WO1993002227A1 - Process and apparatus for treating fluoride containing acid solutions - Google Patents

Abstract

A fluoride-containing acid pickling solution which also contains metal salt impurities and free acid is regenerated by a process which includes treating the pickle solution by an acid sorption unit (ASU), neutralizing the resulting metal salt solution to produce a precipitate of metal impurities, which is removed, and then subjecting the salt solution to cation exchange to produce a solution containing hydrofluoric acid. The hydrofluoric acid is returned to the pickle tank, as is the free acid recovered in the ASU.

Description

PROCESS AND APPARATUS FOR TREATING FLUORIDE CONTAINING ACID SOLUTIONS

FIELD OF THE INVENTION

This invention relates generally to the treatment of fluoride containing acid solutions, for example, metal "pickling" solutions. BACKGROUND OF THE INVENTION

Pickling is the chemical removal of surface oxides or scale from metals by immersion in an aqueous acid solution. Acid solutions usually containing fluoride ions and often also containing a strong acid such as nitric acid are employed for pickling stainless steels, titanium and other metals that are corrosion resistant.

Pickling solutions become contaminated with dissolved metals through use. As the metal concentration increases, the free acid concentration decreases and pickling efficiency drops. Additions of fresh concentrated acid are made from time to time to rejuvenate the solution, but eventually it becomes spent and must be discarded.

Although many mineral acids such as sulfuric, hydrochloric and nitric are relatively inexpensive, hydrofluoric acid is considerably more expensive, so that disposal of pickle liquors containing fluoride ions represents a significant loss in terms of the value of the contained fluoride.

Also, disposal of spent pickling solutions is becoming increasingly difficult and expensive. It is no longer considered environmentally acceptable to discharge spent pickling solution directly into municipal sewers or watercourses and the availability of deep well disposal sites is becoming limited. Transport of spent pickling solution is also becoming difficult and costly, as spent pickling solution is classified as a hazardous substance, the transport of which is strictly controlled.

Many pickling operations neutralize spent pickle liquors with an alkali such as sodium hydroxide (caustic soda) or calcium hydroxide (lime) . In the case of fluoride containing pickle liquors, calcium hydroxide is usually utilized. Calcium fluoride is only slightly soluble, so that fluoride ions are removed simultaneously with the metal ions, which are precipitated. The cost of these neutralizing chemicals is considerable and can contribute appreciably to the overall cost of pickling the metal. Recently, the disposal of the resulting sludges has become a particular concern. These sludges are also considered hazardous waste and, as such, their disposal has become severely restricted and very expensive.

It is becoming widely recognized that a more sensible approach to the problem of disposal of hazardous solid waste is to reclaim the metal values. In the case of metal hydroxide sludges, pyro-metallurgical technology for converting them back to metals is well understood and is being practised today. This approach is particularly attractive for stainless steel pickling operations since sludges emanating from these operations typically contain appreciable quantities of chromium and nickel, which possess significant potential economic value. Unfortunately, the presence of fluoride in these sludges is considered deleterious to the sludge recovery process. As a result, it is not generally feasible to reclaim sludges emanating from pickling operations employing hydrofluoric acid.

In order to minimize acid consumption and disposal costs, considerable efforts are made to minimize the quantity of spent pickling solution generated by maximizing the dissolved metal concentration in the solution before it is discarded. At the same time, efforts are made to minimize the free acid concentration of the spent pickling solution to minimize the quantity of neutralization chemicals required in its ultimate treatment. Such measures, while successfully minimizing disposal costs, are not conducive to maximizing pickling productivity and quality. The time required to pickle a metal is inversely related to the free acid concentration; that is, it will require more time to pickle steel if the free acid concentration is low than if it is high. The metal concentration can also affect the pickle time. For example, higher iron concentrations significantly increase the necessary pickle time for nitric acid/hydrofluoric acid pickle liquors for stainless steel.

There are additional problems with operating a pickle process with a high metal concentration. The solubility of iron salts is limited and precipitation can occur. For example, iron fluoride will crystallize from nitric acid/hydrofluoric acid stainless steel pickle liquors when the concentration exceeds about 4% w/w. These crystals adhere strongly to tank walls and pipes and are difficult and time consuming to remove, causing a considerable amount of line down-time. Pickle liquors are known to cause considerable fuming. These acid fumes are highly corrosive and toxic. Adequate ventilation must be provided to minimize corrosion of machinery and buildings and to maintain an acceptable working environment for plant personnel. In addition, a considerable amount of acid is lost to the fumes, thereby increasing the acid consumption. These fumes must be scrubbed before admission to the atmosphere and there are capital and operating costs associated with this scrubbing process. A higher metal salt concentration in the bath will increase the vapour pressure of the acid and increase the amount of fuming.

In summary, in order to minimize waste problems, pickle baths should be operated at low final acid concentration and high final metal concentration. Unfortunately, this increases pickling times and causes operational problems such as iron salt crystallization. The increased pickle times can be partially overcome by increasing the temperature, but this will increase the amount of fuming. DESCRIPTION OF THE PRIOR ART

Various processes have been employed to purify or regenerate (ie. recover) spent pickling solution to reduce the cost of disposal of spent pickling solution and reduce purchases of fresh acid. For example, spray roasters have been employed for hydrochloric acid and more recently nitric/hydrofluoric acids. Ion exchange/sorption and dialysis systems have been introduced over the past few years for this purpose. Strong base anion exchange resins with quaternary amine functionality have the ability to sorb acids from solution, while excluding metallic salts of those acids. The process, known as "acid retardation", is reversible, in that the acid can be readily de-sorbed from the resin with water. It is thus possible, by alternately passing contaminated acid and water through a bed of this resin, to separate the free acid from the metal salt. A similar phenomenon occurs with anion exchange membranes. It is possible to utilize anion exchange membranes in the so-called "diffusion dialysis" process to separate free acid from the metal salts.

Both acid retardation and diffusion dialysis systems, which may both be considered to be acid sorption systems, have been employed to purify pickling solutions of dissolved metallic contamination.

Contaminated pickling acid flows from the pickle bath to the acid sorption unit (ASU) . The acid is removed by the ASU and the metal salt bearing byproduct solution exits from the ASU. Water is used to elute the acid from the ASU and the resulting acid product flows directly back to the pickle bath.

Both the acid sorption processes have the advantage of being simple and low cost. In addition, with these processes, it is possible to operate the pickle tank at any desired concentration of dissolved metal and free acid, so that pickling performance can be optimized.

The major disadvantage of these acid sorption processes is that they generate a by-product or waste stream consisting of a mildly acidic salt solution of the metal being dissolved in the pickling process. This by¬ product stream must be further treated, usually by neutralization with base, in order to render it harmless to the environment. In the case of stainless steel pickling, where hydrofluoric acid is employed, this by- product stream contains an appreciable quantity of fluoride since some of the metals are strongly complexed by fluoride. The by-product is usually neutralized with lime to remove the fluoride ions as well as the metals. As discussed above, the presence of fluoride obviates the possibility of pyro-metallurgically reclaiming the metal values from the sludge.

Regular additions of concentrated makeup acid are required to replace acid neutralized through metal dissolution. Even when a recovery system of this type is employed it is normally not possible to reclaim more than about 50% of the fluoride values in the spent pickling solution in the case of pickling of stainless steel with nitric/hydrofluoric acid.

Munns ("Iron Control in Hydrometallurgy" pages 537-548) suggested a process to recover the fluoride values from the ASU by-product in which the fluoride/nitrate salt by-product from the ASU is neutralized with sodium hydroxide to precipitate out the metal impurities. After filtration, the fluoride/nitrate salt solution is processed by cation exchange to convert sodium fluoride and sodium nitrate to hydrofluoric/nitric acid, which after evaporative concentration could be recycled.

Iri the Munns paper, no reference is made to the anion composition of the streams — only metals and free acid. It would be expected that the ratio of fluoride to nitrate in the salt solution would be the same as in the pickling bath and that the ion exchange unit would therefore not be effective in removing the sodium because of the high concentration of nitric acid that would be produced by the ion exchange unit as the sodium is exchanged for hydrogen.

Munns did not consider the difficulty in treating such a concentrated_ solution. The amount of dilution that would occur during the cation exchange treatment of such a highly concentrated solution would make the process impractical.

Munns does not consider regeneration of the cation exchanger. Sulfuric or hydrochloric acid is normally employed to regenerate cation exchange resins. This could result in contamination of the pickle bath with foreign anions (i.e. sulfate or chloride) if the resin is not well rinsed following regeneration.

Finally, Munns suggested the use of sodium hydroxide as a neutralization agent. This would not be effective in terms of fluoride recovery because of the limited solubility of sodium fluoride. Munns also does not address the question of che'mical consumption. The cost of purchasing sodium hydroxide for neutralization as suggested by Munns would be prohibitive. BRIEF DESCRIPTION OF THE INVENTION An object of the present invention is to provide an improved process and apparatus for treating a fluoride- containing acid solution to reclaim fluoride values.

In the process aspect of the invention, a fluoride-containing acid solution is treated by an acid sorption unit to reclaim free acid and produce a metal salt solution. The acid solution contains, in addition to fluoride ions, free acid, a strong acid and at least one dissolved metal salt selected from the group consisting of iron, chromium and titanium salts, in which the molar ratio of fluoride to anion of said strong acid is less than about 1. The molar ratio fluoride to anion of the strong acid in the metal salt solution is greater than about 1. The solution is neutralized with a base solution, the fluoride salt of which is appreciably soluble, to produce a precipitate of the metal of the metal salt, and a salt solution of the base, containing fluoride ions. The precipitate is removed and the salt solution is treated by a cation exchanger to cause cations from the base solution to exchange for hydrogen ions, and produce a solution containing hydrofluoric acid, which is recovered. While the process may be used to recover hydrofluoric acid from any acid solutions of appropriate composition, as has been indicated above, a primary application of the process is in regenerating fluoride containing metal pickling solutions. The invention provides a means of reclaiming a high proportion of the fluoride values from spent pickling solutions and producing a metal hydroxide sludge containing a low fluoride content which can be further processed to reclaim the metal values. At the same time, the process allows the pickling bath to be operated at high free acid and low dissolved metal levels to optimize the pickling performance.

The fact that the molar ratio of fluoride to anion of the strong acid changes from less than^' about 1 in the pickling solution to greater than about 1 in the metal salt solution means that the effectiveness of treatment in the cation exchanger is high.

The invention also provides an apparatus for carrying out the process. BRIEF DESCRIPTION OF DRAWINGS

In order that the invention may be more clearly understood, reference will now be made to the accompanying drawings which illustrate particular preferred embodiments of the invention by way of example, and in which: Fig. 1 is a diagrammatic illustration of a prior art arrangement for purifying pickling solutions of dissolved metallic contamination, using an ASU; Fig. 2 is a diagram illustrating the process and apparatus of the present invention;

Fig. 3 is a graph illustrating metal solubilities at different pH levels; Fig. 4 is a graph illustrating the effect of pH on the passage of fluoride ions through a particular reverse osmosis membrane;

Fig. 5 illustrates a modification of the process shown in Fig. 1 in which a reverse osmosis unit is used to concentrate the salt solution prior to its delivery to the cation exchange unit of the process;

Fig. 6 shows another embodiment of the invention, in which an electrolytic cell is employed to recover acid and base; Fig. 7 is a diagram of anion concentration profiles for a cation exchange unit that may be used in the process;

Fig. 8 is a further diagram illustrating an embodiment of the invention in which an evaporator is used to remove water from the system;

Fig. 9 is a graph illustrating efficiency of fluoride recovery as a function of the temperature in the neutralization step of the inventive process; and,

Figs. 10 to 13 are views similar to Fig. 2 illustrating further embodiments of the invention.

In the drawings, the same reference numerals are used to denote the same parts throughout the various views.

DESCRIPTION OF THE PRIOR ART Fig. 1 shows an example of an arrangement that has been used in the past to purify pickling solutions of dissolved metallic contamination. A pickle bath is indicated by reference numeral 1 and contains contaminated pickling acid indicated by reference numeral 2. Acid is withdrawn from the bath and flows to an acid sorption unit (ASU) 3. The acid is removed by the ASU and the metal salt bearing by-product solution exits from the ASU as indicated at 9. Water represented by arrow 7 is used to elute the acid from the ASU and the resulting acid product is returned to the pickle bath as indicated at 8. DESCRIPTION OF PREFEWRKn EMBODIMENTS Referring first to Figure 2, the principal components of the apparatus provided by the invention are the acid sorption unit (ASU) 3, a chemical neutralization tank 4 and a cation exchange unit 6. The ASU may be of the acid retardation or diffusion dialysis type. Contaminated fluoride containing pickling solution 2 bearing free acid and metal salt contamination is fed to the ASU 3. Water 7 is used to recover an acid product 8 from the ASU, from which a portion of the metal contamination has been removed. The acid product is returned directly to the pickle bath. A by-product solution 9 containing dissolved metal salt, including fluorides, and a small quantity of free acid is directed to a chemical neutralization tank 4.

A base solution 10 which forms soluble fluoride salts (such as sodium hydroxide, potassium hydroxide or ammonium hydroxide) , is added to the by-product solution to raise the pH of the solution to pH 7-11 to neutralize the free acidity and precipitate the metals as hydroxides according to equation (1). The final pH of the solution should be sufficiently high to break the metal fluoride complexes and liberate the fluoride ions to the solution. As can be seen from Figure 3 if the pH is too low, the solubility of the nickel hydroxide becomes appreciable. On the other hand, if the pH is too high, chromium hydroxide, being amphoteric, will re-dissolve, increasing the solubility of chromium in the solution.

FeF₃ + 3K0H > 3KF + Fe(OH)₃ (1)

Neutralization of the solution in this manner liberates the anions, including the fluorides, as dissolved alkali metal salts. For example, in a nitric acid/hydrofluoric acid stainless steel pickling operation, if potassium hydroxide is employed for neutralization, a sludge of iron hydroxide, chromium hydroxide and nickel hydroxide is produced along with a solution of potassium fluoride and potassium nitrate.

It has been found that the temperature at which the neutralization reaction of equation (1) is carried out is very important in maximizing the fluoride recovery efficiency of the process and minimizing the amount of fluoride remaining in the precipitate. It is important to minimize the amount of fluoride remaining in the precipitate to ensure that it is suitable for pro- metallurgical reclamation. Specifically, it has been found that neutralization should be conducted at an initial temperature higher than ambient temperature (22°C) and preferably at least 60°C, for maximum fluoride recovery. Experiments that were conducted in relation to this aspect of the invention are discussed below under the heading "Examples". The sludge 11 resulting from neutralization is separated from the liquid by suitable means such as gravity sedimentation or filtration and de-watered by suitable means such as a filter press 5, producing a filtrate containing dissolved potassium nitrate and potassium fluoride. The de-watered sludge may be washed with water 12 in order to remove any residual fluoride content. The wash water , bearing the fluoride, could be combined with the filtrate, which would tend to dilute the filtrate. The salt solution 15 which results from neutralization, including the filter cake washings, and containing the dissolved fluoride salt, is directed to an ion exchange unit 6. The ion exchange unit contains a bed of particulate strong, sulfonic acid type cation exchange resin such as Amberlite IR 120, manufactured by the Rohm and Haas company of Philadelphia. The ion exchange resin is initially in the regenerated or hydrogen form. Upon passing through the ion exchange unit, the base cation (eg. Na⁺, K⁺ or NH₄+) is exchanged for hydrogen, thereby producing a third acid solution according to equation (2) .

KF + RH > RK + HF (2)

where, RH represents the cation exchange resin in the regenerated, hydrogen form.

As discussed above, although various neutralization agents can be employed, there is an additional advantage in using potassium hydroxide in lieu of sodium hydroxide for neutralization that is not recognized by the prior art. The strong acid cation exchange resins employed in this invention have a higher selectivity for potassium than sodium. This means potassium can be taken up by an ion exchanger more efficiently than sodium, which allows for the production of a purer hydrofluoric acid product.

The ion exchange equilibrium is favourable for the production of hydrofluoric acid. Hydrofluoric acid, being a weak acid, exists largely in its undissociated form, so that there are very few free hydrogen ions present in the solution to compete with the base cation for the exchange. As a result, it is possible to produce_, relatively concentrated solutions of hydrofluoric acid containing extremely low levels of cationic contamination. The hydrofluoric acid solution 13 produced by the ion exchange unit is recycled back to the pickle 1 bath for reuse as indicated by line 13.

Pickling solutions frequently contain strong acids in addition to the hydrofluoric acid. A strong acid is defined as one which exists virtually entirely in the ionized or dissociated state. If the pickling solution contains nitric acid for example, the salt solution delivered to ion exchange unit 6 will contain anions of that acid i.e. nitrate. A higher concentration of nitric acid will tend to improve the performance of the pickling process. Ideally, the quantity of anion in the strong acid will be greater than the amount of fluoride so that the molar ratio of fluoride to strong acid anion will be less than 1. If nitrate ions are present, nitric acid is simultaneously produced by the ion exchange unit with hydrofluoric acid according to equation (3). Other acids can be similarly produced.

KN0₃ + RH > RK + HNO₃ (3)

Since nitric acid is a much stronger acid than hydrofluoric acid, the equilibrium for uptake of salt cation (e.g. K⁺) is not so favourable because the free hydrogen ions will compete with the salt cation for exchange sites. As a result, the concentration and purity of nitric acid or other strong acids that can be produced is somewhat less than with hydrofluoric acid. Where a mixture of hydrofluoric acid and stronger acid such as nitric is treated, the uptake will consequently be better at higher fluoride to strong acid anion ratios.

Consideration may be given to the omission of the ASU from the system and neutralization of the contaminated pickle solution directly. However, use ^*of an

ASU in the system as taught by this invention, is highly advantageous. Utilization of the ASU allows the pickle bath to operate at high free acid levels and low metal levels without significantly increasing the consumption of base. This in turn reduces the loading on the ion exchange unit and consequently its required size and chemical regeneration requirements. The benefits of operating a pickle process at high acid and low metal levels have been outlined above. Use of the ASU has additional benefits however which are not so readily apparent.

Normally in mixed acid metal pickling solutions such as nitric/sulfuric acid, the ratio of anions in the metal salt byproduct will be approximately the same as in the pickle bath. For example, normally, if the ratio of two different anions is less than 1 in the feed, the ratio would also be less than 1 in the metal salt bi-product. Surprizingly, in nitric/hydrofluoric acid stainless steel pickle liquors it has been found that the ratio of fluoride to nitrate ions in the metal salt byproduct solution from the ASU and therefore the neutralized by¬ product solution exceeds that of the pickling solution. This has heretofore not been known. In order to optimize the performance of the pickling bath, the bath composition is adjusted so that the nitric acid concentration is relatively high. As a result, the equivalent ratio of fluoride to nitrate in the pickle solution is usually less than 1 (eg. 0.8). At the same time, in the ASU byproduct this ratio is usually greater than 1 (eg. 1.4). The acid then produced by the ion exchange unit will be predominantly hydrofluoric acid with a relatively small proportion of nitric. From an ion exchange equilibria stand-point, this is a more favourable situation, as already discussed. Thus use of an ASU in the manner taught by this invention, not only reduces the quantity of base which is required, which might normally be expected, but it results in an unexpected improvement in the performance of the ion exchange unit due to the change in the fluoride to nitrate ratio in the ion exchange feed.

This change in anion ratio in the ASU likely occurs because iron, chromium, titanium and many other metals that may be pickled form very strong fluoride complexes, so that most of the fluoride in the pickle bath exists as metal fluoride complexes which are rejected by the ASU resin or membrane and recovered in the metal salt by-product stream. Nitrate, on the other hand, does not complex to any significant degree, so that it exists predominantly as nitric acid, which is readily taken up by the ASU and recovered by water elution in the ASU acid product. Other strong acid anions, such as chloride and sulphate do not complex strongly with metals such as iron, chromium and titanium, which form fluoride complexes, so that mixtures of these anions with fluoride would behave in a like manner to fluoride/nitrate solutions. This complexation phenomenon is not universally beneficial, however. It has been found for example that in zirconium nitric/hydrofluoric acid pickle baths, zirconium fluoride complexes are not rejected efficiently by the ASU so that there is no benefit in employing an ASU. Other examples where metal/anion complexation occurs with deleterious effects on the ASU, include copper, lead and zinc in hydrochloric acid containing solutions.

Upon exhaustion, the ion exchange resin must be regenerated with a strong acid according to equation (4) . Sulphuric acid or nitric acid or hydrochloric acid regenerant solution 17 at a concentration of 0.5-5 N, for example, is delivered to the ion exchange unit for this purpose. Regeneration will yield a salt solution 14 of the acid employed for regeneration and the original base employed in the neutralization step. For example, if potassium hydroxide is employed for neutralization and sulfuric acid is employed for regeneration, a potassium sulfate waste solution will be produced. This third salt solution is considerably less objectionable to the environment than the original spent pickling solution and in most cases can be discharged directly, after minor adjustment of pH with further addition of base, if necessary.

2RK + H₂S0₄ > K₂S0₄ + 2RH (4)

While any of these acids can be utilized with good regeneration efficiency, nitric acid is preferred since its use avoids the possibility of contaminating the pickle bath with foreign anions such as sulfate or chloride. This could occur for example if the rinsing of the ion exchange resin after regeneration is inadequate. Nitric acid is not normally recommended for regenerating ion exchange resins because of the risk of resin oxidation, but in the present invention this risk is avoided by employing very dilute nitric acid i.e. at less than 2 molar concentration and preferably at less than about 1 molar. While this concentration would be considered too dilute to be used in most ion exchange applications it is effective in this application because the monovalent salt cations (e.g. potassium) are not held too tightly to the resin. While the basic cation exchange process is frequently employed in the demineralization of water and. is well known to those skilled in the art, the concentration of ions to be exchanged in the present invention far exceeds that normally treated by ion exchange. For example, the maximum total dissolved solids level normally considered treatable by an ion exchange demineralizer is about 1000 ppm or 1 g/L (as CaC0₃) . With this invention the ion exchange unit is required to treat solutions typically containing 40 g/L (as CaC0₃) - forty times greater in concentration than normally considered treatable. For this reason one skilled in the art would not normally consider utilizing ion exchange in this manner. There are two reasons why ion exchange is not normally employed for treating such concentrated solutions:

The first reason is that the regenerant chemical consumption would be very high to treat such concentrated solutions. Means are outlined below on how the cost of chemicals in operating the ion exchange system is minimized by this invention, thereby overcoming this limitation.

The second reason is that a substantial amount of dilution occurs when the small volume of concentrated fluid being treated is passed in and out of the ion exchange resin vessel. It has been unexpectedly found that the ion exchange unit employed in the present invention can actually be used to concentrate the fluoride contained in the second salt solution while exchanging the cations. This concentrating effect helps to compensate for the amount of dilution that occurs as the second salt and third acid solutions are passed in and out of the ion exchange resin vessel and ultimately reduces the amount of water that must be removed from the system. This concentrating effect has not previously been reported.

Figure 7 shows anion concentration profiles for a cation exchange unit in which a salt is converted to an acid. The concentration on the Y axis is normalized against the feed concentration (C/C₀), while the volume on the X axis is shown in terms of fractions of a resin bed volume (BV) .

The salt solution to be treated is delivered into the bottom of the resin bed, which prior to the treatment is filled with water. Ideally, there is no mixing of the water in the bed with the feed solution, so that the concentration profile for the salt anion would resemble line A. Note that the concentration should break through at approximately 0.38 bed volumes, which corresponds to the interstitial volume of a bed of uniform spheres, and instantly rise to the feed concentration. In practice, there will be some intermixing of the water initially in the bed with the feed, so that the concentration profile for the anion would normally more closely resemble line B. The inflection point of the breakthrough curve will be observed slightly later e.g. 0.5 BV, due to other void volumes in the resin bed such as flow distribution apparatus. Note that the concentration of the anion should still rise to a maximum concentration approximately equal to the feed concentration. This is because normally the anion is not affected by the cation exchange resin. Typical concentration profiles for cation exchange resin beds are shown in U.S. Patent No. 4,673,507 (Brown).

Curve C shows the fluoride concentration profile for treatment of the second salt solution E listed in Table 1. Note that the inflection point for the breakthrough profile occurs at approximately 0.75 BV, well after the normally expected 0.5 BV. Note also that the concentration reaches approximately 1.15 times the feed concentration. It would seem that the ion exchange resin is adsorbing the fluoride that first enters the resin bed, thereby causing the delay in the breakthrough. This fluoride is later desorbed from the resin, fortifying the solution and causing the fluoride concentration collected from the resin bed to exceed the initial level.

Thus it can be seen that with the present invention, the cation exchange unit has the effect of concentrating the fluoride ions, in addition to its primary function of converting the fluoride salt to hydrofluoric acid.

After collecting the acid product from the cation exchange unit, it is necessary to recover the entrained void of second salt solution from the resin bed with water. In order to maximize recovery of this solution from the resin bed it is necessary to employ more than the theoretical 0.38 BV of water. This will result in some dilution. The diluted solution recovered from the cation exchange resin bed can then be combined with more second salt solution and treated in subsequent ion exchange treatment cycles. The concentrating effect discussed above, will then help to offset this dilution.

The volume of the acid product 8 produced by the ASU is typically approximately the same as the volume of contaminated pickling solution treated by the ASU. The volume, of the acid solution produced by the ion exchange unit 13 therefore represents a surplus amount of liquid added to the pickling bath. This excess water must be removed to avoid overflowing the liquid level in the bath.

Many pickling processes operate at elevated temperatures so that there is a certain amount of natural water evaporation from the liquid surface. This will help to compensate for the excess water. Under many circumstances the surface evaporation rate will be insufficient to meet the water balance. In this case, it will be necessary to supplement the surface evaporation by further concentration means. Various known concentration techniques such as evaporation, electrodialysis or reverse osmosis, may be employed to remove water from the pickle bath or concentrate the acid produced by the ion exchange unit or the ASU to reduce the volume of solution being recycled. It is also possible to employ a concentrator to concentrate the salt solution prior to ion exchange treatment. In this case, the materials of construction for the concentrator will be less expensive, as a relatively neutral fluoride salt is less corrosive than hydrofluoric acid. If an evaporator concentrator is used, there will be less carry-over of fluoride into the recovered water condensate when treating fluoride salt than acid.

As noted above, an evaporator could be installed at a variety of locations. As shown in Figure 8, a particularly effective location for the evaporator 22, is on the spent pickle liquor feed to the ASU. Providing the solubility limit of the iron salt is not exceeded, this location has several advantages: - the flow rate of solution 24 requiring treatment with the ASU is reduced, which reduces the size of the ASU since the flow rate of the ASU acid product and ion exchange acid product are reduced, the amount of water that must evaporated is also reduced, with a corresponding reduction in evaporator capital and operating costs removal of energy for the evaporation may result in cooling of the solution, which helps to reduce oxidation of the resin or membrane employed in the ASU by nitric acid. The condensed water stream from the evaporator 23 may contain low concentrations of nitric and/or hydrofluoric acid since both these acids have relatively high vapor pressures. For this reason the water stream should not be discharged directly to waste without proper pretreatment. The water can be utilized instead, as a replacement for fresh water 7 by the ASU as shown and/or as a replacement for fresh water 12 for washing the filter cake.

Reverse osmosis is a concentration process which is a very low energy consumer. There are a number of important considerations involved in applying reverse osmosis in conjunction with this invention.

The efficiency of the reverse osmosis process for removal of water from fluoride solution is highly dependent on pH. Figure 4 shows the percentage passage of fluoride ions through a Dupont Permasep B-10 reverse osmosis membrane as a function of pH. From this it can be seen that for removal of water with less than 7.5% passage (i.e.loss) of fluoride, the pH of the solution to be treated must be greater than pH = 7.

The stability of reverse osmosis membrane is also dependent on pH. For example, the Dupont Permasep® B-10 membranes, manufactured by E. I. Dupont de Nemours, which are an aramid hollow fibre polymer, have a recommended operating pH range of 4 - 10. Taking into consideration good fluoride rejection and good membrane life, the pH of the fluoride solution to be concentrated should be in the range of 7-10.

The hydrofluoric acid solutions in the pickle bath or those produced by the ion exchange unit or either of the product or byproduct solutions produced by the ASU will have a pH significantly less than 7 and even less than 1, if appreciable levels of a strong acid such as nitric are present. It is therefore not feasible to employ reverse osmosis to concentrate any of these solutions. On the other hand, the second salt solution produced as a result of neutralization of the ASU byproduct is at the ideal pH (i.e. pH = 7-11) for concentration by reverse osmosis.

Figure 5 shows a reverse osmosis unit 19 used to concentrate the second salt solution containing fluoride 15. The concentrated salt solution rejected by the reverse osmosis membrane 20 is then treated by the ion exchange unit 6. The water permeate rom the reverse osmosis unit 21 will contain a low concentration of fluoride. This stream can be discharged to waste after proper pretreatment or utilized as a replacement for fresh water 7 by the ASU and/or as a replacement for fresh water 12 for washing the filter cake. By utilizing the reverse osmosis unit in this manner the overall water imbalance in the system is substantially reduced. As discussed above, various bases can be employed in the neutralization step, providing that soluble fluoride salts are yielded by the neutralization. Ammonium hydroxide is inexpensive and produces ammonium fluoride which has a high solubility. Ammonium hydroxide has two principal disadvantages however: First of all, ammonium ions form a complex with nickel, which prevents precipitation of the nickel hydroxide. Secondly, the ammonium salt produced by the ion exchange unit (ie. the third salt) would also contain the nickel that leaves the neutralization step. Further treatment of this stream would be required prior to final discharge. Treatment of ammonium containing wastewaters is somewhat difficult because of its complexing action.

Sodium hydroxide is somewhat more expensive than ammonium hydroxide but is very effective in precipitating nickel and other metals without complexing any of the metals. The major limitation of sodium hydroxide is that the solubility of the sodium fluoride that is produced is somewhat limited ie. approximately 1 M in water. For this reason, the concentration of fluoride in the metal salt byproduct must be less than this level to avoid crystallization of the sodium fluoride and contamination of the metal hydroxide sludge with fluoride ion, which as already stated, is objectionable.

Potassium hydroxide is particularly effective, as potassium fluoride is very soluble in water and potassium does not complex with the metals. In addition, it forms highly conductive sulfate nitrate and chloride salts, which significance will explained below. Unfortu¬ nately, potassium hydroxide is relatively expensive and its use would adversely impact on the economics of the process.

Figure 6 illustrates the fact that it is possible to employ an electrolytic cell 18 to convert the potassium salt waste 14 from the cation exchange unit 6 back to potassium hydroxide 10 for reuse in the neutralization step and acid 17 (either sulfuric or hydrochloric) for reuse in regeneration of the cation exchange unit 6. This significantly reduces or eliminates the cost of purchasing chemicals for the process and reduces or eliminates the discharge of salt to waste. The energy costs may be less than the cost of the chemicals so that the economics of the process are more favourable.

The use of a monopolar electrolytic cell for the purpose of splitting a salt into acid and base is well known to those skilled in the art and is described by Millington and Nott. An electrolytic cell of the electrodialysis type employing bipolar membranes can also be utilized to split a salt into acid and base as discussed in U.S. Patent 4,504,373. The voltage requirements for the bipolar membrane cell are lower than for the monopolar cell. Bipolar cells have also been used to recover spent regenerants from ion exchange units. In this case both the acid and base produced from the cell are re-used by the ion exchange unit for regeneration purposes. Bipolar membrane electrolytic cells have been employed for the regeneration of fluoride containing solutions such as nitric/hydrofluoric stainless steel pickle liquors as in U.S. Patent 4,740,281 (Chlanda) and 4,943,360 (Sugisawa) . In both these processes, the pickling solution is first treated by electrodialysis unit to recover a portion of the free acid content. The de-acidified salt solution emanating from the electrodialysis unit is then neutralized with a base. The fluoride/nitrate salt solution produced from the neutralization step is processed directly by the bipolar membrane electrolytic cell. These processes have a number of disadvantages compared to the present invention, including the following:

A significant amount of electrical energy is expended in the electrodialysis cell used for recovery of the free acid content of the pickling solution. With the present invention the electrical energy requirements for the acid sorption unit are minimal.

The electrodialysis unit does not recover the free hydrofluoric acid contained in the pickling solution to any significant extent. This is because hydrofluoric exists in the un-ionized HF state in these solutions. This increases the quantity of base required for neutralization and the size of the bipolar cell. With the present invention, the acid sorption unit can recover 80-90% of the free hydrofluoric acid. The membranes in both the electrodialysis and bipolar electrolytic cells of the Chlanda/Sugisawa process come in contact with nitric acid which may be in excess of 2 molar. Nitric acid at this strength has a tendency to oxidize the membranes. In the present invention, the membranes come in contact with sulfuric or hydrochloric acid which are not oxidizers or nitric acid at a concentration of less than 2 molar and preferably less than about 1 molar. Therefore the life of the membranes is extended. The chance of anion contamination occurring in the electrolysis cell of the present invention due to diffusional transport of anions across the membranes or via a membrane perforation is probably of even greater concern than contamination in the ion exchange unit. By employing nitric acid for regeneration of the ion exchanger this problem is avoided. There is an additional reason to limit the acid concentration when one considers the recovery of acid and base by an electrolytic cell. The electrical efficiency of the electrolytic cell is limited to a large extent by the back diffusion of hydrogen ions across the anion exchange membranes employed in the cell. At acid concentrations of greater than 2 molar the electrical efficiency is prohibitively low.

The fluoride solutions processed by the electrolytic cell in the Chlanda/Sugisawa process are much less conductive than the sulfate nitrate or chloride solutions processed by the cell in the present invention, so that energy requirements (i.e. voltage) are lower for the present process.

The present invention seeks to overcome the limitations of these prior processes.

Another embodiment of the invention will now be described. This particular embodiment which is shown in Figure 10 is similar to that shown in Figure 2 except that the ion exchange unit 6 ' employs cation exchange membranes in lieu of particulate cation exchange resins, as heretofore described. It is not necessary to provide an acid regenerant 17 in this case. Regeneration is accomplished, in effect, by applying a direct electric current to the ion exchange unit. The ion exchange unit is therefore an electro-membrane cell as shown in more detail in Figure 11 consisting of an anode 20 and a cathode 21 as well as a cation exchange membrane 19, which divides the cell into its two compartments — the cathode compartment 22 and anode compartment 23. The cation exchange membrane may be chosen from a variety of commercial products including Nafion, manufactured by E.I. Dupont. When a DC current is applied to the cell, the anode reacts with the water to produce oxygen gas and hydrogen ions (H⁺), while the cathode reacts with water to produce hydrogen gas and hydroxyl ions (OH^") . The fluoride bearing salt solution 15 is admitted to the anode compartment 23. Free cations present in the anode compartment, which in .this __ case could be primarily potassium and hydrogen, pass through the cation membrane 19 into the cathode compartment 22. Potassium ions passing across the membrane will result in the production of potassium hydroxide in the cathode compartment, while hydrogen ions will result in the neutralization of hydroxyl ions and production of water. As a result of these reactions base (e.g. potassium hydroxide) can be collected from the cathode compartment 10, either on a batch or continuous basis. The base so produced can be employed in the neutralization tank 4 as shown in Figure 10. At the same time, the initial charge of fluoride containing salt in the anode compartment will be gradually converted to acid so that eventually a solution containing hydrofluoric acid 13 can be collected from the anode compartment. Collection of the acid is preferably done an a batch basis to maximize the efficiency of conversion of said fluoride bearing salt to acid. Since the fluoride ions will associate with hydrogen ions as they are generated at the anode, there will not be a significant quantity of free hydrogen ions available to pass across the cation membrane until the fluoride salt has been largely converted to hydrofluoric acid. Conversion of any strong acid salt present will be much less efficient, however, since its anion will not associate with the hydrogen ions to any significant extent. As a result, the efficiency of the process drops off sharply soon after conversion of the hydrofluoric acid has been completed and conversion of the strong acid begins. Since in the present invention the ratio of fluoride to anion of the strong acid (e.g. nitrate) in the salt solution is greater than 1, the overall efficiency of the process will remain reasonably good, providing it is not necessary to convert too high a proportion of the strong acid anion. The process efficiency would be extremely low if the molar ratio of fluoride to strong acid anion was much less than 1. This is the reason that similar prior art processes, where the molar ratio of fluoride to strong acid anion may be more than 1 , such as those of Chlanda and Sugisawa, employ a salt splitting process with anion exchange membranes as well as cation exchange and bipolar membranes. In these cases the acid is collected from the compartment on the opposite side of the anion membrane from that into which the salt has been admitted. In the present invention the acid is collected from the same compartment as that into which the salt has been admitted.

In a variation of this latest embodiment, a bipolar ion exchange membrane can be used instead of electrodes to split the water molecules. This is shown in Figure 12. A single electrode pair (20 and 21) is required at the end of a series of bipolar membrane 24 and cation membrane 19 pairs. The fluoride containing salt is fed to the compartment 23' between the cathode side of the bipolar membrane and the cation membrane. Acid product is collected from the same compartment 23' , while potassium hydroxide is collected from the compartment 22' between the anion side of the bipolar membrane and the next cation membrane.

Many anode materials and certain types of bipolar membranes are not stable when in contact with high concentrations of hydrofluoric acid. This includes bipolar membranes currently manufactured by WSI Technologies of St. Louis, Missouri. In these cases, it is necessary to isolate the anode or bipolar membrane from the hydrofluoric acid as shown in Figure 13 with a second cation membrane 19 ' and fill the chamber formed between the anode or bipolar membrane with an acid which is compatible with the anode or bipolar membrane, such as sulfuric, hydrochloric or nitric acid. The bipolar membrane 24, the second cation membrane 19' and the additional isolating chamber formed 25 are shown in addition to the first cation membrane 19. In this case, hydrogen ions are passed from the anode or bipolar membrane 24 through the acid solution contained in the isolating chamber 25 to the second cation membrane 19' and then through the second cation membrane into the compartment containing the fluoride solution 23' . The potassium and hydrogen ions pass across the first cation membrane 19, as before into the base compartment 22".

Thus, in these latest embodiments, a cation exchange membrane is employed to cause cations from the base solution to exchange for hydrogen ions and produce a solution containing hydrofluoric acid.

Because the concentration of nitrate in the salt is low according to this invention, the concentration of nitric acid produced by the electromembrane cell is relatively dilute, typically less than 1 molar. Moreover, the membranes are only in contact with nitric on an intermittent basis, the remainder of the time the nitrate exists as a salt. As a result, oxidation of the membranes by nitric acid is not a serious problem. This is in contrast to prior processes such as those of Chlanda and

Sugisawa, where the membranes are in continuous contact with nitric acid at a higher concentration.

Example 1

A sample of typical spent stainless steel pickle liquor with a composition as shown in Table 1 and denoted as solution (a) was treated with a commercial acid sorption unit employing the acid retardation technique

(known as the APU and manufactured by Eco-Tec Inc. of

Pickering, Ontario, Canada) . The composition of the recovered acid product solution (b) and byproduct solution

(c) are also shown in Table 1. Normally, the acid product would be recycled to the pickle bath for reuse and the byproduct would be disposed of as waste liquid. The relative volume of each solution in relation to 1 litre of pickling solution (stream (a), first acid solution) are also shown in Table 1. The by-product solution (c) was neutralized with

5N potassium hydroxide to pH = 8.5. The resulting sludge was separated from the liquid with a filter press and the filter cake was washed with water. The wash water was combined with the filtrate, forming solution (d) and having a composition as shown in Table 1. Approximately 80% of the total fluoride values originating in solution (c) were recovered in solution (d).

Stream (d) was then concentrated by reverse osmosis. The reverse osmosis unit employed a hollow fibre aramid membrane (Dupont Permasep B-10) and operated at approximately 1200 psi. The composition of the final reject solution produced by the reverse osmosis unit is shown in Table 1 as solution (e) . The permeate stream from the reverse osmosis unit is not shown in Table 1, however it contained less than 2% of the dissolved fluoride and nitrate salts recovered in the reject stream.

Solution (e) was then treated by an ion exchange unit employing a sulfonic acid type strong acid ion exchange resin. The resin was subsequently regenerated with IN hydrochloric acid to convert the resin back to the hydrogen form. After 15 loading and regeneration cycles the solutions were collected and analyzed. The composition of solution (e) after treatment by the ion exchange unit is shown as solution (f) in Table 1. It can be seen that approximately 84% of the fluoride and 93% of the nitrate contained in solution (d) as potassium salts were collected as acids in solution (f). The remainder of the fluoride and nitrate were for the most part left in the resin bed after the entrained void of solution (e) was displaced from the bed with water. It will be recognized that this recovery efficiency could be improved if a greater volume of water were employed for this displacement, the compromise being a reduction in the concentration of solution (f).

Solution (f) would be recycled back to the pickle bath, although it is recognized that if large quantities of this solution were recycled it may be necessary to artificially supplement the pickle bath surface water evaporation losses.

Recycle of streams (b), the ASU product and (f), the third acid solution, represents 86% recovery of the fluoride and 94% recovery of the nitrate in the contaminated pickling solution (a) which would normally have gone to waste if no regeneration system were utilized. In other terms, it represents 68% recovery of the fluoride and 69% recovery of the nitrate which would normally have gone to waste if an ASU were employed by itself, according to the prior art.

Table 1

Experiments were also conducted to determine the effect of temperature on the neutralization reaction in the process of the invention. Neutralization was carried out at different temperatures. The resulting precipitate was filtered out and then repulped (i.e. washed) with water several times. A mass balance was conducted on the fluoride based upon a chemical analysis of the original solution, the filtrate and the precipitate washings. Figure 9 shows the percentage recovery of fluoride in the salt solution as a function of the initial neutralization temperature (i.e. the temperature excluding the temperature increase which occurred during the reaction) . It can be seen that fluoride recovery can be increased from about 91% at 22°C to about 97% at 60°C. There is no further improvement beyond 60°C. In each case, the water employed to wash the precipitate was at approximately 20°C. This phenomenon is not simply a result of the increased solubility of potassium fluoride at elevated temperatures that one would normally expect, however. Another experiment was carried out, in which the neutralization reaction was performed at about 22°C and the precipitate was subsequently washed with hot (60°C) water in an attempt to re-dissolve the fluoride salt. In this case, the recovery efficiency was only 88%.

A further experiment was performed whereby the neutralization reaction was performed at 22°C and the. precipitate slurry resulting from the neutralization was subsequently heated and digested at a temperature of 60°C for 1 hour and then filtered and washed with 20°C water.

In this case, the recovery efficiency was still only 90%.

These experiments indicate that in order to maximize the fluoride recovery efficiency, the neutralization reaction must be conducted at an initial temperature in excess of ambient (22°C) and preferably about 60°C or higher. Example 2 An electromembrane cell was assembled as shown in Figure 11. The cation exchange membrane was Nafion

324, the anode was Ebonex an iridium oxide-coated, titanium sub-oxide supplied by Ebonex Technologies of Emeryville, California and the cathode was stainless steel. The cathode chamber initially contained 0.86 N KOH. The anode chamber was initially charged with 3 litres of a potassium fluoride/potassium nitrate solution containing [F] = 16.67 g/L and [K] = 61.2 g/L which corresponding to a fluoride to nitrate ration of 1.26. A DC current of 11 amps was passed through the cell, corresponding to a current density of 100 ASF. After 12 hours, the anode compartment contained 1.63 litres of solution. The potassium concentration was analyzed as [K] = 16.4 g/L. This represents a 85% conversion of- salt to acid at a current efficiency of 81.6%. Example 3 An electromembrane cell was assembled as shown in Figure 13. The cation exchange membranes were Asahi Glass CMV and the bipolar membrane was manufactured by WSI Technologies of St. Louis. The anode chamber, cathode chamber and the base chamber initially contained 1 N KOH. The chamber formed between the bipolar membrane and second cation membrane 19' contained IN HN0₃. The salt chamber 23' was initially charged with 1 litre of a potassium fluoride/potassium nitrate solution containing [F] = 17.72 g/L and [K] = 58.2 g/L which corresponds to a fluoride to nitrate ratio of 1.44. A DC current of 8.25 amps was passed through the cell, corresponding to a current density of 100 ASF. After 4.5 hours_r this compartment contained 0,875 litres of solution. The potassium concentration was analyzed as [K] = 11.7 g/L. This represents a 79.9% conversion of salt to acid at a current efficiency of 86%.

It should of course be noted that the preceding description relates to particular preferred embodiments of the invention only and that many modifications are possible within the broad scope of the invention. For example, while reference has been made to use of this invention in conjunction with pickling processes, it may be also used in conjunction with other processes employing fluoride containing acid solutions and reference to pickling is not to be considered a limitation of the invention.

Claims

I CLAIM;

1. A process for treating a fluoride-containing acid solution which comprises the steps of: providing a said acid solution which contains, in addition to fluoride ions, free acid, a strong acid and at least one dissolved metal salt selected from the group consisting of iron, chromium and titanium salts, and in which the molar ratio of fluoride to anion of said strong acid is less than about 1; treating said acid solution by an acid sorption unit to recover free acid and produce a metal salt solution in which the molar ratio of fluoride to anion of said strong acid is greater than about 1; neutralizing said metal salt solution with a base solution, the fluoride salt of which is appreciably soluble, to produce a precipitate of the metal of said at least one metal salt, and a salt solution of said base, containing fluoride ions; removing said precipitate; treating the salt solution produced in said neutralization step by a cation exchanger to cause cations from said base solution to exchange for hydrogen ions, and produce a solution containing hydrofluoric acid; and, recovering said solution containing hydrofluoric acid.

2. A process as claimed in claim 1, wherein said acid sorption unit includes an anion exchange material with quaternary amine functionality, and wherein said free acid is recovered from said anion exchange material by contacting with water.

3. A process as claimed in claim 1, wherein said base solution is potassium hydroxide.

4. A process as claimed in claim 1, wherein said cation exchanger includes a particulate cation exchange resin capable of being regenerated with a regenerant acid solution, which yields a further salt solution, and wherein the process further comprises the steps of: treating said further salt solution in an electrolytic cell to produce a further acid solution and a base solution; utilizing said further acid solution in said regeneration of the cation exchanger; and, utilizing said base solution in said neutralization step.

5. A process as claimed in claim 1, comprising the further step of removing water from at least one of the solutions being treated, by means of a concentrator.

6. A process as claimed in claim 5, wherein said concentrator is a reverse osmosis unit and is used to remove water from said salt solution following said neutralization step.

7. A process as claimed in claim 5, wherein said concentrator is an evaporator and is used to remove water from said acid solution prior to treatment by said acid sorption unit.

8. A process as claimed in claim 2, wherein said at least one dissolved metal salt is selected from the group consisting of iron, chromium and titanium salts.

9. A process as claimed in claim 1, wherein said neutralization step is carried out at a temperature greater than about 22°C.

10. A process as claimed in claim 1, wherein said neutralization step is carried out at a temperature of at least about 60°C.

11. A process as claimed in claim 1, wherein the cation exchanger comprises an electrolytic cell having an anode and a cathode and a first cation exchange membrane which defines two compartments within the cell, one containing the anode and the other containing the cathode, and wherein the process comprises the further steps of feeding said salt solution to the compartment on the anode side of said membrane, collecting said hydrofluoric acid from the same compartment, collecting base from the compartment on the cathode side of said membrane, and utilizing the base in said neutralization step.

12. A process as claimed in claim 11, wherein said electrolytic cell employs a bipolar membrane adjacent to said cation membrane.

13. A process as claimed in claim 11, wherein a second cation change membrane is employed between said first cation membrane and said anode.

14. A process as claimed in claim 12, wherein a second cation exchange membrane is employed between said first cation membrane and said bipolar membrane.

15. A process as claimed in claim 4, wherein the regenerant acid is nitric acid at a concentration of less than about 2 molar.

16. A process as claimed in claim 1, wherein the nitrate concentration in said salt solution is less than about 1 molar.

17. A process for regenerating a fluoride-containing acid pickling solution contained in a pickle tank, comprising the steps of: providing a said acid solution which contains, in addition to fluoride ions, free acid, a strong acid and at least one dissolved metal salt selected from the group consisting of iron, chromium and titanium salts, and in which the molar ratio of fluoride to anion of said strong acid is less than about 1; withdrawing acid solution from said pickle tank; treating acid solution withdrawn from said pickle tank by an acid sorption unit to recover free acid and produce a metal salt solution in which the molar ratio of fluoride to anion of said strong acid is greater than about 1; recycling said recovered free acid to the pickle tank; neutralizing said metal salt solution with a base solution, the fluoride salt of which is appreciably soluble, to produce a precipitate of the metal of said at least one metal salt, and a salt solution of said base, containing fluoride ions; removing said precipitate; treating the salt solution produced in said neutralizing step by a cation exchanger to cause cations from said base solution to exchange for hydrogen ions, and produce a solution containing hydrofluoric acid; delivering said solution containing hydrofluoric acid to said pickle tank.

18. An apparatus for treating a fluoride-containing acid solution which also contains, in addition to fluoride ions, free acid, a strong acid and at least one dissolved metal salt selected from the group consisting of iron, chromium and titanium salts, in which the molar ratio of fluoride to anion of said strong acid is less than about 1, the apparatus comprising: an acid sorption unit for treating said acid solution to recover free acid and produce a metal salt solution in which the molar ratio of fluoride to anion of said strong acid is greater than about 1; means for neutralizing said metal salt solution with a base solution, the fluoride salt of which is appreciably soluble, to produce a precipitate of the metal of said at least one metal salt, and a salt solution of said base, containing fluoride ions; means for removing said precipitate; and, a cation exchanger for treating said salt solution produced by said neutralization means to cause cations from said base solution to exchange for hydrogen ions, and produce a solution containing hydrofluoric acid.

19. An apparatus as claimed in claim 18, wherein said acid sorption unit includes an anion exchange material with quaternary amine functionality, and wherein said acid is recoverable from the anion exchange material by contacting with water.

20. An apparatus as claimed in claim 18, wherein said cation exchanger includes a particulate cation exchange resin capable of being regenerated with a regenerant acid solution, and wherein the apparatus further comprises means for delivering a said regenerant acid solution to said exchanger, and for withdrawing a further salt solution from said exchanger, and wherein the apparatus further comprises an electrolytic cell for treating said further salt solution to produce a further acid solution and a base solution; means for delivering said further acid solution to said cation exchanger for said regeneration; and, means for delivering said base solution to said neutralization means.

21. An apparatus as claimed in claim 18, further comprising a concentrator for removing water from at least one of said solutions being treated.

22. An apparatus as claimed in claim 21, wherein said concentrator is a reverse osmosis unit and is used to remove water from said salt solution following said neutralization step.

23. An apparatus as claimed in claim 21, wherein said concentrator is an evaporator and is used to remove water from said acid solution prior to treatment by said acid sorption unit.

24. An apparatus as claimed in claim 18, wherein the cation exchanger comprises an electrolytic cell having an anode and a cathode and a first cation exchange membrane which defines two compartments within the cell, one containing the anode and the other containing the cathode, and wherein the process comprises the further steps of feeding said salt solution to the compartment on the anode side of said membrane, collecting said hydrofluoric acid from the same compartment, collecting base from the compartment on the cathode side of said membrane, and utilizing the base in said neutralization step.

25. An apparatus as claimed in claim 24, wherein said electrolytic cell employs a bipolar membrane adjacent to said cation membrane.

26. An apparatus as claimed in claim 24, wherein a second cation exchange membrane is employed between said first cation membrane and said anode.

27. An apparatus as claimed in claim 25, wherein a second cation exchange membrane is employed between said first cation membrane and said bipolar membrane.