sign in sign up
<<
Home , Cyanogen

Ind. Eng. Chem. Res. 1988, 27, 1157-1162 1157 volves now more defining conditions. Since the evidence Greek Symbols 2 for tristability is not conclusive at 208 °C, we do not pursue a= dimensionless exothermicity, Cb1E 1'A/RT8 this avenue. (-Ml) = heat of reaction 'A = adiabatic temperature rise per 1 vol % cbl Acknowledgment Subscript Work was supported by the Israel Academy of Sciences i = species i and Humanities. Registry No. Pt, 7440-06-4; C2H4, 74-85-1. Nomenclature Literat ure Cited a = stoichiometric coefficient av = surface-to-volume ratio Froment, G. F.; Bischoff, K. B. Chemical Reactor Analysis and C;, Cb; = surface and bulk concentration, vol % Design; Wiley: New York, 1979; Chapter 2. Da Damkohler number Froment, G. F.; Hosten, L. In Science and Technology; = Anderson, J. R., Boudart, M., Eds.; Wiley: New York, 1984; Vol. EifR = activation energy, K 2, p 97. F = steady-state function Harold, N. P.; Luss, D. Ind. Eng. Chem. Res. 1987a, 26, 2092; 1987b, h = heat-transfer coefficient 26, 2099. K, = adsorption equilibrium coefficient Harold, N. P.; Sheintuch, M.; Luss, D. Ind. Eng. Chem. Res. 1987, k, k0 = reaction rate constant and preexponential factor, 26, 783. respectively Hershkowitz, M.; Kenney, C. N. Can. J. Chem. Eng. 1983, 61, 194. kc = mass-transfer coefficient Mandler, J.; Lavie, R.; Sbeintucb, M. Chem. Eng. Sci. 1983, 38, 979. l = apparent reaction order Schmidt, J.; Sheintuch, M. Chem. Eng. Commun. 1986, 46, 289. m = order of reaction Sheintuch, M.; Schmidt, J.; Rosenberg, S. submitted for publication = in Ind. Eng. Chem. Res. 1988. n inhibition order Tsotsis, T. T .; Haderi, A. E.; Schmitz, R. A. Chem. Eng. Sci. 1982, Q = heat input by oven 37, 1235. r = reaction rate T., Tb, T. = ambient, apparent bulk, and solid temperature, Received for review July 1, 1987 respectively Revised manuscript received January 1, 1988 .x = degree of conversion Accepted January 21, 1988

The Effect of and Iron Complexation on Removal of by

Mirat D. Gurol* Department of Civil Engineering and Environmental Studies Institute, Drexel University, Philadelphia, Pennsylvania 19104 Timothy E. Holden Versar, Inc., 6850 Versar Center, Springfield, Virginia 22151

Oxidation of cyanide by ozone in basic aqueous solution was studied in the absence and presence of copper and iron. Free cyanide was oxidized via a fast reaction under mass-transfer-limited conditions. Copper catalyzed the oxidation of cyanide further by entering into an oxidation-reduction reaction. Oxidation of cyanide consumed equal moles of ozone and produced equal moles of cyanate. Upon prolonged ozonation, cyanate was converted to dioxide. Complexation of cyanide with iron hindered the oxidation reaction. Under the experimental conditions, oxidation of each mole of iron-complexed cyanide to required the consumption of more than 30 mol of ozone. Cyanate was not detected during ozonation of iron-cyanide complex.

Complete destruction of free cyanide ion in aqueous of 3 mol/L. Khandelwal et al. (1959) however, reported solution has been reported unanimously by numerous re­ that increasing the Cu(II) concentration did not markedly searchers (Gurol and Bremen, 1985; Rowley and Otto, increase the reaction rate. 1980; Zeelvalkink et al., 1979; Matsuda et al., 1975a,b; Sondak and Dodge (1961a,b) studied the reaction in the Balyanskii et al., 1972; Sondak and Dodge, 196la,b; presence of copper, iron, manganese, and vanadium and Khandelwal et al., 1959; Seim, 1959; Walker and Zabban, claimed no noticable effect of any of these metals on the 1953). However, there is disagreement as to the effects reaction rate. On the other hand, iron complexes of of metal complexation of cyanide on the reaction rate. For cyanide in plating wastes were found resistant to ozonation example, copper as Cu(II) was found to catalyze the oxi· by Mauk et al. (1976) and Streebin et al. (1981), who dation of cyanide (Matsuda et al., 1975a,b; Khandelwal et recommended coupling UV radiation with ozonation for al., 1959). It was believed that cu++ formed an oxide with complete destruction of iron-cyanide complexes. ozone; subsequently, cu++ was regenerated by a reaction This study was undertaken to understand the effects of of the oxide with cyanide (Khandelwal et al., 1959). complexation of cyanide with copper and iron on the ki· Matsuda et al. (1975a,b) observed a maximum in the re­ netics and the mechanism of cyanide oxidation by ozone action rate which corresponded to a Cu(II) concentration in aqueous solution. It was hoped that the findings would

0888-5885/ 88/2627-1157$01.50/0 © 1988 American Chemical Society 1158 Ind. Eng. Chem. Res., Vol. 27, No. 7, 1988

help to resolve the apparent conflicts in the literature. 4.0 pH : 11.5, T :2o' c Experimental Methods a: 1.0 L/mln (0,llnf : 32mg/L Ozone was contacted with cyanide solutions in a 500-m.L 3.5 glass reactor operated in semicontinuous fashion, i.e., (cu]. = 0.96mM continuous with respect to the flow and batch with respect to solution. The details of the experimental setup and the procedure were presented elsewhere (Gurol and 3.0 30 Bremen, 1985). The free-cyanide solutions were prepared by dissolving NaCN in a 0.01 M phosphate buffer. NaCN and CuCN 2.5 were dissolved in the buffered solutions to obtain cop­ per--cyanide complexes. Since the original ratio of CN/Cu i was 4:1, Cu(CN).3- was the predominant species in solu­ .s 2.0 .~ tions before ozonation (Rothbaum, 1957; Penneman and ..,..c: Jones, 1956). Solutions of iron-cyanide complexes were <.) prepared by dissolving K Fe(CN) in the buffered solu­ 0 3 6 c 1.5 3 .2 tions. Fe(CN)6 - was the predominant species in unozo­ nated solutions (Busey, 1965; Broderius, 1973). The final f c.,.. pH of all the test solutions was adjusted to 11.5 to reduce c 0 1.0 the chance of HCN formation. NaOH was used for pH <.) adjustments. During the course of the ozonation of cyanide solutions, samples were taken at various time intervals. The samples .5 5 were analyzed for total cyanide and cyanate concentrations. Cyanide ion in solutions of NaCN was measured by ti­ tration with AgN03 (Standard Methods for the Exami­ nation of and Wastewater, 1980) or by an Orion cyanide electrode. The solutions of copper--cyanide com­ Time (min) plexes were treated with EDTA under acidic conditions Figure l. Oxidation of free cyanide and copper-complexed cyanide at 50 °C for decomplexation of cyanide (Frant et al., 1972). by ozone. The concentration of cyanide was measured colorimet­ rically after it reacted with chloramine-T at neutral pH function of time of ozonation is depicted in Figure 1. No (Standard Methods for the Examination of Water and ozone could be detected in either sofotion, indicating the Wastewater, 1980). This technique yielded greater than mass-transfer-limited nature of the process for both solu­ 99% average recovery on 31 standard solutions. tions. Gurol and Bremen (1985) studied the reaction be­ The concentration of cyanide in solutions of iron com­ tween ozone and free cyanide by a stopped-flow spectro­ plexes was measured by monitoring the light absorbance photometer. The results indicated that the rate of the oxidation of free cyanide at pH 11.5 would be much faster of Fe(CN)63- at 410 nm (Broderius, 1973). However, copper added to iron-cyanide solutions interfered with the ab­ than the absorption of ozone from the gas phase into the sorption at this wavelength. Therefore, these solutions liquid phase. were exposed to UV radiation at acidic pH in the presence The disappearance of free cyanide followed a straight of sodium sulfide for decomplexation (Sekerka and Lech­ line, as would be expected from a mass-transfer-limited ner, 1976). A Rayonet photochemical reactor equipped process with constant mass-transfer coefficient (kLa) in a with 16 low-pressure mercury lamps, each emitting 1.5 W bubble contactor as follows (Gurol, 1985): of 254-nm UV light, was operated for 30 min for each (rate)cN remove) = (rate of ozone absorption) I z (1) sample of 50 mL. After the addition of bismuth nitrate ); - ) and pH adjustment to 11.5 (Sekerka and Lechner, 1976), (rate of ozone absorption) = Q8{(03 11 (03 8 rrl (2) cyanide ion was measured by the cyanide electrode. Acid )mf 41) distillation in the presence of MgC12, as described in (rate of ozone absorption) = Qg(03 1 - e· (3) Standard Methods for the Examination of Water and Wastewater (1980) for decomplexation of , was where z is the number of moles of ozone consumed per also tested for the copper and the iron complexes. How­ mole of cyanide oxidized, Qi is the gas flow rate, (Os)in and ever, this technique was not used routinely since it was (03)eff are the concentrations of the ozone in the influent time consuming, it required large sample sizes, and it did and the effluent gas streams, and not produce significantly more accurate results than the

4.0 ' 0 PH: 11.5, T:2o' c Q : 1.0 L/mln

C01)inf: 32mg/L :::! [cu].a o.e6mM ! 4.0 3.0 @ 30 (01lelf) .. " (!) j @) w/ and w/o Cu °' ' c .. ' ' 0 ! ''t:... :! ...... Cyanate ;:;:; 0"' 2.0 Cyanate_./' 0 , ( w/Cu 20 ,; 3.0 wloCu ..... , c 0 ...... o ... -2 ...... :! £ ''" c - ...... "'1l .. i" pH : 11.5 ,, g ~ ... ,§ g" 1.0 T:20' C - --- 10 u" 2.0 0 0 : 0.5L/min c " "0 -~ (01)1n!=,4mg/L ... ~ 0 c., (,) c 0 u 0 10 20 30 40 60 80 70 80 90 1.0 Time (min) Figure 3. Oxidation of cyanate by ozone.

suggested that the decrease in the concentration of soluble 0 2 3 4 5 6 7 8 9 10 copper and/or the change in the speciation might be re­ Time (min) sponsible for the reduction in the reaction rate. Figure 2. Oxidation of cyanide to cyanate by ozone. Cyanate. Cyanate, CNO-, was produced as an inter­ mediate product of the oxidation reaction, as shown in Figure 2. One mole of CNO- was produced per mole of not enhanced above its physical value, which was measured cyanide oxidized, in both the absence and presence of in the absence of the chemical reactions. Accordingly, the copper. It is believed that the oxidation of cyanide takes reaction regime prevalent during the oxidation of cyanide place according to a stepwise scheme which involves the was an "intermediate reaction regime" (Danckwert.s, 1970), formation of gaseous cyanogen, (CNh, as follows (Migrd­ and the reactions take place mainly in the bulk water. ichian, 1947; Williams, 1948; Deltombe and Pourbaix, The presence of copper increased the rate of cyanide 1955): disappearance significantly (Figure 1). Comparison of the initial rates indicates a 5-fold increase in the presence of 2CN- oxidation (CN)2 ~ CN- + CNO- (5) copper under the experimental conditions. This increase in the rate of cyanide removal is due to an enhancement Cyanogen was also identified in this study in ozonated in the mass transfer of ozone, as is obvious in Figure 1. solutions of free and copper-complexed cyanide by di­ During the oxidation of free cyanide, ozone was detected recting the effluent gas from the reactor into a 0.5 M in the effluent gas immediately, and its concentration NaOH solution in which cyanogen gas was trapped. reached a steady-state value of about 24 mg/ L. However, Cyanide which was produced upon hydrolysis of cyanogen during the oxidation of copper-complexed cyanide, no according to the above reaction scheme was measured by ozone was detected in the effluent gas for the first 3 min, the cyanide electrode. A KI trap was placed between the indicating that all the ozone introduced to the solution was reactor and the NaOH solution to remove ozone from the absorbed and reacted. The breakthrough occurred only effluent gas. Since cyanogen was measured only at trace after 85% of the cyanide was oxidized. Mass balance levels (1-2 mg/L), and the rate of-formation of cyanate calculations revealed the consumption of 1 mol of ozone was observed to be equal to the rate of oxidation of cyanide per mole of cyanide oxidized. . (Figure 2), the first reaction in the above reaction scheme The enhancement in the rate of ozone absorption in the is expected to be the slowest, and hence the rate-limiting presence of copper can be caused by a physical effect of step. copper on the mass-transfer coefficient, such as an increase In the absence of ozone, the hydrolysis of cyanate to in the interfacial area "a" due to surface accumulation carbonate and , as suggested by Seim (1959), did (Gurol and Nekouinaini, 1985). This possibility was not occur over a time period of 24 h. Nevertheless, cyanate checked by measuring the mass-transfer coefficient for was oxidized further by ozone. Balyanskii et al. (1972} in the presence and absence of copper in cyanide reported for the reaction of ozone with cyanide and cyanate solutions under the same experimental conditions. The empirical and system-specific rate constants which com­ measurements were not significantly different from each bined the kinetics of mass transfer of ozone and of chem­ other to justify the observed enhancement in ozone ab­ ical reactions. On the basis of these rate constants, they sorption in the presence of copper. Oxygen was used in concluded that cyanate is oxidized at one-fifth of the ox­ these experiments because it does not react with cyanide idation rate of cyanide. Bremen (1984) measured the true or copper. These results and further evidence presented reaction rate constants by a stopped-flow spectrophotom­ later in this paper indicate that the observed enhancement eter. The rate constant for the reaction of ozone with was more likely due to a change in the reaction regime from cyanate is reported to be 2 orders of magnitude smaller "intermediate" to "fast" because of the very fast reaction than the rate constant of cyanide oxidation. In Figure 2, of ozone with copper. cyanate was observed to accumulate until cyanide is oxi­ During the oxidation of cyanide, copper released from dized completely; this is consistent with the 2 orders of cyanide complexes formed various precipitates in colors magnitude difference in the reaction rate constants, as of blue, green, brown, and black. The reduction in the observed by Bremen. oxidation rate of cyanide after the third minute corre­ The reaction of ozone with cyanate was studied by using sponded to the beginning of heavy precipitation, which the same experimental setup. The results of a typical 1160 Ind. Eng. Chem. Res., Vol. 27, No. 7, 1988

4 .0 ~------~ •.o lron-cyanldt <:lnh 44 motL in the solution phase because of the accelerated self-de­ composition of ozone at such high pH values. Under the

experimental conditions, the absorption of a minimum of 1.5 "----.-----.----..,.----,.---.----' 5 mol of ozone was required for the oxidation of 1 mol of 0 10 20 30 40 50 60 cyanate. This stoichiometric factor increased with ozo­ Tlme(mln ) nation time probably because of the competition presented Figure 5. Effect of copper on the oxidation rate of iron-cyanide by the reaction products for the oxidant. These observa­ complex. tions support an indirect reaction mechanism in which the 1 1 hydroxyl radicals generated upon decomposition of ozone, profiles for the Cu/ CN ratios of 0, / 4, and / 2 are shown rather than molecular ozone, act as the predominant ox­ in Figure 5 for the same experimental conditions. The idizing agents (Hoigne and Bader, 1976). Contrary to the amount of cyanide which was removed from the solution reports by Matsuda et al. (1975a,b), copper did not affect during the first 15 min of ozonation increased by a factor the rate of this indirect reaction (Figure 3). Ozone con­ of 4 in the presence of 0.96 mM copper and by a factor of centration in the effluent gas increased immediately to a 8 in the presence of 1.92 mM copper. The rate of removal value which remained ·stable throughout the experiment. reduced after the first 15 min, probably because of the Ozonation of Iron-Complexed Cyanide. Figure 4 precipitation of copper from the solution. During the presents evidence that the complexation of cyanide to iron reaction period, no buildup of cyanate was observed. hinders the oxidation of cyanide by ozone to a great extent. These experimental observations indicate that the This supports the earlier reports by Mauk et al. (1976) and iron-cyanide complexes in metal-plating wastewaters may Streebin et al. (1981). not be oxidized by ozone within a reasonable period of Increasing the ozone concentration in the applied gas time. However, in the presence of copper, e.g., copper­ did not affect the oxidation rate of iron-complexed cyanide. plating wastewaters, ozonation might still be an econom­ After 90 min of ozonation, only about 25% of the cyanide ically feasible treatment alternative. was oxidized. Chemical analysis throughout this time The Mechanism of Copper Catalysis. Cu(Il) is be­ period showed no buildup of cyanate in the solution. In­ lieved to be unstable in the presence of cyanide. According dependent experiments proved that iron did not interfere to Cotton and Wilkinson (1972), Cu(Il) is reduced to Cu(I) with cyanate analysis. Accordingly, the following expla­ a.s illustrated below: nation might be offered for the absence of cyanate in 4CN- + 2Cu++ - 2CuCN + (CN) (9) ozonated iron-cyanide complexes. 2 3 As in the case of cyanate, cyanide in Fe(CN)6 - might When coke plant wastewater was distilled in the presence be oxidized by the hydroxyl radicals which are produced of CuC12 for cyanide an.alysis, low results were observed upon self-decomposition of ozone. The observation of a by Barton et al. (1978). The authors believed that the requirement of more than 30 mol of ozone/mo! of cyanide oxidation of Cu(I) to Cu(ll) by oxygen in air consequently in iron complexes supports this statement. Hence, the caused the oxidation of cyanide as follows: reaction might take place according to the following 2Cu++ + 3CN- + 20H- - 2CuCN + CNO- + H 0 (10) scheme: 2 Bernardim (1973) observed the catalytic effect of Cu(I) on (6) the oxidation of cyanide on granular activated carbon. k, Coleman (1968), who studied the oxidation of cyanide by 3 Fe(CN)6 - + OH· - CNO- (7) Cr(VI), reported that the reaction occurred only in the presence of copper. k 2 CNO- + OH· - products (8) These observations suggest that the catalytic effect of copper on the oxidation of cyanide by ozone is due to an Comparison of Figures 3 and 4 leads to the conclusion that independent oxidation-reduction reaction between Cu(Il) k2 > k1• Therefore, CNO- is expected to react as soon as and cyanide. In order to test this hypothesis, Cu(II) as it is formed from cyanide. The rate constant k1 might be CuS04 was added in increments to cyanide solutions in determined by the relatively slow dissociation rate of Fe­ various amounts at pH 11.5 in the absence of ozone. After (CN)63- complex into CN-. 1 h of stirring at each concentration, the samples were Solutions of K3Fe(CN) 6 were ozonated in the presence analyzed for cyanide and cyanate. The results are sum­ of various amounts of copper (added as CuS04) to deter­ marized in Table I. Cu(II) oxidized free cyanide and mine whether copper will also increase the rate of removal cyanide in complexation with Cu(I), and cyanate was of cyanide which is in complexation with iron. The cyanide produced at approximately one-to·one ratio, as postulated Ind. Eng. Chem. Res., Vol. 27, No. 7, 1988 1161

Table I. Oxidation of Cyanide by Copper (Concentrations In mM) free cyanide copper-cyanide iron- cyanide [Cu} / [CN] cyanide remaining cyanate produced cyanide remaining cyanate produced cyanide remaining cyanate produced 0 3.84 0 3.84 0 3.84 0 1/ 4 3.05 0.46 3.22 0.41 3.84 0 1/ 2 2.71 0.78 3.27 0.43 3.47 0 1/ 1 2.88 1.00 3.06 0.61 2.04 0 2/ 1 2.80 0.99 3.18 0.58 0.87 0

Table II. Carbon Balance on Ozonated Cyanide Solutions free cyanide copper-cyanide iron-cyanide components, mg/ L of C blank I• IIb 1• nb TC 18 51 40 54 41 56 56 TC-TCblanlc 0 33 22 36 23 38 38 TOC <2 NM' <2 NM <2 NM 13 TOC-TOCb1ank 0 NM 0 NM 0 NM 11 % C02 produced 67 64 71 % CN remaining 0 0 29 % CNO remaining 0 0 0 % C accountable 67 64 100

•I, before ozonation. b II, after prolonged ozonation.