effect also (Stumm, 1984). Further, unless all the redox pairs exhibit reversible behavior, and exist equilibrium with each other a measured redox potential is not adequate to use as a master variable (Frevert, 1984). 4. Conclusions The results of this monitoring study show that the measurement of Eh using Pt electrode is of considerable use to explain the temporal variation of redox sensitive species of As, Cr, Fe and Se, notwithstanding the drawbacks associated with the theory and measurement of Eh. Among the different redox couples studied, the couple Fe(III)/Fe(II) is useful and can be considered as a master redox couple of the groundwater. However, the measured Eh value cannot be used for geochemical modeling, and to calculate the activity of the redox species. Other redox couples i.e.  As(V)/As(III), NO 3 =NO2 , O2/H2O and Se(VI)/Se(IV) are not useful for the calculation of groundwater Eh in the study area. Measurement of Eh should be complemented by other redox indicators such as DO, Fe(II), Mn(II), H2S, CH4, etc. in order to have meaningful interpretation of redox conditions and to avoid ambiguity. With the improved methods available for speciation analysis, we do not see difficulties in the analysis. However, sample preservation may pose some limitations, because sample matrix plays a key role in the stability of elemental species. For shallow groundwater samples collected in this study, the ratio of As(V)/As(III), Se(VI)/Se(IV) and Cr(VI)/Cr(III) increased during post-monsoon seasons, due to the increased oxidizing nature of groundwater after rainfall. This has been supported by the measurement of Eh, DO and the redox species of the studied elements. The mobilization of the elements As, Cr and Se due to groundwater recharge reported in this study has environmental significance, because of the potential toxicity of these oxyanions. It is generally believed that groundwater recharge will improve the quality in terms of the total dissolved solids, but the findings of this study show the possibility of mobilization of trace elements due to recharge. In addition to the redox couples studied here, other redox couples can be considered for studying groundwater redox reactions occurring in different geological strata. Acknowledgments The first author thanks Dr. S.C. Dhiman, Chairman, CGWB, Mr. S. Kunar, Member (SAM), CGWB, Faridabad and Mr. D.S.C. Thambi, Regional Director, CGWB, SECR, Chennai for granting permission to publish this paper. The authors thank the Associate Editor, and the two anonymous reviewers for their comments on an earlier version of this article that have helped to improve this article. References Ahmed, K.M., Bhattacharya, P., Hasan, M.A., Akhter, S.H., Alam, S.M.M., Bhuyian, M.A.H., Imam, M.B., Khan, A.A., Sracek, O., 2004. Arsenic enrichment in groundwater of the alluvial aquifers in Bangladesh: an overview. Appl. Geochem. 19, 181–200. Appelo, C.A.J., Postma, D., 2005. Geochemistry, Groundwater and Pollution, second ed. A.A. Balkema, New York (pp. 415–480). Applin, K.R., Zhao, N., 1989. The kinetics of Fe(II) oxidation and well screen encrustation. Groundwater 27, 168–174. Bachmann, T.M., Friese, K., Zachmann, D.W., 2001. Redox and pH conditions in the water column and in the sediments of an acidic mining lake. J. Geochem. Explor. 73, 75–86. Barcelona, M.J., Holm, T.R., Schock, M.R., George, G.K., 1989. Spatial and temporal gradients in aquifer oxidation-reduction conditions. Water Resour. Res. 25, 991–1003. Berner, R.A., 1963. Electrode studies of hydrogen sulfide in marine sediments. Geochim. Cosmochim. Acta 27, 563–575. Berner, R.A., 1981. A new geochemical classification of sedimentary environments. J. Sediment. Res. 51, 359–365.

197

Bradley, P.M., 2003. History and ecology of chloroethene biodegradation: a review. Bioremed. J. 7, 81–109. Canfield, D.E., Thamdrup, B., 2009. Towards a consistent classification scheme for geochemical environments, or, why we wish the term ‘suboxic’ would go away. Geobiology 7, 385–392. Cao, X., Chen, Y., Wang, X., Deng, X., 2001. Effects of redox potential and pH value on the release of rare earth elements from soil. Chemosphere 44, 655–661. Champ, D.R., Gulens, J., Jackson, R.E., 1979. Oxidation-reduction sequences in groundwater flow systems. Can. J. Earth Sci. 16, 12–23. Central Ground Water Board (CGWB), 2003. Groundwater resources and developmental potentials of Kancheepuram district, Tamil Nadu, pp. 1–81. Chapelle, F.H., Haack, S.K., Adriaens, P., Henry, M., Bradley, P., 1996. Comparison of Eh and H2 measurements for delineating redox processes in a contaminated aquifer. Environ. Sci. Technol. 30, 3565–3569. Chen, W-F., Liu, T-K., 2003. Dissolved oxygen and nitrate of groundwater in Choshui Fan-Delta, Western Taiwan. Environ. Geol. 44, 731–737. Chen, W.-F., Sung, M., 2009. The redox potential of hot springs in Taiwan. Terr. Atmos. Ocean. Sci. 20, 465–479. Cherry, J.A., Shaikh, A.U., Tallman, D.E., Nicholson, R.V., 1979. Arsenic species as an indicator of redox conditions in groundwater. J. Hydrol. 43, 373–392. Christensen, T.H., Bjerg, P.L., Banwart, S.A., Jakobsen, R., Heron, G., Albrechtsen, H.-J., 2000. Characterization of redox conditions in groundwater contaminant plumes. J. Contam. Hydrol. 45, 165–241. Clesceri, L.S., Greenberg, A.E., Eaton, A.D., 1998. Standard Methods for the Examination of Water and Wastewater, twentieth ed. American Public Health Association, Washington, DC. Edmunds, W.M., Miles, D.L., Cook, J.M., 1984. A comparative study of sequential redox processes in three British aquifers. In: Erikksson, E. (Ed.), Proceedings of Hydrochemical Balances of Fresh Water Systems Uppsala Symposium, vol. 150. IAHS publication, pp. 55–70. Frevert, T., 1984. Can the redox conditions in natural waters be predicted by a single parameter? Schweiz. Z. Hydrol. 46, 269–290. Gezahegne, W.W.A., Planer-Friedrich, B., Merkel, B.J., 2007. Obtaining stable redox potentials in gneiss groundwater and mine water: difficulties, meaningfulness, and potential improvement. Hydrogeol. J. 15, 1221–1229. Grenthe, I., Stumm, W., Laaksuharju, M., Nilsson, A.-C., Wikberg, P., 1992. Redox potentials and redox reactions in deep groundwater systems. Chem. Geol. 98, 131–150. Groffman, A.R., Crossey, L.J., 1999. Transient redox regimes in a shallow alluvial aquifer. Chem. Geol. 161, 415–442. Gruau, G., Diaa, A., Olivie-Lauqueta, G., Davranchea, M., Pinay, G., 2004. Controls on the distribution of rare earth elements in shallow groundwaters. Water Res. 38, 3576–3586. Grundl, T., 1994. A review of the current understanding of redox capacity in natural, disequilibrium systems. Chemosphere 28, 613–626. Grundl, T.J., Macalady, D.L., 1989. Electrode measurement of redox potential in anaerobic ferric/ferrous chloride systems. J. Contam. Hydrol. 5, 97–117. Hermann, R., Neumann-Mahlkau, P., 1985. The mobility of zinc, cadmium, copper, lead, iron and arsenic in groundwater as a function of redox potential and pH. Sci. Total Environ. 43, 1–12. Holm, T.R., Curtiss, C.D., 1989. A comparison of oxidation–reduction potentials calculated from the As(V)/As(III) and Fe(III)/Fe(II) couples with measured platinum-electrode potentials in groundwater. J. Contam. Hydrol. 5, 67–81. Horowitz, A.J., Lum, K.R., Garbarino, J.R., Hall, G.E.M., Lemieux, C., Demas, C.R., 1996. Problems associated with using filtration to define dissolved trace element concentrations in natural water samples. Environ. Sci. Technol. 30, 954–963. Hsu, C.-H., Han, S.-T., Kao, Y.-H., Chen-Wuing Liu, C.-W., 2010. Redox characteristics and zonation of arsenic-affected multi-layers aquifers in the Choushui River alluvial fan, Taiwan. J. Hydrol. 391, 351–366. Ioka, S., Sakai, T., Igarashi, T., Ishijima, Y., 2011. Determination of redox potential of sulfidic groundwater in unconsolidated sediments by long-term continuous insitu potentiometric measurements. Environ. Monit. Assess. 178, 171–178. Jackson, R.E., Patterson, R.J., 1982. Interpretation of pH and Eh trends in a fluvialsand aquifer system. Water Resour. Res. 18, 1255–1268. Johannesson, K.H., Tang, J., 2009. Conservative behavior of arsenic and other oxyanion-forming trace elements in an oxic groundwater flow system. J. Hydrol. 378, 13–28. Kehew, A.L., Passero, R.N., 1990. pH and redox buffering mechanisms in a glacial drift aquifer contaminated by landfill lechate. Ground Water 28, 728–737. Kim, M.-J., Nriagu, J., Haack, S., 2002. Arsenic species and chemistry in groundwater of southeast Michigan. Environ. Pollut. 120, 379–390. Kumar, A.R., Riyazuddin, P., 2006. Studies on the efficiency of hydrogen selenide generation with different acid media by continuous flow hydride generation atomic absorption spectrometry. Microchim. Acta 155, 387–396. Kumar, A.R., Riyazuddin, P., 2008. Determination of arsenic(III) and total inorganic arsenic in water samples using variable tetrahydroborate(III) and acid concentrations by continuous-flow hydride-generation atomic absorption spectrometry. Int. J. Environ. Anal. Chem. 88, 255–266. Kumar, A.R., Riyazuddin, P., 2011a. Chromium speciation in a contaminated ground water: redox processes and temporal variability. Environ. Monit. Assess. 176, 647–662. Kumar, A.R., Riyazuddin, P., 2011b. Speciation of selenium in groundwater: seasonal variations and redox transformations. J. Hazard. Mater. 192, 263–269. Levy, D.B., 2007. Oxidation-reduction chemistry of lechuguilla cave seepage. J. Cave Karst Stud. 69, 351–358.

198

A. Ramesh Kumar, P. Riyazuddin / Journal of Hydrology 444–445 (2012) 187–198

Lindberg, R.D., Runnells, D.D., 1984. Groundwater redox conditions: an analysis of equilibrium state applied to Eh measurements and geochemical modeling. Science 225, 925–927. Masscheleyn, P.H., Delaune, R.D., Patrick Jr., W.H., 1990. Transformations of selenium as affected by sediment oxidation-reduction potential and pH. Environ. Sci. Technol. 24, 91–96. Matia, L., Rauret, G., Rubio, R., 1991. Redox potential measurement in natural waters. Fresenius J. Anal. Chem. 339, 455–462. Mok, W.M., Riley, J.A., Wai, C.M., 1988. Arsenic speciation and quality of groundwater in a lead–zinc mine, Idaho. Water Res. 22, 769–774. McMahon, P.B., Chapelle, F.H., 2007. Redox processes and water quality of selected principal aquifer systems. Ground Water 46, 259–271. Nicholson, R.V., Cherry, J.A., Reardon, E.J., 1983. Migration of contaminants in groundwater at a landfill: a case study. 6. Hydrogeochemistry. J. Hydrol. 63, 131–176. Nordstrom, D.K., 1977. Thermochemical redox equilibria of Zobell’s solution. Geochim. Cosmochim. Acta 41, 1835–1841. Nordstrom, D.K., Jenne, E.A., Ball, J.W., 1979. Redox equilibria of iron in acid mine waters. In: Jenne, E.A. (ed.), Chemical Modeling in Aqueous Systems: Speciation, Sorption, Solubility and Kinetics, Symp. Ser. No 93, Washington, DC, pp. 51–59 (Am. Chem. Soc.). Parkhurst, D.L., Appelo, C.A.J., 1999. User’s guide to PHREEQC (version 2)—A computer program for speciation, batch-reaction, one-dimensional transport, and inverse geochemical calculations. Water Resources Investigations Report 99-4259, US Geological Survey, Denver, Colorado. Parks, J.L., McNeill, L., Frey, M., Eaton, A.D., Haghani, A., Ramierez, L., Edwards, M., 2004. Determination of total chromium in environmental water samples. Water Res. 38, 2827–2838. Rios-Mendoza, L.M., Macias-Zamora, J.V., Zirino-Weiss, A.R., 2003. Iridium: an option for measurement of the redox potential. Cienc. Mar. 29, 509–520. Rose, S., Long, A., 1988a. Monitoring dissolved oxygen in groundwater: some basic considerations. Ground Water Monit. R. 8, 93–97. Rose, S., Long, A., 1988b. Dissolved oxygen systematics in the Tucson basin aquifer. Water Resour. Res. 24, 127–136. Runnels, D.D., Lindberg, R.D., 1990. Selenium in aqueous solution: the impossibility of obtaining a meaningful Eh using a platinum electrode, with implications for modeling of natural waters. Geology 18, 212–215. Ryu, J.-I., Dahlgren, R., Gao, S., Tanji, K.K., 2004. Characterization of redox processes in shallow groundwater of Owens dry lake, California. Environ. Sci. Technol. 38, 5950–5957. Sacher, F., Raue, B., Klinger, J., Brauch, H.-J., 1999. Simultaneous determination of Cr(III) and Cr(VI) in ground and drinking waters by IC-ICP-MS. Int. J. Environ. Anal. Chem. 74, 191–201. Sato, M., 1960. Oxidation of sulfide ore bodies, 1, geochemical environments in terms of Eh and pH. Econ. Geol. 55, 928–961.

Smedley, P.L., Edmunds, W.M., 2002. Redox patterns and trace element behavior in the east midlands Triassic sandstone aquifer, UK. Ground Water 40, 44–58. Smedley, P.L., Kinniburgh, D.G., 2002. A review of the source, behavior, and distribution of arsenic in natural waters. Appl. Geochem. 17, 517–568. Sracek, O., Bhattacharya, P., Jacks, G., Gustafsson, J.-P., Bromssen, M.V., 2004. Behavior of arsenic and geochemical modeling of arsenic enrichment in aqueous environments. Appl. Geochem. 19, 169–180. Stefansson, A., Arnorsson, S., Sveinbjornsdottir, A.E., 2005. Redox reactions and potentials in natural waters at disequilibrium. Chem. Geol. 221, 289–311. Stumm, W., 1984. Interpretation and measurement of redox intensity in natural waters A comment to T. Frevert’s paper ‘‘Can the redox conditions in natural waters be predicted by a single parameter?’’ Schweiz Z. Hydrol. 46, 291–296. Stumm, W., Morgan, J.J., 1996. Aquatic Chemistry, third ed. John Wiley & Sons, New York. Subramanian, K., 1988. Determination of chromium(III) and chromium(VI) by ammonium pyrrolidinecarbodithioate-methyl isobutyl ketone furnace atomic absorption spectrometry. Anal. Chem. 60, 11–15. Theis, T.L., Singer, P.C., 1974. Complexation of iron(II) by organic matter and its effect on iron(II) oxygenation. Environ. Sci. Technol. 8, 569–573. Vershinin, A.V., Rozanov, A.G., 1983. The platinum electrode as an indicator of redox environment in marine sediments. Mar. Chem. 14, 1–15. Viollier, E., Inglett, P.W., Hunter, K., Roychoudhury, A.N., Van Cappellen, P., 2000. The ferrozine method revisited: Fe(II)/Fe(III) determination in natural waters. Appl. Geochem. 15, 785–790. Walton-Day, K., Macalady, D.L., Brooks, M.H., Tate, V.T., 1990. Field methods for measurement of groundwater redox chemical parameters. Ground Water Monit. Rev. 10, 81–89. Wang, S.-W., Liu, C.-W., Jang, C.-S., 2007. Factors responsible for high arsenic concentrations in two groundwater catchments in Taiwan. Appl. Geochem. 22, 460–476. Whalley, C., Hursthouse, A., Rowlatt, S., Iqbal-Zahid, P., Vaughan, H., Durant, R., 1999. Chromium speciation in natural waters draining contaminated land, Glasgow, UK. Water Air Soil Pollut. 112, 389–405. White, A.F., Peterson, M.L., Solbau, R., 1990. Measurement and interpretation of low levels of dissolved oxygen in groundwater. Ground Water 28, 584–590. Whitfield, M., 1969. Eh as an operational parameter in estuarine studies. Limnol. Oceangr. 14, 547–558. Whitfield, M., 1972. The electrochemical characteristics of natural redox cells. Limnol. Oceangr. 17, 383–393. Whitfield, M., 1974. Thermodynamic limitations on the use of platinum electrode in Eh measurements. Limnol. Oceangr., 857–865. Zheng, Y., Stute, M., van Geen, A., Gavrieli, I., Dhar, R., Simpson, H.J., Schlosser, P., Ahmed, K.M., 2004. Redox control of arsenic mobilization in Bangladesh groundwater. Appl. Geochem. 19, 201–214.