Distribution of iron and manganese in the Seine river estuary: approach with experimental laboratory mixing Baghdad Ouddane,* Mohamed Skiker,† Jean Claude Fischer and Michel Wartel Universite´ des Sciences et Technologies de Lille, Laboratoire de Chimie Analytique et Marine, CNRS UPRESA 8013 Baˆt. C8 (2e`me e´tage), 59655 Villeneuve d’Ascq Cedex, France. E-mail:
[email protected] Received 10th May 1999, Accepted 26th July 1999
The physico-chemical behaviour of iron and manganese has been observed during many surveys covering various hydrodynamic conditions in the Seine river estuary system. The results obtained confirm the non-conservative behaviour of these two metals. Generally, dissolved iron exhibits non-conservative removal and shows a rapid decrease in low salinity; it is moved from fresh waters with high concentrations to saline waters with very low concentrations. This can be attributed to the flocculation processes as confirmed by laboratory experiments. Dissolved manganese versus salinity curves exhibit a peak concentration in the low salinity zone. Laboratory mixing experiments have been undertaken comparing iron and manganese adsorption/desorption from suspended material versus salinity, using a series of water samples collected in the up-river and marine regions in order to assess the importance of particulate material and salinity on iron and manganese distributions. The salinity was controlled by varying the marine to fresh water ratio. The reaction kinetics aspect is developed in more detail for manganese in the last series of remobilization experiments starting from a stock of suspended particles collected in the upstream river site (Caudebec) in mixtures of waters, according to time and salinity. This study has allowed us to show that iron and manganese behaviour in the Seine estuary is strongly influenced: (i) by the high turbidity zone and by the presence of calcium carbonate which could stabilise the Mn() form; and (ii) by the increase of salinity, calcium, magnesium and suspended matter concentrations and by complex formation.
Introduction The behaviour of trace metals is, in some cases, related to the concentration of dissolved and particulate iron and manganese. This is not the only factor which can play an important role in controlling trace metal concentrations in the aquatic environment but the chemical forms (speciation) and the different reactions (complexation, oxidation–reduction, adsorption–desorption, dissolution-precipitation or coprecipitation) are also important. It is well known that manganese and iron are present in aquatic systems in different forms, in both dissolved and particulate phases, and the distribution between these species is fundamental to understanding the behaviour of these two metals and their influence on other trace metals.1–3 Manganese and iron oxides are known to take up a variety of transition elements and heavy metals, either by adsorption on the oxide surface or directly by incorporation into the crystal structure.4,5 In estuarine media, the distribution of dissolved manganese and iron is complicated by the strong gradients of several physico-chemical properties like salinity, turbidity, temperature, dissolved oxygen concentration, pH, Eh, and the character and the concentration change of particles which result from the mixing of fresh and saline waters.6 The behaviour of these two metals in estuaries is influenced by all these parameters and the various processes involved in estuaries, such as adsorption–desorption, precipitation–dissolution, sedimentation–re-suspension and flocculation–coagulation. Manganese is one of the most reactive and mobile elements and has a geochemically important metal cycle in the environ†Present address: University Mohamed Premier, Faculte des Sciences, Oujda, Morocco.
ment, particularly in estuaries.7,8 It is strongly affected by the changes in oxidation state in reponse to its redox chemistry7–10 and can be found in three oxidation states (, and ) in aquatic media. Thus, when conditions are sufficiently reducing, Mn is relatively soluble and geochemically stable as Mn(); under less reducing regimes conversion to Mn() and Mn() leads to transfer to solid phases. The behaviour of dissolved manganese in estuaries is generally non-conservative,7,9,11,12 with the exception of certain estuaries in which Mn is conserved.13–16 The distribution of iron in estuaries is well documented and the dissolved iron is found to be largely removed, the major fraction being on suspended particulate material through colloid flocculation.13,17,18 Its biochemistry in sea-water has been extensively studied;19–21 the chemistry of iron is strongly related to the action of micro-organisms in natural waters, and its bioavailability is governed both by chemical speciation and by biological process.22 The concentration and form of dissolved iron has an important role in the limitation of the production of phytoplankton23,24 and the complexation of iron with organic ligands is now evident;25–28 the stability of these complexes can explain the high concentration of dissolved iron in some fresh waters. Thus, the purpose of this study is to elucidate the distribution of dissolved and particulate manganese and iron in the Seine estuary, and to expose the processes which control their behaviour. Laboratory mixing experiments of the suspended particles from the Seine River with sea-water have been carried out. In these experiments, the parameters pH, Eh and oxygen concentration were controlled and experiments completed within two days in order to investigate the kinetics of desorption–adsorption as function of salinity. J. Environ. Monit., 1999, 1, 489–496
489
Fig. 1 Localisation of sampling stations in the Seine river estuary.
Materials and methods 1 Sampling The Seine river estuary (Fig. 1) is known to be highly contaminated by wastes derived from industrial sources, as 30% of French industrial activity is located in the Seine Basin (75 000 km2 area), from agricultural operations, which correspond, in this region, to 40% of the total economic activity in France, and finally from sewage inputs, since 30% of the whole French population lives in the Seine Basin.29 The hydrosedimentological processes in the Seine estuary are very complex; they are enhanced by a macrotidal regime which controls suspended material transport, deposition and erosion, as discussed by Avoine.29 The average water discharge is 450 m3 s−1, with an average inter-annual variation from 200 to 650 m3 s−1; the lowest discharge was 60 m3 s−1 and the highest one is more than 2200 m3 s−1. The salinity intrusion may extend up to 50 km upstream of Le Havre (near Vieux Port, Fig. 1) during periods of low river discharge, and up to 20 km during flood conditions.30 The limit of tide variations can be observed 100 km upstream of Le Havre at Poses (Fig. 1). The estuary is characterised by a high turbidity zone (up to several g l−1), usually found between Tancarville and Le Havre; this is due to the large tide range (3 m at neap tide to 7.5 m at spring tide) coupled with relatively shallow depths that maintain and amplify the friction asymmetry of the tide wave as it propagates upstream.29 A total of four surveys were carried out during the period 1995–1997. The first survey (1995) was carried out in specific conditions of high discharge (2019–2079 m3 s−1) (the regular average flow is 450 m3 s−1) under spring tide conditions along the estuary at 10 stations (station 1–10; Fig. 1) on January 31, and at a fixed station during a tide cycle at hourly intervals (station 8, Fig. 1) on February 1. The second survey (1996) was carried out along the estuary at 14 stations (1–14; Fig. 1) under spring tide conditions on February 17. The third one (1996) was performed at a fixed station (station 8; Fig. 1) and the last five samples were taken in the Bay (marine zone) in order to have the marine end station; this campaign was carried out under neap tide conditions and during a period of low discharge (201 m3 s−1) on May 10–11. The last survey 490
J. Environ. Monit., 1999, 1, 489–496
was accomplished at 9 stations in February 1997 under spring tide conditions. Field sampling was performed essentially throughout the estuarine mixing zone in approximate coincidence with the local high water. Sampling positions were chosen to cover the full salinity range. In all cases the water samples were collected aboard a boat, from a depth of 1 m below the water surface (surface samples) and also from ~1 m above the bottom (the deep samples) and 3 m off the side of the boat using a Teflon pumping system [ Teflon tube connected to a pneumatic suction-pump internally coated with polytetrafluoroetylene (PTFE ) (ASTI, Courbevoie, France)]. Sample bottles (high density polyethylene), Nuclepore filters (polycarbonate membrane), filtration equipment and other plasticware were leached in 1 nitric acid (Suprapur, Merck, Darmstadt, Germany) for 1 week and thoroughly rinsed several times with Milli-Q water (Millipore, Bedford, MA, USA) prior to use. In addition, sample bottles were rinsed on site with water samples. Approximately 2 l volumes of unfiltered water were collected and placed into polyethylene bottles for the determination of total suspended particulate matter (SPM ). Each sample (~10 l ) was filtered on site directly through 0.4 mm cleaned Nuclepore filters (diameter: 147 mm) under a horizontal laminar flow hood. For the trace and the major elements, portions of filtered water were collected in high density polyethylene flasks and acidified to pH ~1 with concentrated nitric acid (Suprapur, Merck). These flasks were sealed in plastic bags and stored at 4 °C in the dark until analysis. Further portions of filtered water (500 ml ) were collected in special centrifugation Nalgene flasks (high density polyethylene) for the analysis of iron and manganese after a preconcentration procedure. Suspended material on the filters was not rinsed in order to avoid loss of exchangeable metals, but kept in acid cleaned Petri dishes. Back in the laboratory, the filters were dried in a horizontal laminar flow hood (Class 100 clean air) and weighed under clean conditions. 2 Methods Dissolved iron and manganese were analysed using the gallium coprecipitation method adapted to estuarine waters by Ouddane et al.31 and analysis by inductively coupled plasma
atomic emission spectrometry (ICP-AES; Varian Axial Liberty Series II with SPS-5 autosampler; Varian Analytical Instruments, Mulgave, Australia). Briefly, a solution of 5 g l−1 of gallium was prepared by dissolving high purity (99.999%) Ga metal (Prolabo, Paris, France) in nitric acid (Suprapur, Merck); 500 ml of this solution was added to 500 ml of samples. After mixing, the pH was adjusted to ~9.5 using 10 sodium hydroxide (NaOH ) in order to initiate gallium and magnesium hydroxide precipitation. The precipitate was then centrifuged at 3400 rpm for 30 min using an X340 Prolabo centrifuge. The supernatant solution was rejected and the precipitate thus obtained was redissolved with 200 ml of concentrated nitric acid before dilution to 5 ml with Milli-Q water. The solutions were analysed using ICP-AES; three replicates were determined for each sample and relative standard deviations were below 5%. For each analytical series analytical blanks were controlled, the concentration of each element being less then 1% of the mean concentration for all of the series samples. The accuracy of the analytical method has been tested using the CASS-2 and SLEW-1 dissolved trace metal standards (Seawater and Estuarine Water Reference Material for Trace Metals, supplied by the National Research Council of Canada, Quebec, Canada). The results obtained for iron and manganese in these two certified sea- and estuarine waters were in good agreement with the certified values. Particulate metals were determined by means of ICP-AES following total acid digestion with a mixture of boiling concentrated acids, HCl+HNO +HF, in a PTFE flask. The digests 3 were evaporated to dryness and the resulting residues were dissolved in 2% nitric acid. For the validation of the digestion procedure and the analytical method, MESS-1, BCSS-1 and PACS-1 Sediment Standards Reference Material for metals supplied by the National Research Council of Canada have been used. Determinations of pH, temperature and salinity were made on site in unfiltered samples; pH measurements were performed with an Orion (Boston, MA, USA) pH meter. Salinity and temperature were measured with a WTW apparatus, Model LF191. Salinity was also measured in the laboratory by means of a conductimetry cell using the Practical Salinity Scale.32 Oxygen concentrations were determined by Winkler titration. In the solid fraction (suspended particulate material ) the fractionation of manganese was accomplished by sequential extraction procedures reported previously33,34 with some modification in digestion of the residual fraction. Four fractions were identified; exchangeable and associated with carbonates, easily reducible phases (Mn oxide or hydroxide, amorphous Fe oxide), organic-bound and residual, the reagents used for each fraction were the same used in Tessier et al.33 procedure. Electron spin resonance (ESR) spectra were recorded at room temperature on a Varian E-109 spectrometer operating at a microwave frequency of ~9.54 GHz, with field sweep of 4000 G centred at 3500 G, a scan time of either 8 or 16 min, and a 100 kHz field modulation amplitude of 0.63–1.25, depending on the spectral linewidth. The magnetic field at the sample was calibrated with DPPH (diphenylpicrylhydrazyl ). Mixing experiments were carried out with the end stations river and sea-waters on the boat with a stock of river particles collected in Caudebec (Station 1), the river end station.
Results and discussion 1 Dissolved iron The concentration of dissolved iron as a function of the salinity is presented in Fig. 2. For all the four surveys, the decrease in the concentration from the fresh waters to the seawaters indicates a removal by a concave curvature of the data; the removal of iron is apparently the repetitious process in
estuaries.8 This behaviour is similar to that reported in the majority of other estuaries.35–38 The non-conservative behaviour was explained by the aggregation of colloidal iron during estuarine mixing,39 or by an association between dissolved iron and dissolved organic matter. Fox and Wofsy17 have explored the relationship between soluble iron and humic substance in estuaries and they deduced that the dissolved iron and the dissolved humic acid have the same behaviour but are not chemically associated. In the Forth estuary Balls et al.40 interpreted the elimination by flocculation and precipitation of iron oxides and hydroxides as being a direct result of increasing pH and ionic strength. Dai and Martin41 have established the partitioning between the truly dissolved and the colloidal fractions of iron in Arctic estuaries; they concluded that the colloidal fraction represents 80% of the total dissolved fraction and is the reactive fraction in the estuarine mixing zone, since the truly dissolved fraction behaves conservatively. This conservative behaviour of dissolved iron was displayed in the Clyde estuary by Muller et al.;16 the variation of iron with salinity has been elucidated by a simple dilution of fresh water with a high iron concentration and sea-waters with much lower iron concentrations. The iron concentrations in the Seine river estuary follow an exponential decrease, with a rapid, pronounced non-conservative loss (>75%) from fresh water to sea-water during each survey, and the concentration levels observed (70–320 n) were comparable to those found in other large world rivers and estuaries (see Table 1). In the most saline water (Bay of Seine) the dissolved iron concentrations recorded (10–15 n) were comparable to the values usually found in open seawater (~1–10 n).37,42 A lower concentration of dissolved Fe was observed in the May 1996 survey [Fig. 2(C )] and, exceptionally, much higher dissolved iron levels were found [more than 1050 n in the river waters, Fig. 2(A)] for the first survey in 1995, which was carried out in conditions of high discharge. These high levels can be explained by a dissolution of previously precipitated iron and soil erosion, as a result of high rainfall and succeeding drainage of the basin to the river. An alternative explanation is that the short residence time was not sufficient for the aggregation of colloidal iron, the stabilisation of the iron organic complex or iron colloidal state in the dissolved phase by dissolved organic matter.15,25,26 This justification is validated by the results of high dissolved organic carbon observed, in the same period in the Seine river estuary by Motamed and Texier.43 2 Dissolved manganese The results of dissolved manganese for the different cruises are plotted as a function of salinity in Fig. 2. Except for the May 1996 survey, the profiles indicate an appreciable nonconservative behaviour of dissolved Mn in the Seine estuary with a maximum at the low salinities. These results confirm the previous data found in the same estuary by Chiffoleau et al.44 and by Ouddane et al.12 The non-conservative behaviour was detected in a several estuaries in the world: the St. Lawrence estuary,10,45 Newport estuary,46 Rhine and Scheldt estuaries,47–49 Tamar estuary,7,50 Severn estuary,51 the Delaware estuary,8 the Geum estuary,52 the Mullica53 and the Forth estuary.54 In contrast, manganese has been found to be essentially conservative in certain estuaries, such as the Beaulieu estuary13,15 and the Clyde estuary.16 In the Tamar Estuary, Knox et al.50 explained a concentration maximum of dissolved Mn by a benthic source. Since Mn remobilization is at least partly determined by redox potential, as suggested in the Rhine and the Scheldt estuaries by Duinker et al.9,35,48 and Wollast et al.,47 benthic fluxes are expected to be enhanced in the warmer summer months, when biological activity is increased and the oxygen concentration J. Environ. Monit., 1999, 1, 489–496
491
Fig. 2 Dissolved iron concentration and dissolved manganese concentration (n) as function of salinity (A, January 1995; B, February 1996; C, May 1996; D, February 1997). Table 1 Dissolved iron concentrations in various world rivers Sites
Iron/n
Authors
Ebro Tsengwen Escaut Lena Elbe St Laurent Clyde Seine
39 143–411 196–501 644±214 895 984 2811–3616 70–320
Guieu et al.38 Hung36 Paucot and Wollast49 Guieu et al.38 Duinker et al.35 Bewers and Yeats45 Muller et al.16 This work
is generally lower.55 When Mn remobilisation occurs in suboxic sediments, metals associated with the Mn–oxyhydroxide coatings may also be released. This can result in increased interstitial water concentrations of these metals and diffusion into the overlying water column.9,35 Addition of metals to estuarine waters may also occur when the bottom sediments 492
J. Environ. Monit., 1999, 1, 489–496
are disturbed and interstitial waters are injected into the water column.9,35,52 This is caused by the spring–neap tidal cycle, which controls the extent of benthic anoxia. Non conservative behaviour has also been attributed, in some cases, to anthropogenic inputs in some industrialised estuaries.54 In a recent paper12 we attempted to identify the parameters influencing the reactive behaviour of Mn in the Seine river estuary in order to find out what possible controlling mechanisms could explain Mn behaviour in this estuary. In brief, the results for dissolved manganese obtained lead us to conclude that the observed increase in dissolved Mn concentration (significant Mn-remobilisation) occurs due to: (i) the solid–solution partitioning of Mn being very sensitive to salinity by the formation of stable dissolved chloride and sulfate complexes; (ii) competitive exchange equilibrium between dissolved Ca2+ and Mg2+ and the Mn embedded in calcium carbonate’s lattice as a solid solution. The control by Ca2+ and Mg2+ ions has been experimentally confirmed by Garnier et al.56 in order to explain
the sorption/desorption process of Co and Mn in the Lena River Estuary. 3 Seasonal variability of Fe and Mn For the May 1996 survey, the concentrations of dissolved Fe were relatively low compared with other surveys and the concentrations of dissolved Mn were lower than those usually measured; this was probably due to the high biological activity in this period (spring phytoplankton blooms) which contributes to the uptake of Mn by the biological material. Lubbers et al.57 have reported a significant removal of dissolved Mn from the solution by Phaeocystis sp. Shumilin et al.3 have demonstrated the interaction of Mn and Fe with a number of phytoplankton cells. Loijens and Wollast58 have demonstrated, by experimental incorporation of radioactive traces metals in the water column, that approximately 40% of the uptake of Mn is due to active transfer from the dissolved phase to the living particulate matter. The evidence of spring phytoplankton blooms is indicated by the results of the dissolved silicon concentration which had a very high depletion in this period, in contrast to other cruise periods when silicon concentrations generally showed a clear linear decrease with salinity. The distribution of Fe and Mn between the dissolved and the particulate phases can be expressed by a distribution coefficient K , which represents another way of visualizing the d combined effect of heterogenous reactions on the solid–liquid distribution of a trace metal.6,58,59 It is defined as the ratio of the concentration of the metal per unit mass of particulate material to the concentration of the metal per unit mass (or volume) of water. This coefficient is highly sensitive to particle type and there is evidence to suggest that some phytoplankton have a greater affinity for some metals than lithogenic particles.60,61 A study of the distribution of an element between the dissolved and particulate phases is very delicate; several phenomena can be involved. In addition to the heterogeneity of the composition of the suspended matter and the diversity of the various phases which constitute it can be added the chemical form of the element in the dissolved and particulate phases. The study can be carried out by considering the distribution of the element between its concentration in the solid phase and the dissolved phase, which is defined by a distribution coefficient (K ) given by the following expression: d Particulate–concentration (m/m) K= d Dissolved–concentration (m/v) Fig. 3 shows the evolution of the logarithm of the partition coefficients of Mn and Fe with respect to salinity. The seasonal behaviour of iron and particulate manganese is very net. The first observation is the higher values in the May, 1996, campaign compared with February, 1996; this confirms a transfer of the elements from the dissolved to the particulate phase. In the low salinities (May, 1996) we can recognise the phenomenon of precipitation of iron on contact of fresh water; with marine water moreover, the strong biological activities that were shown during the discussion of the dissolved phase can be seen. The coefficient of partition of manganese in May, 1996, decreases with salinity in the Seine estuary; this reduction results from a less important Mn sorption when salinity increases and it can be interpreted as competition between this element and other components of the dissolved phase. Indeed, Garnier et al.56 showed by laboratory studies using radiotracer markers that the sorption of manganese in water estuaries is marked by competition with the Ca2+ and Mg2+ of the dissolved phase. High values of the coefficient of partition in May, 1996, of Mn and Fe show the close connection between the enrichment of these elements in the particulate phase and the strong biological activity. This seasonal phenomenon, at least for Mn, was emphasised in the English Channel
Fig. 3 Distribution coefficient of Mn and Fe as function of salinity.
waters. In this site Wartel et al.60 have shown that manganese is mainly in the form of carbonate and that dissolved manganese is controlled by the support of the solid phase, which consists of coccolithophores (pelagic alga with calcareous hull ), this type of alga being abundant in surface water of the North-West Atlantic in the summer period. Manganese has a strong affinity for the particulate phase, and the coefficient of partition is ten times higher in the May, 1996, campaign than in February of the same year. This evolution profile of the K of Mn clearly shows the seasonal d behaviour of this element in the estuary of the Seine, a phenomenon already observed in waters of the English Channel60 and checked with laboratory experiments by Loijens and Wollast,58 who studied the exchanges between the dissolved and particulate phases by the use of radiotracers and, more particularly, the transfer kinetics of these elements. In the case of the particles in the Seine, we propose to study the exchanges between the dissolved and particulate phases, and we will particularly examine the case of manganese to explain its non-conservative behaviour in the Seine estuary. Experimental studies of Mn affinity To explain the non-conservatism of manganese in the estuary of the Seine, in particular the increase of Mn concentrations in the mix zone of fresh and saline water, several mixing experiments were undertaken in the laboratory in order to simulate the behaviour of this element which were published by Ouddane et al.12 The first series of experiments was carried out in order to study the affinity of manganese for various sort of solids (clay, humic substance and calcite); this study used electron paramagnetic resonance (EPR). The second series of experiments consisted of simulations examining the effects of salinity by addition of various salts present in seawater, such as magnesium or calcium sodium chloride, to fresh water from the River Seine containing a constant weight of particulate material. The dissolved salt concentrations were increased in order to reach their in sea-water levels. To complete this work, in this study we have conducted another, third, series of tests, centred on the remobilization of Mn from the suspended particles, from the synthetic calcite doped with manganese and from a mixture of particles and doped calcite. In the last series of experiments we have studied J. Environ. Monit., 1999, 1, 489–496
493
the remobilization of Mn and Fe, starting from a stock of suspended particles collected in the river site (Caudebec) in mixtures of waters, according to time and according to salinity (mixture of filtered fresh water and sea-water of variable proportions). Exchange reaction water–particle studies The mobility of metals in natural water depends largely on the interactions between the processes of transport and the heterogeneous phenomena of mixtures. Adsorption and precipitation delay the phenomenon of transfers to the solute phase; on the other hand, the solubility of metals increases when complexation with inorganic or inorganic ligands in the aqueous phase is possible. The characteristics of the suspended matter (composition, speciation, specific surface, concentration at sites on surfaces, concentration in metal complexes, etc.) are the determining factors in the distribution and the behaviour of trace metal in the estuarine medium. 1 Interaction between calcite–particle. With the aim of remobilizing manganese starting from the solid phase, we carried out a series of studies by taking into account the interaction of the particles from various origins and the influence of salinity on the behaviour of the suspended particles. The particles from river origin changed structure and composition with increase in the ionic force (salinity), variations of the physicochemical conditions (temperature, oxygen, pH, Eh) and with the mixture of the particles from marine origin.62,63 Experimentally, the remobilization tests were carried out an particles from the Seine river (Caudebec) and synthetic calcite freshly prepared in the laboratory and doped with Mn (14.56 mmol kg−1, the average level of concentration that found in suspended particles) while following the protocol of McBride.64 In these experiments we plotted the dissolved manganese concentration according to the time of contact between water and the particles added. The results of remobilization of manganese starting from the particles of the Seine, doped calcite and from a mixture of both (particles of the Seine+doped calcite) are given in Fig. 4. The concentration of the solid used was about ~50 mg l−1 in all the experiments and the pH of the solutions did not change significantly (approximately 7.85±0.15) in a similar way to the dissolved oxygen concentration (8.25± 0.25 mg l−1). One observes the same profile of curves; first of all a very fast release of Mn in the solution followed by a plateau. The curve obtained shows the speed of the reactions of exchanges between the solution and the particles. Broadly, one can note that dissolved manganese passes from 18.2 n to more than 72.8 n at the end of 2 h of contact in the case of Seine water and doped calcite and at the end of 6 h in the
Fig. 4 Manganese remobilisation as function of time contact (A, from the Seine particles; B, from doped calcite by Mn; C, from a mixture 151 (m/m) of the Seine particles and doped calcite.
494
J. Environ. Monit., 1999, 1, 489–496
case of the water mixtures with natural Seine particles. This reduction of desorption rate can be explained by the influence of organic matter covering the natural particles. When we mixed the particles of the Seine and the calcite doped (1+1) in the same volume of water, the concentration of dissolved manganese remobilized (Fig. 4, curve C ) was 218.4 n, a content definitely higher than the expected value of ~145.6 n. This excessive increase finds its explanation in the change of the physicochemical characteristics of the particles and calcite, in particular the surface loadings which change at the time of the mixture of the two solids with loads of opposite surface charge. It is known that the electrophoretic mobility of artificial calcite is often positive and those of the particles of river origin (argillaceous and calcareous) is negative.65 The surface charges are very sensitive to the variations in the pH of the surrounding aqueous medium.66 Moreover, selective adsorption of some ligands (inorganic or organic) on the surface of the sites modifies the surface charge to a significant degree.67 This phenomenon can occur in the estuaries, as described in experiments by Biscan and Dragcevic,65 which show that artificial calcite charged positively is transformed immediately when it is mixed with estuarine water into negatively charged calcite. These authors affirm that in general all the positively charged particles in natural water see their loads being transformed quickly into negative charge by interaction with particles of opposite charge and coagulation. 2 Effect of salinity on the exchange reactions between water– particle. Salinity has an important effect on the surface charge and thus the stability of the colloidal particles. In order to evaluate the role of salinity in the remobilization of Mn from the suspended particles, we undertook a last series of remobilization experiments according to time and salinity. In these tests, we used ~500 mg l−1 of suspended matter in order to attempt a possible remobilization of iron in a similar way to manganese, using a stock of particles collected in Caudebec (river reference) and placed in contact with water at variable salinities with a constant agitation and temperature. The range of salinity was obtained by use of mixtures of fresh water (Caudebec) and sea-water (bay of the Seine) saturated with oxygen, the parameters pH and dissolved oxygen concentrations being followed in all the experiments. The results of the remobilization of manganese and iron according to time for several salinities are given in Fig. 5. (a) Desorption reaction. For manganese, we can see that for all the ranges of salinity chosen, there is a fast increase in dissolved Mn concentrations in the first hour (desorption or dissolution) followed by a slower decrease, a phenomenon which was not observed during the previous experiments, the quantity of suspended matter used being always approximately ~50 mg l−1. The reactions of adsorption–desorption depend on the nature and the quantity of the suspended particle.61 Franklin and Morse68 have studied the adsorption of manganese in the presence of a synthetic quantity of calcite in distilled water and sea-water and observed a more significant adsorption with the increase in the quantity of calcite. For iron, in contrast to manganese, the fast release (in the first hour) of iron to the solution followed by an adsorption (or precipitation) is observed only for the low salinities (0.3–7). The absence of iron remobilization over these salinities can be explained by a rapid elimination immediately after its dissolution. These results are in agreement with the kinetic iron studies of Meyer,39 who showed that iron agglomerates in the form of colloids during the mixing of fresh and saline waters (in estuaries) while following two stage kinetics. A first, fast, stage (a few minutes) with first-order kinetics, relating to the interaction of iron with high molecular mass organic matter followed by a second, slower, stage with second-order kinetics,
Fig. 5 Manganese and iron remobilisation experiments as function of time of contact at different salinity (by mixing of the end stations in various ratios).
corresponding to the formation of colloidal particles, which would depend on the temperature and turbidity (quantity of suspended particles). These experimental results confirm the affinity of iron for the particulate material, as we showed in the determination of the coefficient of partition (K ). For the manganese remobilizd ation (Fig. 5), one observes in general, for the whole salinity range studied, the same profile of curves, i.e., a fast increase in the manganese concentrations in the first hour and then a decrease. The desorption of manganese starting from the suspended particles can partly explain the maximum of dissolved manganese observed in low salinities at the time of the various campaigns carried out in the Seine estuary. This mobility is related to the form in which manganese is present in the solid phase. We previously showed that manganese is mainly in the carbonated fraction.12,69,70 Several authors71,72 have already shown the importance of this fraction in the distribution of manganese in solution; Li et al.73 showed that MnCO , rhodochrosite, or double salts such as kutnokorite, 3 CaMn(CO ) , formed in the aquatic environment, play an 3 2 important part in the control of the dissolved manganese concentration. We have proved that manganese in the particles of the Seine is present in the fraction carbonated in the form of a solid solution (Mn Ca CO ), which could be a source of x y 3 desorption or dissolution, or by exchange reactions between calcium and dissolved magnesium. (b) Adsorption reaction. The decrease of the manganese concentrations after 1 h of contact (Fig. 5) takes place in two stages, a fast steep decrease in the first hours followed by a slow phase from the third hour. In order to detect the reaction order and the influence of salinity on the adsorption kinetics, we traced the reciprocal dissolved manganese concentration according to the time of contact at various salinities. We
observe initially a fast increase which corresponds to the first step and a quasi-linear variation corresponding to the second step, which shows that this last reaction follows kinetics of the second order, the slopes being equal to the rate constant k. The square root of this rate constant decreases linearly with the increase of salinity (R2=0.928), showing that the transfer of manganese from the dissolved phase to the particulate phase decreases according to salinity. This reduction in adsorption as a function of salinity is also indicated by the percentage of the manganese remobilized and adsorbed on particles after 48 h decreasing linearly as a function of salinity (R2=0.971), the decrease being due mainly to the presence of calcium and magnesium competing with Mn2+ for the adsorption sites of calcite. This phenomenon was observed by Kanel and Morse,74 and by Shanbhag et al.,75 in the adsorption studies of americium phosphate on calcite, and more recently by Garnier et al.56 These last have evaluated with experiments the part played by calcium and magnesium in the adsorption–desorption reactions of Co and Mn on particles in the Lena estuary by using radiotracers. Moreover, this reduction in adsorption of manganese can be explained by the amount of its free form in solution (available Mn2+); indeed we have shown in the Seine estuary70 that dissolved free Mn2+ is present at more than 95% in fresh waters and decreases to less than 50% in sea-water. In conclusion, manganese in Seine water has a double behaviour, it can be transferred from the particulate to the dissolved phase by desorption or dissolution of suspended matter during a change of ionic strength. Its elimination from the dissolved phase takes place in two stages, a fast stage and a relatively slow stage whose kinetics depend on salinity. Reaction rates of the desorption of manganese in the suspended particles are larger than those of the adsorption reactions. In a natural medium both adsorption and desorption reactions are simultaneous; in low salinities, the majority of dissolved manganese results from the river contributions, which are added to the desorption of this element from the particulate phase. The reduction in the Mn content according to salinity results in a dilution by marine water and by adsorption phenomena.
Acknowledgement This work was supported by The Seine Normandie region (programme SEINE AVAL).
References 1 2 3 4 5 6 7 8 9 10 11 12 13 14
A. Tessier, F. Rapin and R. Carigan, Geochim. Cosmochim. Acta, 1985, 49, 183. C. S. Shieh and I. W. Duedall, Mar. Chem., 1988, 25, 121. A. N. Shumilin, V. V. Aniekiev, N. A. Goryachev, A. P. Kassatkina and S. M. Fazlullin, Mar. Chem., 1993, 43, 217. L. S. Balistrieri and J. W. Murray, Geochim. Cosmochim. Acta, 1986, 50, 2235. D. J. Burdige, Earth Sci. Reviews, 1993, 35, 249. G. E. Millward and A. Turner in Trace Elements in Natural Waters, ed. B. Salbu and E. Steinnes, CRC Press, Boca Raton, Florida, USA, 1995, pp. 223–245. A. W. Morris, A. J. Bale and R. J. M. Howland, Estuar. Coastal Mar. Sci., 1982, 14, 175. T. M. Church, Mar. Chem., 1986, 18, 393. J. C. Duinker, R. F. Nolting and D. Michel, Thalassia Jugosl., 1982, 18, 191. P. A. Yeats and P. M. Stain, Estuar. Coastal Shelf Sci., 1990, 31, 11. D. R. Ackroyd, A. J. Bale, R. J. M. Howland, S. Knox, G. E. Millward and A. W. Morris, Estuar. Coastal Shelf Sci., 1986, 23, 621. B. Ouddane, E. Martin, A. Boughriet, J. C. Fischer and M. Wartel, Mar. Chem., 1997, 58, 189. L. M. Holliday and P. S. Liss, Estuar. Coastal Mar. Sci., 1976, 4, 349. V. Subramanian and B. D’Anglejan, J. Hydrol., 1979, 29, 341.
J. Environ. Monit., 1999, 1, 489–496
495
15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45
496
R. M. Moore, J. D. Burton, P. J. L. Williams and M. L. Young, Geochim. Cosmochim. Acta, 1979, 43, 919. F. L. L. Muller, M. Tranter and P. W. Balls, Estuar. Coastal Shelf Sci., 1994, 39, 105. L. E. Fox and S. C. Wofsy, Geochim. Cosmochim. Acta, 1983, 47, 211. L. E. Fox, Nature, 1988, 333, 442. J. H. Martin, R. M. Gordon and S. E. Fitzwater, Nature, 1990, 354, 156. J. Wu, G. W. Luther III, Limnol. Oceanogr., 1994, 38, 1119. K. S. Johnson, R. M. Gordon and K. H. Coale, Mar. Chem., 1997, 57, 137. K. W. Bruland and M. L. Wells, Mar. Chem., 1995, 50, 1. N. M. Price, B. A. Ahner and F. M. M. Morel, Limnol. Ocean., 1994, 39(3), 520. S. Takeda, A. Kamatani and K. Kawanobe, Mar. Chem., 1995, 50, 229. M. Gledhill and C. M. G. Van den Berg, Mar. Chem., 1995, 50, 51. C. M. G. van den Berg, Mar. Chem., 1995, 50, 139. R. F. Nolting, L. J. A. Gerringa, M. J. W. Swagerman, K. R. Timmermans and H. J. W. de Baar, Mar. Chem., 1998, 62, 335. A. E. Witter and G. W. Luther III, Mar. Chem., 1998, 62, 214. J. Avoine, J. Geol. Soc., 1986, 144, 135. J. Avoine, D. Boust and J. F. Guillaud, Flux et comportements des contaminants dissous et particulaires dans l’estuaire de la Seine, Rapp. P.V. Re´un. Cons. Int. Explor. Mer, 1986, 186, 392. B. Ouddane, M. Skiker, J. C. Fischer and M. Wartel, Analusis, 1997, 25, 308. UNESCO, Technical Papers in Marine Science, Background Papers and Supporting Data on the Practical Salinity Scale, 1981, vol. 37, pp. 1–144. A. Tessier, P. G. C. Campbell and M. Bisson, Anal. Chem., 1979, 51, 844. A. Tessier and P. G. C. Campbell in Metal Speciation: Theory, Analysis and Application, ed. J. R. Kramer and H. E. Allen, Lewis, Chelsea, UK, 1988, pp.183–194. J. C. Duinker, M. T. J. Hillebrand, R. F. Nolting and S. Wellershaus, Neth. J. Sea Res., 1982, 15, 141. J. J. Hung, Terr. Atmos. Ocean. Sci., 1990, 1/3, 257. J. M. Martin, D. M. Guan, F. Elbaz-Poulichet, A. J. Thomas and V. V. Gordeev, Mar. Chem., 1993, 43, 185. C. Guieu, W. W. Huang, J. M. Martin and Y. Y. Yong, Mar. Chem., 1996, 53, 255. L. M. Meyer, Geochim. Cosmochim. Acta, 1982, 46, 2527. P. W. Balls, R. E. Laslett and N. B. Price, Neth. J. Sea. Res., 1994, 33, 1. M. Dai and J. M. Martin, Earth Planet. Sci. Lett., 1995, 131, 127. A. R. Flegal, G. J. Smith, G. A. Gill, A. Sanudo-Wilhelmy and L. C. D. Anderson, Mar. Chem., 1991, 36, 329. B. Motamed and H. Texier in Hydrodynamique et Transport Se´dimentaire, Rapport 1995/FIN-1, Programme Seine Aval, 1996, pp. 21–31. J. F. Chiffoleau, D. Cossa, D. Auger and I. Truquet, Mar. Chem., 1994, 47, 145. J. M. Bewers and P. A. Yeats, Nat. Can. (Que.), 1979, 106, 149.
J. Environ. Monit., 1999, 1, 489–496
46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62
63 64 65 66 67 68 69 70 71 72 73 74 75
D. W. Evans, N. H. Cutshall and D. A. Wolfe, Estuarine Coastal Mar. Sci., 1977, 5, 71. R. Wollast, G. Billen and J. M. Duinker, Estuarine Coastal Mar. Sci., 1979, 9, 161. J. C. Duinker, R. Wollast and G. Billen, Estuarine Coastal Mar. Sci., 1979, 9, 727. H. Paucot and R. Wollast, Mar. Chem., 1997, 58, 229. S. Knox, D. R. Turner, A. G. Dickson, M. I. Liddicoat, M. Whitfield and E. I. Butler, Estuarine Coastal Shelf Sci., 1981, 13, 357. A. W. Morris, Mar. Pollut. Bull., 1984, 15, 57. J. T. Byrd, K. W. Lee, D. S. Lee, R. G. Smith and H. L. Windom, Estuaries, 1990, 13, 8. L. Yan, R. F. Stallard, R. M. Key and D. A. Crerar, Chem. Geol., 1990, 85, 369. P. W. Laslett and P. W. Balls, Mar. Chem., 1995, 48, 311. F. Dehairs, W. Baeyens and D. Van Gansbeke, Estuarine Coastal Shelf Sci., 1989, 29, 457. J. M. Garnier, J. M. Martin, J. M. Mouchel and K. Sioud, Mar. Chem., 1996, 53, 269. G. W. Lubbers, W. W. C. Gieskers, P. del Castilho, W. Salomons and J. Bril, Mar. Ecol. Prog. Ser., 1990, 59, 285. M. Loijens and R. Wollast, Wat. Pollut. Res. Rep., 1992, 28,309–320. A. Turner, Mar. Chem., 1996, 54, 27. M. Wartel, M. Skiker, Y. Auger, A. Boughriet, E. Puskaric and P. Guegueniat, Mar. Chem., 1991, 36, 85. A. Turner, G. E. Millward, A. J. Bale and A. W. Morris, Estuarine Coastal Shelf Sci., 1993, 36, 1. J. M. Martin, J. M. Mouchel and J. Biscan, in Biogeochemical Processes at the Land-Sea Boundary, ed. P. Lasserre and J. M. Martin, Elsevier, Amsterdam, The Netherlands, 1986, pp. 53–71. W. Stumm, Aquatic Surface Chemistry. Chemical Processes at the Particle-Water Interface, Wiley, New York, USA, 1987. M. B. McBride, Soil Sci. Soc. Am. J., 1979, 43, 693. J. Biscan and U. Dragcevic, Mar. Chem., 1993, 43, 127. A. C. M. Bourg, in Chemistry and biology of Solid Waste, ed. W. Salomons and W. Fo¨rstner, Springer-Verlag, Berlin, Germany, 1988. B. M. Moudgil and S. Behl, J. Colloid Interface Sci., 1991, 146, 1. M. L. Franklin and J. W. Morse, Mar. Chem., 1983, 12, 241. A. Boughriet, B. Ouddane and M. Wartel, Mar. Chem., 1992, 37, 149. A. Boughriet, B. Ouddane, J. C. Fischer, M. Wartel and G. Leman, Wat. Res., 1992, 26, 1359. J. J. Sawland and J. W. Murray, Earth Planet. Sci. Lett., 1983, 64, 213. G. L. De Lange, Geochim. Cosmochim. Acta, 1986, 50, 2543. Y. H. Li, J. Bischoff and G. Mathieu, Earth Planet. Sci. Lett., 1969, 7, 265. J. De Kanel and J. W. Morse Geochim. Cosmochim. Acta, 1978, 42, 1335. P. M. Shanbhag and J. M. Morse, Geochim. Cosmochim. Acta, 1982, 46, 241.
Paper 9/03721G
