AU1333892A - Method for removing and/or recovering mercury - Google Patents

Description

METHOD FOR REMOVING AND/OR RECOVERING MERCURY

The present invention relates to a method for removing and/or recovering mercury from natural water or wastewater.

Mercury is a potentially very toxic pollutant to the aqueous environment. The levels of mercury in wastewater need to be considerably reduced prior to discharge. Typically mercury levels need to be reduced from mg/1 levels to low ppb levels, generally 10 to 200 ppb or less. The usual procedure for treating mercury containing wastewater is to precipitate the mercury chemically. The most common approach is to precipitate the mercury with sulphide. This step results in the formation of a sludge containing the sparingly soluble mercury sulphide along with other precipitated metals which can then be dumped as landfill. This precipitation technique, although effective for reducing mercury content of wastewater, has several disadvantages. The mercury that is precipitated is not recoverable and the disposal of the sulphide sludge itself may pose significant environmental problems.

It is known that some ion exchange resins extract mercury, and thus it should be possible to devise a method whereby the mercury is selectively extracted from wastewater to very low levels and then recovered to yield a suitably concentrated pure form of mercury which can be used elsewhere. Mercury is a relatively rare metal having many industrial uses and so the recovery of mercury from wastewater by ion exchange has some attraction.

Previous work has examined mercury removal and recovery with a sulphonic acid cation exchanger and an iminodiacetic acid chelating exchanger. In these experiments, it was found that both resin types were very effective at removing mercury from sulphuric acid solutions. The iminodiacetic acid resin selectively adsorbed mercury at low pH in the presence of other metals. Mercury was recoverable from the sulphonic acid resin by elution with a strong mineral acid. The recovery of mercury could not be achieved from the iminodiacetic acid resin with mineral acid but could be achieved with a solution of a salt, such as sodium chloride, which forms anionic complexes with mercury.

Unfortunately, it was also found that in the presence of chloride, mercury uptake by the iminodiacetic acid resin was impaired. The reason for this is that in the presence of chloride, mercury forms the complex anion (HgCl₄)²". Although this resin could not as effectively adsorb (HgCl_<)^2" when compared to Hg²⁺, it did have some affinity for the complex anion. This is probably due to the fact that the protonated imine groups on the resin were able to act as anion exchange sites and so adsorb some of the complex anion.

Mercury is commonly present in a wastewater as the complex anion (usually as the chloride) and so an ion exchange process needs to be developed around extracting mercury specifically as a complex anion. It is an object of the present invention to provide an ion exchange process in which the uptake of mercury is achieved by using anion exchange resins.

We have investigated the uptake of anionic complexed mercury under acid conditions by some selected anion exchange resins. Leakage of mercury from all of the resins tested was well into the low ppb range indicating that anion exchangers are very efficient at adsorbing this form of mercury. Furthermore, it was found that the adsorbed mercury could be removed from the loaded resin by elution with a complexing agent.

According to the present invention there is provided a method for removing and/or recovering mercury from natural water or wastewater contaminated with mercury, characterized in that the method comprises the steps of:

(a) extracting the mercury from the natural water or wastewater by adsorption onto an anion exchange resin; and

(b) removing the adsorbed mercury from the loaded anion exchange resin by elution with a complexing agent selected from a sulphur- containing inorganic anion and an organic ligand containing one or more donor atoms selected from oxygen, nitrogen and sulphur.

The mercury may be optionally recovered from the eluate and the eluate may be optionally recycled as the eluant to step (b). The recovered mercury may be optionally converted into metallic mercury or a desired mercury compound.

The term "wastewater" as used herein includes aqueous effluents and other wastewaters containing mercury which are discharged from industrial processes. The mercury to be removed and/or recovered by the method of the invention is present in the natural water or wastewater as a complex anion. Typically the complex anion is (HgCl₄)²".

The mercury is extracted using any suitable technique known per se for use in ion exchange processes, e.g. by passing the natural water or wastewater through a bed of settled anion exchange resin. The anion exchange resins suitable for use in the present invention can be broadly classified into two types, namely, strong base and weak base exchange resins. Strong base resins comprise functional groups that are quaternary ammonium groups and remain ionised over all pH ranges, while weak base resins usually comprise primary, secondary or tertiary amines that are readily converted to their free base forms and hence lose their ion exchange properties above pH 7. Examples of strong base anion exchange resins include those containing benzyltrimethyl ammonium functional groups, for example, Amberlite IRA 900,

Duolite A161, Diaion PA-318, Lewatit MP-500 and Dowex MSA-1. Examples of weak base anion exchange resins include those containing tertiary amine functional groups, for example, Amberlite IRA 93, Amberlyst A-21, Duolite A378, Diaion WA-30, Lewatit MP-62 and Dowex MWA- 1. An example of a very weak base anion exchange resin is Reillex 425 which is composed of polyvinyl pyridine.

[The following registered trademarks are used throughout the description Amberlite IRA 900, Duolite A161, Diaion PA-318, Lewatit MP-500, Dowex MSA-1, Amberlite IRA 93, Amberlyst A-21, Duolite A378, Diaion WA-30, Lewatit MP-62, Dowex MWA-1 and Reillex 425]

Preferably the anion exchange resin is a strong base as such resins exhibit much lower leakages of, and a higher capacity for, mercury. More preferably, the anion exchange resin contains benzyltrimethyl ammonium functional groups, such as, for example, Amberlite IRA900 which is capable of removing mercury from a complex acidic wastewater resulting in mercury leakages of less than 1%.

The adsorbed mercury may be removed from the loaded anion exchange resin by elution with a complexing agent that binds more strongly with mercury than the resin itself. The complexing agent is selected from a sulphur containing inorganic anion, such as, for example, sulphide, sulphite or thiosulphite and an organic ligand containing one or more donor atoms selected from oxygen, nitrogen or sulphur, such as, for example, ethylenediamine, ethylenediaminetetracetic acid (EDTA), thiourea, methionine, thiazoleamine, diaminodiphenyl- sulphide or thioglycolic acid.

A particularly preferred complexing agent is ethylenediamine which can be used to successfully elute mercury from both strong and weak base resins. With a weak base resin, 3-5 bed volumes of 1.5M ethylenediamine will achieve complete mercury elution. This is more efficient than a strong base resin where complete mercury elution occurs after 5-6 bed volumes. With the weak base resin, a slow growing precipitate forms in eluate fractions containing high mercury concentrations following ethylenediamine elution.

Sodium sulphite may also be used to elute mercury from both strong and weak base resins. Although sodium sulphite is not as effective as ethylenediamine in eluting mercury from the resin, it is relatively cheap, easy to handle and readily available. In both strong and weak base resins, a slow growing precipitate forms in eluate fractions containing high mercury concentrations. Complete elution of mercury from a weak base resin with 1M sodium sulphite occurs within 7 bed volumes, while with the strong base resin complete elution occurs within 15 bed volumes. Sodium sulphide and EDTA can also elute mercury from both types of resins, but not as effectively as sodium sulphite or ethylenediamine.

In a particularly preferred embodiment of the present invention, a strong base anion exchange resin is used in combination with sodium sulphite as the complexing agent.

After elution with a complexing agent, the mercury may be recovered from the eluate. Where a precipitate forms, as in the case of ethylenediamine or sodium ■ sulphite elution, the mercury complex can be recovered directly. The precipitates are generally slow forming and depend on the concentration of mercury in the eluate. The precipitate can be isolated from the eluate after a time and the eluate can then be recycled back as eluant to step (b).

Once the mercury complex is collected as a precipitate, further steps may need to be taken to obtain a useful product. The mercury from the precipitate could be obtained by breaking the precipitate down thermally. Thermogravimetric analysis (TGA) indicates that the sulphite precipitate is a better candidate for thermal decomposition, than the ethylenediamine precipitate.

Mercury from the sulphite precipitate may also be converted to mercurous chloride which can be processed commercially into calomel or another mercury product by dissolving the precipitate in hydrochloric acid. The acid will destroy the sulphite and remove it from solution through the evolution of sulphur dioxide yielding a reasonably pure solution of mercuric chloride. Mercury can also be recovered by precipitation as Hg(I) or Hg(0) after the addition of a suitable reductant. There are many reagents capable of reducing Hg(II). One of the most common reagents used to reduce mercuric compounds is stannous chloride. If the stannous chloride is added in excess, then the product will be Hg(0), but if not in excess, then Hg(I) will be precipitated. Ascorbic acid also reduces mercuric solutions and reduction has also been observed with hydrazine solution.

The mercury in the eluate may also be precipitated by reduction using electrolysis. It is also possible to recover dissolved mercury as the metal from both ^• ethylenediamine and sodium sulphite solutions, which are both decomposable by electrolysis.

Of these methods of mercury recovery, isolation of the mercury ethylenediamine or mercury sulphite precipitates in the eluate with eluate recycling is the most preferred option. Mercury can then be recovered from the precipitate by a variety of means depending on the desired mercuric compound.

The invention is further exemplified and illustrated by reference to the accompanying drawings, in which:

Figure 1 is a graph comparing the breakthrough curves for mercury through Amberlite IRA900 and Amberlite IRA93 resins (feed [Hg] = 400 mg/1);

Figure 2 is a graph showing the elution curve for mercury loaded Amberlite IRA93 resin with 1.5M ethylenediamine (flow rate = 8.8 ml/hr/ml); Figure 3 is a graph showing the elution curve for mercury loaded Amberlite IRA900 resin with 1.5 M ethylenediam_.ne (flow rate = 10.2 ml/hr/ml);

Figure 4 is a graph showing the elution curve for mercury loaded Amberlite IRA93 resin with 1M Na₂S0₃ (flow rate = 9.4 ml/hr/ml);

Figure 5 is a graph showing the elution curve for mercury loaded Amberlite IRA900 resin with 1M Na₂S0₃ (flow rate = 12.6 ml/hr/ml);

Figure 6 is a graph showing the elution curve for mercury loaded Amberlite IRA93 resin with 1M Na₄EDTA at pH 12.7 (flow rate = 13.5 ml/hr/ml).

The following examples further describe and illustrate the method of the invention. It will be understood, however, that the invention is not limited by these examples.

EXAMPLE 1 - MERCURY UPTAKE TRIALS WITH VARIOUS ANION

EXCHANGERS.

Several different types of anion exchange resins were tested for mercury uptake. These were Amberlite IRA900 which is a strong base resin and Amberlite IRA93, Amberlyst A-21 and Reillex 425 which are weak base resins.

Before use, each resin was first regenerated with NaOH, then converted to the chloride form with NaCl or HC1.

The uptake of Hg by each resin was tested by passing a Hg solution containing about 40 mg/1 Hg in H₂S0₄ solution containing about 2000 mg/1 Cl (as NaCl) at pH 1.5 through a bed (3-5 ml wet settled resin (wsr)) of the resin. Mercury was analysed by either flame atomic absorption spectrophotometry (AAS) in the case of high [Hg] (in excess of 20-50 mg/1) or by the cold vapour flameless AAS procedure using a home made set up for low level [Hg] .

The results of each load trial conducted on each resin are shown in Tables 1,2,3 and 4.

The distribution coefficients for the four anion exchangers were determined in the presence of 400 mg/1 Hg solution and 2000 mg/1 Cl plus S0₄ at pH 1.5. The resins were tested in the Cl form.

In each case, 0.1 g of dry resin was shaken overnight with 200 ml of the Hg solution. After equilibration the Hg was determined in the solution phase and Dg calculated using the following formula;

Dg = (moles Hg on resin/g resin)/(moles Hg in solution/ml solution).

The result is as follows:

RESIN Dg LOG Dg

Amberlite IRA93 (weak base) 2150 3.3

Amberlyst A-21 (weak base) 2400 3.4 Reillex R-425 (weak base) 2604 3.4

Amberlite IRA900 (strong base) 2686 3.4

It can be concluded from these results that both strong base and weak base anion exchangers are able to remove anionic complexed Hg from water very effectively. Breakthrough curves for the strong base exchanger Amberlite IRA900 and the weak base exchanger Amberlite IRA93 were obtained using a 400 mg/1 Hg solution (at pH 1.5 in the presence of chloride). The results are shown in Tables 5 and 6, while a graphical comparison of the two curves is shown in Figure 1.

Analysis of the breakthrough curves for both resins shows that the mercury leakage from the strong base resin is much lower than the corresponding leakage from the weak base resin. In addition, the weak base resin breaks through earlier than the strong base resin indicating a lower Hg capacity. The results indicate that the strong base resin is superior to the weak base resin with respect to Hg uptake.

The strong base resin Amberlite IRA900 has also been tested for its ability to extract mercury from a wastewater obtained from a zinc refinery. The wastewater was scrub tower effluent ex Al pellet column treatment. This waste typically contains high levels of zinc and aluminium and lower levels of other metals including mercury in an acidic sulphuric acid medium containing a variety of other anions (pH 1.6). Some mercury was added to the waste to bring the total Hg concentration to 9.09 mg/1. When this waste feed was passed through a 4.4 ml bed of Amberlite IRA900 resin at 18.5 ml/hr/ml wsr the Hg leakage over 265 bed volumes varied between 0.018 and 0.061 mg/1, averaging at 0.028 mg/1 which is well below the desired target of 0.05 mg/1.

It would seem from this preliminary result that Amberlite IRA900 strong base resin is able to effectively adsorb Hg, even from a complex electrolytic zinc produced wastewater. EXAMPLE 2 - INITIAL ATTEMPTS AT RECOVERY OF Hg

Anion exchangers can usually be regenerated by passing a strong solution of a salt containing an anion that in excess will displace other anions loaded on the resin. Alternatively in the case of the weak base resins, a dilute alkali solution can be passed through the resin which results in the resin being converted to the free base form which effectively switches off the anion exchange resin.

Samples of Hg loaded Amberlite IRA93 weak base resin were initially eluted with a variety of common eluants to see what effect they would have on the recovery of Hg from the resin after 10 bed volumes elution.

The elution results are disappointing. 2M NaCl at pH 4, 1M NaCl at pH 13, 1M Nal at pH 7, 1M Nal at pH 13, 1M Na₂S0₄ at pH 7 or 11, and 2M NH₄C1 at pH 5 all gave less than 7% recovery. 2M NH₃ was best at 33%. Even alkaline solutions of NaCl and Nal have failed to displace significant amounts of Hg from the resin. These results indicate that Hg is holding on to the resin functional group with great strength. The fact that the Hg is not displaced even under alkaline conditions indicates that Hg is also coordinating directly with the nitrogen group in the resin.

Chloride elutions of Amberlite IRA900 strong base resin under both slightly acid (pH 5) and alkaline (pH 13) conditions resulted in poor Hg recovery also. Amberlite IRA900 remains ionized even under alkaline conditions and the mercury chloride complex must be held very strongly also by this resin. The NH₃ solution was partially successful in eluting Hg from the Amberlite IRA93 resin. A further series of elutions was carried out where samples of Hg loaded Amberlite IRA93 resin were eluted with 2M, 4M and 6M NH₃ solution respectively. For the 4M and 6M elutions, a white precipitate was observed to form as the NH₃ solution was passed through the resin bed. Although the NH₃ solution can elute Hg from the resin, it is still not very effective since the recovery with a 6M solution was around 60%.

The Hg loaded Amberlite IRA93 resin was then eluted with a 2M solution of ethyl amine. The result was very similar to the NH₃ elutions with a Hg recovery from the resin of 61% after 10 B.V's. elution. Once again a white precipitate was observed.

It was concluded that in order to recover Hg from the resin, the resin should be eluted with an eluant that forms a stronger complex with Hg than the resin functional group. Clearly NH₃ and EtNH₂ are partially effective, but are not sufficiently powerful to completely elute all Hg from the resin.

EXAMPLE 3 - ETHYLENEDIAMINE TRIALS

A 2M solution of ethylenediamine (1,2-diaminoethane, H₂NCH₂CH₂NH₂) was prepared, and this solution was used to elute a portion of Hg loaded Amberlite IRA93 resin. The pH of the ethylenediamine solution was 12.7. After 10

B.V's, the Hg measured value for recovery from the resin was 104%. An elution curve was obtained and this is shown in Figure 2. The curve is extremely good with most of the Hg eluted off the resin within 5 B.V's. During the elution, in the strongest [Hg] containing fractions there was observed the formation of a slow growing crystalline precipitate. Some crystalline precipitate was isolated from the solution and analysed. Analysis indicated that the precipitate had the formula C₂H₈N₂Cl₂Hg. The compound is a dichloride complex of mercury with ethylenediamine and has subsequently been observed to only form in solutions where [Hg] exceeds 6000-10000 mg/1. The precipitate has also been produced from mercuric chloride solutions to which ethylenediamine has been added.

The effect of other end chain diamines to remove Hg from Amberlite IRA93 resin was also investigated. The C₂ and C₃ diamines are very effective for Hg removal while the C₄-C₆ diamines are much less effective. It is indicated that the Hg coordinates directly to both amines in the molecule to form a ring structure. As the ring size increases, the complex formed becomes less stable and Hg is not as effectively eluted from the resin.

Ethylenediamine can also elute Hg from a strong base resin. A typical elution curve with Amberlite IRA900 resin and 1.5M ethylenediamine is shown in Figure 3. The elution curve produced is broader than the weak base resin curve but is still quite acceptable. Once again, most of the Hg is removed from the resin after 5 B.V's.

The optimum ethylenediamine concentration for the removal of Hg from both the weak and strong base resins was determined. These experiments showed that 1M to 1.5M is the optimum ethylenediamine concentration to use for elution of Hg from these resins. The weak base resin is particularly sensitive to changes in ethylenediamine concentration whilst the strong base resin performs equally as well with 1M ethylenediamine as with 2M.

A portion of IRA93 resin was put through three successive load (with 400 mg/1 Hg and 2000 mg/1 Cl in H₂S0₄ at pH 1.5) and elute ( ith 1.5M ethylenediamine) cycles in order to check that the resin could retain mercury capacity under such conditions.

In each cycle the resin was loaded up to breakthrough capacity then eluted with 5 bed volumes of ethylenediamine. The results are shown in Table 7.

It is clear from these results that the resin suffers no loss in mercury capacity as a result of elution with ethylenediamine. The effectiveness of the ethylenediamine eluant is shown by the fact that 5 bed volumes resulted in complete elution of mercury from the resin.

EXAMPLE 4 - SODIUM SULPHITE TRIALS

A sample of Hg loaded Amberlite IRA93 weak base resin was eluted with 10 B.V's. of 1M Na₂S0₃ at the ambient pH of 10.7. The recovery of mercury from the resin was almost complete indicating that S0₃ ²" can indeed elute Hg from the resin. A typical elution curve using 1M Na₂S0₃ is shown in Figure 4. During the elution of the resin, in the highest [Hg] fractions was observed a precipitate which formed in a similar fashion to the precipitate that formed during the ethylenediamine elutions in Example 3 above.

Analysis of the precipitate has indicated that it is Na₂Hg(S0₃)₂. This precipitate has been found to form when the mercury concentration exceeds around 6000 mg/1.

SO3²" will also elute Hg from Amberlite IRA900 strong base resin. A typical elution curve for this resin is shown in Figure 5. Unlike the ethylenediamine elution of this resin, a precipitate also forms when this resin is eluted. A cyclical load/regeneration trial was performed with Amberlite IRA900 strong base resin. The resin was put through three successive trials where it was first loaded with Hg solution, then eluted with 10 bed volumes of IM Na₂S0₃ solution (pH 10). The results of the trial are shown in Table 8.

These results are encouraging because they indicate that the Amberlite IRA900 resin can exchange for Hg in the S0₃ ²" form. The resin has not lost any capacity for Hg (the resin was loaded up to Hg breakthrough) which has remained in the range of 5.1 to 5.5 meq (per 5 ml wsr). It was feared initially that the resin may not adsorb any Hg after being converted to the S0₃ ^2" form. However, the resin does take up mercury in this form. When acidic feed media is passed through the resin bed any adsorbed S0₃ ²" may be converted to S0₂. Also the selectivity of the resin for (HgCl₄)²" may be much higher than that of S0₃ ²" hence the resin can adsorb Hg in this form. Not all the mercury was recovered from the resin in the 10 B.V's. elution indicating that some had accumulated on the resin after each cycle. Elution curves indicate that complete recovery of mercury is only possible after 15 B.V's. elution. During the third elution, the column became blocked with Na₂Hg(S0₃)₂ precipitate which was observed to form in voluminous amounts during each elution.

For both resin types, the optimum Na₂S0₃ concentration was determined. These results show that IM Na₂S0₃ is strong enough for the elution, and that S0₃ ^2" is almost as effective an eluting agent as ethylenediamine.

EXAMPLE 5 - OTHER ELUTING AGENTS

During the search for effective eluting agents for the removal of Hg from anion exchange resins several other substances were also found to be reasonably effective.

Ethylenediaminetetraacetic acid (EDTA) was one reagent that was also found to elute Hg from a weak base resin. EDTA's mode of action is similar to that of ethylenediamine where a stable complex is formed between the Hg and the EDTA. Unfortunately, the concentration of EDTA required for efficient elution of Hg from the resin is quite high. An elution curve for Amberlite IRA93 weak base resin using IM Na₄EDTA (the tetra sodium salt of EDTA is the only salt that will completely dissolve at this concentration, and even at IM the solution is quite syrupy) at pH 12.7 is shown in Figure 6. The elution curve obtained is very broad with complete elution of Hg not occurring until about 18 B.V's. of eluant has been passed through the column.

Na₂S will also elute Hg from a weak base resin. An elution of Amberlite IRA93 with 0.5M Na₂S at pH 7 was performed. Initially, as expected the elution produced a black precipitate (HgS) which passed through the column with the effluent. After about a further 30 minutes collection time the black precipitate completely dissolved as more Na₂S was run through the column. The mercury remained in solution under alkaline conditions, but when acid was added the black precipitate reformed (with the evolution of H₂S). Also when the solution was diluted with distilled water (10:1) the precipitate again reformed. HgS is soluble in concentrated (0.5M is concentrated) alkaline sodium sulphide solutions. Another elution was carried out with 0.5M Na₂S at pH 13 and a similar result to the above was obtained where the mercury initially precipitated off the resin then later redissolved in the excess sulphide solution. EXAMPLE 6 - RECOVERY OF Hg

Hg can be recovered directly from either ethylenediamine (weak base resin only) or Na₂S0₃ eluates as the precipitate that forms in high [Hg] fractions of the eluate. The precipitates are generally slow forming (several hours is required for maximum formation depending on the concentration of Hg in the eluate) . The precipitate can be isolated from the eluate after a time and the eluate can then be recycled back as eluant for the next regeneration of loaded resin.

This has been demonstrated with Na₂S0₃ regeneration of Amberlite IRA900 strong base resin. Table 9 shows the results of a cyclical trial involving loading Amberlite IRA900 resin up with Hg (as (HgCl₄)²" ) and then eluting with recycled Na₂S0₃ regenerant from the previous elution. Where a precipitate (of Na₂Hg(S0₃)₂) had formed in the eluate, the precipitate was removed by filtration and a small amount of fresh IM Na₂S0₃ was added to the eluate to form the eluant for the next elution.

Table 9 shows that a precipitate was observed to form in the eluate when the [Hg] was above about 6000 mg/1. The result also indicates that the eluate could be recycled back as regenerant after a small charge with some fresh Na₂S0₃ and still elute most of the Hg from the resin. After collection of precipitated Hg complex, the regeneration effluent could also be recycled.

Once the Hg is collected as a precipitate, further steps may need to be taken to obtain a useful product. The Hg from the precipitate could be obtained by breaking the precipitate down thermally. Thermogravimetric analysis (TGA) of Na₂Hg(S0₃)₂ precipitate indicates that at 150°C the precipitate loses 30% of its weight and up to 60% at 220°C leaving a grey residue. Hg was probably lost as elemental Hg(vap) to the atmosphere, and could be collected by condensation. The Hg-ethylenediamine precipitate was more resistant to thermal decomposition, losing only 30% of its weight at 280°C. This result indicates that the sulphite precipitate would be a good candidate for thermal decomposition, but not the ethylenediamine precipitate.

The mercury in the eluate could be precipitated by reduction using electrolysis. Some electrolysis trials were carried out on Hg-ethylenediamine and Hg-sulphite solutions using a variable voltage/current DC power supply connected to two strips of platinum foil about 1cm in width (the electrodes). The electrodes were dipped into a beaker containing about 30ml of the solution to be tested and a suitable constant voltage and current were applied.

When a 1.5M ethylenediamine solution is electrolysed (12V) a colourless gas was evolved at the cathode which smelt very similar to acetylene. The presence of ammonia was also detected.

Electrolysis of a 1.5M ethylenediamine solution containing 28,300 mg/1 Hg and IM NaCl was carried out at 5V and 200 mA for 40 minutes. As with the above solution a colourless gas was evolved at the cathode (and possibly also at the anode) along with the deposition of mercury metal (which coated the cathode after a time and began to drop off in globules). The voltage dropped slowly with time. After electrolysis, analysis of the solution [Hg] showed a 75% drop in solution [Hg] . The pH of the solution had also dropped from 12.0 to 11.1 and a distinctive ammonia smell indicating decomposition of ethylenediamine was observed. Electrolysis of a IM Na₂S0₃ solution resulted in the formation of a colourless odourless gas at the cathode. S0₄ ²" was probably produced at the anode.

Electrolysis of a filtered (to remove Na₂Hg(S0₃)₂ complex precipitate) Hg-S0₃ ²" solution at 4-5V and 200 mA resulted in evolution of a colourless odourless gas at the cathode along with formation of a black precipitate in the beaker followed by coating of the cathode with mercury metal.

Table 1; Hg UPTAKE WITH AMBERLITE IRA-93 RESIN lg of Amberlite IRA-93 resin in the Cl form was loaded with a synthetic solution containing Hg, Cl and S0₄ at pH 1.5.

Feed details; [Hg] = 38.8 mg/1 [Cl] = 1910 mg/1 [SO = 1825 mg/1

Resin bed volume = 3.4 ml wsr (lg dry resin) Average flow rate = 13 ml/hr/ml wsr

Total leaked (mg) 0 945 1045

SUBSTITUTE SHEET Table 2: Hg UPTAKE WITH AMBERLYST A-21 RESIN lg of Amberlyst A-21 resin in the Cl form was loaded with a synthetic solution containing Hg, Cl and S0₄ at pH 1.5.

Feed details;

Resin bed volume = 4.4 ml wsr (lg dry resin) Average flow rate = 13 ml/hr/ml wsr

SUBSTITUTESHEET Table 3: Hg UPTAKE WITH REILLEX 425 PV4P RESIN lg of reillex 425 polyvinylpyridine resin in the Cl form was loaded with a synthetic solution containing Hg, Cl and S0₄ at pH 1.5.

Feed details; [Hg] = 43.7 mg/1 [Cl] = 1980 mg/1 [S0₄] = 1600 mg/1

Resin bed volume = 3 ml wsr (lg dry resin) Average flow rate = 16 ml/hr/ml wsr

RESULT

SUBSTITUTESHEET Table 4: Hg UPTAKE TRIAL WITH AMBERLITE IRA900 RESIN lg of IRA900 in the Cl form was loaded with a synthetic solution containing HCl, H₂S0₄ and HgS0₄ at pH 1.5.

Feed details;

Resin bed volume = 4 ml wsr (lg dry resin) Average flow rate = 15 ml/hr/ml wsr

SUBSTITUTESHEET Table 5: BREAKTHROUGH CURVE OF AMBERLITE IRA93

4 ml wsr of Amberlite IRA93 was loaded with a 400 mg/1 Hg solution (2000 mg/1 Cl was also present) at pH 1.5. Fractions of effluent was collected and analysed for leaked mercury.

Resin bed volume 4 ml wsr Feed [Hg] 400 mg/1

Average flow rate 25 ml/hr/ml wsr

EFFLUENT RESULTS

Table 6: BREAKTHROUGH CURVE OF AMBERLITE IRA900

6.3 ml wsr of Amberlite IRA900 was loaded with a 400 mg/1 Hg solution (2000 mg/1 Cl was also present) at pH 1.5. Fractions of effluent was collected and analysed for leaked mercury.

Resin bed volume = 6.3 ml wsr Feed [Hg] = 407 mg/1

Average flow rate = 15.5 ml/hr/ml wsr

EFFLUENT RESULTS

Table 7: CYCLICAL TRIAL WITH IRA93 RESIN AND ETHYLENEDIAMINE ELUTION

Table 8: CYCLICAL TRIAL RESULTS WITH IRA900 RESIN AND IM Na₂S0₃ ELUTION (10 B.V. 's)

Cycle Hg on resin Hg loaded on to Hg on resin before Hg recovered by

# initially (mg) resin (mg) elution (mg) elution (mg)

1 0 510 510 468 (92%) 2 42 480 522 417 (80%)

3 105 440 545 *

(* Column blockage with N∑^HgfSO_j^ formation).

Table 9: RESULTS OF REGENERANT LIQUOR RECYCLE TRIAL USING Na₂S0₃ AS REGENERANT AND IRA900 STRONG BASE RESIN

CYCLE

Claims (19)

1. A method for removing and/or recovering mercury from natural water or wastewater contaminated with mercury, characterized in that the method comprises the steps of:

(a) extracting the mercury from the natural water or wastewater by adsorption onto an anion exchange resin; and

(b) removing the adsorbed mercury from the loaded anion exchange resin by elution with a complexing agent selected from a sulphur- containing inorganic anion and an organic ligand containing one or more donor atoms selected from oxygen, nitrogen and sulphur.

2. A method as claimed in Claim 1, characterized in that the mercury is recovered from the eluate.

3. A method as claimed in Claim 2, characterized in that the eluate is recycled as the eluant to step (b).

4. A method as claimed in Claim 2 or Claim 3, characterized in that the recovered mercury is converted into metallic mercury or a desired mercury compound.

5. A method as claimed in any one of the preceding claims, characterized in that the anion exchange resin is a strong base anion exchange resin or a weak base anion exchange resin.

6. A method as claimed in Claim 5, characterized in that the strong base anion exchange resin is a resin containing a benzyltrimethyl ammonium functional group.

7. A method as claimed in Claim 6, characterized in that the resin containing a benzyltrimethyl ammonium functional group is selected from Amberlite IRA 900 (Registered Trade Mark), Duolite A161 (Registered Trade Mark), Diaion PA-318 (Registered Trade Mark), Lewatit MP- 500 (Registered Trade Mark) and Dowex MSA-1 (Registered Trade Mark).

8. A method as claimed in Claim 5, characterized in that the weak base anion exchange resin is a resin containing a tertiary amine functional group.

9. A method as claimed in Claim 8, characterized in that the resin containing a tertiary amine functional group is selected from Amberlite IRA 93 (Registered Trade Mark), Amberlyst A-21 (Registered Trade Mark), Duolite A378 (Registered Trade Mark), Diaion WA-30 (Registered Trade Mark), Lewatit MP-62 (Registered Trade Mark) and Dowex MWA-1 (Registered Trade Mark).

10. A method as claimed in Claim 5, characterized in that the weak base anion exchange resin is composed of polyvinyl pyridine.

11. A method as claimed in Claim 10, characterized in that the weak base anion exchange resin composed of polyvinyl pyridine is Reillex 425 (Registered Trade Mark).

12. A method as claimed in any one of the preceding claims, characterized in that the sulphur-containing inorganic anion is selected from sulphide, sulphite and thiosulphate.

13. A method as claimed in any one of Claims 1 to 11, characterized in that the organic ligand containing one or more donor atoms selected from oxygen, nitrogen and sulphur is ethylenediamine, ethylenediaminetetraacetic acid (EDTA), thiourea, methionine, thiazoleamine, diaminodiphenylsulphide or thioglycolic acid.

14. A method as claimed in any one of Claims 2 to 13, characterized in that the mercury is recovered as a mercury complex from the eluate by precipitation.

15. A method as claimed in any one of Claims 2 to 13, characterized in that the mercury is recovered as metallic mercury from the eluate by reduction using electrolysis.

16. A method as claimed in any one of Claims 4 to 14, characterized in that the recovered mercury is converted into metallic mercury by thermal decomposition.

17. A method as claimed in any one of Claims 4 to 14, characterized in that the recovered mercury is converted into Hg (I) or Hg (0) by the addition of a reducing agent.

18. A method as claimed in Claim 17, characterized in that the reducing agent is selected from stannous chloride, ascorbic acid and hydrazine solution.

19. A method as claimed in any one of Claims 4 to 14, characterized in that the desired mercury compound is mercurous chloride or calomel.