SEASONAL VARIATION OF TOTAL DISSOLVED ARSENICAND ARSENIC SPECIATION IN A POLLUTED SURFACE

WATERWAY

ANDREW G. GAULT∗, DAVID A. POLYA and PAUL R. LYTHGOEDepartment of Earth Sciences, University of Manchester, M13 9PL, Manchester, UK

(∗author for correspondence: fax +44 161 275 3947; e-mail: agault@fs1.ge.man.ac.uk)

Received 12 June 2001; accepted in revised form 5 November 2001

Abstract. Seasonal differences in the dissolved arsenic concentration and speciation in a contami-nated urban waterway in northwest England have been determined using a coupled ion chroma-tography-inductively coupled plasma-mass spectrometry (IC-ICP-MS) technique. Waters sampledin the vicinity of an industrial works during relatively dry conditions in April 2000 were foundto contain total arsenic concentrations (�As) of up to 132 µg L−1, more than an order magnitudegreater than the 4 µg L−1 maximum found in December 2000. The difference in �As between theApril and December sampling periods is speculated to be largely due to the irregular anthropogenicsupply of arsenic to the watercourse. For both sampling periods, the dissolved arsenic was exclusivelyinorganic in nature and had an As(V)/�As ratio of between 0.6 and 0.8. Analysis of samples takendownstream of the industrial site, after the confluence with a relatively As-poor stream, revealedthat As(III), As(V) and �As concentrations were lower than would be expected from conservativemixing. The As(V)/�As ratio was also observed to decrease markedly. The loss of arsenic fromsolution is thought to be due to adsorption on the iron oxyhydroxide-rich sediment observed to coatthe riverbed downstream of the confluence. The reduction in the As(V)/�As ratio is believed to bedue to the more rapid adsorption of As(V) compared to that of As(III). Deviations from conservativebehaviour were more marked during the relatively dry April 2000 sampling period and suggest theincreased importance of adsorption processes controlling arsenic availability during this time.

Key words: Accrington, adsorption, arsenic, mixing, seasonal, speciation

1. Introduction

Interest in the biogeochemical cycling of arsenic in the aquatic environment hasburgeoned in recent years. This is, in part, due to public health concerns sur-rounding gross arsenic contamination of drinking water supplies in West Bengal(Das et al., 1996), Bangladesh (Nickson et al., 1998, 2000) and other countries(Chen et al., 1994; Smith et al., 2000), but also to increasing concerns that arsenicin drinking water may have long-term deleterious health effects even at concen-trations in the part per billion range (Smith et al., 1992). In particular, seasonalvariations in both the concentration and speciation of dissolved arsenic in a range ofaquatic environments have attracted a significant amount of study. Temporal vari-ations in total arsenic concentrations in drinking-water wells (Frost et al., 1993) andbiologically mediated seasonal variations in the appearance of methylarsenicals inestuarine (Howard et al., 1984, 1995) and lacustrine (Anderson and Bruland 1991;

Environmental Geochemistry and Health 25: 77–85, 2003.© 2003 Kluwer Academic Publishers. Printed in the Netherlands.

78 ANDREW G. GAULT ET AL.

Sohrin et al., 1997) waters are well-documented. Little work, however, has beenreported on temporal changes in inorganic arsenic speciation in urban waterways.We report here a study of the seasonal variations in the geochemical behaviourof arsenic species in a polluted waterway in northwest England using a coupledion chromatography-inductively coupled plasma-mass spectrometry (IC-ICP-MS)technique.

2. Study site

Accrington is an industrial town situated in northwest England (Figure 1). Thewaterway studied has historically been subjected to anthropogenic inputs froman inorganic chemicals factory and associated spoil heaps. Environment Agencyanalyses of the Tinker Brook waters downstream of the industrial works over thepast 5 years have revealed highly variable dissolved arsenic concentrations withlevels reaching 880 µg L−1 on occasion (K. Gallagher, personal communication).Tinker Brook flows into White Ash Brook, which is distinguished by an ochreouscoating on its stream sediments both above and below the confluence. This ochreis absent from Tinker Brook. The watercourse is relatively fast flowing with an

Figure 1. Location of study site.

ARSENIC IN A POLLUTED SURFACE WATERWAY 79

average depth of approximately 0.5–1 m. A small ochreous discharge to White AshBrook, approximately 100 m downstream of the confluence, was observed duringboth site visits, however, it was only sampled in December 2000. The underlyinggeology of the area is dominated by sandstones of the Upper Carboniferous LowerCoal Measures (Wright et al., 1927), however, there is little exposure of bedrockdue to extensive industrial activities.

3. Materials and methods

3.1. SAMPLING AND ANALYSIS

Sampling in April and December 2000 was carried out by the procedure describedby Gault et al. (2001a). All laboratory analyses were undertaken within 48 h ofsampling.

Identification and quantitation of the dissolved arsenic species present was per-formed by IC-ICP-MS. The sample was injected into an anion exchange column(ANX3206, CETAC Technologies, USA) using 20 mM (NH4)2CO3 as the eluent toeffect species separation. The column was interfaced to a PlasmaQuad II ICP-MS(Fison’s, UK) which acted as the arsenic specific detector through monitoring ofthe signal at mass to charge (m/z) = 75. Further details regarding the techniqueare described by Gault et al. (2001b).

4. Results and discussion

4.1. WATER CHEMISTRY

4.1.1. Bulk constituentsThe concentrations of selected chemical components for the two sampling visits arelisted in Tables Ia and b. The waters sampled on both occasions were circumneutral,well-oxygenated and dominated by Na, Ca, SO4 and Cl.

Conservative mixing of components from White Ash Brook and Tinker Brookwould result in downstream concentrations given by

Cmix = XTBCTB + (1 − XTB)CWAB, (1)

where XTB is the volume fraction of the Tinker Brook contribution to the flowdownstream of the confluence and C refers to the concentration of the subscriptedwaterbody. It follows that XTB may be calculated from measured solute concentra-tions downstream and upstream of the confluence by

XTB = Cmix − CWAB

CTB − CWAB. (2)

80A

ND

RE

WG

.GA

ULT

ET

AL

.

TABLE I

Averaged concentrations of selected chemical constituents in waters sampled from White Ash Brook down-stream (mix) and upstream (WAB) of confluence with Tinker Brook (TB), Accrington, UKa

Constituent WAB TB Mix Seepage XTB �C

(a) 4 April 2000

pH 6.9 ± 0.4 7.1 ± 0.4 6.7 ± 0.4 – –

Eh (mV) 277 ± 4 373 ± 4 334 ± 4 – –

Na (mg L−1) 73 ± 2 17 ± 2 40 ± 2 0.58 ± 0.04 −0.7 ± 3.3

Ca (mg L−1) 62 ± 3 37 ± 2 48 ± 3 0.58 ± 0.13 −0.3 ± 3.5

Mg (mg L−1) 14 ± 1 5.1 ± 0.2 9.0 ± 0.4 0.56 ± 0.07 0.1 ± 0.8

K (mg L−1) 5.9 ± 0.5 4.0 ± 1.2 6.0 ± 1.5 −0.05 ± 0.78 1.2 ± 1.7

Fe (mg L−1) 1.5 ± 0.3 0.4 ± 0.2 0.7 ± 0.2 0.77 ± 0.29 −0.22 ± 0.30

SO42− (mg L−1) 87 ± 4 28 ± 1 54 ± 2 0.56 ± 0.07 0.8 ± 4.8

Cl− (mg L−1) 40 ± 2 22 ± 1 32 ± 2 0.44 ± 0.11 2.2 ± 2.2

NO3− (mg L−1) 2.8 ± 0.1 4.4 ± 0.2 3.6 ± 0.2 0.48 ± 0.12 −0.1 ± 0.2

As(III) (µg L−1) 1.0 ± 0.2 48 ± 3 16 ± 2 0.31 ± 0.04 −12 ± 3

As(V) (µg L−1) n.d. 84 ± 7 5.4 ± 1.5 0.06 ± 0.02 −42 ± 6

�As (µg L−1) 1.0 ± 0.1 132 ± 8 21 ± 3 0.15 ± 0.02 −54 ± 8

As(V)/�As – 0.64 ± 0.06 0.26 ± 0.08 – –

AR

SEN

ICIN

APO

LL

UT

ED

SUR

FAC

EW

AT

ER

WA

Y81

(b) 6 December 2000

pH 7.7 ± 0.4 7.9 ± 0.4 7.8 ± 0.4 8.2 ± 0.2 – –

Eh (mV) 308 ± 4 370 ± 4 300 ± 4 101 ± 5 – –

Na (mg L−1) 14 ± 1 3.4 ± 0.2 7.4 ± 0.2 44 ± 1 0.63 ± 0.05 −0.1 ± 0.6

Ca (mg L−1) 54 ± 2 26 ± 2 37 ± 1 201 ± 5 0.61 ± 0.08 0.3 ± 2.5

Mg (mg L−1) 9.7 ± 0.2 3.2 ± 0.1 5.7 ± 0.1 50 ± 1 0.62 ± 0.03 −0.02 ± 0.29

K (mg L−1) 4.7 ± 0.4 2.5 ± 0.3 3.9 ± 0.3 48 ± 3 0.37 ± 0.20 0.6 ± 0.5

Fe (mg L−1) 2.1 ± 0.2 0.9 ± 0.1 1.2 ± 0.1 n.d. 0.75 ± 0.18 −0.2 ± 0.2

SO42− (mg L−1) 71 ± 3 20 ± 2 39 ± 2 45 ± 3 0.62 ± 0.06 0.1 ± 3.6

Cl− (mg L−1) 28 ± 1 14 ± 1 20 ± 1 117 ± 6 0.59 ± 0.10 0.4 ± 1.4

NO3− (mg L−1) 6.7 ± 0.3 4.6 ± 0.2 5.0 ± 0.2 0.6 ± 0.1 0.80 ± 0.18 −0.4 ± 0.4

As(III) (µg L−1) 0.2 ± 0.1 0.7 ± 0.1 0.7 ± 0.1 32 ± 3 0.98 ± 0.24 0.20 ± 0.11

As(V) (µg L−1) n.d. 2.8 ± 0.6 1.0 ± 0.2 0.6 ± 0.2 0.34 ± 0.09 −0.78 ± 0.40

�As (µg L−1) 0.2 ± 0.1 3.5 ± 0.6 1.7 ± 0.2 33 ± 3 0.45 ± 0.09 −0.57 ± 0.41

As(V)/�As – 0.79 ± 0.21 0.57 ± 0.10 0.02 ± 0.01 – –

a Stated propagated errors do not exactly match analytical errors due to rounding. n.d. = not detected; � As = totalarsenic concentration.

82 ANDREW G. GAULT ET AL.

The consistency of XTB values so calculated using the major components Na, Ca,Mg and SO4 (displayed in bold in Tables Ia and b) suggest that these componentshave indeed mixed conservatively (Filipek et al., 1987) and that the XTB valuescalculated, 0.57 ± 0.04 and 0.62 ± 0.03 for April and December, respectively,accurately reflect the relative volumetric flow rates of White Ash Brook and TinkerBrook.

That apparent XTB values obtained using other components are variably dif-ferent from those above indicates that these components exhibit a degree of non-conservative behaviour. For each of these components the net effect of thesenon-conservative processes, �C, may be calculated by

�C = (Cmix − CWAB) − XTB(CTB − CWAB) (3)

and are shown in Tables Ia and b. For �C = 0, conservative mixing is obeyed,thus use of the conservative mixing value for XTB in Equation (3) will indicatethe extent of removal or addition of a constituent to the waters downstream of theconfluence.

The magnitude of the negative values of �C calculated for Fe relative to themeasured Fe concentrations suggest that it is removed from solution downstream ofthe confluence for both sampling periods. Although the large errors associated withthese values weakens this interpretation, the thin ochreous precipitate, expectedto be composed of Fe oxyhydroxides, observed to coat the bottom of White AshBrook would seem to support this hypothesis.

4.1.2. Arsenic behaviourIC-ICP-MS analysis revealed that the dissolved arsenic was exclusively composedof inorganic species. Tinker Brook supplies the majority of arsenic to the waterway,primarily as As(V). Upon mixing with the arsenic-poor White Ash Brook, arsenicis lost from solution as evidenced by the negative �C calculated for both samplingvisits. This is believed to be caused by adsorption onto the Fe oxyhydroxide-coatedWhite Ash Brook sediment. The reduction in the As(V)/�As ratio is thought to bedue to the greater degree of As(V) removal from solution rather than As(V) re-duction, a process that is thermodynamically unfavourable under these conditions.This shift towards the lower oxidation state has previously been attributed to thefaster rate of As(V) adsorption on the Fe-rich layer of White Ash Brook sediment(Gault et al., 2001a).

During the December visit, an input of As(III) was observed downstream ofthe confluence (Table Ib). The ochreous seepage (possibly sourced from an un-derlying coal seam) observed downstream of the confluence contained 33 µg L−1

As, almost completely as As(III) when analysed in December. Within an hour ofsampling, filtration and acidification of the waters from this site, an orange Fe-rich precipitate began to form, hence lowering the dissolved iron in the seepagewater sample to below detection limits. Since the water was not analysed untilthe following day it is likely that a significant proportion of the dissolved arsenic

ARSENIC IN A POLLUTED SURFACE WATERWAY 83

was scavenged from solution by this newly formed phase. Thus, although the flowvolume of this seepage is considerably smaller than White Ash Brook, it maymake an important contribution to the arsenic burden of the stream, particularlywhen the concentration of arsenic in the waters of Tinker Brook is relatively low.In addition, dissolution of the orange precipitate surrounding the seepage wouldfurther enhance the dissolved arsenic levels of the system.

4.2. TEMPORAL VARIATIONS

A general reduction in major element concentration of the system is observed inDecember relative to April. This is almost certainly due to dilution of the systemcaused by the marked increase in precipitation at the time of this sampling visit.The dramatic 37-fold reduction in Tinker Brook dissolved arsenic levels cannot,however, be explained by meteoric dilution alone. The erratic nature of TinkerBrook arsenic concentrations in the past (Environment Agency, unpublished data)is likely to be caused by variations in the industrial input to the stream. Hence, thedrop in arsenic concentration recorded in December is expected to be related to acorresponding decrease in anthropogenic discharge to the watercourse.

A larger proportion of arsenic is removed from solution downstream of theconfluence in the April survey than in December (72 and 24%, respectively). Theextent of the reduction in the As(V)/�As ratio is also greater during the Aprilsampling survey (Tables Ia and b). This implies that adsorption processes are moreimportant in attenuating dissolved arsenic concentrations during periods of dryweather.

5. Conclusions

The spatial and temporal variations in the arsenic species distribution of a pollutedsurface waterway have been quantified using an IC-ICP-MS technique. Althoughthe considerable difference in total dissolved arsenic concentration between theApril and December sampling periods may in part be a seasonal effect related toincreased precipitation and streamflow during the winter-time, the variation is morelikely to be due to the irregular anthropogenic supply of arsenic to the system. Themarked drop in dissolved arsenic levels upon mixing with the relatively uncon-taminated White Ash Brook waters and the coincident reduction in the As(V)/�Asratio can be ascribed to adsorption on the Fe oxyhydroxide-rich sediment that char-acterises White Ash Brook and the differential adsorption kinetics of inorganic Asspecies on this material respectively. The greater downstream decline in dissolvedarsenic concentration and As(V)/�As ratio recorded during the April visit sug-gests that adsorption processes are more effective at regulating the aqueous arseniclevels during drier conditions, perhaps reflecting the greater ratio of adsorption sitesurface area to aqueous bulk volume present at these times.

84 ANDREW G. GAULT ET AL.

Acknowledgements

AGG acknowledges the receipt of a NERC/CASE PhD studentship (04/99/FS/184)with CETAC Technologies. The authors are grateful to two anonymous SEGHreviewers for their constructive comments that improved the manuscript, AlastairBewsher for practical assistance with IC analysis and Kevin Gallagher for provisionof Environment Agency historical data of Tinker Brook arsenic concentrations.

References

Anderson, L.C.D. and Bruland, K.W.: 1991, Biogeochemistry of arsenic in natural waters: Theimportance of methylated species, Environmental Science and Technology 25, 420–427.

Chen, S.-L., Dzeng, S.R., Yang, M.-H., Chiu, K.-H., Shieh, G.-M. and Chien, M.W.: 1994, Arsenicspecies in groundwaters of the blackfoot disease area, Taiwan, Environmental Science andTechnology 28, 877–881.

Das, D., Samanta, G., Mandal, B.K., Chowdhury, T.R., Chanda, C.R., Chowdhury, P.P., Basu,G.K. and Chakraborti, D.: 1996, Arsenic in groundwater in six districts of West Bengal, India,Environmental Geochemistry and Health 18, 5–15.

Filipek, L.H., Nordstrom, D.K. and Ficklin, W.H.: 1987, Interaction of acid mine drainage withwaters and sediments of West Squaw Creek in the West Shasta mining district, California,Environmental Science and Technology 21, 388–396.

Frost, F., Frank, D., Pierson, K., Woodruff, L., Raasina, B., Davis, R. and Davies, J.: 1993, A sea-sonal study of arsenic in groundwater, Snohomish county, Washington, USA, EnvironmentalGeochemistry and Health 15, 209–214.

Gault, A.G., Polya, D.A. and Lythgoe, P.R.: 2001a, Arsenic speciation in contaminated urbanwaters, Accrington (UK) using IC-ICP-MS, in: Cidu, R. (ed), Proceedings of the 10th Interna-tional Symposium on Water–Rock Interaction, Villasimius, Sardinia, A.A. Balkema, Rotterdam,pp. 1083–1086.

Gault, A.G., Polya, D.A. and Lythgoe, P.R.: 2001b, Hyphenated IC-ICP-MS for the determinationof arsenic speciation in acid mine drainage, in: Holland, J.G. and Tanner, S.D. (eds), PlasmaSource Mass Spectrometry: The New Millenium, The Royal Society of Chemistry, Cambridge,pp. 387–400.

Howard, A.G., Arbab-Zavar, M.H. and Apte, S.: 1984, The behaviour of dissolved arsenic in theestuary of the River Beaulieu, Estuarine, Coastal and Shelf Science 19, 493–504.

Howard, A.G., Comber, S.D.W., Kifle, D., Antai, E.E. and Purdie, D.A.: 1995, Arsenic speciation andseasonal changes in nutrient availability and micro-plankton abundance in Southampton Water,UK, Estuarine, Coastal and Shelf Science 40, 435–450.

Nickson, R.T., McArthur, J.M., Burgess, W.G., Ahmed, K.M., Ravenscroft, P. and Rahman, M.:1998, Arsenic poisoning of Bangladesh groundwater, Nature 395, 338.

Nickson, R.T., McArthur, J.M., Ravenscroft, P., Burgess, W.G. and Ahmed, K.M.: 2000, Mechan-ism of arsenic release to groundwater, Bangladesh and West Bengal, Applied Geochemistry 15,403–413.

Smith, A.H., Hopenhayn-Rich, C., Bates, M.N., Goeden, H.M., Hertz-Picciotto, I., Duggan, H.M.,Wood, R., Kosnett, M.J. and Smith, M.T.: 1992, Cancer risks from arsenic in drinking water,Environmental Health Perspectives 97, 259–267.

Smith, A.H., Arroyo, A.P., Guha Mazumder, D.N., Kosnett, M.J., Hernandez, A.L., Beeris, M.,Smith, M.M. and Moore, L.E.: 2000, Arsenic-induced skin lesions among Atacameño people

ARSENIC IN A POLLUTED SURFACE WATERWAY 85

in northern Chile despite good nutrition and centuries of exposure, Environmental HealthPerspectives 108, 617–620.

Sohrin, Y., Matsui, M., Kawashima, M., Hojo, M. and Hasewaga, H.: 1997, Arsenic biogeochem-istry affected by eutrophication in Lake Biwa, Japan, Environmental Science and Technology 31,2712–2720.

Wright, W.B., Sherlock, R.L., Wray, D.A., Lloyd, W. and Tonks, L.H.: 1927, The geology ofthe Rossendale anticline, Memoirs of the Geological Survey, H.M. Stationery Office, London,200 pp.