ORIGINAL ARTICLE

Geological sources of boron and fluoride anomaliesin Silurian–Ordovician aquifer system, Estonia

Marge Uppin • Enn Karro

Received: 16 December 2010 / Accepted: 17 September 2011 / Published online: 1 October 2011

� Springer-Verlag 2011

Abstract The objective of this study was to examine the

possible natural sources of fluorides and boron in Silurian–

Ordovician (S–O) aquifer system, as the anomaly of these

elements has been distinguished in groundwater of western

Estonia. Water–rock interactions, such as dissolution and

leaching of the host rock, are considered to be the main

source of high fluoride and boron concentrations in

groundwater. Altogether 91 rock samples were analysed to

determine if high F- and B levels in groundwater could be

attributed to certain aquifer forming rock types. Fluorine

and boron contents in limestones and dolomites vary from

100 to 500 mg/kg and 5 to 20 mg/kg, reaching up to 1,000

and 150 mg/kg in marlstones, respectively. K-bentonites,

altered volcanic ash beds, are rich in fluorine (400–

4,500 mg/kg) and boron (50–1,000 mg/kg). Thus, clay-rich

sediments, providing ion-exchange and adsorption sites for

F- and B, are the probable sources of both elements in S–O

aquifer system in western Estonia.

Keywords Boron � Fluorine � Carbonate rocks �Silurian–Ordovician aquifer system � Estonia

Introduction

The occurrence of fluorine (F) and boron (B) in ground-

water has drawn worldwide attention due to their important

physiological role in the health of man. Fluorine is a

ubiquitous element in continental crust, with an average

concentration of 625 ppm (Edmunds and Smedley 2005). It

is the most electronegative and reactive element which

occurs in water primarily as a negatively charged ion—

fluoride (F-) (Hem 1985). Fluoride is mainly released into

the groundwater through weathering and leaching of the

fluorine-containing minerals in rocks. The most common

fluorine-bearing mineral in the geological environment is

fluorite (CaF2). Fluorine is also abundant in various other

rock-forming minerals, for example, apatite, micas,

amphiboles, and certain clay minerals (Hem 1985; Saxena

and Ahmed 2003; Sujatha 2003; Edmunds and Smedley

2005; Rafique et al. 2008; Naseem et al. 2010). Because of

similar ionic radii, fluoride can easily replace hydroxyl ion

(OH-) in many rock-forming minerals. High fluorine

contents are associated with volcanic activities. Thus,

elevated concentrations of F- are often found in geother-

mal waters (Kundu et al. 2001; Desbarats 2009). Besides,

fresh volcanic ash may be rather rich in fluoride, and ash

interbedded with other sediments could contribute later to

high fluoride concentrations in groundwater in such areas

(Hem 1985). Therefore, high F- contents in water could be

found in the areas, where F-bearing minerals are abundant

in the rocks.

Small doses of fluoride have beneficial effect on teeth by

hardening the enamel and reducing the incidence of caries.

On the other hand, concentrations of fluoride above

1.5 mg/l in drinking water may cause dental fluorosis

(Grobler et al. 1986; ADA 2001; Billings et al. 2004; WHO

2008). According to the EU directive 98/83/Council

directive 98/83/EC (1998) and Estonian requirements for

drinking water quality (Joogivee 2001), the limit value for

fluoride is 1.5 mg/l.

Boron (B) is widely distributed in nature with average

concentration of 10 ppm in the earth’s crust (Yazbeck et al.

2005). It is a minor dissolved constituent of groundwater

where it is often associated with F- (Queste et al. 2001;

M. Uppin (&) � E. Karro

Department of Geology, Institute of Ecology and Earth Sciences,

University of Tartu, Ravila 14A, 50411 Tartu, Estonia

e-mail: marge13@ut.ee

123

Environ Earth Sci (2012) 65:1147–1156

DOI 10.1007/s12665-011-1363-7

Earle and Krogh 2004; Desbarats 2009). In aqueous solu-

tions, boron is mostly present as B(OH)4- ion and undis-

sociated boric acid B(OH)3 (Mather and Porteous 2001;

Gonfiantini and Pennisi 2006; Pennisi et al. 2006), and the

occurrence of these species in environment is controlled by

the pH. Boron is mainly released into the groundwater

through the weathering of rocks, seawater intrusion into

aquifers, and volcanic activity (Hem 1985; Pennisi et al.

2006). Therefore, the leaching of rocks and soils is

responsible for the most of the boron dissolved in natural

waters. Boron undergoes a number of interactions with the

aquifer-forming rocks, among which the process of

adsorption and desorption of boron to clay mineral surfaces

plays a crucial role (Gonfiantini and Pennisi 2006; Pennisi

et al. 2006).

High boron levels in drinking water are considered to

bear risks for human health. The World Health Organisa-

tion guideline value for boron is 0.5 mg/l (WHO 2008).

However, the limit value for boron in drinking water is

1.0 mg/l in Estonia (Council directive 98/83/EC 1998;

Joogivee 2001). The long-term experiments with animals

have shown that excessive content of boron leads to

reduced fertility and sterility, low weight of foetus, me-

tabolistic disorders and acute neurological effects (Heindel

et al. 1992; Price et al. 1996; WHO 2008). However,

human data are lacking, the only available data show that

acute exposure is associated with short-term irritant effects

of the upper respiratory tract (Garabrant et al. 1985;

Wegman et al. 1994).

Silurian–Ordovician (S–O) aquifer system provides 30%

of Estonian public water supply, being the most important

drinking water source in central and western Estonia.

Previous hydrochemical studies have shown that high

fluoride and boron values in S–O aquifer system are mostly

present in western Estonia (Karro et al. 2006, 2009; Karro

and Uppin 2010). The natural concentrations on boron and

fluorides in groundwater reach up to 2.1 and 7.2 mg/l,

respectively. The health risks in Estonia arising from

drinking water are mainly due to the high levels of fluoride

and boron in groundwater (Saava 1998; Indermitte et al.

2006; Indermitte 2010). A strong relationship between the

fluoride content in drinking water and the prevalence of

dental fluorosis among 12-year-old schoolchildren in

Estonia has been found (Indermitte et al. 2007).

The groundwater chemistry is controlled among the

other factors by the lithological composition of water-

bearing rocks. Presence of fluorine and boron-bearing

minerals in the host rock and their interaction with water

are considered to be the main causes for fluoride and boron

enrichment in groundwater. The aim of this paper is to

present the results of geochemical and mineralogical study

of Silurian and Ordovician carbonate rocks as the natural

fluorine and boron sources in Estonia, focusing on the

occurrence and variability of F and B contents in different

rock types. Furthermore, the other possible geological

reasons of the fluoride and boron anomalies in S–O aquifer

system in western Estonia are discussed. As the collected

rock samples originate from different geological units of

Silurian and Ordovician System, and in places the number

of samples was taken along the vertical geological cross

section; the spatial distribution of fluorine and boron in

Silurian and Ordovician rocks has not been studied in this

research. To perform the throughout and representative

geochemical mapping of F and B, denser sampling set

should be designed.

Study area

Estonia is situated in the north-western part of the East-

European Platform. Its sedimentary beds, lying on the

southern slope of the Baltic Shield, are declined south-

wards at about 3–4 m per km. The Estonian Paleoprote-

rozoic basement is overlaid by Neoproterozoic (Ediacaran)

and Palaeozoic (Cambrian, Ordovician, Silurian and

Devonian) sedimentary rocks (Fig. 1) covered by Quater-

nary deposits. Both Ordovician and Silurian sequences in

Estonia consist mainly of shallow water carbonates—

limestone, dolomite, and marlstone with clayey interlayers.

Exceptionally, the basal part of Ordovician sequence is

represented by terrigenous sediments—silty and clayey

sandstones and graptolite argillites.

Two main structural elements of the Baltic Ordovician

Basin are distributed in Estonia: the marginal confacies belt

(North Estonian Confacies Belt) and the central confacies

belt (Central Baltoscandian Confacies Belt, including the

Livonian Tongue). A transitional zone between the above

mentioned belts has also been distinguished (Fig. 2a). The

marginal belt was dominated by shallow water carbonate

sediments with a lot of discontinuity surfaces, while the

relatively deeper water central belt comprised predomi-

nantly clayey sediments such as clayey limestones and

marlstones (Nestor and Einasto 1997). The Ordovician

System in Estonia is divided into three series: Lower-,

Middle- and Upper Ordovician, which are subdivided into

different regional stages (Table 1).

During the Silurian, relatively shallow water sedimen-

tary conditions prevailed in Estonia. Based on the facies

changes of the Silurian rocks, the Mid-Estonian and

South-Estonian confacies belts have been distinguished.

The Mid-Estonian Confacies Belt is dominated by various

limestones and dolomites. The belt covers the islands of

West-Estonian Archipelago and the western and central

parts of mainland Estonia. The South-Estonian Confacies

Belt consists mostly of marlstones. The composition of

sediments, mainly clay content, reflects the sea level

1148 Environ Earth Sci (2012) 65:1147–1156

123

changes of the Silurian Basin. The Silurian transgression

maximum coincides with Juuru, Adavere and Jaani Stage,

which corresponds to maximum amount of clay in car-

bonates. The proportion of clay content increases to

southwards or south-westerly (Fig. 2b) (Nestor and Ein-

asto 1997). The Silurian System is divided into four series

(Pridoli, Ludlow, Venlock, Llandovery) and ten regional

stages (Table 1).

The Silurian and Ordovician stratigraphic record in

Estonia and neighbouring areas contains numerous K-rich

altered volcanic ash beds—K-bentonites. The thickness of

bentonites ranges from some mm up to 10 cm, rarely up to

20–30 cm and more (Bergstrom et al. 1995; Kiipli et al.

2001; Hints et al. 2008). The composition of bentonite clay

matrix in the Baltic Basin is typically mixed-layer illite–

smectite with some amount of kaolinite (Kiipli et al. 2007;

Hints et al. 2008; Somelar et al. 2010), except of the

bentonites of Pirgu age which are characterized by chlorite-

smectite type mixed-layer minerals (Hints et al. 2006). The

interlayers of K-bentonite have been recorded in Jaagarahu,

Jaani, Adavere and Raikkula Stages of Silurian System.

Ordovician K-bentonite beds in Estonia are present only in

the Upper Ordovician Series: Kukruse, Haljala, Keila and

Pirgu Stages (Bergstrom et al. 1995; Kiipli et al. 2001,

2008; Hints et al. 2008; Somelar et al. 2010).

Ordovician and Silurian carbonate rocks (Fig. 1,

Table 1) form the S–O aquifer system, which is an

important and often the only source of drinking water in

central and western Estonia and on islands of the West-

Estonian archipelago. The upper part of the S–O aquifer

system with a thickness of 30 m is extremely cavernous,

with numerous cracks and fissures. Water in the fissure

systems and karst cavities of the carbonate bedrock flows

relatively fast. Silurian and Ordovician carbonate rocks

have fragmentary 1–2-m thick water-conducting zones

with parallel lamination and an abundance of fissures,

where groundwater flows in a lateral direction. These zones

are separated from each other by 5–10-m thick layers in

which groundwater flows predominantly in vertical

fissures. The caverns and fissures decrease in the deeper

Fig. 1 Location of the study

area with sampling sites

(number of samples are marked

in brackets) and geological

cross section of Estonia

Environ Earth Sci (2012) 65:1147–1156 1149

123

part of the S–O aquifer system and the aquifer system

transforms into aquitard (Perens and Vallner 1997).

The S–O aquifer system has a characteristic of HCO3–

Ca–Mg and HCO3–Mg–Ca water type with total dissolved

solids (TDS) mainly below 0.6 g/l in its upper 30–50-m

thick portion. In coastal areas and greater depths, the

content of Cl- and Na? in groundwater increases and

HCO3–Cl–Na–Mg–Ca water type with TDS between 0.3

and 1.5 g/l is widespread (Perens et al. 2001).

Materials and methods

Ninety-one samples were collected from drillcores and out-

crops during 2008 and 2009 to study the chemical

composition of the aquifer forming Silurian and Ordovician

rocks (Fig. 1). The samples were analysed for fluorine, boron,

and other major and minor chemical constituents (SiO2,

Al2O3, Fe2O3, MgO, CaO, Na2O, K2O, TiO2, P2O5, MnO,

Cr2O3, Ba, Sr, Cu, Ni, Co). Rock samples were crushed and

pulverized to 200 mesh and dried at 105�C prior to analysis.

The chemical composition of the rocks was determined using

standard ICP-ES and ICP-MS techniques, content of fluorine

was analysed by specific ion electrode at ACME Analytical

Laboratories Ltd., Vancouver, Canada.

The relative content of terrigenous (Ter), dolomite (Dol)

and calcite (Cal) fractions in rocks was calculated to

determine how F and B are related to different rock types.

The content of Ter, Dol and Cal component in carbonate

rocks was calculated according to the following formulas

taking into consideration that terrigenous material con-

tained on average 2.5% MgO (Kaljo et al. 1997):

Ter ¼ Al2O3 þ SiO2 þ Fe2O3 þ K2Oþ TiO2ð Þ � 1:025

Dol ¼ 4:57� MgO� 0:025� Terð ÞCal ¼ 1:786� CaO� 0:304� Dolð Þ

The results obtained were recalculated normalizing the sum

of Ter ? Cal ? Dol to 100%.

To identify the main minerals (quartz, calcite, dolomite,

apatite, K-feldspar, albite, illite, kaolinite, chlorite, pyrite,

biotite) of Silurian and Ordovician carbonate rocks, 10

unoriented powdered whole rock preparations were ana-

lysed by means of X-ray powder diffractometry (XRD).

XRD analyses were carried out at the Department of

Geology, Tartu University. The purpose of the XRD study

was to compare the calculated mineralogical composition

with the results of XRD determinations. The whole-rock

mineralogical composition by XRD fairly agrees with

calculated values. Thus, calculation methodology provided

by Kaljo et al. (1997) could be used in this study.

For data processing, interpretation and visualization

MapInfo professional 6.5, AquaChem 3.70 and Grapher 8

were used.

Results and discussion

Altogether the composition of 91 rock (79 carbonate rock

and 12 K-bentonite) samples all over the Estonia was

studied to determine the chemical composition of the

sedimentary rocks forming the S–O aquifer system

(Fig. 1). Carbonate rocks in this research are classified as

stated by Vingisaar et al. (1965) as follows: limestones,

dolomites, clayey limestones and dolomites, marlstones,

domerites, and clay. This simplified classification is based

on the clay content in carbonaceous rocks, and is used in all

tables and diagrams in this paper.

Fig. 2 a Late Ordovician facies belts (after Ainsaar and Meidla

2001). The hatched area is the transition zone according to Polma

(1982), dotted line is the northern limit of the distribution of

Ordovician sedimentary rocks. b Terrigenous sediments in Silurian

basin of Estonia (after Jurgenson 1988). Distribution of Silurian rocks

is marked with grey colour

1150 Environ Earth Sci (2012) 65:1147–1156

123

A positive correlation between dissolved fluoride, boron

and pH has been observed in Estonian groundwater (Karro

and Rosentau 2005; Karro et al. 2006), thus, it can be

assumed that both fluorine and boron have similar origin

and undergo similar geochemical reactions in S–O aquifer

system of Estonia. Previous groundwater studies have

shown that high fluoride (up to 7.2 mg/l) and boron

(2.1 mg/l) concentrations are widespread in western Esto-

nia (Karro and Rosentau 2005; Karro et al. 2006, 2009;

Karro and Uppin 2010). Groundwater in S–O aquifer sys-

tem is mainly HCO3–Ca–Mg type and owing to the high

Ca2? contents, quite low amounts of F- may be mobilised.

The highest F- concentrations prevail in wells, which

produce the water with low Ca2? content. The pH of

groundwater and the contents of Na? and Cl- increase with

depth and the groundwater changes towards HCO3–Na–Cl

chemical type. Thus, geochemically favourable conditions

for high dissolved F- in water prevail in deeper portions of

S–O aquifer system (Karro et al. 2006).

According to the results of the rock analyses, the same

pattern in distribution of F and B in aquifer-forming rocks

is not evident. The F and B contents in the rocks depend on

the composition and the type of the rock and show wide

concentration range all over the study area.

The average concentrations and some other descriptive

statistics of boron and fluorine in different rock types are

given in Table 2. The results of the present research show

that the total fluorine and boron content of limestones and

dolomites are 70–1,001 and 3–54 mg/kg, respectively,

being slightly higher in dolomites. Cal–Dol–clay ternary

diagram could be useful to illustrate how F and B con-

centrations vary in different rock types. Generally, the

increase of the clay content is followed by the increase in

the F and B concentrations in the carbonate rocks (Fig. 3).

K-bentonite samples as non-carbonate rocks are not pre-

sented in ternary diagrams.

The fluorine and boron contents in marlstones and

domerites reach up to 1,030 and 160 mg/kg, respectively,

Table 1 Stratigraphy and lithological characteristics of Estonian Ordovician and Silurian systems (after Raukas and Teedumae 1997)

System Series Stage Characteristic rocks Samples

Silurian Pridoli Ohesaare Domerite, limestone 49

Kaugatuma Marlstone, limestone 8

Ludlow Kuressaare Marlstone, clayey limestone

Paadla Limestone, dolomite 28; 50

Wenlock Rootsikula Clayey dolomite, limestone 3; 6

Jaagarahu Limestone, dolomite 20; 26; 27

Jaani Marlstone, domerite

Llandovery Adavere Marlstone, domerite 2; 55; 56; 58–66

Raikkula Limestone, dolomite 12; 13; 14; 23; 24; 25; 37

Juuru Clayey limestone, marlstone 15; 19; 21; 31

Ordovician Upper Ordovician Porkuni Limestone, dolomite 28; 93; 94

Pirgu Clayey limestone, marlstone 48; 88–92

Vormsi Clayey limestone, marlstone 87; 95; 96

Nabala Clayey limestone 33; 85; 86

Rakvere Limestone, marlstone 83; 84

Oandu Marlstone, limestone 1; 7; 10; 52; 54; 81; 82

Keila Marlstone, clayey limestone 30; 38–47; 51; 53; 79; 80

Haljala Clayey limestone, marlstone 4; 29; 32; 77; 78

Kukruse Limestone, clayey limestone 74; 75; 76

Middle Ordovician Uhaku Clayey limestone 72; 73

Lasnamagi Clayey limestone 9; 17; 71

Aseri Limestone

Kunda Limestone, sandy limestone 5; 69; 70

Volkhov Glauconitic limestone 11; 67; 68

Lower Ordovician Billingen Glauconitic limestone

Hunneberg Glauconitic sandstone

Varangu Clay

Pakerort Siltstone, sandstone

Environ Earth Sci (2012) 65:1147–1156 1151

123

thus being higher compared to limestones and dolomites

(Table 2). K-bentonites, a clay-rich sediments formed by

the weathering of volcanic ash deposits, are rich in total

fluorine (363–4,578 mg/kg) and boron (57–1,124 mg/kg).

Excluding the K-bentonite samples from the calculation,

the average concentrations of boron and fluorine in Silurian

and Ordovician carbonate rocks in Estonia are 36.0 and

459.5 mg/kg, respectively. As a comparison, the average

concentration for boron in earth’s crust is 10 mg/kg

(Yazbeck et al. 2005) and for fluorine 625 mg/kg (Edm-

unds and Smedley 2005).

Although fluorite (CaF2) is the most common fluorine-

bearing mineral, being abundant in granitic rocks (Hem

1985; Kundu et al. 2001; Saxena and Ahmed 2003; Sujatha

2003; Reddy et al. 2010), it is a minor accessory mineral in

carbonate rocks and its solubility in fresh water is low.

Biotite may be rich in fluorine and is common mineral in

granitic areas as well. In present research, small amount of

biotite was found only in terrigenous material, especially in

K-bentonite samples. Thus, dissolution of fluorite and

biotite is not the noteworthy source of fluorides in carbo-

naceous groundwater of Estonia.

Beside the fluorite and biotite, apatite and fluorapatite

are also considered to be possible sources of fluorides in

groundwaters in granitic areas (Hem 1985; Rafique et al.

2008; Desbarats 2009; Reddy et al. 2010). Although apatite

is present in carbonate rocks as well, it is unlikely the

major source of fluorides in groundwater. The S–O aquifer

system has a characteristic of HCO3–Ca–Mg and HCO3–

Mg–Ca water type, with low B and F concentrations. In

western Estonia, where high fluoride concentrations pre-

vail, the content of Cl- and Na? in groundwater increases

and HCO3–Cl–Na–Mg–Ca water type with increased pH

value becomes widespread (Perens et al. 2001). The solu-

bility of apatite is rather low and decreases with increasing

pH values (Leybourne et al. 2008; Desbarats 2009).

Consequently, Ca-phosphate minerals, containing fluorine

Table 2 Descriptive statistics for boron and fluorine concentrations in carbonaceous rock types

Rock type n Boron concentrations in rocks (mg/kg) Fluorine concentrations in rocks (mg/kg)

Mean Median Min. Max. Mean Median Min. Max.

Limestone 18 8.94 5.00 3.00 54.00 261.83 205.00 70.00 690.00

Dolomite 10 10.50 8.50 3.00 32.00 377.30 308.00 187.00 1,001.00

Clayey limestone 17 22.82 24.00 3.00 43.00 356.35 350.00 187.00 500.00

Clayey dolomite 8 24.63 24.00 9.00 38.00 443.88 425.00 264.00 660.00

Marlstone 12 89.00 73.00 33.00 160.00 771.33 760.00 525.00 1,029.00

Domerite 5 67.00 54.00 27.00 121.00 828.00 830.00 680.00 1,030.00

Clay 9 130.89 129.00 97.00 168.00 712.67 704.00 600.00 870.00

K-bentonite 12 438.17 309.00 57.00 1,124.00 2,813.42 3,069.00 363.00 4,578.00

Fig. 3 Ternary diagram showing F and B content in various rock

types. Classification of rocks is based on the lithological diagram of

carbonate rocks (Vingisaar et al. 1965): A limestone, B dolomite,

C marlstone, D domerite, E clay

1152 Environ Earth Sci (2012) 65:1147–1156

123

in their mineral lattice, may be rather one of the minor

geological sources of fluorides in S–O aquifer system. On the

other hand, substituted apatites with high fluorine are more

soluble than purer apatites (Edmunds and Smedley 2005).

Thus, to understand the mechanism of apatite dissolution,

additional leaching experiments should be carried out.

Remarkable amount of boron and fluorine in rocks is

considered to be related to clay minerals. Ion-exchange and

adsorption are the most typical processes through which

boron and fluorine could be bounded to clay minerals (Hem

1985; Saxena and Ahmed 2003; Sujatha 2003; Edmunds

and Smedley 2005; Gonfiantini and Pennisi 2006; Pennisi

et al. 2006; Rafique et al. 2008; Naseem et al. 2010). The

average clay content in the rocks of Silurian and Ordovician

sequences tends to increase from north to south and south-

west in the Baltic Basin (Nestor and Einasto 2007). Due to

the increase of clay content in aquifer-forming rocks and

prolonged residence time of groundwater, Ca is replaced by

Na through ion-exchange processes and pH value of

groundwater slightly increases (up to 8.5) in West-Estonia.

As a result, the formed alkaline Na-rich groundwater of

HCO3–Cl–Na–Mg–Ca chemical type favours the leaching

of boron and fluorine into the groundwater. Furthermore,

clay content is generally higher in Silurian carbonate rocks

compared to Ordovician ones. Former studies have shown

that the sea stand was higher in the Silurian compared to the

late Ordovician (Kiipli and Kiipli 2006), thus Silurian rocks

are expected to be more clayey. Long-term leaching of

those clayey rocks might provide the elevated fluoride and

boron content in groundwater of western Estonia.

According to this study, boron and fluorine are most likely

leached out from clayey material like marlstones, domerites

and K-bentonites. The CaO/Al2O3 and K2O/Al2O3 ratios

reflect the relationship between carbonaceous and terrigenous

component, and between terrigenous compounds (K-feldspar

and clay) in rock samples, respectively (Fig. 4a, b). K-ben-

tonite samples are not included to data used in these diagrams.

Higher values of CaO/Al2O3 ratios refer to carbonate rock

samples, lower ones to terrigenous material, including clay. It

can be seen that the values of CaO/Al2O3 ratios rapidly rise

in limestone and dolomite samples where the contents of B

and F are low, generally below 20 and 400 mg/kg, respec-

tively. Lower CaO/Al2O3 ratios (up to 25) refer to terrige-

nous material with elevated B (20–160 mg/kg) and F

(350–1,000 mg/kg) concentrations. Unlike the typical lime-

stones and dolomites, apatite-rich samples exhibit also high

fluorine content (450–1,000 mg/kg) (Fig. 4a).

Using the K2O/Al2O3 ratio as the indicator of terrige-

nous material, the lower ratios are the characteristic for a

finer grain size, higher ratios for coarser grain size. Fig. 4b

shows that the limestones and dolomites have K2O/Al2O3

ratios varying from 0.4 to 0.8 with low B (up to 20 mg/kg)

and F (up to 350 mg/kg) contents. K2O/Al2O3 ratios up to

0.5 are common to clayey rocks with higher B and F

concentrations. Exceptionally, rock samples rich in apatite

do not coincide with the rest of the limestone and dolomites

samples in the diagrams showing higher F concentrations

(Fig. 4b). Altogether, elevated B and F contents are clearly

concentrated in the terrigenous fraction of the rock

(Fig. 4a) and furthermore, are bounded to clay mineral

phases (Fig. 4b). The B content increases exponentially

with the clay amount, represented mainly by illitic clay

minerals (illite, illite–smectite). The abundance of illitic

clay minerals (0.7–32.5% of the whole rock) in rock

samples has been determined with XRD analyses. Thus,

clay-rich sediments, providing ion-exchange and adsorp-

tion sites for F- and B, may contribute to elevated con-

centrations of those elements in groundwater. The

geochemical behaviour of F is more complicated compared

to B due to the presence of Ca-phosphate (apatite and

fluorapatite) minerals, which occur both in terrigenous as

well as in authigenic fraction. As it was mentioned above,

elevated F concentrations may be also found in limestones

and dolomites poor in clay minerals (Fig. 4a, b). In such

samples, a positive correlation between fluorine and

phosphate (P2O5), which is a likely indicator of apatite and

fluorapatite, can be observed (Fig. 4c).

Analysing the chemical composition of aquifer-forming

rocks and water–rock interactions is an essential part of

groundwater studies, as the chemical composition of the

groundwater reflects the mineralogy of the host rock. In

addition to apparent relationships between fluorine and

boron concentrations and rock type, the influence of other

factors such as the residence time of groundwater, spatial

variations in pH values, Ca2? and Na? concentrations

(chemical type of water) play the important role in the

leaching of F and B into the groundwater.

Groundwater is an important source of drinking water

supply all over the world, including Estonia. Both fluorine and

boron are considered to be toxic in large quantities to human

body, acquired largely from drinking water and food (Zohouri

and Rugg-Gunn 2000; WHO 1998). Therefore, there is a need

for examining probable natural sources of F- and B in

groundwater. The results of current study give the first over-

view of the different rock types as the natural sources of these

elements in S–O aquifer system and could help to work out the

strategies for safe drinking water supply in Estonia.

Conclusions

Anomalously high fluoride and boron concentrations (up to

7.2 and 2.1 mg/l, respectively) are common in groundwater

in western Estonia, where main drinking water resource is

carbonaceous S–O aquifer system. Leaching of the host

rocks is the major natural source of F- and B in

Environ Earth Sci (2012) 65:1147–1156 1153

123

groundwater. Therefore, chemical composition of ninety-

one rock samples from S–O aquifer system forming rocks

was determined. The results of this study show that fluorine

and boron in carbonate rocks are clearly bounded to clay

mineral phase, as F and B content in marlstones reach up to

1,000 and 150 mg/kg, respectively, being even higher in

K-bentonites. F and B concentrations in limestones and

dolomites do not generally exceed 500 and 20 mg/kg,

respectively. Consequently, clay-rich sediments such as

marlstones, domerites and K-bentonites are the most

probable sources of high fluoride and boron levels in

groundwater. Ca-phosphate minerals are generally rich in

fluorine. Although, the solubility of apatite is rather slow, it

is unlikely the remarkable source of fluorides in ground-

water. Fluorite, the most common fluorine-bearing mineral,

and biotite are not widespread minerals in carbonate rocks.

Thus, dissolution of fluorite and biotite cannot be the

sources of fluorides in S–O aquifer system.

Acknowledgments This study has been carried out with the finan-

cial support of Estonian Science Foundation grant No 7403. The

authors thank Dr. Leho Ainsaar, Dr. Enli Kiipli and MSc. Tonu Pani

for the assistance with rock samples selection. Prof. Kalle Kirsimae is

acknowledged for the fruitful discussions in field of mineralogy. The

XRD analyses were carried out with the assistance of Dr. Peeter

Somelar at the Department of Geology, Tartu University. The chem-

ical analyses of rock samples were performed at ACME Analytical

Laboratories Ltd., Vancouver, Canada.

References

ADA (2001) Position of the American Dietetic Association: The

Impact of fluoride on health. J Am Diet Assoc 101:126–132

Fig. 4 The F and B concentrations plotted against the content of the major element oxides and their ratios in aquifer-forming rocks

1154 Environ Earth Sci (2012) 65:1147–1156

123

Ainsaar L, Meidla T (2001) Facies and stratigraphy of the Middle

Caradoc mixed siliciclastic-carbonate sediments in eastern

Baltoscandia. Proc Estonian Acad Sci Geol 50:5–23

Bergstrom SM, Huff WD, Kolata DR, Bauert H (1995) Nomencla-

ture, stratigraphy, chemical fingerprinting, and areal distribution

of some Middle Ordovician K-bentonites in Baltoscandia. GFF

117:1–13

Billings RJ, Berkowitz RJ, Watson G (2004) Teeth. Pediatrics

113:1120–1127

Council directive 98/83/EC (1998) On the quality of water intended

for human consumption. Official Journal L 330, 05/12/1998

P. 0032–0054

Desbarats A (2009) On elevated fluoride and boron concentrations in

groundwaters associated with the Lake Saint-Martin impact

structure, Manitoba. Appl Geochem 24:915–927

Earle S, Krogh E (2004) Geochemistry of Gabriola’s groundwater.

Shale 7:35–42

Edmunds M, Smedley P (2005) Fluoride in natural waters: occur-

rence, controls and health aspects. In: Selnius O, Alloway B,

Centeno JA, Finkleman RB, Fuge R, Lindh U, Smedley P (eds)

Essentials of medical geology—impacts of the natural environ-

ment on public health. Elsevier, Amsterdam, pp 301–329

Garabrant DH, Bernstein L, Peters JM, Smith TJ (1985) Respiratory

effects of borax dust. Br J Ind Med 42:831–837

Gonfiantini R, Pennisi M (2006) The behaviour of boron isotopes in

natural waters and in water–rock interactions. J Geochem Explor

88:114–117

Grobler SR, van Wyk CV, Kotze D (1986) Relationship between

enamel fluoride levels, degree of fluorosis and caries experience

in communities with a nearly optimal and high fluoride level in

the drinking water. Caries Res 20:284–288

Heindel JJ, Price CJ, Field EA (1992) Developmental toxicity of boric

acid in mice and rats. Fundam Appl Toxicol 18:266–277

Hem JD (1985) Study and interpretation of the chemical character-

istics of natural water. US Geological Survey Water-Supply

Paper 2254

Hints R, Kirsimae K, Somelar P, Kallaste T, Kiipli T (2006)

Chloritization of late Ordovician K-bentonites from the northern

Baltic Palaeobasin-influence from source material or diagenetic

environment. Sediment Geol 191:55–66

Hints R, Kirsimae K, Somelar P, Kallaste T, Kiipli T (2008)

Multiphase Silurian bentonites in the Baltic Palaeobasin.

Sediment Geol 209:69–79

Indermitte E (2010) Exposure to fluorides in drinking water and

dental fluorosis risk among the population of Estonia. Disserta-

tiones Geographicae Universitatis Tartuensis 41, Tartu Univer-

sity Press

Indermitte E, Saava A, Kull A (2006) The survey of drinking water

supply in Estonia from the point of view of public health. In:

Mander U, Brebbia CA, Tiezzi E (eds) The Sustainable City IV.

Urban Regeneration and Sustainability. WIT Press, Southamp-

ton, pp 817–826

Indermitte E, Saava A, Russak S, Kull A (2007) The contribution of

drinking water fluoride to the risk of dental fluorosis in Estonia.

In: Brebbia C (ed) Environmental Health Risk IV. WIT Press,

Southampton, pp 161–170

Joogivee kvaliteedi-ja kontrollinouded ning analuusimeetodid (2001)

(The quality and monitoring requirements for drinking water and

methods of analysis), RTL 2001/100/1369 (in Estonian)

Jurgenson E (1988) Deposition of the Silurian Beds in the Baltic.

Valgus, Tallinn (in Russian)

Kaljo D, Kiipli T, Martma T (1997) Carbon Isotope event markers

through the Wenlock-Pridoli sequence at Ohesaare (Estonia) and

Priekule (Latvia). Palaeogeogr Palaeoclimatol Palaeoecol

132:211–224

Karro E, Rosentau A (2005) Fluoride levels in the Silurian–

Ordovician aquifer system of western Estonia. Fluoride

38:307–311

Karro E, Uppin M (2010) Siluri-ordoviitsiumi veeladestu fluoriidide

ja boori geoloogilised allikad (Fluoride and boron in Silurian–

Ordovician aquifer system). Keskkonnatehnika 3:25–27 (in

Estonian)

Karro E, Indermitte E, Saava A, Haamer K, Marandi A (2006)

Fluoride occurrence in publicly supplied drinking water in

Estonia. Environ Geol 50:389–396

Karro E, Marandi A, Vaikmae R, Uppin M (2009) Chemical

peculiarities of the Silurian–Ordovician and Cambrian-Vendian

aquifer systems in Estonia: an overview of hydrochemical

studies. Est J Earth Sci 58:342–352

Kiipli E, Kiipli T (2006) Carbonate distribution in the East Baltic

deep shelf in the late Ordovician-early Silurian. GFF

128:147–152

Kiipli T, Mannik P, Batchelor RA, Kiipli E, Kallaste T, Perens H

(2001) Correlation of Telychian (Silurian) altered volcanic ash

beds in Estonia, Sweden, and Norway. Norw J Geol 81:179–193

Kiipli T, Kiipli E, Kallaste T, Hints R, Somelar P, Kirsimae K (2007)

Altered volcanic ash as an indicator of marine environment,

reflecting pH and sedimentation rate—Example from the

Ordovician-Kinnekulle bed of Baltoscandia. Clay Clay Min

55:177–188

Kiipli T, Soesoo A, Kallaste T, Kiipli E (2008) Geochemistry of

Telichian (Silurian) K-bentonites in Estonia and Latvia. J Volc-

anol Geoth Res 171:45–58

Kundu N, Panigrahi MK, Tripathy S, Munshi S, Powell MA, Hart BR

(2001) Geochemical appraisal of fluoride contamination ofgroundwater in the Nayagarh District of Orissa, India. Environ

Geol 41:451–460

Leybourne MI, Peter JM, Johannesson KH, Boyle DR (2008) The

Lake Saint-Martin bolide has a big impact on groundwater

fluoride concentrations. Geology 36:115–118

Mather JD, Porteous NC (2001) The Geochemistry of boron and its

isotopes in groundwaters from marine and non-marine sandstone

aquifers. Appl Geochem 16:821–834

Naseem S, Rafique T, Bashir E, Bhanger MI, Laghari A, Usmani TH

(2010) Lithological influences on occurrence of high-fluoride

groundwater in Nagar Parkar area, Thar Desert, Pakistan.

Chemosphere 78:1313–1321

Nestor H, Einasto R (1997) Ordovician and Silurian carbonate

sedimentation basin. In: Raukas A, Teedumae A (eds) Geology

and mineral resources of Estonia. Estonian Academy Publishers,

Tallinn, pp 192–204

Pennisi M, Bianchini G, Muti A, Kloppmann W, Gonfiantini R (2006)

Behaviour of boron and strontium isotopes in groundwater–

aquifer interactions in Cornia Plain (Tuscany, Italy). Appl

Geochem 21:1169–1183

Perens R, Vallner L (1997) Water-bearing formation. In: Raukas A,

Teedumae A (eds) Geology and mineral resources of Estonia.

Estonian Academy Publishers, Tallinn, pp 137–145

Perens R, Savva V, Lelgus M, Parm T (2001) The Hydrogeochemical

Atlas of Estonia (CD version). Geological Survey of Estonia,

Tallinn

Polma L (1982) Comparative lithology of the Ordovician Carbonate

Rocks in the Northern and Middle East Baltic. Valgus, Tallinn

(in Russian)

Price CJ, Marr MC, Myeos CB, Seely JC, Heindel JJ, Schwetz BA

(1996) The developmental toxicity of boric acid in rabbits.

Fundam Appl Toxicol 34:176–187

Queste A, Lacombe M, Hellmeier W, Hillermann F, Bortulussi B,

Kaup M, Ott K, Mathys W (2001) High concentrations of

fluoride and boron in drinking water wells in the Muenster

Environ Earth Sci (2012) 65:1147–1156 1155

123

region: results of a preliminary investigation. Int J Hyg Environ

Health 203:221–224

Rafique T, Naseem S, Bhanger MI, Usmani TH (2008) Fluoride ion

contamination in the groundwater of Mithi sub-district, the Thar

Desert, Pakistan. Environ Geol 56:317–326

Raukas A, Teedumae A (1997) Geology and mineral resources of

Estonia. Estonian Academy Publishers, Tallinn

Reddy DV, Nagabhushanam P, Sukhija BS, Reddy AGS, Smedley PL

(2010) Fluoride dynamics in the granitic aquifer of the Wailapally

watershed, Nalonga District, India. Chem Geol 269:278–289

Saava A (1998) Health hazards due to drinking water. Proc Latvian

Acad Sci 52:162–167

Saxena VK, Ahmed S (2003) Inferring the chemical parameters for the

dissolution of fluoride in groundwater. Environ Geol 43:731–736

Somelar P, Kirsimae K, Hints R, Kirs J (2010) Illitization of Early

Paleozoic K-bentonites in the Baltic Basin: decoupling of burial-

and fluid-driven processes. Clay Clay Min 58:388–398

Sujatha D (2003) Fluoride levels in the groundwater of the south-

eastern part of Ranga Reddy district, Andhra Pradesh, India.

Environ Geol 44:587–591

Vingisaar P, Oraspold A, Einasto E, Jurgenson E (1965) Karbonaatk-

ivimite uhtne klassifikatsioon ja legend (Classification of

carbonate rocks). Eesti NSV Teaduste Akadeemia, Tallinn (in

Estonian)

Wegman DH, Eisen EA, Hu X, Woskie SR, Smith RS, Garabrandt D

(1994) Acute and chronic respiratory effects of sodium borate

particulate exposures. Environ Health Perspect 102:119–128

WHO (1998) Environmental Health Criteria 204: Boron. World

Health Organization, Geneva

WHO (2008) Guidelines for drinking-water quality, 3rd edn. World

Health Organization, Geneva

Yazbeck C, Kloppmann W, Cottier R, Sahuquillo J, Debotte G, Huel

G (2005) Health impact evaluation of boron in drinking water: a

geographical risk assessment in Northern France. Environ

Geochem Health 27:419–427

Zohouri FV, Rugg-Gunn AJ (2000) Sources of dietary fluoride intake

in 4-year-old children residing in low, medium and high fluoride

areas in Iran. Int J Food Sci Nutr 51:317–326

1156 Environ Earth Sci (2012) 65:1147–1156

123