Biosorption of Lanthanides Using Select Marine Biomass

Contamination of toxic metals in the aquatic environment is one of the most debated problems in the world with industrial development. Thus, the minimization and recovery of harmful pollutants such as heavy metals in natural environment is very significant [1]. Various treatment technologies such as ion exchange, precipitation, ultrafiltration, reverse osmosis and electrodialysis have been used for the removal of heavy metal ions from aqueous solution [2]. However, these processes have some disadvantages, such as high consumption of reagent and energy, low selectivity, high operational cost, and difficult further treatment due to generation of toxic sludge [3].


Introduction
Contamination of toxic metals in the aquatic environment is one of the most debated problems in the world with industrial development.Thus, the minimization and recovery of harmful pollutants such as heavy metals in natural environment is very significant [1].Various treatment technologies such as ion exchange, precipitation, ultrafiltration, reverse osmosis and electrodialysis have been used for the removal of heavy metal ions from aqueous solution [2].However, these processes have some disadvantages, such as high consumption of reagent and energy, low selectivity, high operational cost, and difficult further treatment due to generation of toxic sludge [3].
Among environmentally friendly technologies for the removal of heavy metals from aquatic effluent, biosorption has attracted increasing research interest recently [4][5].The major advantages of biosorption are its high effectiveness in reducing the heavy metals and the use of inexpensive biosorbents [6].Biosorption studies using various low cost biomass as adsorbents have been currently performed widely for the removal of heavy metals from aquatic effluent [7][8][9][10][11][12][13][14][15][16][17][18].Among many biosorbents, marine seaweed can be an excellent biosorbent for metals because it is well known to concentrate metals [19][20].Seaweeds are reported to accumulate hydrocarbons (as well as metals); and they are exposed to the ubiquitous presence of organic micropollutants and can work as suitable biomonitors [21].Furthermore, it is considered that the shell (usually treated as waste material) can be also an promising adsorbent.The shell has an internal structure comprised of three distinct layers.The innermost layer (i.e., hypostracum) consists of aragonite; the middle layer (i.e., ostracum), which is the thickest of the three, consists of various orientations interbedded with protein molecules (conchiolin); and the outermost layer (i.e., periostracum) consists of chitin, which is represented as (C8H13NO5)n [22].Particularly, it is considered that protein (called It is known that alginate is an exopolymer extracted mainly from brown algae (and various bacteria) that has been used both as immobilization material and as biosorbent of several heavy metals [29].Then, biosorption studies using seaweed have been generally concentrated on brown algae so far [30][31].Green and red algae as well as brown algae were also used for biosorbent of REEs in the present work.
Considering the above-mentioned, laboratory model experiments for confirming the efficiency of marine biomass (seaweed and shell) as sorbent for REEs was designed in present work.Furthermore, the surface morphology of the marine biomass used in this work was determined by SEM (Scanning Electron Microscope) before and after metal adsorption.
The crystal structure, and the specific surface area of the shell biomass were also determined by by XRD (X-ray powder diffraction), and BET (Brunaeur, Emmet and Teller) and Langmuir method, respectively.

The seaweed biomass
Many kinds of seaweeds samples (10 species of green algae, 21 species of brown algae and 21 species of red algae) were taken along several coasts in Niigata Prefecture (referred to the figure in our previous paper [32]) since April, 2004.Among seaweed species, the seaweeds for biosorbent used in this work were Sargassum hemiphyllum (brown algae), Ulva pertusa (green algae) and Schizymenia dubyi (red algae).Each seaweed sample was washed in the surrounding seawater to remove attachment at sampling place.After transport back to the laboratory, the seaweed was first washed with tap water and ultrapure water thoroughly and then air-dried for 2-3 days.Afterwards, it was dried overnight in an electric drying oven (Advantec DRA 430DA) at maximum temperature of 55 °C to avoid degradation of the binding sites, the biomass was ground.Sizes of biomass ranging from 0.5 mm to 1 mm were obtained by passing through sieves (SANPO Test Sieves).
Based on Diniz and Volesky's study [31], each sieved biomass sample was loaded with Ca 2+ in a solution of 50 mmol・dm -3 Ca(NO3)2 (biomass concentration of 10 g・dm -3 ) for 24 h under gentle agitation in order to remove the original cations on seaweed.Later, the biomass was washed with ultrapure water to remove excess Ca 2+ until the mixture was reached approximately pH 5. Finally, the washed biomass was dried again overnight at 50°C in an electric drying oven, and stored in desiccators (containing silica gel as a desiccant) before use.

The shell biomass
Buccinum tenuissimum shellfish used for shell biomass were collected at fishermen's cooperative association.After being separated from the meat by boiling, organism shells were washed thoroughly with ultra-pure water after washed with tap water repeatedly.After drying, the shells were ground and sieved through a sieve (SANPO Test Sieves) to remove particles having size more than 500µm.Sieved material was used for adsorption experiments.Afterwards, a part of this sieved materials was heated for 6 h at 480°C or 950°C in an electric furnace (ISUZU Muffle Furnace STR-14K, Japan).Moreover, adequate ultrapure water was added to a part of heat-treatment (950°C, 6h) samples, and heated at 100 °C on a hotplate for evaporation to near dryness (removing water), and finally dried in an electric drying oven at 60 °C.

Sorption experiment for lanthanides using seaweed and shell biomass
The following sorption experiments were performed using the above-mentioned marine biomass.Experimental conditions (i.e., pH, contact time and biosorbent dose rate) in this work were optimized and determined based on our preliminary experiments [e.g., 21] and other literatures [24,31].The pH of each solution was adjusted by using 0.1 moldm -3 NH3aq / 0.1 moldm -3 HNO3.

The seaweed biomass
Samples of 0.4 g of the biomass were contacted with 200 cm 3 of solution containing known initial each lanthanide (La, Eu or Yb) concentration ranging from 0.1 to 4 mmol・dm -3 .Afterwards, the suspensions were shaken for 24 h in a water bath at ambient temperature (～25 °C) at pH 4.

The shell biomass
Each sample of 0.2 g was contacted with 100 cm 3 of multi-element standard solution (prepared by XSTC-1) including known initial lanthanide concentration (10 to 500 µgdm -3 ) in a 200 ml conical flask.Afterwards, the suspensions were shaken for 30 min in a water bath at room temperature at pH 5.
Following with each sorption experiment, the suspension containing biomass and lanthanides standard solution was filtered through a 0.10 µm membrane filter (Advantec Mixed Cellulose Ester, 47mm) to remove lanthanides that have been adsorbed into the biomass, and the concentration of these metals in the filtrate was determined with ICP-MS or ICP-AES.
The metal uptake by the marine biomass was calculated using the following mass balance equation [33]: where q = metal uptake (mmolg -1 ); Ci = initial metal concentration (mmoldm -3 ); Cf = equilibrium metal concentration (mmoldm -3 ); V = volume of the solution (dm 3 ); and W = dry mass of seaweed (g).
The removal efficiency (RE, %) of the biosorbent on the metal in the solution was determined by the following equation [24]:

Langmuir and Freundlich isotherm model
Langmuir adsorption isotherm model was applied based on Tsui et al. [24] in this study, and the model assumes monolayer sorption onto a surface and is given as below.
The Freundlich equation is widely used in the field of environmental engineering, and was applied based on based Dahiya et al. [10][11].Freundlich isotherm can also be used to explain adsorption phenomenon as given below.
where KF and n are constants incorporating all factors affecting the adsorption capacity and an indication of the favorability of metal ion adsorption onto biosorbent, respectively.It is shown that 1/n values between 0.1 and 1.0 correspond to beneficial adsorption.That is, qe versus Cf in log scale can be plotted to determine values of 1/n and KF.
Biosorption of Lanthanides Using Select Marine Biomass 105

Results and discussion
3.1.The seaweed samples

Metal sorption capacity at different species of seaweed
The equilibrium sorption isotherms of La, Eu and Yb by three kinds of Ca-loaded seaweed biomass are shown in Fig. 1.Sorption experiment of Eu using U.p. could not be conducted in this work due to the lack of sample.The adsorption data obtained in this work were analyzed using Langmuir and Freundlich equations.The correlation coefficients (R 2 ) of Langmuir and Freundlich isotherms for La, Eu and Yb using three kinds of seaweed biomass are shown in Table 1 along with other parameters.
From this table, it is found that R 2 value for each datum is comparatively large for both isotherms.The value of 1/n less than unity indicates better adsorption and formation of relatively stronger bonds between adsorbent and adsorbate [10].That is to say, favorable adsorption for La, Eu and Yb by these seaweed biomass used in this work is presented.Furthermore, it is noted that R 2 values for these data are particularly large for Langmuir isotherm than for Freundlich isotherm.This result suggests that the adsorption on these samples mainly occurred by monolayer reaction.Therefore, the curves obtained from nonlinear regression of the data by Langmuir isotherm are also shown in Fig. 1.Langmuir and Freundlich parameters for biosorption of lanthanides on three kinds of Ca-Table 1. loaded biomass at pH 4.0 ( ---represents the missing data due to the lack of sample)

Sargassum hemiphyllum Schizymenia dubyi Ulva pertusa
From Table 1, it is found that the data of qe for Eu and Yb by Sargassum hemiphyllu (brown algae) obtained in this study are similar to the result of species of Sargassum conducted by Diniz and Volesky [31], although qe for La is slightly small.Moreover, it is noteworthy that both Langmuir parameters: qe and A for La by Ulva pertusa (green algae) is large.Then, the comparison of sorption isotherms of La among three kinds of seaweed biomasses is shown in Fig. 2. From this figure, the sorption capacity of La by U. p. (green algae) is considerably large compared to that by other algae: S. h.(brown algae) and S. d. (red algae).It is generally known that brown seaweed is superior to red and green seaweed in metal sorption capacity for heavy metals such as Cd, Pb, Cu [34][35]; and such a high value of qe (as observed in this work) using green algae have not been reported so far.In other words, it is significant outcome to find that Ulva pertusa (green algae) can be a promising biosorbent for removing La.
According to our previous work [32], in case of U, the mean concentration is the highest in brown algae and is the lowest in green algae among phyla (i.e., green, red and brown algae).However, as for the mean concentration of light REE (LREE) such as La, a slightly higher concentration is found in green algae; whereas the concentration of heavy REE (HREE) such as Yb or Lu in green algae is smaller than that in brown algae (as shown in the figure in our previous paper [32]).
Then, large sorption capacity of La by Ulva pertusa may be related to the character (or the constituent) of green algae; and it is possible that "La adsorption on Ulva pertusa" is due to "metal-specific", although further studies is needed to confirm the peculiarity by investigating the other combinations of metals and seaweed species.
◆ : Sargassum hemiphyllum, ■ : Schizymenia duby, ▲ Ulva pertusa.The curves obtained from non-linear regression of the data by Langmuir isotherm are also shown (S. h.; solid curve, S. d.; broken curve and U. p.; dotted curve).Data are mean±standard deviation (n=3).According to SEM observation, the surface of S. d. (red algae) seems to be relatively flat, whereas S. h.(brown algae) and U. p. (green algae) have more extensive surface area, although the specific surface area of the seaweed biomass could not be measured due to their small specific surface area.Furthermore, by comparing SEM pictures in Fig. 3 with that in Fig. 4, it is found that the morphology of S. h. and U. p. surface has hardly changed even  From the above observation, two kinds of biomass: Sargassum hemiphyllum and Ulva pertusa should be predicted to withstand the repeated use; and hence it can be a good adsorbent for lanthanides.

Removal efficiency and binding mechanism of seaweed biomass
The removal efficiency (RE) of 3 kinds of seaweed biomass as a function of initial metal concentrations (Ci) for 3 lanthanides is shown in Fig. 5((a): La, (b): Eu, (c): Yb).With increasing Ci, the RE generally decreased exponentially; and at high Ci, similar RE (i.e., about 40%) occurred for each lanthanide even with any biomass.These data are well fitted into an exponential function (R 2 ranging from 0.866 to 0.994) shown in Fig. 5; and the equations and R 2 for each lanthanide in each biomass are shown in Table 2.
From the viewpoint of recovering trace metals from aqueous environment such as seawater, the removal efficiency at low concentration of metal is particularly important.The coefficient before exponential function in each equation in Table 2 represents the value of RE at low Ci near approximately zero mmoldm -3 .From Table 3, the coefficient for each lanthanide in Sargassum hemiphyllum and especially that for La in Ulva pertusa is large.This implies that U. p. could be an efficient adsorbent for La as well as S. h. for lanthanides in aqueous environment such as seawater.
The amount (mmolg) of adsorbed lanthanide and released Ca from three kinds of Caloaded seaweed biomass is shown in Tables 3-5.Based on the data in these tables, relationship between the uptake of each lanthanide ion and calcium ion released from each biomass is shown both in terms of mill equivalent per gram (meqg -1 ) in Fig. 6.Good and linear relationship is generally found for these samples between the uptake of each lanthanide and Ca released from these biomasses into the solution as shown in Fig. 6.Particularly, in case of S. h. and U. p., the slope of the line is about one with the y-intercept of the graphs almost passes through the origin.It indicates that ion-exchange process is found to be the main mechanism responsible for the sorption of lanthanide ion onto the seaweed as Tsui et al. [24] and Diniz & Volesky [31] also pointed out.X-ray powder diffraction (XRD) patterns of the four kinds of Buccinum tenuissimum shell biomass samples are shown in Fig. 7.The crystal structure of the shell biomass was transformed from aragonite (CaCO3) into calcite (CaCO3) phase by heat-treatment (480°C, 6h).Moreover, the crystal structure of the shell biomass was mainly transformed into calcium oxide (CaO) by heat-treatment (950°C, 6h); and was mainly into calcium hydroxide (Ca(OH)2) by adding water after heat-treatment (950°C, 6h).SEM pictures of the four kinds of sieved shell biomass samples are shown in Fig. 8. Comparing Fig. 8(b) with Fig. 8(a), comparatively clear crystal with a lot of big particles may be observed by heat-traetment (480°C, 6h).It is suggested that ground original sample contains a lot of organic materials such as protein, and most of organic matter seem to disappear by heat-traetment.Moreover, fine crystal particle is not observed in Fig. 8 (c).This may be attributable to the phenomena that many crystals were connected largely with each other due to high-temperature sintering.Meanwhile, relative clear crystal (sizes are mostly 1.0-4.0µm) is observed in Fig. 8 (d).Ca release / meq g -1

Schizymenia dubyi
Metal uptake / meq g -1 (c)   Furthermore, the measurement of specific surface area of the four kinds of sieved samples was performed in this study; and the results are shown in Table 6 along with the main crystal structure of these samples.Remarkably decrease of specific surface area (i.e., from 3.32m 2 /g to 0.390m 2 /g for BET, or from 5.35 m 2 /g to 0.612 m 2 /g for Langmuir) was found after heat-treatment (480°C, 6h).It is suggested that the crystal structure transformation (i.e., from aragonite (CaCO3) into calcite (CaCO3) phase) and also the difference of the surface morphology can be closely related to the remarkable decrease of specific surface area of the shell biomass.On the other hand, the surface area of "heat-treatment (950°C, 6h) sample" was 1.88m 2 /g for BET or 3.10m 2 /g for Langmuir respectively; and that of "heat-treatment (950°C, 6h) and water added sample" was 6.37m 2 /g for BET or 9.91m 2 /g for Langmuir, respectively.

Comparison for sorption capacity lanthanides by four kinds of sieved biomass
The comparison for sorption capacity of lanthanides by four kinds of sieved Buccinum tenuissimum shell samples is shown in Fig. 9.In this experiment, the initial lanthanides concentration was taken as 100µgdm -3 .From this figure, it is found that all kinds of sieved samples showed excellent sorption capacity under this experimental condition.However, the sorption capacity in sample (b) (i.e., the main phase is calcite) decreases slightly relative to that of the original material (i.e., (a): the phase is aragonite) and others.The decrease Sample Main crystal structure Specific surface area (a)Ground original sample Aragonite (CaCO3) 3.31m 2 /g (BET) 5.35m 2 /g (Langmuir) (b) Heat-treatment (480°C) sample Calcite (CaCO3) 0.390m 2 /g (BET) 0.612m 2 /g (Langmuir) (c) Heat-treatment (950°C) sample Lime syn.(CaO) 1.88m 2 /g (BET) 3.10m 2 /g (d) Heat-treatment (950°C) and water added sample Portlandite (Ca(OH)2) 6.37m 2 /g (BET) 9.91m 2 /g (Langmuir) The crystal structures and the specific surface areas of four kinds of sieved Buccinum Table 6.tenuissimum shell biomass of sorption capacity in sample (b) may be attributable to the remarkable decrease (i.e., by a factor of less than one eighth) of specific surface area of the biomass.
Prieto et al. [36] pointed that the sorption capacity of calcite is considerably lower than that of aragonite for Cd.In case of lanthanides, similar tendency of sorption capacity were suggested from our work.

Effect of competitive ions on the sorption of lanthanides
The percentage removal of lanthanides under the presence of common ions (Ca 2+ , Mg 2+ , Na + and K + ) at different concentrations 50, 100 and 200 mgdm -3 is shown in Fig. 10.From this figure, the remarkable decrease of sorption capacity of lanthanides was not observed.Even when the concentrations of common ions are 200 mgdm -3 , the percentage removal of light REE (LREE) such as La or Ce decreased slightly (2-3%), whereas the removal decreased about 5% for heavy REE (HREE) such as Yb or Lu.This implies that the shell biomass can be an efficient adsorbent for lanthanides in aqueous environment such as seawater, although it requires further investigations to apply the shell biomass to use as an adsorbent for lanthanides more practically.

Characteristics of Buccinum tenuissimum shell biomass after adsorption of metals
X-ray diffraction (XRD) patterns of four kinds of sieved samples after adsorption of metals are shown in Fig. 11.Similar to the XRD patterns before adsorption of metals, aragonite and calcite were found as the main crystal structure in (a): Ground original sample and (b): Heattreatment (480°C, 6h) sample, respectively.However, the decrease of peak and increase of noise were also observed in both patterns, particularly in the ground original material as shown in Fig. 11(a).Bottcher [37] pointed out that the natural powdered aragonite was transformed to mixed rhombohedral carbonates by the reaction with (Ca, Mg)-chloride solutions.Therefore, there is the possibility that the transformation of aragonite occurred by the reaction with lanthanides in our experiment.
Moreover, according to XRD analysis, the main crystal structure of (c): "Heat-treatment (950°C) sample" was transformed from calcium oxide (CaO) to the mixture of calcium hydroxide (Ca(OH)2) and calcite (CaCO3) after exposing metals; and that of (d): "Heattreatment (950°C) and water added sample" was transformed from calcium hydroxide(Ca(OH)2) to calcite (CaCO3) after adsorption of metals.These changes may be due to the reaction with water or carbon dioxide in atmosphere.SEM pictures of four kinds of shell biomass after adsorption of metals are shown in Fig. 12.By comparing SEM pictures in Fig. 8 with that in Fig. 12, it is found that the morphology of sample (a) and (b) has hardly changed even after exposing metals.From this observation, these sieved samples should be predicted to withstand the repeated use; and hence it can be a good adsorbent.
In contrast to sample (a), clear crystal structure (sizes are mostly 0.25-2.0µm)was observed in sample (b) even after adsorption of metal.In case of Cd conducted by Kohler et al. [38], the difference of procedure for reaction with metals between aragonite and calcite was suggested.According to their work, the precipitation of several distinct types of crystals was observed after exposing metals in the case of aragonite.Then, it is anticipated that similar phenomenon were occurred by adsorption of lanthanides in case of our samples.On the other hand, the surface of sample (c) and (d) after exposing metals have changed largely compared to that before adsorption of metals (Fig. 8).This is in good accord with the results of XRD patterns.Particularly, remarkable transformation was observed in the morphology of sample (d).The reaction of sample (d) with metal is supposed to proceed rapidly.

Adsorption isotherms of lanthanides by Buccinum tenuissimum shell biomass
The adsorption data obtained for lanthanides using Buccinum tenuissimum shell biomass were analyzed using Langmuir and Freundlich equations.The correlation coefficient (R 2 ) of Langmuir and Freundlich isotherms for lanthanides using ground original shell biomass is shown in Table 7 along with other relevant parameters.
From this table, it is found that R 2 value for lanthanides is comparatively high.It indicates the applicability of these adsorption isotherms satisfactorily for lanthanides in this sample.The dimensionless parameter Hall separation factor (RL) for lanthanides is in the range of 0<RL<1, which means that the sorption for lanthanides by this shell biomass is favorable.Furthermore, the negative value of ΔG indicates that the sorption is spontaneous.The higher R 2 value for Freundlich model rather than for Langmuir isotherm (0.638-0.886 for Langmuir isotherm and 0.844-0.932for Freundlich one) suggests that the adsorption on this sample is due to multilayer coverage of the adsorbate rather than monolayer coverage on the surface.It is noted that the value of 1/n less than unity indicates better adsorption mechanism and formation of relatively stronger bonds between adsorbent and adsorbate [10].That is to say, favorable adsorption for lanthanides by this shell biomass is presented.
On the other hand, R 2 and other parameters of Langmuir and Freundlich isotherms for lanthanides using "heat-treatment (480°C) sample" is shown in Table 8.It is noteworthy that R 2 value for REEs in this sample is still more large (0.947-0.982 for Langmuir isotherm and 0.948-0.975for Freundlich one), compared with the original ground sample ( Futhermore, this result indicates the stronger the monolayer adsorption (the surface adsorption) on the heat-treatment sample relative to on the original sample (before heattreatment).Judging from the value of RL or 1/n in Table 4, the heat-treatment (480°C) sample also exhibits the favorable property for lanthanides adsorption.

Figure 2 .
Figure 2. Comparison of sorption isotherms for La among three kinds of Ca-loaded seaweed biomass at pH 4.0

Figure 9 .
Figure 9.Comparison for sorption capacity of lanthanides by four kinds of sieved Buccinum tenuissimum shell samples.

Figure 10 .
Figure 10.Effect of common ions (Ca 2+ , Mg 2+ , Na + and K + ) on the removal efficiency of lanthanides using ground original sample.

Table 2 .
efficiency (RE) and initial concentrations (Ci) of different lanthanides in the sorption system Equations and correlation coefficients (R 2 ) to describe the relationships between removal

Table 3 .
Amount of adsorbed lanthanide and released Ca by Ca-loaded Sargassum hemiphyllum biomass

Table 4 .
Amount of adsorbed lanthanide and released Ca by Ca-loaded Schizymenia dubyi biomass Amount of adsorbed lanthanide and released Ca by Ca-loaded Ulva pertusa biomass

Table 7
).Coefficient of Langmuir and Freundlich isotherms for lanthanides using original Buccinum

Table 7 .
tenuissimum shell biomass Coefficient of Langmuir and Freundlich isotherms for lanthanides using Buccinum tenuissimum