Science of the Total Environment 327(2004) 41–51

0048-9697/04/$ - see front matter� 2003 Elsevier B.V. All rights reserved.doi:10.1016/j.scitotenv.2003.10.011

Heavy metals in drinking waters from Mount Amiata(Tuscany,Italy). Possible risks from arsenic for public health in the Province

of Siena

Gabriella Tamasi, Renzo Cini*

Department of Chemical and Biosystem Sciences and Technologies, University of Siena, Via Aldo Moro 2, I-53100 Siena, Italy

Received 18 April 2003; accepted 3 October 2003

Abstract

Concentrations of As, Al and some heavy metals(V, Cr, Mn, Fe, Co, Ni, Cu, Zn, Cd, Pb) were measured indrinking waters from Siena and Grosseto districts, South Tuscany, Italy. The analysis, performed mostly byelectrothermal activated atomic absorption spectroscopy equipped with graphite furnace, and in some cases high-resolution inductively coupled plasma mass spectrometry, indicated that concentrations of the elements were generallyfar below the maximum allowed concentration(MAC). However, the concentration of As in some of the waters atsources or at the terminals of the water webs was relatively high(largest value, 14.4(2) mgyl) when compared to theMAC value (10 mgyl, December 25, 2003; Italian Law). Relatively high concentrations of some metals had beenAs

detected in a few samples from the ends of the distribution webs, when compared to values at sources. These effectsare probably due to leaching from metal pipes. A general ‘metal index’(MI) for drinking water, which takes intoaccount possible additive effects ofN heavy metals on the human health that helps to quickly evaluate the overallquality of drinking waters, is introduced in this paper as MIs8 wC y(MAC) x. Samples from Ermicciolo springis1,N i i

and Siena water web had MI values of 1.1 and 1.3, respectively, showing that the quality of drinking water in townis somewhat worse than that at one of the main sources, at least regarding the 12 elements taken into account.� 2003 Elsevier B.V. All rights reserved.

Keywords: Heavy metals; Arsenic; Vanadium; Drinking water; Atomic absorption spectroscopy

1. Introduction

The interest on the effects on humans and otheranimals of heavy metals and metalloids like arsenictaken in through drinking and use of thermalwaters has increased in recent years(Smith et al.,2002; Nordstrom, 2002; Senn and Hemond, 2002;

*Corresponding author. Tel.:q39-0577-234368; fax:q39-0577-234177.

E-mail address: cini@unisi.it (R. Cini).

Fewtrell et al., 2001; Lind and Glynn, 1999;Manay et al., 1999; Ramessur et al., 2001; Flaten,2001; Gauthier et al., 2000; Veeramachaneni et al.,2001; Pedersen and Permin, 1988). The definitionof the maximum allowable concentration(MAC)values for certain elements(recently classified astrace elements) in spring, drinking, thermal andsurface waters has been and still is the subject ofchemical and biological research, and of politicaldebate, in several countries. For instance, the value

42 G. Tamasi, R. Cini / Science of the Total Environment 327 (2004) 41–51

of MAC (arsenic, AW 74.92, Group 15, VA) inAs

the USA was decreased from 50 to 10mgyl inMay 2001 (starting January 22, 2006) (US EPA,Office of Water, 2000, 2001). Notwithstanding,the debate on this parameter is still open. TheEnvironmental Protection Agency(EPA), theNational Research Council(NRC) and severalresearch groups stated that chronic effects onhumans may be caused by prolonged consumptionof water with a concentration of As as low as 5mgyl (EPA) or even 3 mgyl (NRC) (Hogue,2001a,b). Arsenic is usually relatively abundant insulfide-containing mineral deposits such as pyriteand ‘soft’-metal mineralizations, as well as inhydrous iron oxides; therefore, monitoring ofwaters in volcanic areas and in districts that arerich in heavy-metal sulfide ores can be importantto assess possible health risks.Increasing interest has grown on the content of

vanadium in drinking and thermal waters of Italy(WHO, 2000; Giammanco et al., 1996, 1998) inthe last few years. Starting on November 1999 theItalian Health Ministry fixed the MAC (vanadi-V

um, AW 50.94, Group 5, VB) value at 50mgylfor drinking water (Decreto del Presidente dellaRepubblica, 1988, n. 183, Decreto Legge, 2001,n. 31, Decreto Ministeriale, 1999, Italian law).The reason of concern came after finding that Vconcentrations were unusually high in drinkingwaters from the Etna volcanic massif in Sicily(University of Catania and Italian Health Ministry,1995; Giammanco et al., 1996, 1998). Concentra-tions of V as high as 50–130mgyl for drinkingwaters from Mount Etna area were measuredthrough a study performed at the University ofCatania, the Italian NRC and the Superior Instituteof Health(SIH) of Rome.Another metal that is attracting increasing inter-

est because it can affect the quality of drinkingwaters is Co(cobalt, AW 58.93, Group 9, VIIIB).Several reports indicated that prolonged consump-tion of the element from waters may cause der-matitis, asthma and failures in the cardio-vascularsystem, at least when ingested at 5–10mgyday(Barceloux, 1999; Nordberg, 1994). It is notewor-thy that the MAC value is not yet defined byCo

law in Italy (Decreto del Presidente della Repub-blica, 1988, n. 183).

Previously reported geological and mineralogi-cal analyses indicate that the geological evolutionof Tuscany produced volcanic bodies and volcan-ites and led to a large-scale geothermal activitythat is still present in southern Tuscany and MountAmiata area. Several areas have been mined formillennia and contain Mn, Fe, Cu, Zn, Sn, Sb, As(Carmignani and Kligfield, 1990; Serri et al., 1993;Grassi and Netti, 2000; Dall’Aglio et al., 2001;RIMIN Spa-ENI, 1990; Protano et al., 1998). Thestructure of the aquifer at Mount Amiata has acapacity of approximately 2=10 m of water and9 3

the regime overall annual withdrawal in the longterm is estimated at approximately 65=10 m6 3

(Amministrazione Provinciale di Siena, 1994).On the basis of these premises and in accord

with our interest on metal ion–biomolecule com-plexes(Defazio and Cini, 2002; Cini and Tamasi,2002; Cini and Pifferi, 1999; Cini et al., 2001),we determined the concentrations of V, Co, Asand other elements in the drinking waters from theSiena countryside, focusing on the Mount Amiataarea, which is volcanically active and rich insulfide-containing mineral deposits. We report hereon the highlights of this work.

2. Experimental

2.1. Reagents and materials

All the chemicals used for the analytical deter-mination were ultrapure compounds. Water forstandard preparations was distilled twice througha glassware device, further distilled through a fullquartz device, then stored in polyethylene contain-ers (this type of water is hereafter named uppw).The standard solutions contained 1.000 gyl of theelements Al, V, Cr, Mn, Co, Ni, Zn, As, Cd andPb, and were purchased from Merck(Darmstadt,Germany).The standard solutions of Fe and Cu were

prepared by dissolving 0.1000 g of ultrapure metal(Merck) in ultrapure nitric acid(65%, Merck) andthen diluted to 100 ml. All solutions containedHNO (0.5 M) and were stored in polyethylene3

bottles. The standard solutions for Al, V, Cr, Mn,Fe, Co, Ni, Cu, As and Pb were combined mixtures

43G. Tamasi, R. Cini / Science of the Total Environment 327 (2004) 41–51

of three or four elements, whereas those for Znand Cd were single metal solutions.Other reagents and solvents were HCl 37%

(analytical grade, a.g., Merck); NH 25% (a.g.,3

Merck); NaOH (a.g., Carlo Erba, Milan, Italy);BaCl Ø2H O (a.g., Baker, Milan, Italy); AgNO2 2 3

(a.g., Carlo Erba); Mg(NO ) Ø6H O (Suprapur,3 2 2

Merck); Na H EDTAØ2H O, C H N O NaØ2 2 2 10 14 2 8 2

2H O (EDTA: N,N,N,N-tetraacetate-1,2-diamino-2

ethane; a.g., Baker); indicator-buffer pH 10,eriochrome black T(Merck); indicator hydroxy-naphthol blue(Aldrich, Milan, Italy).Class A glassware(Continental, Supreme) was

used for all the determinations. The estimated errorfor the burette was 0.01 ml. Porcelain cruciblesand porcelain gooches(Haldenwanger, Berlin)were used for the gravimetric determination ofsulfate and chloride. Whatman filtering paper(42-ashless) was used for the analysis of sulfate. Theclass A glassware and polyethylene vessels weresoaked in HNO (0.2%) for 24 h and then rinsed3

with uppw before performing the metal trace deter-minations. A micropipette Nichipet EX(Nichiryo,Flanders, USA), 10–100 ml, and polyethyleneDiamond D200, 200ml (Gilson, Middleton, USA),tips were used throughout the work. All sampleswere filtered using sterile FP 030y0.20 CA-S,0.20-mm (Schleicher & Schuell, Duren, Germany)¨disks, previously washed by fluxing approximately10 ml of the acidified sample.As standard reference material for the trace

elements in water was used the SRM 1643d fromthe National Institute of Standards and Technology(NIST, Gaithersburg, MD 20899-2322, USA).

2.2. Collection and pretreatment of samples

Samples were collected in October 2000–Octo-ber 2002 from both springs and public aqueducts.For each site three distinct collections were carriedout (Fall 2000, Spring 2001, Spring or Fall 2002).The water samples were stored in polyethylenebottles, pre-washed with concentrated nitric acid(f30% vyv), and after collection, nitric acid(0.2%) was added as a preservative as requestedby the standard methods(Italian Health Ministry,2000).

The temperature of each sample was measuredat the sampling site using a Crison T-637 thermom-eter (Crison Strumenti, Carpi, Modena, Italy),equipped with a platinum thermocouple, Pt100DIN("0.1 8C). The pH measurements were carriedout at 25 8C using an E603 Metrohm pH-meter("0.01 log unit), equipped with a Metrohm Her-isau (Herisau, Switzerland) combined glass elec-trode. The pH-meter was calibrated against Crisonbuffer standard solutions(pH 4.01("0.02), 7.00("0.02), 9.21("0.02), at 258C). The conductiv-ity measurements(25 8C) were carried out througha Conductimeter GLP 32(Crison Strumenti)equipped with a conductivity cell(Sensors Crison;C, 1.02). The device was calibrated against theKCl standard solutions(1413("12) mSycm and12.88 ("0.11) mSycm; Crison). All samples forthe analysis of metals were acidified with ultrapurenitric acid (f1 mlyl) and stored in sealed poly-ethylene bottles maintained at 48C.Some of the samples were concentrated before

the analysis of V, Cr, Mn, Co, Ni, Cu for a moreaccurate determination of these trace elements.These samples(400 ml) were slowly heated inDuran-glass beakers by using 100-W lamps placedat approximately 5 cm above the sample surface.The evaporation process was carried out in a dust-free room until reaching a final sample volume of10.0 ml. Uppw samples also were concentratedwith the same procedure. The concentrations ofthe above-mentioned six metals in each concen-trated sample were then blank corrected.

2.3. Atomic absorption spectroscopy and high-resolution inductively coupled plasma massspectroscopy

The analyses were carried out using atomicabsorption spectroscopy(AAS) with electrotherm-ic atomization in graphite furnace for the deter-mination of the total content of Al, V, Cr, Mn, Fe,Co, Ni, Cu, Zn, As, Cd and Pb. The instrumentused was Aanalyst100(Perkin-Elmer) spectropho-tometer(continuum background correction) locat-ed at the Department of Chemical and BiosystemSciences and Technologies. The machine wasequipped with an HGA-800 furnace and an AS-72autosampler. All lamps used were hollow cathode

44 G. Tamasi, R. Cini / Science of the Total Environment 327 (2004) 41–51

Table 1Concentration of the elements(mgyl) examined in this work,in the Standard Reference Material(Water SRM 1643d, NIST)via atomic absorption spectroscopy

Element Water SRM 1643d(NIST)

Certified Found

Al 127.6"3.5 134.5(7)V 35.1"1.4 36.7(1)Cr 18.53"0.2 19.4(1)Mn 37.66"0.83 39.3(2)Fe 91.2"3.9 94(2)Co 25"0.59 26.0(1)Ni 58.1"2.7 59.8(1)Cu 20.5"3.8 23(1)Zn 72.48"0.65 77(2)As 56.02"0.73 54.2(3)Cd 6.47"0.37 6.3(2)Pb 18.15"0.64 18.8(3)

The estimated standard deviations are in parentheses.

multi-element lamps, except for V, As, Cd and Pb.Measurement of As content was carried out usingan electrodeless discharge lamps system(EDLSystem2) for the frequency modulation. Thegraphite tubes used for the analyses were usuallypyrolitic-type without platform (Perkin-Elmer),except in a few cases when measurements werecarried out with L’vov concave platform graphitefurnace tubes. No appreciable difference betweenthe results from the two types of furnace tubeswas detected. Specific interferences from matrixeffects were not observed in all of the samplesand the continuum background correction wassufficient for the detection of all the specificabsorptions. A solution of ultrapure magnesiumnitrate as modifier matrix was used for the analysisof Al and V (Mg(NO ) , 0.01 mgyml, for 20 ml3 2

of sample). No modifier matrix was used for theanalysis of all the other elements.A stream of ultrapure argon(Sol, Pisa, Italy) at

3.5 bar was used for the cooling and washingprocedures of the graphite tube.Temperature program, monochromator width slit

and eventual use of modifier matrix were obtainedby applying the trial and error technique for eachelement, with standard solutions as reference. Thestandard parameters published by SIH(ItalianHealth Ministry, 2000) and Perkin-Elmer(Perkin-

Elmer Editor, 1984, 1997a) were used as startingvalues. The parameters used for all the elementsanalyzed are listed in Table IS(Supporting Mate-rial). The software used for data collection andprocessing wasWINLAB version 3.2(Perkin-ElmerEditor, 1997b).Some samples were analyzed by Polymed(Sam-

buca, Florence, Italy, via AAS) and Mikro-analytishes Labor Pascher(Remagen-Bandorf,Germany, via HG-ICP-MS) for comparativepurposes.To check whether our analytical results agreed

with standard reference materials, SRM 1643d wasanalyzed utilizing the same analytical proceduresas that for the unknown samples. The observedvalues are in good agreement with the certifiedvalues(Table 1).

2.4. Gravimetric and volumetric analysis

2.4.1. Sulfate and chloride analysisThe analysis of sulfate and chloride anions was

based on standard methods(IRSA (CNR), 1994).

2.4.2. Total-hardness analysisThe analysis was based on standard techniques

(IRSA (CNR), 1994) as follows. One milliliter ofammonia(25%) was added to 50 ml of sample.The solution was treated with one specific pelletof indicator-buffer(Section 2.1) for the determi-nation of the total hardness of water. A red-winecolor was obtained. A volumetric titration throughEDTA (0.0100 M) was carried out up to the finalpoint, which was evidenced by a dark-green color.

2.4.3. Calcium-hardness analysisThirty drops of NaOH(1:1 mym) were added

to 50 ml of sample. The solution was shaken for2 min in order to precipitate Mg(OH) . Hydroxy-2

naphthol blue was added under stirring until apink-violet color was obtained. The suspensionwas rapidly titrated by adding EDTA(0.0100 M)to the end point, characterized by a sky-blue color.The suspension was let aside for 1 min in order tocheck the stability of the color.

45G. Tamasi, R. Cini / Science of the Total Environment 327 (2004) 41–51

Table 2Identification code, location site and selected physical and chemical data of the samples

Sampling site FC T pH x TH CCa CMg CSO4 CCl

(lys) (8C) (mSycm) (8F) (mgyl) (mgyl) (mgyl) (mgyl)

1 Prato di Sant’Agostino – 15.5 6.9 804.0 29.3 90.0 17.3 77.8 18.4Siena

2 Piazza Castello – 12.0 7.2 94.8 3.1 7.2 3.1 – –Piancastagnaio

3 Cerro del Tasca Spring 10.0 10.1 7.0 81.0 2.2 5.9 1.8 4.1 7.1Piancastagnaio

4 Saragiolo – 8.0 7.1 110.2 3.1 7.2 3.1 – –Piancastagnaio

5 Ermicciolo Spring Vivo D’Orcia 120 7.9 7.3 94.8 2.5 6.6 2.0 4.4 7.0Castiglione D’Orcia

6 Fonte Dell’Oro Spring Vivo D’Orcia 2.0 8.4 7.1 76.0 3.7 10.9 1.9 7.7 –Castiglione d’Orcia

7 Acqua Gialla I Spring Vivo D’Orcia 1.0 8.2 6.6 78.0 3.3 10.6 1.4 5.3 –Castiglione D’Orcia

8 Acqua Gialla III Spring Vivo D’Orcia 0.5 7.8 6.3 79.0 2.5 8.5 0.9 6.6 –Castiglione D’Orcia

9 Via Roma – 9.0 7.0 84.5 2.6 6.2 2.5 – –Abbadia San Salvatore

10 Saragio Spring 20.0 7.9 6.7 91.6 2.8 8.2 1.5 5.6 6.0Abbadia San Salvatore

11 Pian dei Renai well 60.0 10.0 7.0 83.0 3.3 9.0 2.5 12.1 5.8Abbadia San Salvatore

12 Fiora Spring 650 11.2 7.2 100.0 2.8 7.6 2.3 4.1 7.0Santa Fiora (Grosseto)

The samples are from the public water web unless otherwise specified. The estimated errors are: flow capacity(FC,"0.1–1.0lys); temperature(T, "0.1 8C); pH ("0.1); conductivity (x, "0.1 mSycm); total hardness(TH, "0.1 8F); concentrations ofspecies(C , "0.1 mgyl).x

3. Results and discussion

3.1. Analytical data

The location of the sampling sites and ancillaryexperimental data are reported in Table 2 and Fig.1, whereas the concentrations of selected elementsare listed in Table 3. The standard deviations,which were based on six replicates, are reportedin parenthesis throughout the text and the tablestogether with the actual MAC values approved bythe Italian law.

3.1.1. Ermicciolo spring and Pian dei Renai wellThe Ermicciolo spring(5) is located at Vivo

d’Orcia (Castiglione d’Orcia Municipality), 60 kmsouth of Siena, in the north declivity of MountAmiata, at 870 m above sea level. The mean flowcapacity (FC) of the spring was 120 lys (total

hardness, TH, 2.48F (8Fsmg of CaCO equiva-3

lenty100 ml); C , 6.6; C , 2.0; C -1.5 mgyCa Mg SO4

l). After mixing with minor flows from othersources(Fonte dell’Oro, FC, 2 lys; Acqua GiallaI, FC, 1 lys; Saragio, FC, 20 lys, TH, 2.88F; C ,Ca

8.2; C , 1.5 mgyl) and from a well(Pian deiMg

Renai, FC, 60 lys; TH, 3.38F; C , 9.0; C , 2.5;Ca Mg

C , 1.8 mgyl), the water of this spring feeds theSO4

Acquedotto del Vivo, which supplies Siena(inpart) as well as most of its southern territorybetween the spring and the town. The mean con-centration of the selected analyzed elements forthe Ermicciolo spring was(Table 3) V, 0.7(1); Cr,-0.01; Mn, 0.20(5); Co, 1.4(1); Cu, 3.5(2); Zn,130.7(3); As, 4.2(2), Cd,-0.10; Pb,-1.0mgyl.The content of all the elements was below therespective MAC values. However, the concentra-tion of As was relatively high(MAC 10 mgylAs

after December 2003, 50mgyl at present). Previ-

46 G. Tamasi, R. Cini / Science of the Total Environment 327 (2004) 41–51

Fig. 1. The maps represent(a) province of Siena as part of the entire Tuscany;(b) the Mt. Amiata area with the locations of mostof the sample collection sites,m.

ously reported works claim that the prolongeduptake of drinking water at C 3mgyl signifi-As

cantly increases the risk of cancer in adult popu-lation (Hogue, 2001a,b). Moreover, C (valueAs

found by Agenzia Regionale Protezione AmbienteToscana(ARPAT) (ARPAT, 1999)) was as highas 14.0mgyl in samples collected on October 1999at Ermicciolo. These results recommend a carefulmonitoring of the spring as well as a propertreatment of its water, in order to decrease theconcentration of As before the water enters theaqueduct web. The search for effective and inex-pensive adsorbents of arsenic-containing com-pounds in waters is increasing in attentionworldwide (Xu et al., 2002).Smaller springs close to Ermicciolo, like Fonte

dell’Oro (6) (C , 2.0(2) mgyl), Acqua Gialla IAs

(7) (2.3(3)), Saragio (10) (3.7(3)), had lowerconcentrations of As. The water from the Pian deiRenai well(11) had C 3.9(4) mgyl almost equalAs

to that from Ermicciolo. The latter water had

elevated Fe(4934(180) mgyl; MAC , 200) andFe

Mn (48(2); MAC , 50). This water undergoes aMn

complex treatment for decreasing the content ofFe and Mn before entering the water web(man-aged by Acquedotto del Fiora, Spa, Grosseto,Italy). The concentration of Ni(11.3(5) mgyl,MAC , 20) was relatively high, but no specificNi

treatment is conducted for this element. Neverthe-less, Ni concentration at the water web(see, forinstance, samples 1, 9) was much lower than theDL (detection limit), and the concentration of Vwas also low(0.30(5) mgyl). The concentrationof Pb for all the waters from these sources wasless than or equal to 1.3mgyl (MAC , 10).Pb

3.1.2. Cerro del Tasca springThe Cerro del Tasca spring(3) is located in

Mount Amiata, Piancastagnaio Municipality, northof the town of Piancastagnaio, at 870 m above sealevel. The FC value was approximately 10 lys(TH, 2.2 8F; C , 5.9; C , 1.8 mgyl) and theCa Mg

47G

.Tam

asi,R

.C

ini/

Scienceof

theTotal

Environm

ent327

(2004)41–51

Table 3Concentration of the elements(mgyl) in the drinking waters examined in this work via atomic absorption spectroscopy

Sample Element Al V Cr Mn Fe Co Ni Cu Zn As Cd PbDL 1.0 5.0 0.3 0.3 1.0 1.0 2.0 0.5 0.5 1.0 0.10 1.0MAC 200 50 50 50 50 – 20 1000 3000 10 5 10

1a 8.9(5) 1.0(1) -0.007 0.8(1) 113(5) -0.025 -0.05 2.8(4) 18.5(9) 3.9(2) -DL 2.0(2)2 5.3(2) 1.4(1) 0.8(1) 0.9(2) 10(1) 0.008(2) 0.150(2) 24(2) 200(4) 7.4(2) 0.003(1) 0.071(2)3a 2.8(3) 1.8(1) 0.8(1) -0.007 4.0(2) 0.15(1) 0.30(3) 1.0(1) 26.8(5) - -DL 1.3(2)4 4.8(2) -DL 0.6(1) 3.2(1) 43(5) -DL -DL -DL 187(3) 9.0(4) -DL -DL5a 7.7(4) 0.7(1) -0.01 0.20(5) 85(2) 1.4(1) -0.08 3.5(2) 130.7(3) 4.2(2) -DL -DL6 8.5(7) -DL -DL -DL 194(12) -DL -DL 1.5(2) 254(27) 2.0(2) -DL -DL7 4.0(3) -DL -DL 26(1) 534(17) -DL -DL -DL 221(9) 2.3(3) -DL -DL8 14.3(5) -DL -DL 27.3(3) 697.7(1) -DL -DL 4.5(1) 28(2) -DL -DL -DL9 16.4(6) -DL 0.5(1) 6.6(3) 120(11) -DL -DL -DL 23.5(4) – -DL -DL10 5.1(2) -DL -DL -DL 67(6) -DL -DL 2.8(2) 48(4) 3.7(3) -DL -DL11a 6.9(3) 0.30(5) -DL 48(2) 4934(180) 0.06(1) 11.3(5) 5.0(5) 109(4) 3.9(4) -DL -DL12a 3.9(5) 1.8(1) 0.5(1) -0.007 18.7(1) -0.025 0.8(1) -0.012 0.7(1) 14.4(2)b -DL -DL

The estimated standard deviations are in parentheses. DL, detection limit; MAC, maximum allowable concentration, according to the enforced law(Decreto delPresidente della Repubblica, 1988, n. 183, Decreto Legge, 2001, n. 31, Decreto Ministeriale, 1999, Italian law). Three samples were collected for each site, and sixreplicates were carried out for each sample(see text).

The samples were concentrated before the analysis. The concentration factors are: 1, 40; 3, 41; 5, 28; 11, 31; 12, 40.a

The value has been determined from samples collected at a supplier of the water web at Santa Fiora, Castle, approximately 0.5 km from the spring.b

48 G. Tamasi, R. Cini / Science of the Total Environment 327 (2004) 41–51

water is mixed with lower flow springs nearby.The concentration of V for 3 was 1.8(1) mgyl(MAC , 50), which corresponds to the highestV

value found among the Mount Amiata sourceslocated in the Siena district, but is considerablyless than C found previously for some drinkingV

waters from Mount Etna(as high as 150mgyl)(University of Catania and Italian Health Ministry,1995; Giammanco et al., 1996, 1998).The concentration of As in the water collected

at the terminal of the water web at Piazza Castello(2, Piancastagnaio Municipality) was 7.4(2) mgyl, higher than the value found for Ermicciolo spring(5, see above). The concentrations of Cr, Co andNi in the water from Cerro del Tasca spring(3)were 0.8(1), 0.15(1) and 0.30(3) mgyl, respective-ly. Similarly, the concentrations of all the otherelements analyzed were far below the respectiveMAC values(Table 3). The arsenic concentrationin samples from the terminal of the water web atSaragiolo(4 km north–west of Piancastagnaio, 4)was 9.0(4) mgyl, in good agreement with thevalue found in sample 2. The quality of thesewaters is good based on the content of the metalsanalyzed, but the concentration of As is relativelyhigh and should be decreased.The concentrations of Zn(200(4)) and Cu

(24(2) mgyl) found at the delivery site of thewater web located at Piazza Castello(2) werehigher than those measured at the sources, proba-bly because of leaching of these elements fromthe metal pipes.

3.1.3. Santa Fiora springThe Santa Fiora spring(12) has the largest FC

value(650 lys) among the Mount Amiata sources.It is located in the Santa Fiora Municipality,Grosseto district, at 687 m above sea level(TH,2.8 8F; C , 7.6; C , 2.3 mgyl). This springCa Mg

supplies drinking water to a large part of thepopulation of the district and in part feeds thewater web of Piancastagnaio Municipality(Siena).The concentrations of all the metals analyzed werefar below the respective MAC values. For exam-ple, the concentrations of V, Cr, Co and Ni were,respectively, 1.8(1), 0.5(1), -0.025 and 0.8(1)mgyl. However, the concentration of As at aterminal of the water web in Santa Fiora was

14.4(2) mgyl, which is the highest value of thiselement found in this study. Except for the Ascontent, sample 12 water is high quality and issimilar to water from Cerro del Tasca spring(3).

3.1.4. Drinking water from the water web of SienaMunicipalityApproximately 35% of the water in the aqueduct

of Siena Municipality comes fromAcquedotto delVivo (sources from Mount Amiata, see above) andapproximately 65% from the sources located inthe Sovicille Municipality (Luco wells, see Fig.1a). Samples collected from this water web camefrom three public delivering sites in downtownSiena (1) (TH, 29.3 8F; C , 90.0; C , 17.3;Ca Mg

C , 77.8 mgyl). The mean concentration of AsSO4

was 3.9(2) mgyl. This level agrees well withconcentrations of As found in the samples fromMount Amiata(see above) and the concentrationof As found by ARPAT in the Luco well waters(ARPAT, 1999). The concentrations of selectedelements(V, 1.0(1); Cr,-0.007; Mn, 0.8(1); Fe,113(5); Co,-0.025; Cu, 2.8(4); Zn, 18.5(9); Cd,-0.10; Pb, 2.0(2) mgyl) were below the respec-tive MAC values. However, the concentration ofFe and Pb was relatively high(MAC , 200 mgyFe

l; MAC , 10) probably due to leaching of bothPb

elements from the pipe and to a high Fe concen-tration from the Luco wells. In fact, the content ofPb at the springsywells as determined by this work(5) or by others(ARPAT, 1999) was always muchlower than that measured in samples collected atthe delivering site of the water web(1).

3.2. The metal index, MI, parameter

An MI for drinking water was preliminarilydefined as

N CiMIs8(MAC)iis1

whereC is the concentration of each element insolution andi is the ith sample.The metalloid As was included in the calculation

of MI. As no MAC is defined by law, the valueCo

of 50 mgyl was assumed. For the elements whoseconcentration has been determined as-DL, Cs

49G. Tamasi, R. Cini / Science of the Total Environment 327 (2004) 41–51

Table 4Values of MI for the samples analyzed in this work

Sample MI

1 1.272 1.003 0.26a

4 1.625 1.116 1.657 3.868 4.579 1.17a

10 1.1011 26.71b

12 1.77

Arsenic is not included.a

Samples collected before the treatment to decrease the con-b

tent of Fe and Mn.

DL was assumed. Therefore, MI was defined toevaluate the quality of the water based on thecontent of 11 heavy metals and As. The higherthe concentration of a metal compared to itsrespective MAC value, the worse the quality ofthe water. If the concentration of a certain elementis higher than the respective MAC value(MI)1),the water cannot be used according to law. Thepresence of several elements with concentrationssmaller than but close to the respective MACvalues will also decrease the overall quality ofwater because of an additive effect. Thus, an MIvalue)1 is a threshold of warning, even in thecase whereC -MAC for all the elements.i i

For instance, sample 5(Ermicciolo spring) hadan MI value of 1.11(Table 4), which came fromthe additive contribution of several elements. Inthis case, As, and to a lesser extent Fe, contributedsignificantly to the index. The MI value for theAcqua Gialla I spring water(7) was 3.86, clearlyindicating a low-quality water. For this spring thehighest contributions to MI came from Fe(534(17) mgyl, MAC , 200) and Mn (26(1);Fe

MAC , 50), whereas the content of As(2.3(3))Mn

was lower than that in Ermicciolo spring water.The Acqua Gialla I spring water is mixed withstream water from Ermicciolo spring to decreaseconcentrations of Fe and Mn allowing the waterto contribute to the water web, as all theC valuesi

were below the respective threshold parameters

(MAC ). As a consequence the water from Ermic-i

ciolo spring(MI, 1.11) was worsened by a com-puted factor of 1.8% because of the addition ofAcqua Gialla I spring water(MI, 3.86), eventhough the latter had a small FC(1 lys) andincreased the stream capacity by 0.8%.The aqueduct of Siena municipality had an MI

1.27, mostly because of the content of Fe(113(5)),As (3.9(2)), Pb (2.0(2) mgyl).

4. Conclusion

This work determined the concentration of 11metals and the metalloid As in the waters fromsome of the springs, wells and water webs of theMount Amiata area and downtown Siena in theperiod 2000–2002. The most significant resultsfrom the investigation were the following:(a) Themetal V was measured and reported for the firsttime in this work. The highest C was 1.8mgylV

(a value much below MAC) in samples fromV

Cerro del Tasca(3) and Santa Fiora(12) springs(Mount Amiata). Therefore, V is not a humanhealth concern as it is in the Mount Etna area(University of Catania and Italian Health Ministry,1995; Giammanco et al., 1996, 1998). (b) Theconcentration of As was relatively high for thesamples analyzed, especially in light of recentwarnings from several reported investigations.Even if C is lower than the MAC , the contentAs As

of As for these drinking waters should bedecreased.(c) The relatively high content of Asfor all the water samples from sources located inthe Mount Amiata area was due to mineral depositsin the aquifer and not derived from human pollu-tion (Dall’Aglio et al., 2001; RIMIN Spa-ENI,1990; Protano et al., 1998). (d) An index, namelyMI, to evaluate the overall quality of drinkingwaters based on 12 elements is proposed as apreliminary basic concept that should be improvedand refined. For instance, the inclusion of manymore elements is mandatory.(e) Significant dif-ferences have been found in the concentrations ofsome elements(mainly Fe, Cu, Zn and Pb)between samples collected at the sources and thosecollected at the terminals of the water webs.The findings suggested the need to increase the

number of sampling sites, especially at the termin-

50 G. Tamasi, R. Cini / Science of the Total Environment 327 (2004) 41–51

als of the water web, and that the pipes and depotsof the water webs should be renewed. Determiningthe concentration of Hg is planned for future workbecause of the toxicity of that metal and the well-known presence of large deposits of cinnabar(HgS) at Mount Amiata, even though preliminaryexperimental determinations have determined aconcentration of 0.028(4) mgyl, far below theMAC (1.0 mgyl). A study devoted to a specia-Hg

tion analysis of the elements is also planned.

Acknowledgments

The authors thank University of Siena for fund-ing and Consorzio Interuniversitario per la Ricercanella Chimica dei Metalli nei Sistemi Biologici(CIRCMSB, Bari) for grants to Gabriella Tamasifor the years 2002 and 2003. Professor R.F. Mar-colongo, University of Siena, is gratefullyacknowledged for stimulating the interest on toxicelements in drinking waters.

References

Amministrazione Provinciale di Siena. Aqueduct Plans for theProvince of Siena(1992–1994). Lalli Editore, Italy, 1994(Italian).

ARPAT (Agenzia Regionale per la Protezione Ambientale perla Toscana). Analytical certificates by Environmental Pro-tection Agency for Tuscany, Siena District Department.Samples collected in the Sovicille Municipality from Luco1, Luco 2, Luco 4, Luco 5 wells on February 9th 1999,March 16th 1999, April 13th 1999, May 18th 1999, June16th 1999, July 20th 1999, August 10th 1999, September14th 1999, October 19th 1999, November 9th 1999. Samplescollected in the Siena Municipality from Ente Uscita andErmicciolo springs on, January 26th 1999, February 23th1999, July 13th 1999, October 12th 1999, December 15th1999(Italian).

Barceloux DG. Cobalt. J Toxicol Clin Toxicol 1999;37:201–206.

Carmignani L, Kligfield R. Crustal extension in the NorthernApennines: the transition from compression to extension inthe Alpi Apuane core complex. Tectonics 1990;9:1275–1303.

Cini R, Chindamo D, Catenaccio M, Lorenzini S, Selvi E,Nerucci F, Picchi MP, Berti G, Marcolongo R. Dissolutionof calcium pyrophosphate crystals by polyphosphates: an invitro and ex vivo study. Ann Rheum Dis 2001;60:962–967.

Cini R, Pifferi C. Supramolecular networks via hydrogenbonding and stacking interactions for adenosine 59-diphos-phate. Synthesis and crystal structure of diaqua(2,29:69,29-terpyridine)copper(II) wadenosine 59-diphosphato(3-)x

(2,29:69,20-terpyridine)cuprate(II) adenosine 59-diphospha-te(1-) hexadecahydrate and density functional geometryoptimization analysis of copper(II)- and zinc(II)-pyrophos-phate complexes. J Chem Soc Dalton Trans 1999;699–710.

Cini R, Tamasi G. Heavy metal and arsenic intake fromdrinking water and autoimmune rheumatic diseases. TUMA2002, University of Urbino, July 15–17, 2002, Italy.

Dall’Aglio M, Barbieri M, Bazzoli P, Venanzi G, Agati L,Mantelli F, Palmieri R. Drinking water quality in theGrosseto Province, Tuscany, Italy. An appraisal on the basisof the study of water rock interaction. Water Rock Inter2001;2:1059–1062.

Decreto Legge(D.L.) February 2nd 2001, n. 31. Accomplish-ment of the EC instruction n. 98y83yEC, related to waterfor human uses(Italian).

Decreto Ministeriale(D.M.) November 10th 1999, art. 1.Maximum allowable concentration for Vanadium in waterfor human uses(Italian).

Decreto del Presidente della Repubblica(D.P.R.) May 24th1988, n. 236. Accomplishment of the European Community(EC) instruction n. 80y778EC related to water for humanuses, art. 15 law April 16th 1987, n. 183(Italian).

Defazio S, Cini R. Synthesis, X-ray structural characterizationand molecular modeling analysis of cobalt(II), nickel(II),zinc(II) and cadmium(II) complexes of the widely usedanti-inflammatory drug meloxicam. J Chem Soc DaltonTrans 2002;1888–1897.

Fewtrell L, Kay D, MacGill S. A review of the science behinddrinking water standards for copper. Int J Environ HealthRes 2001;11:161–167.

Flaten TP. Aluminum as a risk factor in Alzheimer’s disease,with emphasis on drinking water. Brain Res Bull2001;55:187–196.

Gauthier E, Fortier I, Courchesne F, Pepin P, Montimer J,Gauvreau D. Aluminum forms in drinking water and riskof Alzheimer’s disease. Environ Res 2000;84:234–246.

Giammanco S, Valenza M, Pignato S, Giammanco G. Mg, Fe,Mn and V concentrations in the groundwaters of MountEtna(Sicily). Wat Res 1996;30:378–386.

Giammanco S, Ottaviani M, Valenza M, Veschetti E, PrincipioE, Giammanco G, Pignato S. Major and trace elementsgeochemistry in the ground waters of a volcanic area: MountEtna(Sicily). Wat Res 1998;32:19–30.

Grassi S, Netti R. Sea water intrusion and mercury pollutionof some coastal aquifers in the province of Grosseto(south-ern Tuscany—Italy). J Hydrology 2000;237:198–211 (andreferences therein).

Hogue C. A sip of arsenic? Debate over allowable level indrinking water focuses on health effects, removal cost.Chem Eng News 2001a;79:50–55.

Hogue C. Cancer risk from arsenic confirmed. New studystrengthen evidence for lower drinking water standard.Chem Eng News 2001b;79:12.

IRSA (CNR), Analytical methods for water. Volume 100.Polygraphic Institute of the Italian Government Ed., Roma1994(Italian).

51G. Tamasi, R. Cini / Science of the Total Environment 327 (2004) 41–51

Italian Health Ministry, Ottavini M and Bonadonna L, Envi-ronmental Hygiene Laboratory. Analytical methods forwaters for human uses. Vol. II, Part 1, Chemical Methods.Roma 2000(Italian).

Lind Y, Glynn AW. Intestinal absorption of copper fromdrinking water containing fulvic acids and an infant formulamixture studied in a suckling rat model. Biometals1999;12:181–187.

Manay N, Pereira L, Cousillas Z. Lead contamination inUruguay. Rev Environ Contam Toxicol 1999;159:25–39.

Nordberg G. Assessment of risks in occupational cobalt expo-sures. Sci Total Environ 1994;150:201–207.

Nordstrom DK. Worldwide occurrences of arsenic in groundwater. Science 2002;296:2143–2145.

Pedersen LM, Permin H. Rheumatic disease, heavy-metalpigments, and the great masters. Lancet 1988;331:1267–1269.

Perkin-Elmer Editor. AAnalyst 100. Analytical techniques forgraphite furnace atomic absorption spectrometry. Boden-seewerk Perkin-Elmer GmbH, 1984.

Perkin-Elmer Editor. HGA-800, Graphite furnace system.Hardware guide. Bodenseewerk Perkin-Elmer GmbH 1997.

Perkin-Elmer Editor. SoftwareWINLAB version 3.2. AAnalyst100, Atomic absorption spectrometry. Bodenseewerk Perkin-Elmer GmbH 1997.

Protano G, Riccobono F, Sabatini G. La cartografia geochimicadella Toscana meridionale. Criteri di realizzazione e rilevan-za ambientale attraverso gli esempi di Hg, As, Sb, Pb e Cd.Mem. Descr. Carta Geol. d’Italia. 1998; LV: 109–140(Italian).

Ramessur RT, Parry SJ, Ramieawon T. The relationship ofdissolved Pb to some dissolved trace metals(Al, Cr, Mn,and Zn) and to dissolved nitrate and phosphate in afreshwater aquatic system in Mauritius. Environ Int2001;26:223–230.

RIMIN Spa–ENI, General Management of Mines, SouthernTuscany 3. Final Report on the Works Done within theAgreement between the Italian Ministry of Industry andCommerce, and ENI(Hydrocarbon National Organization),1990(Italian).

Senn DB, Hemond HF. Nitrate controls on iron and arsenic inan urban lake. Science 2002;296:2373–2376.

Serri G, Innocenti F, Manetti P. Geochemical and petrologicalevidence of the subduction of delaminated Adriatic conti-nental lithosphere in the genesis of the Neogene-Quaternarymagmatism of Central Italy. Tectonophysics 1993;223:117–147.

Smith AH, Lopipero PA, Bates MN, Steinmans CM. Arsenicepidemiology and drinking water standards. Science2002;296:2145–2146.

University of Catania, Hygiene and Preventive Medicine Insti-tute, and Italian Health Ministry. July 15th 1995; andreferences therein(Italian).

US EPA, Office of Water. Drinking water regulation and healthadvisories, 2000.

US EPA, Office of Water. National primary drinking waterregulation. Arsenic and clarifications to compliance and newsource contaminants monitoring. Register Federal 22 May2001:66.

Veeramachaneni DN, Palmer JS, Amann RP. Long-term effectson male reproduction of early exposure to common chemicalcontaminants in drinking water. Hum Reprod 2001;16:979–987.

WHO. Air quality guidelines for Europe, regional publicationEuropean series, 2000;ch.6.12.

Xu YH, Nakajima T, Ohki A. Adsorption and removal ofarsenic(V) from drinking water by aluminum-loaded Shir-asu-zeolite. J Hazard Mat B 2002;92:275–287 (and refer-ences therein).