In this work, a chloride/hypochlorite leaching procedure was performed for Zn works residues. Sodium chloride and Ca hypochlorite were used as leaching and oxidising agents, severally. Fractional factorial method has been used to prove the chief effects and the interactions among the factors was investigated. The statistical package named Design-Expert 7 ( DX7 ) has been used to plan experiments and subsequent analysis. Parameters and their degrees were reaction clip ( t = 16 and 120 min ) , reaction temperature ( T = 30 and 70 -C ) , solid-to-liquid ratio ( S/L = 1/6 and 1/38 ) , pH ( pH = 0.5 and 2 ) , and Ca ( OCl ) 2 concentration ( C = 0.6 and 3 g/L ) . Statistical analysis, ANOVA, was besides employed to find the relationship between experimental conditions and output degrees. The consequences showed that the reaction temperature and pH were important parametric quantities for both lead and silver extractions but the solid-to-liquid ratio had important consequence merely on lead extraction. Increasing pH reduced leaching efficiency of lead and Ag. However, increasing reaction temperature promoted the extraction of lead and Ag. The ultimate optimal conditions from this survey were t1:16 min, T2:70 -C, ( S/L ) 2:1/38, pH1:0.5, and C1:0.6 g/L. Under these conditions extraction of lead and Ag were 93.60 % and 49.21 % , severally.
Keywords: Zinc works residue ; Chloride/hypochlorite leaching ; Lead ; Silver ; Fractional factorial method.
*Tel. : +98-412-3443802. Facsimile: +98-412-3443443. Electronic mail: moghaddam @ sut.ac.ir, hastyir @ yahoo.com.
Turning environmental concerns and economic necessity are taking the excavation and metallurgical industries around the universe to put in suited techniques for treating the wastes generated and maximising the recycling of resources. More than 80 % of the universe broad primary Zn is produced via a combined roast-leach-purification-electrowinning procedure. During the hydrometallurgical processing of Zn from the roaster, calcine lead and silver study to the leach residue [ 1, 2 ] which can be utilized for the recovery of these two metals. Besides, big measures of Fe waste are produced in the signifier of three chief sorts of residues: ( a ) gothite ( FeOOH ) , ( B ) jarosite ( XFe3 ( SO4 ) 2 ( OH6 ) or ( degree Celsius ) haematite [ 3-6 ] . The toxicity of the waste is chiefly due to the presence of different metals such as lead, Cd, arsenic, Cr, etc. The released residue during the procedure could be recycled for farther processing [ 7-10 ] .
Different processing techniques are practiced to handle the leach residue to retrieve Zn every bit good as lead in the presence of high Fe content [ 2 ] . However, chloride leaching is the most accepted and widely used recovery method [ 11 ] . Chloride leaching procedures have been employed utilizing either NaCl [ 1, 2, 11-15 ] , or MgCl2 and CaCl2 [ 16 ] , or FeCl3 [ 12, 17 ] along with HCl. Recently, Behnajady and Moghaddam [ 18 ] have evaluated the chloride leaching of ZPR in NaCl-H2SO4 media. Based on their plants, it was concluded that utilizing H2SO4 is more cost-efficient than HCl.
Due to the turning environmental concern of the usage of nitrile for gold and Ag processing, involvement on the usage of noncyanide lixiviants, particularly with halides has been renewed late [ 19, 20 ] . Leachants based on halides ( chloride, bromide and iodide ) have been used in the yesteryear to fade out gold and Ag into solution [ 21, 22 ] . Different combinations of oxidizers ( such as hypochlorite, H peroxide, Br and I ) can be used in concurrence with the complexants ( chloride, bromide and iodide ) to supply leaching conditions. One of the options is a chloride solution with an appropriate oxidizing agent. Widely used oxidising agent for gold and silver-chloride system is hypochlorite. Several workers have investigated chloride/hypochlorite solutions to leach gold and Ag from several types of gold ores and dressed ores [ 20, 23 ] , but in our survey chloride ( complexant ) -hypochlorite ( oxidant ) mixtures were used to leach lead and Ag from a ZPR.
ZPRs are SO4-bearing residues and the presence of inordinate degrees of sulfate adversely affects the lead and silver extraction. To extinguish sulfate from the residues and to retrieve water-soluble Zn such as ZnSO4, a H2O rinsing measure has been found to be really effectual [ 1, 2, 11, 24 ] .
The purpose of the present survey was the finding of the effects of the chief factors involved in the chloride/hypochlorite leaching procedure for ZPRs. First the residue was leached with H2O. Then, washed residue was leached in NaCl-H2SO4-Ca ( OCl ) 2 media to pull out lead and Ag. The planning of the experimental work was arranged utilizing fractional factorial method in order to set up, in a really efficient mode, the chief effects and the interactions among the factors were investigated. Consequently, the consequence of operating factors including reaction clip ( T ) , reaction temperature ( T ) , solid-to-liquid ratio ( S/L ) , pH ( pH ) , and Ca ( OCl ) 2 concentration ( C ) on the extraction efficiency of lead and Ag were investigated.
2. Experimental process
2.1. Word picture of the sample
ZPR was obtained from the Bafgh Zn smelting works located at Yazd, Iran. Initially, the ZPR sample was dried, so it was crushed utilizing a ball factory and sieved utilizing a 100 mesh ( 149 µm ) ASTM standard screen. Noted that oppressing and screening were repeated until all atoms became finer than 149 µm.
Mineralogical construction of the residue was identified by X-ray diffraction ( Bruker advanced-D8 ) analysis. Afterwards, the residue was chemically analyzed utilizing X-ray fluorescence ( XRF ) ( Phillips, theoretical account PW2404 ) and atomic soaking up spectrometer ( AAS ) ( Perkin-Elmer, AA300 ) .
2.2. Water leaching
Prior to utilize, the residue was leached with tap H2O at 65 & A ; deg ; C for 60 min at a mush denseness of 250 g/L with mechanical scaremonger at 600 revolutions per minute to better the recovery in the subsequent phase of chloride/hypochlorite leaching. After filtration, the solution was analyzed by atomic soaking up spectrometer for Zn and lead. The residue was dried, crushed by ball factory, sieved, and physically characterized to find the atom size distribution. Then the residue was chemically analyzed for Zn, Fe, lead, and silver utilizing an atomic soaking up spectrometer, and its mineralogical construction was identified by X-ray diffraction analysis.
2.3. Leaching with chloride/hypochlorite solutions
After H2O leaching, the ZPR was subjected to chloride/hypochlorite leaching to pull out lead, and silver. Commercial class salt ( NaCl ) , as a inexpensive agent, was used to fix chloride solution. Industrial grade H2SO4 was used for pH accommodation and proficient class Ca ( OCl ) 2 was the oxidizing agent.
The leaching experiments were performed in a 2L beaker in a thermostatically controlled H2O bath equipped with mechanical fomenter. After pouring 1 L of clear NaCl solution ( 300 g/L ) to the beaker and puting the temperature at the coveted value, a known measure of residue was added and acidified by H2SO4, and eventually Ca ( OCl ) 2 was added while stirring the content of the beaker at 700 revolutions per minute. In all experiments the solution conditions ( pH and temperature ) were controlled utilizing a pH accountant ( Metrohm-827 pH lab ) . After each test, the leach slurry was filtered instantly and the leach solution was analyzed for Pb, and Ag by atomic soaking up spectrometer ( AAS ) .
2.4. Experimental design
In this survey, a two-level fractional factorial design was employed. The 2k factorial design is peculiarly utile in the early phases of experimental work when many factors are likely to be investigated. In fact it is possible to set up an extraneous experimental plane in which both the chief consequence and the interaction among the factors investigated can be evaluated independently. As the figure of factors increase in 2k factorial design, the Numberss of needed tallies quickly outgrow. Fractional factorial designs can frequently place which factors are important by running merely a fraction ( subset ) of a full factorial experiment [ 25, 26 ] .
The most of import phase in the design of an experiment lays in the choice of control factors, hence as many factors as possible should be included and no important variables must be identified at the earliest chance. On the footing of our old experience in related plants and those experimental conditions reported by other research workers for the leaching of similar residues and preliminary trials performed: reaction clip ( T ) , reaction temperature ( T ) , solid-to-liquid ratio ( S/L ) , pH ( pH ) and Ca ( OCl ) 2 concentration ( C ) were chosen as the five factors to be investigated, and two degrees are the exclusory set for each of our five factors. Five selected control factors in two degrees applied in this survey, have been listed in Table 1.
For the fractional factorial design and subsequent analysis, the statistical package named Design-Expert 7 ( DX7 ) was used. Since we were analyzing 5 factors, the 25 full factorial design would necessitate 32 experimental tallies for all combinations of the degrees of the factors were investigated. Sixteen experiments of the full factorial design have been considered alternatively of 32. Table 2 represents the selected experiments for this survey. This design is a 1/2nd fraction, so every consequence will be aliased with one other effects, most of which are ignored by default to avoid unneeded screen jumble. The end product indicates that each two-factor interaction will be confounded with one three-factor interaction, which are by and large non of import. The assumed names construction for design indicates as below:
[ A ] = A [ AB ] = AB + CDE [ BD ] = BD + ACE
[ B ] = B [ AC ] = AC + BDE [ BE ] = BE + ACD
[ C ] = C [ AD ] = AD + BCE [ CD ] = CD + ABE
[ D ] = D [ AE ] = AE + BCD [ CE ] = CE + ABD
[ E ] = E [ BC ] = BC + ADE [ DE ] = DE + ABC
E = ABCD
Replicate trials were non done for the finding of the experimental mistake, for the F values in the ANOVA analysis, because it is possible to find the mistake utilizing interactions that are non important. Self-contained experimental tallies can be realized without replicate trials when a factorial design is applied to experimental informations. In this manner, the high interactions or the interactions with comparatively low values of the effects can be used to gauge the experimental mistake. The procedure of disregarding a factor or interactions once it was deemed insignificant was called pooling. In this instance, holding used a fractional design, all the possible high interactions are saturated by chief and two-interaction effects. However, by consideration of the comparatively low values of some two-interactions, it is possible to utilize them to do an rating of experimental mistake.
3. Consequences and treatment
3.1. Word picture of the sample
Fig. 1a illustrates the atom size distribution of the residue. As it can be seen from the atom size distribution curve, d80 was calculated as 88 ?m. The chemical composing ( XRF analysis ) of the residue has been given in Table 3. The XRD analysis showed that the residue compounds were lead sulfate ( PbSO4 ) , calcium sulfate dihydrate ( CaSO4.2H2O ) , zinc sulfate heptahydrate ( ZnSO4.7H2O ) , iron oxide ( Fe2O3 ) , jarosite ( KFe3 ( SO4 ) 2 ( OH ) 6 ) , quartz ( SiO2 ) , zinc ferrite ( ZnO.Fe2O3 ) , and Fe silicate ( Fe2O3.SiO2 ) . Harmonizing to the atomic soaking up analysis, the residue contains about 7 % Zn, 8.8 % Fe, 14.2 % lead, and 0.106 % Ag.
3.2. Word picture of the washed sample
The atom size distribution of the washed residue is shown in Fig. 1b. As it can be seen from the atom size distribution curve, d80 was calculated as 105 ?m. When the XRD form of washed residue was compared to that of the initial residue for their components, it was concluded that most of the Zn sulfate heptahydrates ( ZnSO4.7H2O ) were taken into leach spirits during H2O rinsing. Since jarosite procedure was used to precipitate Fe in this peculiar works, Fe hydrated oxide in the signifier of jarosite was detected in XRD form of washed residue.
Harmonizing to the atomic soaking up analysis, after H2O lavation, the residue contains about 3 % Zn, 9.1 % Fe, 14.4 % lead, and 0.11 % Ag. Besides Zn and lead concentrations in the filtrate were in the scope of 7500-8500 and 5-15 mg/L, severally. This is consistent with a part of the Zn in a water-soluble compound ZnSO4.7H2O, and the limited solubility of lead sulfate in H2O.
3.4. Leaching in NaCl-H2SO4-Ca ( OCl ) 2 media
Matching leaching efficiencies obtained under the campaigner conditions based on fractional factorial design have been displayed in Table 4. The collected informations were analyzed by package to measure the consequence of each parametric quantity on the optimisation standards. The maximal sum of lead and silver extraction were defined as optimisation standards. Statistical analysis of discrepancy ( ANOVA ) was performed to look into whether the procedure parametric quantities were statistically important or non. The F-value for each procedure parametric quantity indicates which parametric quantity has a important consequence on the leaching efficiencies and is merely a ratio of the squared divergences to the mean of the squared mistake [ 27 ] . Normally, the larger F-value indicates the greater consequence on the leaching efficiency.
The values of the ANOVA analysis for lead extraction after pooling have been given in Table 5. F- value for this status with 95 % assurance degree is 6.61 [ 27 ] . Therefore, the consequences of F-value from Table 5 show that the chief effects including the pH ( pH ) , reaction temperature ( T ) , and solid-to-liquid ratio ( S/L ) , and the interaction between T-S/L and S/L-C were important for lead extraction responses within the degrees and conditions tested. The F-value of these factors and interactions are greater than the extracted F-value from the tabular array ( 95 % assurance degree ) . This means that the discrepancy of these factors and interactions were important compared to the discrepancy of mistake. Another method to find the important factors is by P-values calculated utilizing DX7 package. P values less than 0.05 indicate that the consequence of theoretical account factors are important within the 95 % assurance degree.
The F and P values of the theoretical account were calculated as 9.95 and 0.0102, severally. Since the theoretical account P value is less than 0.05, the theoretical account is important within the 95 % assurance degree. The theoretical account F-value ( 9.95 ) implies that the theoretical account is important, and there is merely 1.02 % opportunity that a theoretical account F-value this big could happen due to resound.
The mean degree response analysis is done by averaging the lead extraction per centum at each degree of each factor and plotting the values in a graphical signifier. The mean degree responses from the secret plans help in optimising the nonsubjective map under survey. The numerical values of the maximal points in these secret plans correspond to the best values of factors. Fig. 2 shows the consequence of governable factors on mean lead extraction per centum.
Harmonizing to the graphs illustrated in Fig. 2, it can be concluded that the reaction clip ( T ) has small consequence on the lead extraction from residue. Since the extractions of metals can be improved at drawn-out periods of leaching, it was intended to happen such dealingss for this residue excessively. Therefore, the leaching trials were carried out in 16 and 120 min but the extraction of lead additions somewhat with increasing reaction clip. However, the disintegration of lead additions significantly by increasing the reaction temperature ( T ) ( lead chloride solubility increases with increasing temperature ) . The addition of the lead extraction with temperature reflects the endothermal heat of lead leaching. Because of the limited solubility of PbCl2, the addition in solid-to-liquid ratio ( S/L ) resulted in a lessening in lead extraction. In fact, lead chloride solubility even at elevated temperatures is non sufficiently high and lead solubility restraints limit the extraction of lead in most commercial chloride leaching procedures ; hence, the chloride leaching of ZPR is better carried out at low mush denseness.
As can be seen from graphs, the pH ( pH ) is a extremely effectual factor. Since activity of the chloride increases with diminishing pH, leaching of lead additions with worsening pH. An addition in the activity of Cl? assists farther disintegration of lead because the increasing chloride activity favours the formation of soluble lead chloride composites. Besides, the lead extraction lessenings by increasing the Ca ( OCl ) 2 concentration ( C ) . This was likely due to the coprecipitation or surface assimilation of dissolved lead. Because the formation of Ca sulfate in a chloride leaching circuit adversely affects the metal extraction.
Ignoring interaction effects for the minute, notice that Fig. 2 shows an betterment at degree 2 for reaction clip ( T ) , reaction temperature ( T ) , and solid-to-liquid ratio ( S/L ) while degree 2 effects for pH ( pH ) and Ca ( OCl ) 2 concentration ( C ) cause a lessening in lead extraction. Hence the optimal degrees for the factors based on the informations are t2T2 ( S/L ) 2pH1C1. Coincidentally, test figure 8 was tested these conditions and produced the highest consequence for the extraction of lead ( 93.11 % ) .
Fig. 3 shows the interaction effects of S/L-C and T-S/L on mean lead extraction per centum. The intersection lines on the left represents the interaction between S/L and C. The 2nd brace of lines represents the consequence of S/L at fixed degrees of T. Observe that the highest value for the braces of lines corresponds to ( S/L ) 2C1 and T2 ( S/L ) 2. Comparing ( S/L ) 2C1 and T2 ( S/L ) 2 to the optimal status for maximal lead extraction, it can be seen that ( S/L ) 2C1 and T2 ( S/L ) 2 are included. Therefore, the interactions S/L-C and T-S/L have no influence on the optimum and no farther alteration is needed.
The consequences of the ANOVA analysis for Ag extraction after pooling have been given in Table 6. F- value for this status with 95 % assurance degree is 5.99 [ 27 ] . Therefore, the consequences of F-value from Table 6 show that the chief effects including the pH ( pH ) and reaction temperature ( T ) , and the interaction between T-pH, T-S/L, t-pH, and pH-C were important for silver extraction responses within the degrees and conditions tested. The chief effects including the reaction clip ( T ) , solid-to-liquid ratio ( S/L ) , and Ca ( OCl ) 2 concentration ( C ) were non a important term, but to show a hierarchal theoretical account they were included in the theoretical account. Model hierarchy maintains the relationships between the chief and interaction effects.
The F and P values of the theoretical account were calculated as 27.97 and 0.0003, severally. Since the theoretical account P value is less than 0.05, the theoretical account is important within the 95 % assurance interval. The theoretical account F-value ( 27.97 ) implies the theoretical account is important, and there is merely 0.03 % opportunity that a theoretical account F-value this big could happen due to resound.
Fig. 4 shows the consequence of governable factors on mean silver extraction per centum. Harmonizing to the graphs, it can be seen that the reaction clip ( T ) , solid-to-liquid ratio ( S/L ) , and Ca ( OCl ) 2 concentration ( C ) have small effects on the leaching of Ag from residue. On the other manus, the pH ( pH ) and reaction temperature ( T ) are extremely effectual factors. These two factors have same effects on lead and Ag leaching ( compare Figures 2 and 4 ) . On the contrary, the increased leaching clip ( T ) from 16 to 120 min, caused a lessening in silver extraction. This may be due to formation and precipitation of Ca sulfate or surface assimilation of composites on solid surfaces [ 20, 23 ] . To minimise this unwanted surface assimilation of silver-chloro species, the leaching clip must be limited to a few proceedingss merely. Besides, the changed solid-to-liquid ratio ( S/L ) from 1/6 to 1/38, caused a lessening in silver extraction. This may be attributed to the fixed pH in high mush denseness and it can be concluded that Ag solubility restrictions can non be a job in the leaching of Ag from ZPR. So chloride/hypochlorite leaching of Ag from ZPR can be carried out in a higher mush denseness which is more eco-friendly than a lower mush denseness.
For maximising silver extraction in the chloride/hypochlorite leaching of ZPR by disregarding interaction effects for the minute, the undermentioned conditions were chosen: t1T2 ( S/L ) 1pH1C2. The experiment matching to the optimal conditions for upper limit Ag extraction has non been carried out during the planned experimental work in Table 4. Therefore, the experiment matching to these optimal conditions performed and the corresponding Pb and Ag extractions were 80.82 % and 50.56 % , severally. If the experimental program given in Table 4 were to be studied carefully, it can be observed that there are good understandings between this and the experiments 4 ( 50.67 % ) and 7 ( 50.86 % ) consequences for silver extraction. Since differences between these experiments are related to undistinguished factors, understandings among silver extraction consequences are logical. Noted that these experiments are the highest consequences for the extraction of Ag ( ~50 % ) .
Fig. 5 shows the interaction effects of T-pH, T-S/L, t-pH, and pH-C on mean silver extraction per centum. Observe that the highest value for the braces of lines corresponds to T2pH1, T2 ( S/L ) 2, t2pH1, and pH1C2. On the one manus, comparing T2pH1 and pH1C2 to the optimal status for Ag extraction, it can be seen that T2pH1 and pH1C2 are included. So, the interactions T-pH and pH-C have no influence on the optimum and no farther alteration is needed. On the other manus, comparing T2 ( S/L ) 2 and t2pH1 to the optimal status for Ag extraction, it can be seen that T2 ( S/L ) 2 and t2pH1 are non included. If the experimental program given in Table 4 were to be studied carefully it can be seen that the experiments matching to these conditions ( t2pH1 and T2 ( S/L ) 2 ) have been carried out during the experimental work as experiments 4 and 7 in Table 4, severally. As antecedently mentioned differences between these experiments are related to undistinguished factors. Under economical considerations, it is desired that the reaction clip and liquid-to-solid ratio should be kept low. For this ground, optimum degrees for the reaction clip ( T ) and solid-to-liquid ratio ( S/L ) were non changed from 1 to 2.
Therefore, because the silver extraction is more of import than the lead leaching ( Ag is cherished metal ) it is desired that those selected parametric quantity degrees to be near to the maximal extraction of Ag. In the position of above, the following were selected as ultimate optimal conditions: t1: 16 min, T2: 70 -C, ( S/L ) 2: 1/38, pH1: 0.5, and C1: 0.6 g/L. It can be seen that experiments matching to ultimate optimal conditions, which is obtained by uniting the two series of optimal conditions in a logical mode, have non been performed during the experiments. So, a verification experiment was performed to verify the decisions drawn based on statistical design. The verification leaching experiments were carried out twice at the same on the job conditions, and the experimental mean consequences under these conditions were 93.60 % and 49.21 % for extraction of lead and Ag, severally. Therefore, it is possible to increase lead and silver extraction per centum significantly utilizing the proposed statistical technique.
The chloride/hypochlorite leaching of a ZPR was investigated in NaCl-H2SO4-Ca ( OCl ) 2 media. The consequence of procedure parametric quantities including reaction clip ( T ) , reaction temperature ( T ) , solid-to-liquid ratio ( S/L ) , pH ( pH ) , and Ca ( OCl ) 2 concentration ( C ) , each in two degrees, was studied with the fractional factorial method. The per centum of lead and silver extraction were optimized individually. Meanwhile, the optimal conditions for coincident maximizing of these two responses were determined. Based on the experimental consequences and their presented analysis, the undermentioned decisions may be highlighted:
( 1 ) The most important parametric quantities impacting the lead extraction were pH ( pH ) , reaction temperature ( T ) , and solid-to-liquid ratio ( S/L ) , severally. pH ( pH ) and reaction temperature ( T ) were besides the most effectual parametric quantities on Ag extraction.
( 2 ) pH ( pH ) and reaction temperature ( T ) have same effects on lead and Ag leaching. Increasing the mush denseness decreases extraction of lead as a effect of precipitation of PbCl2. However, this consequence is different for Ag and leaching of Ag additions with increasing the mush denseness.
( 3 ) Measure values of optimal conditions for lead extraction are 120 min for reaction clip ( T ) , 70 -C for reaction temperature ( T ) , 1/38 for solid-to-liquid ratio ( S/L ) , 0.5 for pH ( denoted as pH ) and, 0.6 g/L for Ca ( OCl ) 2 concentration ( C ) . Under these conditions, the extraction of lead in NaCl-H2SO4-Ca ( OCl ) 2 media was ~93 % .
( 4 ) Measure values of optimal conditions for Ag extraction are 16 min for reaction clip ( T ) , 70 -C for reaction temperature ( T ) , 1/6 for solid-to-liquid ratio ( S/L ) , 0.5 for pH ( denoted as pH ) and, 3 g/L for Ca ( OCl ) 2 concentration ( C ) . Under these conditions, the leaching of Ag in NaCl-H2SO4-Ca ( OCl ) 2 media was ~50 % .
( 5 ) The entire optimal leaching status to maximise lead and silver extraction at the same time were t=16 min, T=70 -C, ( S/L ) =1/38, pH=0.5, and C=0.6 g/L. In such a status, extraction of lead and Ag were 93.60 % and 49.21 % , severally.
( 6 ) It was shown that the leaching in NaCl-H2SO4-Ca ( OCl ) 2 media is an efficient method to pull out lead and Ag from ZPR.