Environ. Sci. Technol. 2001, 35, 4271-4276
113
Cd-NMR and Fluorescence Studies of the Interactions between Cd(II) and Extracellular Organic Matter Released by Selenastrum capricornutum M A R I A G R A S S I * ,† A N D MARINA MINGAZZINI‡ Dipartimento di Chimica Inorganica, Metallorganica e Analitica, Universita` di Milano, Via G. Venezian 21, I-20133 Milano, Italy, and Istituto di Ricerca sulle Acque, C.N.R, Via della Mornera 25, I-20047 Brugherio (MI), Italy
113Cd NMR spectra were measured in solution for a series of adducts between the extracellular organic matter (EOM) of the green alga Selenastrum capricornutum and cadmium(II). From the results it appears that EOM forms complexes with Cd(II), which are in a fast exchange in NMR time scale. Thus the observed shift is the molar average of limit values for the exchanging free and bound cadmium species. Definite support for this dynamic stems from an extensive 113Cd NMR equilibrium analysis. Although in principle a multiple binding mode cannot be excluded, our 113Cd NMR findings and the consideration that carbohydrates are the prevailing constituents of algal releasing in stationary growth phase led us to suggest a carbohydrate type coordination. This hypothesis is also supported by NMR studies on model compounds. However, this is probably an oversimplified view of the binding which does not properly account for the fluorescence results. Fluorescence spectroscopy provided further evidence of the EOM binding process also allowing a discrimination between the fluorophoric groups involved. The addition of Cd(II) affected both fluorescence intensity and peak position in the “humiclike” band (between 340 and 400 nm Ex). Finally, the 113Cd NMR and synchronous fluorescence measurements showed a linear correlation between 113Cd chemical shifts and EOM concentration (as fluorescence intensity at 340 and 360 nm). NMR and fluorescence data suggest the existence of structurally different binding sites (carbohydratelike and humic-like) which appear somehow related. Further, they definitely put in evidence the EOM-Cd interaction together with the specific organic components mostly involved.
Introduction The study of the interaction between heavy metals and biological organisms and derivatives such as algae, plantderived materials, bacteria, yeast, fungi, and lichens is of paramount importance in environmental science and technology for the control of the toxic effects on living systems, * Corresponding author phone: + 39-02-2664203; fax: + 39-022362748; e-mail:
[email protected]. † Universita ` di Milano. ‡ Istituto di Ricerca sulle Acque. 10.1021/es0018769 CCC: $20.00 Published on Web 09/26/2001
2001 American Chemical Society
remedy procedures, and useful industrial and analytical applications (1). Biomass materials have been shown to bind metals with high capacities and selectivities (2) thus making biosorption a highly cost-effective alternative for the treatment of contaminated water and for the recovery of precious metals (3, 4). More recently, biomass based processes have also been exploited for analytical purposes (5, 6). The binding capacity of algal cell (7-12) and humic material (13-17) has been the topic of several structural and mechanistic studies. By contrast, the role played in the binding process by the algal EOM (extracellular organic matter) is still rather unexplored. This material has a complex and composite natureswhich varies as a function of both algal species and growth phasesscontaining a variety of compounds such as amino acids, proteins, polysaccharides, carbohydrates, etc. (18) with a potentially high affinity toward metal ions. In this regard, as early as 1955, complexation and detoxification of copper by extracellular polypeptide from a culture of Anabaena cylindrica was observed (19). Later, by means of potentiometric techniques, the copper complexing properties of EOM released by several algal species were characterized (20). More recently, capillary electrophoresis was used to identify the nature of thiol-polypeptide oligomers produced by some marine diatoms in response to cadmium exposure (21). However, to our knowledge, no direct spectroscopic evidence of EOM-Cd binding has been reported. 113Cd NMR spectroscopy has been widely used as a standard technique for studying the metal surrounding in biological (22) and environmental systems (23) due to the NMR favorable parameters and the extreme sensitivity of the nucleus to its coordination sphere. In previous papers the great potential of synchronous fluorescence spectroscopy to distinguish and measure the EOM components produced in different algal growth phases was shown (24-26). By combining 113Cd NMR and fluorescence spectroscopy we have recently analyzed the Cd(II) binding capacity of EOM released by different species of marine phytoplankton (27). A similar approach was used in this work, aiming to investigate the interaction between Cd(II) and the EOM produced by the freshwater alga Selenastrum capricornutum.
Methods and Materials Chemicals. Cd(ClO4)2‚nH2O, benzoic acid, galactose, and glycine were purchased from Fluka and used without further purification. Preparation of Model Compounds. Galactose, benzoate, and glycinate cadmium adducts were prepared as model compounds. The samples for 113Cd NMR analysis were obtained by adding the ligand to a solution of Cd(ClO4)2 0.05 M in water (in 2:1 molar ratio). The opportune pH corrections were made with diluted HClO4 or KOH. Algal Material. The green alga Selenastrum capricornutum (renamed Raphidocelis subcapitata) from IRSA stock cultures was used as EOM producer organism. Over a few months S. capricornutum was inoculated weekly and grown in EPA (28) medium (at a concentration factor of 2). Standard culturing conditions (29) were adopted, except for lighting. After exponential growth (about 1 week after the inoculation) the cultures were exposed at progressively lower light conditions to extend the aging of living cells to months. This procedure, as already described (25), was adopted as a strategy to favor the natural accumulation in the medium of the EOM compounds released during the stationary growth phase. While no antibiotic treatment was made, extreme care was used to avoid contamination of cultures and EOM samples. After aseptic transfer in sterile medium, each culture was VOL. 35, NO. 21, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4271
opened only once, at the end of culturing. After few months, a series of differently aged cultures (ranging from 0 to about 200 days after inoculation) was simultaneously used for EOM sampling. The cultures were filtered (Millipore HA 0.45 µm filters) under reduced pressure, and the filtered medium was immediately analyzed for the fluorimetric characterization of the EOM. Spectrofluorimetric Characterization and Selection of EOM Samples. Synchronous fluorescence spectroscopy was used as a multicomponent analysis able to distinguish and measure the EOM components produced in different algal growth phases (24-26). The analyses were performed using a SPEX FluoroMax spectrofluorimeter, equipped with a 150-W ozone-free Xenon lamp. Spectra were recorded in an excitation wavelength range of 250-500 nm with 25 nm ∆λ, 4.25 nm band-pass, 2 nm increment, and 1 s integration time. Details on the analytical procedures are given in a previous paper (26). As previously found for a number of phytoplanktonic species (25, 26, 30) two main fluorescence signals are present in the EOM spectra: the “amino acidic-like” component fluorescing in the UV band (peak A, 280 nm Ex max) and the “humic-like” component, which exhibits different peaks between 340 and 410 nm Ex and respectively labeled B (340 nm), C (360 nm), and D (410 nm). The “amino acidiclike” fluorescing component is released only in the exponential growth phase, while the “humic-like” component progressively accumulates in the extracellular medium also in stationary phase. Based on the EOM production trend over time, a series of EOM samples, characterized by a fairly constant “amino acidic-like” and progressively increased “humic-like” components was thus selected, to investigate the role of the latter in Cd(II) binding. The EOM samples were labeled S0-SN (N ) incubation days). The culture medium (S0) was used as a blank. A series of samples was then stored in sterile dark-glass bottles at 4 °C and used within 72 h for Cd(II) addition (S0-S190) or acidic treatment and Cd(II) reaction (S0-S168). Samples ranged from colorless to pale yellow and from pH 7.5 to 8.5, according to the culture aging. No further chemical analysis was performed on this material; therefore, the volume is the parameter hereafter used to express EOM relative concentration. Preparation of the Adducts: EOM-Cd (SN-Cd). These adducts were prepared by reacting the EOM series with a stock solution of Cd(ClO4)2 in H2O in such a way as to keep a fixed Cd(II) concentration and to dilute all the original EOM to the same extent. Typically 1 mL of 0.15 M Cd(II) solution was added to 2 mL of EOM to produce a final concentration of 0.05 M in Cd(II). A blank was prepared similarly by reacting Cd(II) stock solution and algal medium (S0). In a case of the S0-S190 series a clear solution was obtained for the first SN-Cd adducts (N ) 0-10). From S13 to S190, an increasing opalescence was instead observed, which eventually became a heavy precipitate. This material was isolated and characterized by CHN determination, Cd content (by AA spectroscopy, after HNO3 digestion), and IR spectroscopy. It was shown to be CdCO3 contaminated by traces of organic material. The clear solution (S0-S10) and the isolated supernatant (S13-S190) of these SN-Cd adducts were then analyzed. Their pH was measured from pH 6.1 to 5.5, following aging time. pH ) 6.3 was found for the blank S0-Cd adduct. EOM Carbonate-Free (SN*) and Respective Cd(II) Adducts (SN*-Cd). A newly selected set of EOM samples (S0S168) was treated to eliminate the inorganic carbon. Typically, a measured volume of original EOM (5-10 mL), acidified at pH 2 with 1 M HClO4, was heated (50-60 °C) for about 30′, while N2 was bubbled into the solution. The pHs did not vary during the acidic treatment, but an average volume loss (about 10-15%) was generally observed. The original pH 4272
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 21, 2001
and volume were restored with KOH and H2O on the assumption that water was the prevailing volatile component in the system. The resulting EOM carbonate-free samples (SN*), after fluorescence analysis, were reacted with Cd(II) as reported above to produce clear solutions of EOM-Cd adducts (SN*-Cd) series. Equilibrium Studies. The experiments were performed on a carbonate free EOM sample containing a significant amount of exudates. Samples for dilution study (X and Y series) were prepared by reacting the following: 0.2 mL of the stock Cd(II) solution (0.5 M) and 0.8 mL of EOM (X) and 0.2 mL of stock Cd(II) solution, 0.4 mL of the same EOM, and 0.4 mL of H2O (Y). The samples were then progressively diluted with H2O from 1/2 up to 1/8. The analysis of EOM/ Cd ratio variation was performed on the A1-A6 series characterized by 1 mL of final volume, constant Cd(II) 0.1 M, and progressively increased EOM amount. Adducts A1A4 were obtained by reacting 0.2 mL of stock Cd(II) solution, an increasing volume of EOM (from 0.2 to 0.8 mL) and the opportune quantity of water to reach 1 mL final volume. Adducts A5 and A6 were prepared by taking to dryness, under low pressure, 1 and 2 mL of EOM, respectively, and quantitatively redissolving the solid residue in 0.8 mL of EOM. The respective concentrated EOM solutions were then reacted with 0.2 mL of stock Cd(II) solution. On the assumption that water is the only volatile component present in the EOM moiety, these samples contain an amount of exudates which formally corresponds to 1.8 and 2.8 mL of original EOM. For both the experiments a blank S0-Cd adduct was prepared by reacting the stock Cd(II) solution and the pure culture medium, opportunely diluted or concentrated as referred to above. NMR Measurements. Since Cd(II) ligand equilibration was deemed a fast process, no delay was introduced between preparing and analyzing all the adducts of this work. Spectra were recorded using a Bruker Advance DRX 300 Fourier transform spectrometer. 113Cd measurements were performed at 300 K in 5 mm tubes with a coaxial capillary filled with D2O as locking substance. Chemical shifts (δ) are in ppm referred to Cd(ClO4)2 (0.1 M in H2O). By convention, values at a lower frequency relative to the reference are denoted as negative. The resonances of these adducts are independent of 1H decoupling and display short T1 (approximately 1 s). Using pulse width for 30° and relaxation delay 1 s, spectra with good signal-to-noise ratio were obtained in 10′-2 h running time.
Results and Discussion The following data refer to the experiments performed on the (S0-S190)-Cd complexes. The result of 113Cd NMR measurements for this set of adducts is reported in Figure 1. The spectrum generally consists of a single sharp peak at positive value in the range 1.3-6.6 δ. The observed deshielding is small and rather variable (from 1.3 to 1.7 δ) for the first SN-Cd adducts (S7-S70), while it increases, although irregularly, for the latter samples of the series up to 6.6 δ (for S190-Cd adduct). Incidentally, to account for this variability it should be recalled that when a dynamic process involves the nucleus, the observed 113Cd shift is strongly concentration dependent. Because of the precipitation, which had occurred in some cases, the cadmium content varies between the samples thus preventing a rigorous comparative NMR analysis of this EOM series. These results can neither be explained by inorganic ligand effect nor by a pH factor. The shift at 1.6 δ, measured for the blank S0-Cd adduct, accounts for the maximum shielding factor on 113Cd nucleus by inorganic ligands. Following the algal uptake, the concentration of inorganic nutrients in the medium as well as their effect on the 113Cd nucleus should decrease over time. Second, a lowering of pH reduces the
TABLE 1. 113Cd NMR Data for EOM-Cd Dilution Studyb [Cd(II)] M
series X
δ (ppm)a
series Y
δ (ppm)a
0.1 0.05 0.033 0.025 0.0125
X X1 X2 X3 X4
5.0 4.5 4.1 3.9 3.3
Y Y1
2.9 2.7
Y3 Y4
2.4 2.1
a Referred to Cd(ClO ) 0.1 M in H O. b EOM/Cd ratio (V/M) is constant, 4 2 2 8 (X) and 4 (Y), in each series, while the concentration of EOM-Cd adducts is progressively decreased.
FIGURE 1. 113Cd NMR spectra for EOM-Cd adducts. From top to bottom spectra refer to S190-Cd, S166-Cd, and S7-Cd samples; the respective chemical shifts (in ppm referred to Cd(ClO4)2 0.1 M in H2O) are shown above each resonance. A progressive deshielding is observed on 113Cd nucleus according to EOM accumulation in culture medium. binding capacity of a weak protic ligand thus shifting the 113Cd resonance toward zero (i.e. to the unbound cadmium species). In our case pH and 113Cd shift values follow the opposite trend: as the former decreases regularly in the series (S0-Cd, S190-Cd) from pH ) 6.3 to pH ) 5.5, the latter increases from 1.6 to 6.6 δ. Our NMR findings then definitely put in evidence EOM-Cd interactions which vary according to EOM aging time. This result is in principle consistent with a temporal variation either of EOM chemical nature or EOM concentration. Fluorescence spectroscopy proved to be effective in structural identification and in monitoring the variation of the DOM (dissolved organic matter) (31). Since the same approachsused in this work to characterize the EOM samplessshowed spectral features of the extracellular compounds unchanged over time with increasing fluorescence intensity according to the accumulation in the medium, the EOM chemical content was assumed qualitatively constant and quantitatively rising. On these bases the NMR results suggest the presence of an equilibrium, fast in the NMR time scale, between free and bound Cd species, the observed shift being the molar average of the limit values of the individual species (32) and a measure of the binding extent.
Cdfree + L ) CdL
(1)
Moreover, since δ of pure Cdfree in water (i.e. the hexa aquo Cd(II) species) is the reference zero for 113Cd ppm scale, our findings indicate that the limit shift for pure Cdbound is at positive δ value. The 113Cd chemical shift in solution reported in the literature spans an interval of approximately 850 ppm. This parameter is strongly affected by the nature of the ligands coordinated to the metal. In particular, ligands linked through carboxylate shield the nucleus, while ligands linked through hydroxy, sulfur, and nitrogen deshield it (33, 34). Therefore our data suggest that Cd(II) is primarily bound to EOM components via N, S, or OH coordination. These results are in good agreement with our previous finding on Cd-EOM adducts in marine environment where a similar trend (as a function of scalar EOM concentration) was observed (27). Finally, by comparing the fluorescence spectra of the SN samples and the respective SN-Cd adducts, some modifications were observed for the cadmium derivatives (particularly a quenching at 410 nm and the rising of a broad band between 330 and 400 nm). These results might indicate fluorimetric evidence of EOM-Cd interaction. However, they can also
simply be explained in terms of the loss of organic components due to the precipitation and coprecipitation processes. Thus further experimentation is required to highlight this point. Equilibrium Study. It could be argued that the analysis of these materials, based on fluorescence data, might not represent the entire sample, as suggested in the case of humic material (35). Indeed, it has been observed that only a fraction of phytoplanktonic DOM has fluorescence properties (36). Therefore the equilibrium hypothesis was further investigated as follows. In the case of a dynamic process between Cdfree and Cdbound, the position of the equilibrium and the resulting chemical shift should be affected both by dilution and by variation of EOM/Cd ratio. To analyze the first effect, two Cd-EOM adducts (characterized by different amounts of the same EOM) were then prepared and diluted (X and Y series). The results of 113Cd NMR measurements on these samples are reported in Table 1. The data are in agreement with the equilibrium 1: the chemical shifts of the X and Y complexes tend, upon dilution, toward zero ppm (i.e. the chemical shift for Cdfree species in water). In other words, dilution of the complexes promotes their dissociation and the molar ratio of Cdfree species increases. As expected, the variation of δ is EOM concentration dependent. In particular, the decrease in chemical shift is more significant for X (∆δX-X4 ) 1.7) than for Y (∆δY-Y4 ) 0.8) series. These figures correlate well with the EOM concentration ratio (2/1) in the samples. Moreover, the experimental data fit very well a “saturation-like” curve represented by the equation
δ ) δmaxb[Cd]/(1 + b[Cd])
(2)
where δmax correspond to the maximum EOM complexing capacity (curve plateau), b is a coefficient related to the affinity between Cd(II) and EOM and [Cd] is the total cadmium concentration. Incidentally, this Langmuir type equation does not derive from mechanistic study but simply reflects the experimental curves. The equation can be rearranged in the form
([Cd])/δ ) (1 + b[Cd])/δmaxb
(3)
By plotting ([Cd])/δ versus [Cd], Figure 2, the slope of the straight line so obtained gives δ max 5.5 and 3.1 for the X and Y complexes. It is worthwhile noting that these values are not the limit chemical shifts for pure Cdbound (Cd(II) being in excess compared to EOM) but represent the maximum binding capacity, in the 113Cd NMR scale, of the EOM in the samples. Samples for the study of the variation of EOM/Cd (V/M) ratio were prepared by keeping [Cd(II)] and final volume constant and varying the volume of the same EOM. The results of 113Cd NMR measurements on these samples are reported in Table 2. The analysis of these data further supports the hypothesis of a dynamic process between free and bound cadmium species. The chemical shift increases with the ratio VOL. 35, NO. 21, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4273
FIGURE 2. 113Cd NMR results on diluting EOM-Cd complexes (X and Y series). Plot of data in Table 1 following eq 3. X and Y adducts are initially 0.1 M in Cd(II) and contain 0.8 and 0.4 mL of EOM, respectively. From the reverse slope of the lines, X and Y δmax values are obtained.
TABLE 2. 113Cd NMR Data for Variation of EOM/Cd (V/M) Ratioc sample
µL EOM
δ (ppm)a
A1 A2 A3 A4 A5 A6
200 400 600 800 1800b 2800b
1.8 2.9 4.1 5 11 17
a
Referred to Cd(ClO4)2 0.1 M in H2O. b The value refers to the original EOM volume formally added (see Experimental Section). c Cd concentration (0.1 M) and total volume (1 mL) are kept constant, while the EOM volume is progressively increased.
TABLE 3. 113Cd NMRa and Fluorescenceb Data for EOM* and Respective Cd Adducts
SN*
δ
hA (280 nm)c
S0d S7d S61 S98 S111 S126 S161 S168
1.7 1.8 2.4 2.5 3.1 2.8 4.6 5.1
257 378 269 208 552 634 459
hB (340 nm)c
hC (360 nm)c
hD (410 nm)c
pH (SN*)
pH (SN*-Cd)
31 138 145 452 313 758 861
24 112 167 471 382 949 1113
15 290 277 634 795 1147 1467
7.7 8.0 8.2 8.2 8.4 8.2 8.5 8.5
6.2 6.2 6.3 6.3 6.5 6.1 6.4 6.5
a Chemical shift in ppm referred to Cd(ClO ) 0.1 M in H O. b Height 4 2 2 of peak in F.U. [1 F.U.) 1000 cps]. c λ referred to excitation wavelength. d Untreated sample.
EOM/Cd thus indicating a progressively higher amount of Cdbound with respect to Cdfree form. Finally, the 113Cd resonance of the blank S0-Cd adduct appeared almost unaffected by diluting or concentrating in a similar manner the respective S0 sample. This result clearly rules out the culture medium involvement in the experiments previously reported. Correlation NMR Fluorescence. To obtain data suitable for comparative analysis, the set of EOM samples (S0-S168), pretreated to eliminate the Carbonate (SN*), was studied. After the acidic stripping of CO2, the samples were analyzed again by fluorescence spectroscopy to test for modifications in the organic matrix, the respective data being reported in Table 3. A comparison between the fluorescence spectra of SN and SN* samples (Figure 3) shows some variations in band intensities upon treatment. In particular, the main effects are observed at 410 nm (band D: quenching) and at 360 and 340 nm (bands C and B: small enhancement), while band A at 280 nm appears almost unaffected. These findings indicate that the EOM matrix is somehow modified. Nevertheless, the binding capacity of the system is not altered, as proved by the following 113Cd NMR results. Moreover, 4274
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 21, 2001
FIGURE 3. Synchronous fluorescence spectra of an EOM 161 days old (before and after CO2 stripping) and of the respective Cd(II) adduct. Intensity (cps) is plotted versus excitation wavelength (nm). By comparing the original (SN) and the treated (SN*) samples quenching is observed at 410 nm (band D) upon CO2 stripping. By comparing the treated (SN*) sample and the respective cadmium complex (S161*-Cd) enhancement and profile variation (between 325 and 400 nm) are observed upon Cd(II) coordination. since the variations of the fluorescence spectra are very similar in the EOM series, a close similarity in scalar EOM content is assumed between SN and SN* samples. The adducts (SN*Cd), obtained by reacting the SN* samples with Cd(II), were then analyzed by 113Cd NMR and fluorescence spectroscopy. The resultant 113Cd NMR parameters (Table 3) are very similar to those measured for the previous SN-Cd adducts (S0S190 series) and the same considerations apply regarding the following: S0 effect, pH factor, and the nature of exchanging species. Moreover, the close similarity found between the 113Cd NMR parameters of SN-Cd and SN*-Cd series is clear evidence that the binding capacity of EOM is not substantially modified by the acidic treatment. As expected, this series is characterized by a regular variation of 113Cd chemical shift (from 1.7 to 5.1 δ) according to alga aging time. Apparently the S126* sample does not follow the temporal trend and displays a relatively low deshielding effect on 113Cd nucleus. However this behavior is associated with a similarly low content of fluorophoric groups at 340 and 360 nm. Since SN samples are from different cultures, their aging is an approximate measure of the EOM content, the fluorescence intensity being the tool to evaluate this parameter. Finally, the analysis of the fluorescence spectra allows a deeper insight into the nature of the binding process. The addition of Cd(II) produces a clear modification of EOM profiles for all the samples (as shown in Figure 3 in the case of S161*-Cd adduct). As a general trend, while bands A and D appear almost unaffected, variation of intensity (enhancement) and peak position is observed between 325 and 400 nm (bands B and C). Further, by plotting the SN* fluorescence intensities versus 113Cd shifts for the respective Cd(II) adducts, an almost linear correlation is found for bands B and C, while band D correlates rather poorly and no correlation is found for band A. These findings are significant for several reasons. First they show clear evidence of EOM-Cd interaction, even allowing a discrimination between the fluorophoric groups (and respective components) involved in the process. Indeed, the relevance of fluorescence spectroscopy for studying the interaction between DOM and metal ions is well documented (37). Nevertheless, the predominant effect observed on different fulvic acid samples was a quenching of fluorescence in binding heavy metals and paramagnetic ions (38). No modifications were instead found for Cd(II) complexes, and doubts were expressed on the utility of the fluorescence
method in this case (39). Second, our results point to a coordination process between Cd(II) and “humic-like” EOM components (particularly the fractions which accumulate in the algal medium and display fluorescence at λ ) 340 and 360 nm). Finally, the intensities of the same EOM components correlate linearly with the δ of the respective Cd derivatives. This result could be read as an independent variation of δ along the SN*-Cd series and a parallel increasing of “humiclike” fraction with EOM age. Nevertheless it might also suggest that these EOM components are somehow involved in deshielding the Cd nucleus. In this context the two binding processes of this workseither monitored by NMR or by fluorescence spectroscopysseem amenable to the same chemical background embedded in the “humic-like” EOM core. It could be argued that the Cd(II) concentration in our samples is very high and that in the case of a “real” EOM binding event modification in fluorescence properties should be effective at significantly lower cadmium level. Further, the pH of our adducts is not kept constant, and this fact might also significantly affect the fluorescence finding of this system. To clarify these points a single EOM was reacted with Cd(II) to produce adducts with constant EOM and variable Cd(II) concentration. The fluorescence analysis of these samples showed a regular increasing of intensity according to Cd(II) concentration. Particularly, by varying the Cd(II) (from 1.12 × 10-3 to 28 × 10-3 M) a progressive rising of bands B and C was observed, while band A varied irregularly and D was almost unaffected. These results parallel well the findings for the SN*-Cd series and strongly support our assumption of a binding between Cd(II) and the EOM components fluorescing at 340 and 360 nm. Instead, no relevant modifications in fluorescence spectra were observed when the pH of a single adduct was regularly increased from pH ) 5.5 to pH ) 7.5. This fact then definitely rules out a pH effect in our results. Nature of Cd-EOM Binding. To ascertain the nature of the functionality mostly involved in Cd(II) binding, a number of adducts representative of Cd-COO-, Cd-N, and Cd-OH coordination were prepared and analyzed by 113Cd NMR spectroscopy. The measurements were performed at controlled pH because of the expected 113Cd shift dependence for ionizable species. At pH ) 6 benzoate-Cd resonance was detected at -13 δ, as expected for this kind of functionality. Glycine was chosen for modeling cadmium nitrogen interaction because of its solubility in water and the consideration, based on previous research (40), that at pH > 4 the Cd-amino coordination becomes dominant thus making the Cd113 NMR shift vary from negative to positive value. Rather surprisingly, no signal was detected for glycinate-Cd compound in our experimental conditions (probably due to a slower exchange rate for this complex). However, on the basis of 113Cd NMR reference values (40), δ > 30 is expected for this adduct at pH ) 6. Further, line broadening of 113Cd resonance due to intermediate rate exchange, as a function of pH, has been observed with several nitrogen ligands (41). Accordingly, by increasing the pH of our sample no NMR signal was detected up to pH ) 7.5 where a broad resonance appeared at 130 δ, thus indicating both an increased exchange rate and the expected marked deshielding for Cd-N coordination. Finally, galactose-Cd showed at pH ) 6 a sharp 113Cd resonance at 0.9 δ, which parallels quite well the results obtained for EOM-Cd complexes. Incidentally, it is rather difficult to put in evidence Cd-OH interactions in water, and to our knowledge no reference data are available in this regard. In fact, H2O and OH molecules coordinate the Cd(II) very similarly, and the resulting equilibrium, due to mass effect, is strongly driven toward the Cdaquo form (i.e. 0 ppm in 113Cd scale). Further, in our previous attempt to monitor this coordination on
model molecules, no 113Cd deshielding was observed in the case of phenol, while dihydroxy-benzoate ligands produced a definite shielding effect (42). In this context the 113Cd NMR downfield resonance for galactose-Cd adduct is taken as significant evidence for Cd-OH binding. As a result of these measurements we suggest that the hydroxyl containing components are most likely responsible for the observed chemical shift of the EOM-Cd adducts. In other words, although a combining effect of several functionalities cannot be excluded, it is reasonable to assume that the EOM-Cd interactions involve primarily OH groups probably from carbohydrate type species. Conversely, a carbohydrate qualitative analysis, performed by HPLC-MS (high performance liquid chromatography-mass spectrometry) on a EOM sample of this work, after acidic digestion, indicated the presence of rhamnose, arabinose, glucose, and galactose. These results are in good agreement with the chemical characteristics, which indicated carbohydrate compounds to be the main constituents of phytoplanktonic releasing (43). Physiological studies on the extracellular release by marine diatoms provided evidence that carbohydrate material sometimes comprises 80-90% of the total EOM produced in stationary growth phase (44). Moreover, a study by Marchetti et al. (45) showed that polymeric carbohydrates, particularly galactans, are the prevailing components of phytoplankton EOM which originated the marine mucilages in Adriatic sea. More recently, an investigation of the carbohydrate production dynamics by benthic diatoms provided evidence that these species release “colloidal carbohydrates” (water-soluble) during growth, the production reaching a maximum in the stationary phase and also continuing during periods of darkness (46). As a last consideration the EOM-Cd interactions described in this work, either followed via NMR or via fluorescence spectroscopy, seem amenable to the same EOM components. Our 113Cd NMR findings together with considerations based on the chemical characteristics of phytoplanktonic releasing led us to suggest a carbohydrate type coordination. However this is probably an oversimplified model which does not explain the fluorescence results. In particular, how carbohydrate type species can be related to the fluorescent matter at λ > 320 nm is rather obscure and would need further investigation. A separative and nondestructive procedure together with structural analysis might help in clarifying this point. Although the question remains open, the combined use of NMR and fluorescence spectroscopy proved to be effective in monitoring and analyzing the binding capacity of the EOM in this work. A similar approach might be valuable in following the whole sample behavior in other important environmental processes. The two techniques are in fact complementary, the first being very powerful in structural and dynamic analysis and the second being rather sensitive.
Acknowledgments The authors are very grateful to Dr. L. Mosca for the significant contribution in preparing and measuring the cadmium adducts of this work. Prof. F. Mangani, University of Urbino (PS), is also thanked for the EOM carbohydrate analysis. Finally, this research was partially supported by the Ministry of University and Scientific and Technological Research (MURST).
Literature Cited (1) Briesley, C. L. Geomicrobial. J. 1990, 8, 201. (2) Gadd, G. M. In Biotecnology; Rehm, H.-J. Ed.; VCH Verlagsgesellschaft: Weinheim, Germany, 1988; Vol. 6b, p 401. (3) Veglio, F.; Beolchini, F. Hydrometallurgy 1997, 44(3), 301. (4) Kratochvil, D.; Volesky, B. Trends Biotechnol. 1998, 16(7), 291. VOL. 35, NO. 21, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4275
(5) Madrid, Y.; Camara, C. TRAC-Trends Anal. Chem. 1997, 16(1), 36. (6) Yao, H.; Ramelow, G. J. Talanta 1998, 45(6), 1139. (7) Kieffer, E.; Sigg, L.; Schosseler, P. Environ. Sci. Technol. 1997, 31, 759. (8) Drake, L. R.; Lin, S.; Rayson, G. D. Environ. Sci. Technol. 1996, 30, 110. (9) Ke, H. D.; Rayson, G. D. Environ. Sci. Technol. 1992, 26, 1202. (10) Zhang, W.; Majidi, V. Appl. Spectrosc. 1993, 47, 2151. (11) Fourest, E.; Volesky, B. Environ. Sci. Technol. 1996, 30, 277. (12) Majidi, V.; Laude, D. A.; Holcombe, J. A. Environ. Sci. Technol. 1990, 9, 1309. (13) Li, J.; Perdue, M.; Gelbaum, L. T. Environ. Sci. Technol. 1998, 32, 483. (14) Larive, C. K.; Rogers, A.; Morton, M.; Carper, W. R. Environ. Sci. Technol. 1996, 30, 2823. (15) Chung, K. H.; Rhee, S. W.; Shin, H. S.; Moon, C. H. Can. J. Chem. 1996, 74, 1360. (16) Leenheer, J. A.; Brown, G. K.; Maccarthy, P.; Cabaniss, S. E. Environ. Sci. Technol. 1998, 32, 2410. (17) Lu, X.; Johson, W. D.; Hook, J. Environ. Sci. Technol. 1998, 32, 2257. (18) Fogg, G. E. Oceanogr. Mar. Biol. A Rev. 1966, 4, 195. (19) Fogg, C. E.; Westake, D. F. Proc. Int. Assoc. Threor. Appl. Limnol. 1955, 12, 219. (20) Mcknight, D. M.; Morel, F. M. M. Limnol. Oceanogr. 1979, 24(5), 823. (21) Torres, E.; Cid, A.; Fidalgo, P.; Herrero, C.; Abalde, J. Aquatic Toxicol. 1997, 39, 231. (22) Sadler, P. J.; Viles, J. H. Inorg. Chem. 1996, 35, 4490. (23) Grassi, M.; Gatti, G. Annal. Chim. Rome 1995, 85, 487. (24) Ferrari, G. M.; Mingazzini, M. Mar. Ecol. Prog. Ser. 1995, 125, 305. (25) Mingazzini, M.; Colombo, S.; Ferrari, G. M. Sci. Total Environ. 1995, 165, 133. (26) Mingazzini, M. In Mediterranean Ecosystems: Structure and Processes; Faranda, F. M., Guglielmo, L., Spezie, G., Eds.; Springer-Verlag: Berlin, 2000; Chapter 16, p 115.
4276
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 21, 2001
(27) Mingazzini, M.; Grassi, M.; Onorato, L.; Pedretti, E. Biol. Mar. Medit. 1999, 6(1), 704. (28) Short-term methods for estimating the chronic toxicity of effluents and receiving waters to freshwater organisms; EPA 600/4-85/ 014; Cincinnati, OH, 1985. (29) The Selenastrum capricornutum Printz algal assay bottle test; EPA/600/9-78/018; Corvallis, OR, 1978. (30) Mingazzini, M.; Onorato, L. Biol. Mar. Medit. 1998, 5(1), 755. (31) Cabaniss, S. E.; Shuman, M. S. Mar. Chem. 1987, 21, 37. (32) Carper, W. R.; Buess, C. M.; Hipp, G. R. J, Phys. Chem. 1970, 74, 4229. (33) Summers, M. F. Coord. Chem. Rev. 1988, 86, 43. (34) Johansson, C.; Drakenberg, T. Annu. Rev. NMR Spectrosc. 1989, 22, 1. (35) Seitz, W. R. Trends Anal. Chem. 1981, 1, 79. (36) Duursma, E. K. in Optical aspects of oceanography; Jerlov, N. G., Steemann Nielsen, E., Eds.; Academic Press: London, New York, 1974; p 237. (37) Sharpless, C. M.; Mcgown, L. B. Environ. Sci. Technol. 1999, 33, 3264. (38) Senesi, N. Anal. Chim. Acta 1990, 232, 77. (39) Saar, R. A., Weber, J. N. Anal. Chem. 1980, 52, 2095. (40) Wang, S. M.; Gilpin, R. K. Anal. Chem. 1983, 55, 493. (41) Kostelnik, R. J.; Bothner-By, A. A. J. Magn. Res. 1974, 14, 141. (42) Grassi, M.; Gatti, G. Annal. Chim. Rome 1998, 88, 537. (43) Fogg, G. E. Botanica Mar. 1983, 26, 3. (44) Myklestad, S. M. Sci. Total. Environ. 1995, 165, 155. (45) Marchetti, R.; Iacomini, M.; Torri, G.; Focher, B. Acqua-Aria 1989, 8, 883. (46) Smith, D. J.; Underwood, G. J. C. J. Phycol. 2000, 36, 321.
Received for review November 14, 2000. Revised manuscript received July 26, 2001. Accepted July 26, 2001. ES0018769
