CA1076276A - Process for reducing the mercury content of industrial waste waters - Google Patents

Abstract

PROCESS FOR REDUCING THE MERCURY CONTENT OF INDUSTRIAL WASTE
WATERS

Abstract of the Disclosure:
The invention deals with the reduction of the mercury content of industrial waste waters, especially those from a chlorine-alkali electrolysis using the amalgam process.
According to the invention all mercury present in the waste water is converted to the form of Hg++ ions by adding chlorine. Subsequently Fe++ ions are added. The pH of the solution is raised until a redox potential of from - 0.1 to - 0.8 volts, relative to the normal hydrogen electrode, is obtained. The mercury precipitated and the iron oxide hydrate formed may be filtered from the waste water. Residual mercury contents in the filtrate of about 50 mg/m3 may be obtained.

Description

~7~7~
Industrial wasta waters often contain mercury in metallic and/or ionic form which, because of its toxlcity, ls very dangerous for rivers, creeks, lakes etc., because it may be introduced into the human body via fish and other organisms serving as food for man. Therefore, the m~rcury emission of industrial plants must be kept as low as possible. Among the plants emitting mercury along with their waste waters are chlorin~-alkali ~lectrolyses operating according to the amalgam process.
Numerous processes for reducing the mercury content of industrial waste waters are described in the literature. There i5 for example a process according to which mercury ions are reduced to metallic mercury by means of sodium-boron hydride (US. Patent No. 3,764,528). However, this process is not easy to operate, because the reducing agent decornposes at a pH
below 9.5 with development of hydrogen, and the mercury metal which forms in the reaction precipitates in a very finely distributed form and coagulates with difficulty only. A
further serious disadvantage resides in the fact that hydrogen is liberatsd which carries mercury vapors to an extent corre-sponding to the vapor pressure of mercury at the temperature of the process and thus causes secondary pollution o~ the air, unless it is liberated from these vapors by a treatment with d~lute nitric acid. This posterior removal of mercury from the waste gases is an additional, inconvenient process step ~hich, due to the cost of the reducing agent, further di~inishes the pro~itability of the process.
Another process uses sodium sulfide as precipitating 29 agent (Japanese Patent No- 66 7012). The disadvantage of this ~ ' :
. : . :: ..... : .- - ..... ~ . : .

;`: . ' - " ' ' ' -. ' ' . , ' :, :: '., ' ., . '~ ' ' ':. .: : '; .

7~ o~ z~

process resides in the fact that -the mercury sulfide is formed very slowly and that this.formation depends on the pH. Further-more, formation of a solubl~ complex anion of ~ gS2~ 2 compo-sition may occur in the presence of an excess of precipitating agent, thus adversely affecting the efficiency of the process.
Moreover, soluble chlorides which are always present in the waste waters of a chlorine-alkali electrolysis have a nega-ti~e effect on the precipitation of mercury sulfide, and the precipitated mercury sulfide is dispersed i~ the water to be purified in such a finely distributed form that its elimination requires an additional flocculant. And finally, excess floccu-lant has to be car0~ull~ removed from the water in order to pre~rent secondary pollutionO
Alternatively, thio-urea and salt of hydroxylamine have been proposed for precipitating mercury and/or mercury sal.ts from waste waters (Gerrnan Offenlegullgsschrift No. 2,437,779).
Besides relatively long reaction times, flocculants are re-quired also in this case, and th~ residual mercury content of the treated waste water is generallly not belaw 100 ppb.
Another process is known according to which mercury salts contained in waste waters are reduced by means of hydrazine (German Offenlegungsschrift No. 1,958,169), Starting from pure HgCl2 solutions, residual mercury contents of less than 100 ppb may ~e attained in some cases, but only with the addition f flocculants ~for example CaCl2) or of special actiYe oharcoal as additional adsorbe~t, ~hen this process is applicd in the industrial practice, ~or example for the work-up of .
waste waters from a chlorine-alkali electrolysis operating ~:
:~ 29 acccrding to tAe amalgam process, filter combinations with - 3 ~
~ ''` . ' . : . . , :- . . .
- , . . - . . . . ::

-. . . : :

.

HoE 76/~' 806 6if~

sand and active charcoal packings consisting partially of ex-pensive, specially treated coal are required besides the sedimentation tubes in order to attain acceptable results of residual mercury ~less than 100 ppb).
Similar processes using tin(II) ions, hypophosphorous acid, formal~ehyde, metallic iron, zinc, sodium, tin, copper etc.0 as reducing a~ent are mentioned in the literature as well.
However, there are drawbacks such as long reaction times, in-complete reaction, heating required, inacti~ation of the ~etal surfaces by amalgam formation, secondary-pollution of the treated waste waters by residues of agents etcØ
Recently, the technology of ~ercury removal from ~aste waters has switched over to a large extent to processes using ion exchangers (for example Swiss Patent No~ 330,863). In 1~ general, it can be stated that this method is dif`ficult and expensive, because the exchanger compositions are destroyed in most eases by strong oxidants such as chlorine or hypochlorite ions, and furthermore because the problem of exchanger regene-ration has not been solved as yet in a satisfactory manner, so that the resins have to be discarded after several runs already.
Moreover, some exchanger eompositions lose their activity very rapidly when the sodium chloride content of the water to be purifled exoeeds about 10 g /liter, which is often the case in waste waters o~ chlorine-alkali electrolyses. Furthermore, only special types of exchangers are appropriate, bscause in a sodium chloride containing solution the mercury is partially present a~ ~HgC147 2 anion and the active groups of the ex-ohangers are o~ten blocked by foreign ions ? for example S03 or -29 S042 . The residual mercur~ content obtainabLe in these ~ 4~

.: , ,, , . . ~ ... . . . . , .... .... . .. . - . .

~3~

processes is often from 0.1 to 0.~ ppm still (starting values from 1 to 5 ppm only) despite series-connection of 2 ex-changer columns, unless the waste water is a~ter-tr~ated in subsequent absorption towers charged with special acti~e charcoal which cannot be regenerated.
It was therefore an object of the present invention to provide a process for reducing the mercury content of in-dustrial waste waters, especially those from chlorine-alkali electrolyses operating according to the amalgam process, which can be carried out in a simple manner, which is economically interesting and satisfactory in its result. A further object was to see to it ~hat the process is of technological robust-ness and adapted to the use of those chemical agents which are generally present already in chlorine-alkali electrolyses, that is, sodium chloride and sodium hydroxide, which are con-tained in such waste ~aters beside6 traces of mercury.
In accordance with this invention, there is provided a process for reducing the tnercury content of i~dustrial waste waters, especially those from chlorine-alkali electrolyses operating according to the amalgam process, which comprisas converting the total amount of mercury contained in such waters to the ionic, bivalent oxidation state by adjusting a content of from 2 to 50 mg of aOtivQ chlorine per liter of . waste water in a mineral acid medium b~ addition of chlorine, chlorinated water or sodium hypochlorite, subsequently adding iron(II) ions until a content of from 0.1 to 1.5 g of such ions per liter of waste water is attained, and then adjusting a rédox potential of fro~ -0.1 to -0.8 volt, relative to the 29 normal hydrogen electrode, by addition of chemical agents : ~ 5 ~
.

., . ~,.. ~ .,.. ,. , , ~: , . . . . . .

. . :.
-. : . .
, H ~
7~

increasing the pH, thus reducing the mercury and precipitatinglt together with the iron oxide hydrates ~ormed, and finally ~liminating the precipitated products.
The process of the invention surprisingly allows the ob-tention of a satisfactory residual mercury content, even at a more than 14 000 - fold excess of sodium chloride (relative to the weight of mercury contained in a m3 ~ waste water), that is, under extremely unfav~rable conditions. This was not to be expected because the mercury ion, under the conditions as described, is bound in complex form as stable tetrachloro-mercurate(II) an ion. ~urthermore, it was not to be expected that, for the obtention of a satlsfactory residual mercury content within the scope of this in~ention, it is nearly un-important how much mercury is cointained in the waste water to be treated. Thus, it has been observed that even the waste water of a chlorine-alkali electrolysis which contained 162 g of mercury per m3 (adj~sted by addition of HgCl2 solution to this waste water), could be purified to a residual mercury content of only 0.05 g/m3, which corresponds to a demercura-tion degree ef 99.97 ~. The e~ficiency of the process of theinvention i5 therefore ensured also in cases where the waste water to be treated contains an extremely large amount of ~ mercury or sodium chloride.
Further ad~antages of the operation mode of the invention 25 are the followlng: first, except a superficial clarification of the waste waters, no preliminary or intermediate puri-fication steps are required; second, and this is especially important, the oxidant to be used in excess in the first 29 process step has not to be remo~ed before the reduction of the . . : .

- HOE 76~ 806 ~76Z~

mercury, but is inactivated only with the use of a small ex-cess amount of cheap reducing agent or alkali lye which latter one is anyhow at disposal in chlorine-alkali electrolyses.
Thc use of iron(II) salts as reducing agents has the ad-vantage that the almost insoluble mercury (I~ salt formed is precipitated in a practically quantitative manner simultan-eously with the mixture of iron(II) and iron(III) oxide hydrates precipitating from the solution, so that an addition-al flocculant is generally no lon~er required.
The demercuration according to this invention o~ waste wa$ers stemming for example from a chlorine alkali electro-lysis using a streaming mercury cathode and containing mercury in metallic as well as ionic form (the amount of metal being generally from 30 to 80 % of the total mercury content) is carried out in detail as follows: first, all mercury is con-verted to the ionic form, that is, the bivalent oxidation state. This is carried out by acidifying the water flowing into the apparatus by means of hydrochloric or sulfuric acid which are generally at disposal to obtain a pH of 4.5 and by adding chlorine in gaseous form, chlorinated water or sodium hypochlorite to this acidified solution until a content of active chlorine of about 2 to 50, preferably 5 to 30 mg/l i9 attained. The pH is not critical; a p~ of ~rom 1 to 3 being however pre~erred, since higher values require the addition of larger chlorine amounts, and lower values necessitate a rela-ti~ely large amount of acid. Because most of the waste waters of chlorine-alkali electrolyses contain also hypochlorite ions which, on acidifioation, form chlorine or hypochlorous acid, 29 the redo~ potential necessary for the oxidation of the - 7 ~
.

,::,, . .. - , . . :
,, , , . . ; . . ..

.,., : : , ' ', . . ' metallic mercury is sometimes established in ths wat~r to be demercurated by the acidification already, ~o that in these cases ~he addition of an oxidant may be omitted.
The genuine demercuration is carried out by adding an excess of bivalent iron ions to the mercury ions now being present in bi~alent form, and by subsequently establishing the low redox potential in the solution required for the reduction of mercury ions even in small amounts by addition of agents increasing the pH, for example sodium hydroxide or calcium hydroxide, preferably alkali lyes. The anion of the Fe(II) salt is not critical. An amvunt of iron(II) salt solution, for example iron(II) sulfate solution or iron(II)chloride solution, is used which ensures a concentration of from 0.1 to 1.5, pre~erably 0.3 to 1.0 g of iron(II) ions per liter of waste water, and the redox potential in the solution contain-ing the mercury and iron ions is adj~sted to -0.1 to -0.8, pre~erably o.l~ to -0~7, and especially -0.5 to -o.6 9 ~olts, relative to the normal hydrogen electrode. Higher concentra~
tion of iron(II) ions may be used, but they do not bring about any further advantages. The Fe~ content may be deter-mined by titrimetric methods. In order to prevent undesirable increase of potential by atmospheric oxygen and`possible re-oxidation of thc mercury, it may be advantageous to carry out th~ alkalization and the subsequent isolation of the preci-j 25 pitated products under an atmosphere of inert gas, for example nitrogen.
The precipitated mixture of scarcely soluble mercury(I) salt; iron(II) and iron(III) oxide hydrates is ~eparated accor-29 ding to known methods, ~or example by ~iltration, optionally .- .

', .. . .

with the aid of ~sual flocculants remaining in th~ sludge, for example polyacrylamides, which are used in amounts of from 1 to 5 g/m3 of water.
Depending on the mercury content of the sludge which is determined by the specific conditions of the manufacturing plant emitting the waste waters, it may be economic to re cover the mercury, for example by means of distillation pro-cesses or known dissolution processes using nypochlorite ion containing solutions. In the case where work-up is not profit-able the sludge is forwarded to a safe dump~
In the filtrate there are only traces of mercury, andexcept the anions stemming from the iron(II) salt used and OH ions or alkali ions, no forei~l substances are introduced into th~ water treated, because the precipitation of iron in the pH r~nge required for the proces~ of the invention is nearly quantitative. The filtrate flowing off has an average iron content of 0.15 mg~l.
The quantitative relation between the redox potential Eh (measured in ~olts) adjusted in the solution and the re-sidual mercury content of the treated waste water C~g (mea-s~red in ppb) is shown in the diagram of ~IGURE 1 of the accomp~nying drawings.
The following examples illustrate operation and effect of the process of the invention. All tests were carried out ~5 at normal temperature (about 20 C). For the analytical mercury examination, an atomic absorption spectrophotometer (Coleman MAS 50) was used~which had beon specially deQigned for this purpose.

, _ 9 _ .. . ...... .

. :: : , : , ~ , ~ . : . . .. . .

~10~ 76/~ 806 E X A M P L E 1:
In a la~oratory apparatus according to FIGURE 2 of the accompanying drawings, a continuously flowing waste water current of 2.2 liters/h stemming from ~ chlorine-alkali elec trolysis operating according to the amalgam process, the NaCl content of which current had been deliberately increased by addition of brine, was fed to the oxidation vessel (2) via the duct (1). The (total) mercury content of the solution fed in was 8.0 mg/liter, the NaCl content 118.0 g/liter and the pH 12Ø Via the duct (3), such an amount of 31 % hydrochloric acid and via the duct (4), such an amount of chlorinated water containing about 1 g/liter of active chlorine were ~ed to the duct (1) that the pH of the solution (20) in the oxidation vsssel (2) was maintained constant at 1.9, and the redox potential was a constant 1.25 volts, relative to the normal hydrogen electrode. The feed was automatic and controlled by the control valve (5) for the chlorinated water, and the con-trol valve (6) for the hydrochloric acid. Control valve (5~
was operated by the measuring device (7) for the redox poten-tial via the pneumatic line (8); and control valve (6) b~ themeasuring device (9) for the pH via the pnewnatic line (10).
Thus, a content of 5~4 mg/liter of active chlorine was ad-~usted. After a mean residence time of about 1 hour, the waste water (20) left the oxidation vessel (2) provided with an agitator (22) via the duct (l1), to which duct, leading to the precipitation vessel (i4), such an amount of an aqueous iron(II) sulfate solutlon containing 25 g/l of FeS04 .7HzO
was fed that the waste water contained 346 mg/l of Fe2~ ions.
29 ~ia the duct (15), such an amount of 18 % sodium hydroxide ~ 10 ~

' "' ' ' ' ' ' ': ' ' ' ' '. ' '' ' ". ' " "'' ': . ' ' .. : ' : ' ' ' 3L~7~

solution was added to the waste water (21) in the precipita-tion vessel (14) tha$ a constant redox potential of -o.68 volt, relati~e to the normal hydrogen electrode, was established therein. This redox potential in (14) was measured by means of a devioe (not shown) which operated the control valve ( 1~) via a corresponding pneumatic line (not shown). After a mean reside~ce ti.~e of about 1 hour, the suspension (21) obtained left the precipitation vessel ~14) provided with an agitator (12) via the duct (17) and was thus forwarded to the closed suction filter (18)~ The gas zone o~ precipitation vessel (14) and of suction ~ilter ~18) Was filled with nitrogen (not shown). The filtrate left the suction filter (18) via the duct ~19). l`he analysis of the filtrate resulted a mean re-sidual mercury concentration of o,oll mg Hg/liter. The fi.lter residue contained 0.8 ~ Oe Hg, relative to the dry substance.
E X A M P L E 2:
.
In an apparatus according to FIGURE 2, where the oxi-dation vessel had a capacity of 0.20 m3 and the precipitation vessel had a capaoity of 0.11 m3, a continuous waste water current of 0. 11 m3/h coming from a chlorine-alkali electro-lysis operating according to the amalgam process and having an average composition o~ 90.0 g/m3 of Hg (total amount), 29,5 k~r/m3 of NaCl and 3.9 kg/m3 of NaOH was treated in the manner as described. The iron(II) sul~ate solution used had a concentration of about 200 g of ~eS04 per liter of solution anhydrou s ) .
According to the manner described in Example 1, the dates adjusted in the apparatus were the ~ollowing.

;

, ~
. , H ~
~7~27~

Overflow P~-C~pit~

_. active chlorine[~ F82~ vessel .
Cmg/lite~ h Eg/lite~ Eh Lv~
. . ~

2.8 . 10.1 +1.30 503 -o.68 _~_ __ __ The residual merc~ry conte~t in the filtrate was on the average 0.05 mg/liter, while the sludge filtered o~f con-ta.ined 6.3 ~ of Hg9 relative to the dry substance.
E X A M P L E ~-In a pilot-plant apparatus accordi.ng to ~IGURE 2, where the oxidation vessel had a volume of 0.50 m3 and the preci-pitation ~essel had a volume of 0.75 m3, a continuous waste water current o~ 1 m3/h coming from a sodium chl.oride electro-lysis operating according to the amalgam process and having the following composition: ~Ig (total amount) 31.0 g/m3, NaCl 45.1 kg/m3, NaOH 1.Z kg/m3 was treated. The iron(II) sulfate solution had the same concentration as in Example 2.
The following conditions were established according to the man~er as described in the oxidation and precipitation vessel, respectively:

_ __ ~--Oxidation vessel Overflow precipitation pH active chlorine ~h [~ Fe~ vessel ...
Eg/lite~ Cmg/lite~ Eh [~]
_ . _ _ ~ ____

3.3 14.3 ~1.30 726 -o.68 ~ ~ ._ ~

The filtrate flowing of~ had a mean content of 0.04 mg/liter of Hg, and in the sludge, 1,0 % of Hg, relativeto the dry : - ' - - ' ~ ' " ' ' ' ' ' ' ~, . .1,, .. ' .. , , . "
,' ' -,., . '; .' ' ` ' '' .. .. ' ' '. . ' ' '. ' ' .' ~ ' ' ' .

.','. ' ' " " '.' '",', ` ' ' ", ' ,' .,' .. ' . :" ,, .', '' ' ': ' ',, ', ' ' . " '." "'' . . ' "', '.,"";.'''" .' ' ' " , " ' " ','," ' ' ., ' ' " . ' " " ' " ' " ' ' ' ' ": ~ . ," ' . ', " . '' '-' ' - . ' , ' ., '' '.' . ' ' "' . , ,' .. " ." ,"' ', ",, . ' - "' '. , . :' " "'.' ' '': '.
', .'' . ' '.. .' .-' " ' ''' '. ' ' '. ' .': ' " ' ' ' . ' .. " . ' : . .,,, .', ' . .:'". ' ' ' '', ", ., '', '' . ' ' ' '. , :' :':
. ' ""' '.: ' ' ': " , ' '".' , ' ' ' '.' ' ' " . ' ' ,'': ' ' ' , ' ' . ""' ' ' ' ' " , ' ' ' ', " " ' ' ' ' ' ' ' '' ' '''' , . ''.' ' " '' ' ' ', , substance, was detected~

In the apparatus according to ~IGURE 2, a continuous waste water current of 0.11 m3/h containing on the average 11.0 g/m3 of Hg (total amount), 18.~i kg/m3 of NaCl and 0.18 kg/m3 of NaOH was treated. In this case, the reaction con-dition~s were deliberately chosen in such a manner that the redox potential in the precipitation vessel exceeded the pre-ferred range. The corresponding test conditions are listed in the following Table~

Oxidation vessel Ovër~low Precipitation , ~ _ Fe vessel pH active chlorine ~h Cv]
Lmg/l i t e ~ rmg/l i t e ~3 Eh [V~
. . . _ ~ __ ____ .~
2.0 ~ ~ ~I JU 145 -0.33 _ The filtrate contained sti~ 0.13 mg of Hg/liter on the average.

.: , . : ,: :
.~ ~ ..... , , . . - .. . .

Claims (3)

What is claimed is:

1) A process for reducing the mercury content of industrial waste waters, which comprises converting the total amount of mercury contained in such waters to the ionic, bivalent oxidation state by adjusting a content of from 2 to 50 mg of active chlorine per liter of waste water in a mineral acid medium by addition of chlorine, chlorinated water or sodium hypochlorite, subsequently adding iron(II) ions until a content of from 0.1 to 1.5 g of such ions per liter of waste water is attained, and then adjusting a redox potential of from -0.1 to -0.8 volt, relative to the normal hydrogen electrode, by addition of chemical agents increa-sing the pH, thus reducing the mercury and precipitating it together with the iron oxide hydrates formed, and finally eliminating the precipitated products.

2) The process as claimed in Claim 1, which comprises ad-justing the redox potential and subsequently separating the precipitated products in an oxygen-free atmosphere.

3) The process as claimed in Claim 1, wherein the waste waters stem from a chlorine-alkali electrolysis operating according to the amalgam process.