Published online 27 October 2006
Published in J Environ Qual 35:2222-2228 (2006)
DOI: 10.2134/jeq2006.0093
© 2006 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
TECHNICAL REPORTS
Surface Water Quality
Use of Magnesia for Boron Removal from Irrigation Water
Nina Dionisioua,
Theodora Matsib,* and
Nikolaos D. Misopolinosa
a Laboratory of Applied Soil Science, Faculty of Agriculture, Aristotle University of Thessaloniki (A.U.Th.), 54124, Thessaloniki, Greece
b Soil Science Laboratory, Faculty of Agriculture, Aristotle University of Thessaloniki (A.U.Th.), 54124, Thessaloniki, Greece
* Corresponding author (thmatsi{at}agro.auth.gr)
Received for publication March 3, 2006.
 |
ABSTRACT
|
|---|
The risk of B phytotoxicity due to high levels of B in irrigation water can be avoided by removing B from the water, before its use, through adsorption on certain adsorbents, such as magnesia (industrial MgO), if the latter can be proven to be an effective and easy to handle means for B removal. In addition, if such a material is applied as a fertilizer after its use and the adsorbed B is easily released into the soil solution, B phytotoxicity could constitute a potential hazard. The objectives of this work were to: (a) establish the optimum working conditions (equilibration time, solution to adsorbent ratio, and particle size of the adsorbent) for B adsorption, (b) assess the magnitude of B adsorption by magnesia, both in capacity and intensity terms, as well as the influence of temperature, (c) study B desorbability from magnesia, spiked with B at two rates, 5 and 0.5 mg g–1, and (d) compare the results from b and c to those obtained using reagent grade MgO. The results showed that the time to achieve equilibrium depended on the B concentration of the external solution and ranged from 6 h (for B 
10 mg L–1) to 48 h (for B 
50 mg L–1). The percentage of B adsorbed decreased as the volume of external solution to adsorbent increased and a working ratio of 50:1 was selected. For magnesia, B adsorption was particle size dependent with the smallest fraction (<0.1 mm) sorbing more B than the other three fractions studied (0.1–1.0, 1.1–2.0, 2.1–4.0 mm). Boron adsorption was conducted under strongly alkaline pH (10.3 ± 0.2 and 10.4 ± 0.1 for the reagent and magnesia, respectively) and increased with temperature. Both adsorbents exhibited a high B adsorption capacity (Langmuir maximum values were 5.85 ± 0.39 and 4.45 ± 1.31 mg B g–1 for the reagent and magnesia, respectively) comparable to other metal oxides. However, the reagent grade MgO seemed to be superior to magnesia in terms of capacity and strength of B retention. This superiority of the reagent was attributed to its greater surface area (34.7 compared with 5.8 m2 g–1 for magnesia) and to its conversion to Mg(OH)2 during the adsorption process, whereas magnesia remained unaltered, as was evident from X-ray diffractograms. Based on this data, magnesia seems to be an effective means for removing excess B from irrigation water, particularly if a material of fine particle size is used. Boron desorbability after 240 h of desorption time was more pronounced for magnesia reaching up to 55 and 60% of the amount of B added, at the spiked rates of 5 and 0.5 mg g–1, respectively. Although these figures indicate that approximately half of the amount of B added remained adsorbed, they cannot be easily extrapolated to field conditions, and if B-laden magnesia is applied to soils, the possibility of B phytotoxicity cannot be excluded.
|
INTRODUCTION
|
|---|
THE B CONCENTRATION is one of the parameters that determine the suitability of water for irrigation, because high B concentrations can lead to B accumulation in soils and resulting phytotoxicity (Nable et al., 1997). Scofield (1936) proposed permissible B concentrations in irrigation water 1 mg L–1 for sensitive crops, 2 mg L–1 for semitolerant crops, and 3 mg L–1 for tolerant crops, whereas according to Keren and Bingham (1985) the safe B concentration ranges are 0.3 to 1 mg L–1 for sensitive crops, 1 to 2 mg L–1 for semitolerant crops, and 2 to 4 mg L–1 for tolerant crops.
To prevent soil B accumulating to unacceptable levels, excess B should be removed from water before irrigation. Treatments to reduce B excess in irrigation water, wastewater, and seawater have been reported in literature. These treatments include physical methods, such as diluting with water of lower B concentration, reverse osmosis, or chemical methods, such as ion exchange using B-specific resins (Okay et al., 1985), and adsorption using adsorbents like lime, alum, and ferric salts (Waggott, 1969), and compounds of Ca, Mg, and Na, such as Ca(OH)2, MgO, and Mg(OH)2, used separately or in combinations (Okay et al., 1985; Jamis et al., 2001, 2002).
Treatments, such as dilution, reverse osmosis, and ion exchange process, although effective, are probably uneconomic (Waggott, 1969; Okay et al., 1985). Adsorption techniques, on the other hand, seem to be cheap and easy to apply methods, the effectiveness of which depends on the material used as an adsorbent. Specifically, Waggott (1969) reported that B removal from wastewaters with B concentrations 3.3 mg L–1 by lime, alum, and ferric salts proved to be ineffective. Jamis et al. (2001) tested different substances such as MgO, CaO, Ca(OH)2, MgO + CaO, MgSO4 + NaOH, and MgCO3 + NaOH as adsorbents for removing B from solution and concluded that only Ca(OH)2 was effective. Based on these results Jamis et al. (2002) proposed the use of Ca(OH)2 for reducing B in wastewaters containing H3BO3 and H2SO4. They reported that under optimal operating conditions, a wastewater B concentration of 0.7 g L–1 was reduced to less than 50 mg L–1.
In contrast to the findings of Jamis et al. (2002) concerning MgO, there are certain references in the literature that support the high affinity of MgO and Mg(OH)2 for B. Specifically, Okay et al. (1985) reported that MgO was effective in reducing the B content in the discharge waters of B mines by 85%. Polat et al. (2004) found that alkaline fly ashes, derived from coal burning, had a high capacity to sorb B from seawater and explained this result by proposing that on contact of Ca-rich fly ash with Mg-rich seawater, B is coprecipitated with Mg(OH)2. Finally, Rhoades et al. (1970) concluded that arid zone soils can have an appreciable B adsorption capacity in their sand, silt, and clay fractions. This capacity was attributed to Mg(OH)2 clusters or coatings existing on the weathered surfaces of ferromagnesian minerals and micaceous layer silicates.
It is evident from the literature review (Rhoades et al., 1970; Okay et al., 1985; Jamis et al., 2001) that B adsorption by magnesia (industrial MgO) has not been studied extensively. Thus the objectives of the present work were to: (a) establish the optimum working conditions (equilibration time, solution to adsorbent ratio, and particle size of the adsorbent) for B adsorption by magnesia and (b) assess the magnitude of B adsorption by magnesia, both in capacity and intensity terms using parameters of the empirical adsorption equations of Freundlich and Langmuir, as well as the influence of temperature. In addition, since magnesia might be applied as a fertilizer after its use as an adsorbent for B, it was considered worthwhile to evaluate the risk of B phytotoxicity induced by the soil application of used magnesia by studying the desorbability of externally added B from magnesia. Finally, for comparison reasons B adsorption and desorption by reagent grade MgO were also studied.
|
MATERIALS AND METHODS
|
|---|
Materials
Samples of magnesia in four particle size fractions (<0.1, 0.1 to 1.0, 1.1 to 2.0, and 2.1 to 4.0 mm), were obtained from a local factory producing magnesia. This factory uses a blast furnace to convert magnesite (MgCO3) derived from the Gerakini mines in northern Greece to MgO. The analytical reagent MgO (fine powder < 0.005 mm) was obtained from the Merck Co (reagent No 5865). To explain differences in B adsorption capacity of the two materials, surface areas (BET) of the reagent and the four size fractions of magnesia were determined from N2 adsorption/desorption isotherms. The isotherms were obtained using an Automatic Volumetric Sorption Analyzer (Autosorb-1, Quantachrome). Employing the same technique, surface areas (BET) of the reagent and the smallest fraction of magnesia were also determined, after wetting with 0.01 M CaCl2 solution and drying. In addition, to see if any structural changes were occurring to the two substances (reagent and the smallest fraction of magnesia) during the adsorption process, X-ray diffraction patterns of randomly oriented powder specimens were obtained before and after wetting with 0.01 M CaCl2 solution and drying. The diffractometer used was a Philips model PW 1830 instrument, equipped with a Cu target operated at 45 kV and 30 mA and a graphite crystal monochromator.
Boron Adsorption Experiments
To establish the optimum equilibration time and solution/adsorbent ratio for B adsorption, reagent-grade MgO was used in batch adsorption experiments. For the first parameter, a tentative ratio of 50:1 was used and the technique of Elrashidi and O'Connor (1982) was employed. Half-gram subsamples of the material in three replications were placed in 50-mL polypropylene centrifuge tubes with 25 mL of a 0.01 M CaCl2 solution containing various B concentrations (0 to 100 mg B L–1 as H3BO3). The suspensions were equilibrated for 1, 2, 4, 6, 10, 24, 48, and 72 h at a constant temperature of 25 ± 1°C with intermittent shaking. Suspension pH was determined periodically to detect if any pH changes were occurring during the adsorption process. At the end of the equilibration period the tubes were centrifuged, and the supernatant liquid (equilibrium solution) was filtered and analyzed for B by the azomethine-H method (John et al., 1975) using a UV-Vis Spectrophotometer (Lambda 5, Perkin-Elmer). Adsorbed B was calculated as the difference between the amount added and that found in solution at equilibrium. In addition, Mg was determined in the equilibrium solution with an atomic absorption spectrophotometer (Analyst 200, Perkin-Elmer) to detect if any dissolution of the material was occurring during the experiment. The pH of the equilibrium solution was also measured to see if it was at levels that are considered normal for irrigation waters, namely 6.5 to 8.4 (FAO, 1985). The above-mentioned determinations were performed and the same technique was applied in all adsorption experiments described henceforth.
Having established an equilibration time of 24 h, the working solution/adsorbent ratio was selected by using different solution/adsorbent ratios (20:1, 25:1, 50:1, 100:1).
To decide which size fraction of magnesia was the optimum to be used, B adsorption experiments with the four fractions (<0.1, 0.1 to 1.0, 1.1 to 2.0, and 2.1 to 4.0 mm) of the material were conducted. Two initial B concentrations were used (5 and 50 mg B L–1), the solution/adsorbent ratio was 50:1, and the equilibration time was 24 h.
Having established the working conditions with respect to equilibration time (24 h), solution/adsorbent ratio (50:1), and the particle size of magnesia (the smallest fraction), B adsorption experiments were conducted with both substances at two temperatures (25 and 60 ± 1°C). The B concentrations in the external solution were 0, 2, 4, 6, 8, 10, 15, 25, 50, and 100 mg B L–1.
Boron adsorption data, obtained for both substances, were fitted to the following nonlinear equations by nonlinear regression using the Levenberg-Marquardt algorithm:
- (a) the Freundlich equation:
 | [1] |
- (b) the Langmuir equation:
 | [2] |
where x is the amount of B adsorbed (mg g–1), C is the B concentration in the equilibrium solution (mg L–1), k is a parameter corresponding to the amount of B adsorbed when C is equal to 1, n is a parameter that can be considered as a measure of adsorption intensity, K (L mg–1) is a parameter related to the affinity of the adsorbent for the adsorbate, and M (mg g–1) is the maximum B adsorption capacity.
Boron Desorption Experiments
Samples of both the smallest fraction of magnesia and the reagent-grade MgO were spiked with B in the form of a H3BO3 solution at rates equal to 5 and 0.5 mg B g–1 of substance. After B addition, the samples were left to equilibrate for 30 d with periodical wetting and drying, then air-dried and used for the desorption experiments.
For B desorption experiments, subsamples of 0.5 g (4 reps) of both B-spiked substances were placed into 50-mL polypropylene centrifuge tubes with 25 mL of 0.01 M CaCl2 solution. The tubes with the suspensions were maintained at 25 ± 1°C with periodic shaking and determination of the suspension pH. After 24 h the tubes were centrifuged, a 15-mL aliquot of the supernatant solution was removed, filtered, and analyzed for B and Mg. The 15-mL aliquot removed was replaced with 15 mL of the 0.01 M CaCl2 solution, the mixtures were resuspended, and the same procedure, as previously described was repeated 9 more times, resulting in 10 desorption steps and an overall desorption time of 240 h. The experiment ended after 240 h because the B concentration in the supernatant liquid had reached undetectable levels for both materials at the lowest B addition rate. The amount of B desorbed at each step was corrected for B transferred from the previous step.
|
RESULTS AND DISCUSSION
|
|---|
Establishing Equilibration Time, Solution/Adsorbent Ratio, and the Particle Size of Magnesia to be used in Boron Adsorption Experiments
Equilibration time depended on the initial B concentration (Table 1). Specifically, a 6-h equilibration time was adequate to achieve equilibrium for the lower B concentrations (10 mg L–1), whereas a 48-h time was required for the higher B concentrations (50 mg L–1). Since relatively low B concentrations (up to 10 mg L–1) are usually encountered in irrigation waters, a 24 h equilibration time was chosen for the subsequent adsorption experiments. The 24 h equilibration time also provided convenience and speed with respect to laboratory work. Reports in the literature corroborate this choice. For example, Goldberg (1997) states that B adsorption on oxide minerals occurred rapidly, being virtually complete after 1 d of reaction time. The same is reported by Rhoades et al. (1970) who used Mg(OH)2 as an adsorbent of B, the concentrations of which were less than 10 mg L–1 in the external solution. During this adsorption experiment, the pH of the suspension was strongly alkaline (10.3 ± 0.2) and similar pH values of the suspension were observed in all the subsequent adsorption experiments where reagent grade MgO was used.
The percentage of B adsorbed decreased as the ratio of external solution/adsorbent increased (Table 2). Keren and Bingham (1985) reported the same decreasing trend for B adsorption on montmorillonite and certain soils as the solution/adsorbent ratio was increased. In our case, for initial B concentrations in the external solution less than 10 mg L–1 and for the two lowest ratios, 100% of added B was removed from the solution by the adsorbent, whereas for the two highest ratios 
90% of added B was removed (Table 2). This is in agreement with the findings of Rhoades et al. (1970) who studied B adsorption on Mg(OH)2 with similar solution/adsorbent ratios and B concentrations in the external solution less than 10 mg L–1. They reported that in all cases 100% of added B was removed from solution after a 24 h equilibration period. In the present study, the ratio of 50:1 was chosen for the subsequent adsorption-desorption experiments for two reasons. First, at that ratio the study of B adsorption on MgO at relative low B concentrations in the external solution, which as mentioned before is of practical merit, was analytically accurate, and second, the dissolution of MgO was negligible (0.69 ± 0.03%). The dissolution of the reagent-grade MgO remained <1% in all the subsequent adsorption experiments.
View this table:
[in this window]
[in a new window]
|
Table 2. Percentages of B adsorbed by the reagent grade MgO at different solution/adsorbent ratios and at different initial B concentrations in the external solution.
|
|
Among the particle size fractions of magnesia tested, the highest amounts of adsorbed B were obtained for the smallest fraction (<0.1 mm) (Table 3), which had the largest surface area. The surface area of the smallest fraction was 5.8 m2 g–1, whereas that of the others were 0.14, 0.10, and 0.05 for the size fractions of 0.1 to 1.0, 1.1 to 2.0, and 2.1 to 4.0 mm, respectively. This result was expected and is in agreement with the findings of Keren and Talpaz (1984) who reported that B adsorption on montmorillonite increased as the particle size of the mineral decreased, and this was attributed to an increase in surface area and a consequent increase in adsorption sites. During this adsorption experiment, the pH of the suspensions were strongly alkaline (10.4 ± 0.1) and the dissolution of magnesia was negligible (0.37 ± 0.21%). In all the subsequent adsorption experiments where the smallest fraction of magnesia was used, similar suspension pH values were obtained and the dissolution of magnesia remained below 1%.
Boron Adsorption Characteristics of Magnesia and Reagent-Grade Magnesia
As is evident from the subsequent data, both substances, the reagent-grade MgO and magnesia, exhibited a high affinity for B. This was attributed to the fact that the pH during adsorption was strongly alkaline and near 9.24 (the pK for the reaction H3BO3 + H2O 
B(OH)4– + H+), where B adsorption maximum is expected (Goldberg, 1993).
Boron adsorption data for both substances conformed satisfactorily to the nonlinear forms of both the Freundlich and Langmuir equations. The Langmuir equation was slightly superior, as judged by the lower standard error of the estimate (SE), which was used as the criterion of goodness of fit (Table 4). In all cases, the estimated values of the parameters, k and n of the Freundlich and K and M of the Langmuir equations, were significant, as was evident from their low standard errors (Table 4). Figures 1 and 2 also show the good agreement between the observed and predicted values of adsorption data for both substances. Boron adsorption on the reagent was higher than that on magnesia and this was reflected in the calculated capacity terms of the equations, namely the Freundlich k and the Langmuir M. The values of the parameters estimated for the reagent were higher (significantly for the Freundlich k) than those estimated for magnesia, although the values of the Langmuir maximum were high for both substances. Also, a similar significant difference was observed for the estimated Langmuir affinity parameter K (Table 4).
View this table:
[in this window]
[in a new window]
|
Table 4. Parameters of the Freundlich and Langmuir equations for the reagent-grade MgO and magnesia and the standard error of their estimate (SE).
|
|
Such high values of B adsorption maximum obtained in this study are reported in the literature for other oxides. McPhail et al. (1972), studying B adsorption on Fe and Al hydrous oxides, found that Al hydrous oxides adsorbed more B per unit mass than did Fe hydrous oxides. Over the range of B solution concentrations used in this study (0 to 16 mM), B adsorption by the Fe hydrous oxides conformed to one of the linearized forms of the Langmuir equation and the computed maximum was 0.765 mmol g–1 (= 8.3 mg g–1). Also, Keren and Gast (1983) studied B adsorption on hydroxy aluminum material prepared in the laboratory over several pH values and a B concentration range in the external solution of 2.5 to 12.5 mg L–1. For describing B adsorption they used the phenomenological equation developed by Keren et al. (1981) and they reported that the predicted maximum was 0.333 mmol g–1 (= 3.6 mg g–1).
The difference in B retention capacity between the reagent-grade MgO and magnesia was at first attributed to differences in the surface area of the two substances. The surface area of the reagent was 34.7 m2 g–1, whereas that of magnesia was almost 6 times less (5.8 m2 g–1), and this difference was persistent and increased after wetting with 0.01 M CaCl2 solution and drying (the surface areas were 58.1 and 4.1 m2 g–1 for the reagent and magnesia, respectively). Since, however, chemical transformation [MgO to Mg(OH)2] of both substances was suspected, X-ray diffractograms of the substances were obtained, before and after wetting with 0.01 M CaCl2 solution and drying. The diffractograms revealed that the reagent transformed to Mg(OH)2 (brucite) (Fig. 3) on its contact with the 0.01 M CaCl2 solution, whereas magnesia remained unaltered (Fig. 4), although a limited chemical transformation of the magnesia surface cannot be excluded. So the higher affinity of the reagent for B in comparison to magnesia could be better explained by the formation of Mg(OH)2 during the adsorption experiments.
The pH of the equilibrium solutions after their separation from both substances was strongly alkaline but dropped below 8.5 after 24 h. Consequently, if magnesia is going to be used for treating irrigation waters enriched with B, these waters should remain in the open air for a period after treatment and before use for irrigation to let the pH drop to acceptable levels, namely 6.5 to 8.4 (FAO, 1985).
Temperature Effect on Boron Adsorption by Magnesia and Reagent-Grade Magnesia
Plotting B adsorption isotherms for the two substances at the two temperatures studied revealed the endothermic character of the reaction, since the isotherms at the higher temperature were located above the isotherms at the lower temperature. This character was more pronounced at the higher B concentrations (Fig. 5). This was in agreement with the findings of Okay et al. (1985), who reported that after a contact time of 20 min MgO removed 95% of B from aqueous solutions at 72°C, whereas only 10 to 15% of B was removed at 20°C. On the contrary, B adsorption on Al and Fe oxides and hydroxides as a function of pH was reported to be exothermic, for a range of temperatures 5 to 40°C (Goldberg, 1997).
Boron Desorption from Magnesia and Reagent-Grade Magnesia
The results of the B desorption experiments can be used to assess the risk of B phytotoxicity, in the case of magnesia used as a fertilizer, after its use for removing excess B in irrigation waters. At both rates of B addition and for both substances, the greatest part of the total desorbable B was released during the first desorption step (24 h). Specifically, B released at the first step was 25 and 44% of the total desorbable B for the reagent-grade MgO and magnesia, respectively, for a B addition of 5 mg g–1. The corresponding figures for a B addition of 0.5 mg g–1 for the reagent and magnesia were 19 and 20%, respectively. After 24 h, the cumulative amounts of desorbable B continued to increase in the subsequent steps and at the end of the desorption (240 h) reached 16.8 and 54.6% of B added, for the reagent and magnesia, respectively, at the 5 mg B g–1 addition. The corresponding figures for the reagent and magnesia at the 0.5 mg B g–1 addition were 54 and 60%, respectively (Fig. 6). During the desorption experiments and for both materials, the pH of the suspensions was strongly alkaline, specifically 10.3 ± 0.2 for the reagent and 10.5 ± 0.3 for magnesia, and at the end of the experiments the cumulative dissolution of both substances did not exceed 5%.
View larger version (12K):
[in this window]
[in a new window]
|
Fig. 6. Cumulative amounts of desorbed B from reagent-grade MgO and magnesia after B addition of (a) 5 mg g–1 and (b) 0.5 mg g–1 and 30-d contact time. Each step represents a 24-h desorption interval.
|
|
The above observations indicate that both substances retained a significant amount of added B and this was more pronounced at the highest amount of added B, with the reagent being superior to magnesia in B retention. This is in agreement with the previous findings based on the Langmuir parameters which showed that the capacity and the strength of B retention by the reagent were higher than those of magnesia.
Although both substances retained a significant amount of added B, the amount of B released might be toxic to plants in the case of soil application of magnesia. As a criterion for this possibility, B concentration in the external solution at each desorption step was used. So for the reagent-grade MgO at the 5 mg B g–1 addition, B concentration in the external solution ranged between 4.2 and 3.1 mg L–1 during the 72-h interval of desorption, then dropped below 3 mg L–1 but remained around 2 mg L–1 even at the last desorption step. For magnesia at the same B addition B concentration in the solution ranged at undesirable levels (11.5 mg L–1) during the 96-h interval of desorption, then dropped below 3 mg L–1 at the fifth desorption step and below 1 mg L–1 after 144 h. At the 0.5 mg B g–1 addition, B concentration was almost 1 mg L–1, during the 48- and 72-h intervals of desorption for the reagent and magnesia, respectively, and then dropped below 1 mg L–1. Based on these observations and according to the general guidelines given by Scofield (1936) and Keren and Bingham (1985) for the characterization of the quality of irrigation water with respect to B, it seems that soil application of magnesia containing relative low amounts of B does not impose any risk of B phytotoxicity. However, if B-laden magnesia is applied to soils, the risk of B phytotoxicity cannot be excluded. Nevertheless, the rate of B release from magnesia will also be dependent on the final soil pH after mixing magnesia with soil, as well as other soil characteristics, such as texture, clay mineralogy, and organic matter content.
|
CONCLUSIONS
|
|---|
The capacity of magnesia (especially that of the smallest particle size fraction) to sorb B was high and comparable to other oxides and oxy-hydroxides. When B was added to both magnesia and reagent-grade MgO and its desorbability studied, it became evident that although both substances retained a significant amount of added B after a 240-h desorption period, large amounts of B were released during the initial desorption steps. The reagent exhibited a higher affinity for B in comparison to magnesia and this was attributed to differences in surface area, but also to the transformation of the reagent to Mg(OH)2 during the adsorption-desorption processes. The above laboratory findings suggested that magnesia could be used as an adsorbent for removing excess B from irrigation water. However, since the pH of irrigation water becomes strongly alkaline on contact with magnesia, the waters need to stand in the open air for a period after treatment and before irrigation to let the pH decrease to acceptable levels. In addition, if magnesia, after its use for removing excess B from irrigation waters, is applied to soils as a fertilizer caution is needed in respect to B phytotoxicity, particularly if it contains high amounts of B.
|
ACKNOWLEDGMENTS
|
|---|
The valuable assistance of our colleague Vissarion Z. Keramidas in editing this manuscript is greatly appreciated.
|
REFERENCES
|
|---|
- Elrashidi, M.A., and G.A. O'Connor. 1982. Boron sorption and desorption in soils. Soil Sci. Soc. Am. J. 46:27–31.[ISI]
- FAO. 1985. Water quality for agriculture. FAO Irrigation and Drainage Paper 29. FAO, Rome.
- Goldberg, S. 1993. Chemistry and mineralogy of boron in soils. p. 3–44. In U.C. Gupta (ed.) Boron and its role in crop production. CRC Press, Boca Raton, FL.
- Goldberg, S. 1997. Reactions of boron with soils. Plant Soil 193:35–48.[CrossRef]
- Jamis, P., H. Muhr, and E. Plasari. 2001. Separation de l' acide borique par un procede de cristallisation. Technical Report CT 00–56 PROG 302, Chemical Engineering Sciences Laboratory, Nancy.
- Jamis, P., H. Muhr, and E. Plasari. 2002. Boron removal from waste solutions using a multiphase coprecipitation processes. p. 975–981. In Proc. of 15th International Symposium on Industrial Crystallization. 15–18/09/02. Sorrento, Italy.
- John, M.K., H.H. Chuah, and J.H. Neufeld. 1975. Application of improved azomethine-H method to the determination of boron in soils and plants. Anal. Lett. 8:559–568.
- Keren, R., and F.T. Bingham. 1985. Boron in water, soils and plants. Adv. Soil Sci. 1:229–276.
- Keren, R., and R.G. Gast. 1983. pH-dependent B adsorption by montmorillonite hydroxyl-aluminum complexes. Soil Sci. Soc. Am. J. 47:1116–1121.[ISI]
- Keren, R., R.G. Gast, and B. Bar-Yosef. 1981. pH–dependent boron adsorption by Na-montmorillonite. Soil Sci. Soc. Am. J. 45:45–48.[ISI]
- Keren, R., and H. Talpaz. 1984. Boron absorption by montmorillonite as affected by particle size. Soil Sci. Soc. Am. J. 48:555–559.[ISI]
- McPhail, M., A.L. Page, and F.T. Bingham. 1972. Adsorption interactions of monosilisic and boric acid on hydrous oxides of iron and aluminum. Soil Sci. Soc. Am. Proc. 36:510–514.
- Nable, R.O., G.S. Bañuelos, and J.G. Paull. 1997. Boron toxicity. Plant Soil 193:181–198.[CrossRef]
- Okay, O., H. Güclü, E. Soner, and T. Balka. 1985. Boron pollution in the Simav river, Turkey and various methods of boron removal. Water Res. 19:857–862.
- Polat, H., A. Vengosh, I. Pankratov, and M. Polat. 2004. A new methodology for removal of boron from water by coal and fly ash. Desalination 164:173–188.[CrossRef]
- Rhoades, J.D., R.D. Ingvalson, and J.T. Hatcher. 1970. Adsorption of boron by ferromagnesian minerals and magnesium hydroxide. Soil Sci. Soc. Am. Proc. 34:938–941.
- Scofield, C.S. 1936. The salinity in irrigation water. Smithsonian Institute Annual Rep. Smithsonian Inst., Washington, DC.
- Waggott, A. 1969. An investigation of the potential problem of increasing boron concentrations in rivers and water courses. Water Res. 3:749–765.[CrossRef]
TOP
ABSTRACT
INTRODUCTION
