CYANIDE in WATER and SOIL: Chemistry, Risk, and Management - Chapter 5 pps

formation of molecular hydrogen cyanide HCN, oxidation of cyanide to cyanate, and adsorptiononto clays and other soil components.In environmental systems, wastewaters, and wastes, cyanid

5 Physical–Chemical Properties

and Reactivity of Cyanide in

Water and Soil

David A Dzombak, Rajat S Ghosh, and Thomas C Young

CONTENTS

5.1 Free Cyanide 58

5.1.1 Cyanide Ion Bonding 58

5.1.2 HCN Formation and Dissociation 58

5.1.3 HCN Volatilization 60

5.1.4 Free Cyanide Adsorption to Soil and Sediment 61

5.1.5 Free Cyanide Oxidation 62

5.1.6 Free Cyanide Hydrolysis 64

5.2 Metal Cyanides: Aqueous Species 65

5.2.1 Weak Metal–Cyanide Complexes 65

5.2.1.1 Formation 65

5.2.1.2 Dissociation 67

5.2.1.3 Adsorption on Soil and Sediment 68

5.2.1.4 Oxidation 71

5.2.2 Strong Metal–Cyanide Complexes 73

5.2.2.1 Formation 73

5.2.2.2 Dissociation 75

5.2.2.3 Adsorption on Soil and Sediment 76

5.2.2.4 Oxidation–Reduction 78

5.3 Metal–Cyanides: Solid Phase Compounds 79

5.3.1 Simple Cyanide Solids 80

5.3.2 Alkali or Alkaline Earth Metal–Metal Cyanide Complex Solids 80

5.3.3 Other Metal–Metal Cyanide Complex Solids 80

5.4 Cyanate 82

5.5 Thiocyanate 84

5.6 Organocyanides 86

5.7 Summary and Conclusions 88

References 88

The reactivity, fate, and toxicity of cyanide in water and soil is highly dependent on the chemical exist The simplest form of soluble cyanide is the negatively charged cyanide ion, CN−, which is composed of a carbon atom triple bonded to a nitrogen atom (–C≡N) The nature of this triple

bond controls the reactivity of the cyanide anion, including complexation with other metal cations,

57

speciation of the cyanide As outlined inChapter 2, many different soluble and solid forms of cyanide

formation of molecular hydrogen cyanide (HCN), oxidation of cyanide to cyanate, and adsorptiononto clays and other soil components.

In environmental systems, wastewaters, and wastes, cyanide usually is found in free and plexed forms, as HCN and as metal–cyanide complexes Because of a reactive electronic arrangement,cyanide anions can readily form metal–cyanide complexes with most metal cations Most of thesecomplexes exist as soluble species, but many, particularly iron-cyanide complexes, can react furtherwith metal cations to form stable cyanide solids The soluble and solid phase cyanide species thatthis chapter, the specific physical–chemical properties and reactivity characteristics of the differ-ent chemical forms of cyanide are presented Included are examinations of the nature of bonding

com-in and with the cyano group and free cyanide speciation; the properties and reactivities of solublemetal–cyanide complexes; the properties and reactivities of metal–cyanide complex solids; and theproperties and reactivities of cyanate, thiocyanate, and organocyanide compounds

5.1 FREE CYANIDE

Free cyanide consists of the cyanide anion, CN−, and molecular hydrogen cyanide, HCN, bothexisting as water soluble entities The cyanide ion acts as a monodentate ligand with the carbonacting as the donor atom, and also as an ambidentate ligand acting as a donor at both ends ofthe ion [1]

Several structural factors govern the reactivity of free cyanide The triple bonded structure of acyanide anion is comprised of a sigma bond, twoπ bonds, and two empty bonding orbitals [2] The

“s” and the “p” orbitals are filled with maximum number of electrons, while the “d” and “f” orbitalsare empty This configuration allows for a number of bonding arrangements Since halogens alsohave filled “s” and “p” orbitals, the behavior of the cyanide anion is similar to that of halogens [3].The cyanide ion is considered a pseudo-halide in that it can formπ-acceptor covalent bonds with

transition metals [3] It may also share electrons at the triple bond with the Group VI elements oxygenand sulfur, forming cyanate, CNO−, or thiocyanate, SCN−[3], or may act as a strong nucleophile inreactions with organic molecules, for example, nucleophilic addition reactions with aldehydes andketones to form cyanohydrins [4]

The cyanide ion readily forms neutral compounds or anionic complexes with most major metalcations The partially or wholly filled “d” orbitals of transition series metals can form covalent bondswith the empty anti-bonding orbitals of the cyanide ion This involves acceptance of electron densityintoπ orbitals of the carbon atom The cyanide ion is a strong σ donor, which is responsible for the

high stability of some of the metal–cyanide complexes [3]

The cyanide anion protonates in water to form hydrocyanic acid, HCN, the most toxic form of

afor HCN dissociation reaction is 9.24 at 25◦C [5] Thus,

at pH greater than 9.24, cyanide anion dominates free cyanide speciation, while soluble HCN is thedominant species under acidic to neutral pH conditions (pH< 9.24) The free cyanide dissociation

reaction is as follows:

HCN= H++ CN−, pKa= 9.24 at 25◦C, I= 0 (5.1)

−species as a function of pH for a simple aqueoussolution at 25◦C The temperature dependence of the equilibrium constant governing the species

occur most often in water and soil are outlined inChapter 2and examined in more detail here In

cyanide (seeChapters 13and14) The pK

Figure 5.1shows the distribution of HCN and CN

as calculated using the thermodynamic data compiled in Stumm and Morgan [6] Substitution ofthis value, and assuming it is approximately constant for the temperature range 5 to 30◦C, enablescalculation of the temperature dependence of the acidity constant in Equation (5.1):

KT= exp[1.756 × 104K(3.356 × 10−3K−1− T−1) − 21.28] (5.3)

where KT is the equilibrium constant for HCN dissociation at the temperature T (K) of interest.

Combining Equation (5.3) with the mass action equation for the reaction in Equation (5.1), and themass balance equation for free cyanide (molar concentrations in [ ]),

Not given 0.104 to 0.114 0.0043 to 0.0047 Smith and Mudder [2]

strength (I) As is evident in Figure 5.2, temperature has a significant effect on free cyanide cies distribution As temperature decreases, dissociation of HCN decreases, extending the speciesdominance of HCN to higher pH values

spe-5.1.3 HCN VOLATILIZATION

Hydrogen cyanide has a very low boiling point (25.7◦C) and thus is volatile in water underenvironmental conditions The equilibrium air–water partitioning of HCN can be described byHenry’s Law:

of temperature One such relationship, reported by Bodek et al [7], is as follows:

where KH,HCNis the Henry’s Law constant, mm Hg/M and T the Temperature, K Equation (5.8) is reported to be valid for HCN concentrations ranging from 0.01 to 0.5 M and temperatures from 20

to 95◦C.

Free cyanide (CN−, HCN) adsorbs weakly on soils and sediment The cyanide anion can be retained

by soils with anion exchange capacity, but in the pH range 4 to 9 of interest for most soils, HCN isthe dominant form of cyanide and CN−concentrations are very low HCN adsorbs weakly or not all

to inorganic soil components such as iron oxide [8], aluminum oxide, clay, and sand [9] However,HCN has been shown to adsorb significantly to soils with appreciable organic carbon content Themagnitude of cyanide adsorption onto soils tested by Chatwin et al [10] showed excellent correlationwith organic carbon content Higgins and Dzombak [9] further demonstrated the interaction of HCNwith organic carbon in experiments with activated carbon and freshwater sediment They developed

an expression relating sorbed HCN concentration, CS, to aqueous phase concentration, Cw, through

an organic carbon normalized distribution coefficient Koc(=Kd/foc)

CS= KocCwfoc= (6.5 L/gs)Cwfoc (5.9)

where CS is inµg/gs, CW is inµg/L, and focis the fraction of organic carbon in the adsorbent.The experiments upon which this linear relationship is based all involved low concentrations of freecyanide in water (<150 µg/L), which is typical for total cyanide concentrations encountered inenvironmental contamination scenarios Adsorption capacities were not determined in the experi-ments with activated carbon and sediment Literature data on free cyanide adsorption onto activatedcarbon have shown an adsorption capacity of about 1 to 2 mg of free cyanide per gram of carbon,while similar tests performed with soil organic carbon have revealed an adsorption capacity of 0.5 mg

of free cyanide per gram of carbon [11]

Batch and column tests performed by Alesii and Fuller [12] with various soils yielded significantremoval of free cyanide at near-neutral pH values Soil constituents included kaolin clay, chlorite,gibbsite clay, and iron and aluminum oxides Based on the laboratory results discussed earlier, it isunlikely that these inorganic constituents would adsorb free cyanide to an appreciable extent As thesoils used in the experiments by Alesii and Fuller were not sterilized and hence biologically active,

it is more likely that the free cyanide was removed from the system via biodegradation

Dzombak and Morel [13] estimated equilibrium surface complexation constants for the tion of CN−, CNO−, and SCN−on hydrous ferric oxide based on correlations of acidity constantsand surface complexation constants fitted to adsorption data for other inorganic ions The surfacecomplexation reactions and the estimated surface complexation constants for those reactions are

adsorp-−4 M solutions of these ions inhydrous ferric oxide suspensions with TOTFe= 10−3M and ionic strength of 0.01 M are shown inface complexation constants, the predictions provide some idea of the expected adsorption behaviorbased on what has been observed with other inorganic ions Available data for free cyanide adsorp-tion on mineral surfaces, however, indicates that the free cyanide adsorption in Figure 5.3 is likely

to be substantially overpredicted Free cyanide has been observed to exhibit little to no adsorption

on mineral surfaces, including the crystalline iron oxide goethite, across a range of pH [8,9]

Figure 5.3 While the accuracy of these predictions is uncertain due to the estimated nature of the given inTable 5.2 Predicted adsorption versus pH curves for 10

sur-TABLE 5.2

Estimated Surface Complexation Reactions and Constants for

Adsorption of CN − , CNO − , and SCN − on Hydrous Ferric

a SC reaction: ≡FeOH 0 + A−+ H+= ≡FeA 0 + H 2O; K2.

b SC reaction: ≡FeOH 0 + A − = ≡FeOHA −; K3.

Source: Data from Dzombak, D.A and Morel, F.M.M., Surface Complexation

Modeling: Hydrous Ferric Oxide, Wiley-Interscience, New York, NY, 1990

(Table 10.10).

pH

CN – CNO – –

of pH Predictions made using surface complexation constants of Dzombak, D.A and Morel, F.M., Surface

Complexation Modeling: Hydrous Ferric Oxide, 1990; see Table 5.2 TOTFe = 0.001 M, I = 0.1 M Adsorption

of CN−is likely overpredicted.

Free cyanide can be oxidized to cyanate, CNO−, or hydrogen cyanate, HCNO, depending on the

pH [14]:

SCN

Cyanate is protonated only at low pH, as the pKais 3.45 [5]:

is often exploited for rapid treatment of free cyanide in water The general reaction chemistry foralkaline chlorination is as follows [17]:

3Cl2+ 2CNO−+ 6NaOH → 2HCO−3 + N2+ 6Cl−+ 6Na++ 2H2O (5.15)

As indicated in Equation (5.13), cyanide ion is first converted to cyanogen chloride, CNCl, a highlytoxic species Under alkaline conditions, the CNCl reacts rapidly with OH−to form CNO−, and uponfurther chlorination the cyanate decomposes to form the completely benign products bicarbonate,HCO−

3 and elemental nitrogen, N2 In the last step, Equation (5.15), the nitrogen is oxidized, movingfrom an oxidation state of−3 to zero

Gurol and Bremen [19] studied ozonation of free cyanide It was found that the ozone molecule,

O3, reacts primarily with the cyanide ion; its reaction with HCN is negligible Further, it wasdetermined that the presence of free cyanide promotes the formation of free radicals (OH•, HO2•),

and that free radical reactions as well as direct reaction of the free cyanide with ozone contribute tothe overall oxidative destruction of the cyanide Hence, there are numerous initiators and pathwaysinvolved in the oxidation of free cyanide by ozone Some of the reactions identified by Gurol andBremen [19] as involved with the ozonation of free cyanide are as follows:

2 −•, which reacts

further with ozone to give the hydroxyl radical OH• The oxidation of cyanide by ozone is rapid

and pH dependent [19] Solutions of several mM of free cyanide were oxidized within 5 to 30 min

by ozone, with faster rates at higher pH values where more of the free cyanide was in the form of

CN− The end product of ozonation of free cyanide is cyanate The cyanate is further oxidized byozone, but since this is a relatively slow reaction cyanate accumulates in solution until free cyanide

Chatwin et al [10] detected cyanate in effluent from saturated soil columns through whichaqueous solutions containing free cyanide were passed It was hypothesized that the free cyanidewas oxidized to cyanate on the surfaces of clay components of the soil, and that the process wasenhanced with the addition of copper and nickel to the system However, since the soils studied weremicrobiologically active, microbial degradation was likely to have had an equal or possibly greaterrole in the conversion of the cyanide, a factor not addressed by Chatwin et al [10] Based on otherwork showing very limited to no interaction of free cyanide with mineral surfaces [8], it appearsunlikely that abiotic oxidation of free cyanide on mineral surfaces will occur appreciably in naturalsystems.

Free cyanide can also react with and be oxidized by various forms of sulfur, especially sulfides and thiosulfate (S2O2−

poly-3 ), to form thiocyanate (SCN−) In neutral to alkaline solutions, bothpolysulfides and thiosulfate are products of oxidation of sulfide The reactions of polysulfide andthiosulfate with the cyanide ion are as follows [2,21]:

SxS2−+ CN−→ Sx−1S2−+ SCN− (5.22)

S2O2−

3 + CN−→ SO2 −

For thiocyanate, the oxidation states of the S, C, and N are−1, +3, and −3, respectively In the

reaction of polysulfide with free cyanide (Equation 5.22), one polysulfide sulfur atom is reducedfrom its oxidation state 0 to−1, while the cyanide carbon atom is oxidized from +2 to +3 [21]

In the reaction of thiosulfate with free cyanide (Equation 5.23), one thiosulfate sulfur atom changesfrom oxidation state+2 to +4, while the other thiosulfate sulfur atom is reduced from the +2 to

the−1 oxidation state [21] The rate of thiocyanate formation through reaction of polysulfide and

free cyanide is approximately three orders of magnitude greater than through reaction of free cyanideand thiosulfate, depending on pH [21] Thus, in systems with equal amounts of polysulfide andthiosulfate present, the reaction of free cyanide with polysulfide will be the dominant thiocyanateformation route The formation of polysulfide through oxidation of sulfide occurs at a slow rate,however, so available polysulfide is often limited [21] Reaction of thiosulfate with free cyanide thusgoverns the formation of thiocyanate in many systems

As discussed in Section 5.1.2, the cyanide ion reacts with water (H+) to form HCN, with theprotonated species HCN being the dominant form of free cyanide at pH values less than 9.24 at

◦

cyanide with water

Free cyanide can react with molecular water under alkaline conditions and high temperature toyield formate and ammonia:

The reaction proceeds at appreciable rates only at high temperatures, and at fast rates at high ure and pressure, for example, temperatures in the range of 165–180◦C and pressures of 100–150 psiglysis is very slow at room temperature, increasing about threefold for every 10◦C rise in temperature.

temperat-At lower pH values, HCN can also be hydrolyzed, yielding formic acid and ammonia [2]:

Under acidic conditions the reaction is also very slow

[22]; seeChapters 20and22 Wiegand and Tremelling [23] showed that the rate of free cyanide

hydro-25 C (Figure 5.1) At ambient temperatures, this protonation reaction is the primary reaction of free

Alkaline hydrolysis has been exploited for treatment of free and complexed cyanide in waters and sludges [22,24] Alkaline conditions assure that any free cyanide remains dissolved in theform of CN−during the treatment process Temperatures in the range of 140 to 275◦C, and pressures

waste-up to 900 psig are employed in the alkaline hydrolysis process This treatment process is discussed

5.2 METAL CYANIDES: AQUEOUS SPECIES

Many metals form aqueous complexes with cyanide ion by means of π bonding, which occurs

when the participating metal atom donates one or more electrons to CN−, which serves as an tron accepting ligand These soluble metal–cyanide complexes, represented by a general formula[M(CN)x]y−, where, M signifies a metal cation, can be classified into weak and strong metal–cyanidecomplexes, depending upon the strength of the metal–cyanide bonding Use of vibrational spec-troscopy reveals different electronic structures of [M(CN)x]y−complexes [1] Depending on thedifferent modes of vibration, a [M(CN)x]y− species can exist in tetrahedral, square planar, oroctahedral forms These common electronic structures are shown in Figure 5.4

5.2.1.1 Formation

The cyanide anion can form weak metal–cyanide complexes with many transition metals, the mostcommon among them being cadmium, zinc, silver, copper, nickel, and mercury Most of these metalsfall in Groups IB, IIB, and VIIIB of the periodic table The metal–cyanide bonds in these complexesare mostly arranged in tetrahedral or square planar forms with relatively weak bonding energy existingbetween the heavy metal atom and the cyanide ligand as compared to the strong cyanide complexeswith iron, cobalt, and platinum Because weakly-bonded metal–cyanide complexes dissociate underweakly acidic pH conditions (4 < pH < 6), they are commonly termed weak-acid-dissociable

(WAD) complexes [15]

Formation data determined by direct thermodynamic measurements are available for complexes

of nickel(II), copper(I), silver(I), zinc(II), cadmium(II), and mercury(II) [1,5] Equilibrium or stabilityconstants have been calculated from standard thermodynamic data for a broad range of metal–cyanideSehmel [5] for formation of weak metal–cyanide complexes For comparable reaction stoichiometry,the higher the value of the formation equilibrium constant (K), the greater is the energy of formationand the stability of the metal–cyanide complex

FIGURE 5.4 Common electronic structures of metal–cyanide complexes.

in more detail inChapter 22

complexes, however [5].Table 5.3lists the measured and calculated stability constants compiled by

Source: Data from Sehmel, G.A., Cyanide and antimony thermodynamic database for the

aqueous species and solids for the EPA-MINTEQ geochemical code, PNL-6835, Pacific

Northwest Laboratory, Richland, WA, 1989, (Table 5).

The equilibrium constants compiled by Sehmel [5] were selected and included in Table 5.3 ratherthan those reported in some other compilations, for example, the work of Beck [25] and Martell et al.(1993), because the constants reported by Sehmel were calculated in a consistent manner using themost current thermodynamic data from the U.S National Bureau of Standards [26,27] The constantsreported in Beck [25] were calculated using older (1952) NBS thermodynamic data [28] As shown

by Gilgore-Schnorr and Dzombak [29], the key difference is in the value used for the partial molarentropy, So, of the cyanide ion CN−, for which the 1952 [28] value of 28.2 cal K−1mol−1was revised

in 1965 [27] to 22.5 cal K−1mol−1, a value retained in the 1982 (and most current) thermodynamicdata compilation [26] The work of Sehmel [5] was performed for the USEPA, which incorporatedthe metal–cyanide complexation constants in the thermodynamic database of the general chemical

PCO2= 10 – 3.5 atm

ZnT= 2.00
10 – 5 M

Zn+2

Zn(CN)2

FIGURE 5.5 Calculated aqueous speciation of zinc(II) in the presence of excess cyanide ZnT= 10−4.70M,

CNT= 10−2.70M, PCO2 = 10−3.5atm, I = 0.1 M (Source: Theis, T.L and West, M.L., Environ Technol.

OH−in the pH range shown Concentrations of dissolved zinc hydroxide species are very small andtheir influence on zinc speciation is not significant As is evident in Figure 5.5, different zinc cyanidespecies predominate in different pH regions, with Zn(CN)0

2the dominant form from pH 6.2 to 6.7,and Zn(CN)2 −

4 being most abundant above pH 6.7

The kinetics of metal–cyanide complex formation can be slow [1,33] For example, Broderius [33]showed that Ni(CN)2 −

4 formation took three days to reach equilibrium (0.5 to 6.5 ppm CNT), whileformation of copper cyanide complexes took over 100 days to reach equilibrium The slow rate offormation of metal–cyanide complexes, and the potential for oxidation of many weak metal–cyanidecomplexes by atmospheric oxygen, makes it difficult to measure the equilibrium constants for thesecomplexes [1]

5.2.1.2 Dissociation

The dissociation properties of weak metal–cyanide complexes in aqueous solutions depend on theirstability constants, pH, temperature and the redox potential of the solution In general, metal–cyanidecomplex dissociation may be described by

M(CN) n−

where M+is a monovalent metal cation, x is the number of cyanide groups, and n is the ionic charge

of the metal–cyanide complex

metal–cyanide complexes in solution to dissociate The higher the value for K for reactions with the

same stoichiometry, the more limited the dissociation into free cyanide (CN−).

Because of the labile nature of weak metal–cyanide complexes, mildly acidic conditions (pH≈ 4

to 6) can result in the dissociation of many of these complexes, especially nickel and zinc For thisreason, the weak-acid-dissociable cyanide analytical method [15] employs a pH 4.5 buffer whiledistilling the aqueous sample at 125◦C for 1.5 h to break down these complexes completely and

5.2.1.3 Adsorption on Soil and Sediment

Weak metal–cyanide complexes can adsorb on common soil and sediment components such asiron, aluminum, silicon, and manganese oxides, and clays, which in most systems will inhibit theiraqueous transport [8,12,34] However, complexation of metals by cyanide can also serve to hold them

in solution, inhibiting their adsorption and retention The enhancement or inhibition of adsorptiondepends on the metal–cyanide species, the adsorbent, and the solution conditions

Theis and Richter [34] studied the adsorption of the predominant nickel–cyanide anion Ni(CN)2 −

4

on silicon dioxide, SiO2(s) and the iron oxide goethite, FeOOH(s) Batch adsorption experiments

were conducted in 0.01 M NaClO4aqueous solutions containing 10−4.77M total nickel (Ni

T) andamounts of total free cyanide (CNT) of 10−5, 10−4, and 10−3M Calculated plots of the equilibriumdistribution of nickel species as a function of pH in aqueous solution with no solids present are

the speciation plots of Figure 5.6; the plots are for illustrative purposes.) As may be seen there,Ni(CN)2 −

4 is predicted to dominate nickel speciation at pH < 5.5 in the system with CNT =

10−4 M, and at pH < 4.5 in the system with CNT = 10−3 M A speciation diagram for thesystem with CNT = 10−5 M is not shown, but with NiT in excess of CNT in this system Ni2+and NiOH+are the dominant forms of nickel and a maximum of about 10% of the nickel becomesbound to cyanide Results of the batch adsorption experiments conducted with SiO2(s) are shown in

T= 10−5M is very similar tothe adsorption of nickel on SiO2(s) in the absence of cyanide For the series of batch experimentswith CNT = 10−4and 10−3 M, however, adsorption of nickel is inhibited These data indicatethat the nickel–cyanide species Ni(CN)2 −

4 , which dominates nickel speciation in both systems,has no affinity for the SiO2(s) surface The SiO2(s) surface is negatively charged for pH > 2, so

electrostatic repulsion of the negative Ni(CN)2 −

4 species is in part responsible for the absence ofsurface binding For the most part, however, it is the presence of cyanide that inhibits adsorptionFeOOH(s) Once again, the adsorption of nickel observed in the system with CNT= 10−5M is verysimilar to the adsorption of nickel on FeOOH(s) in the absence of cyanide In the systems with CNT=

10−4and 10−3M, adsorption of nickel is enhanced at lower pH values, and inhibited at higher pHvalues The FeOOH(s) surface is positively charged up to about pH 6, or even higher, so electrostaticattraction of Ni(CN)2 −

4 explains in part its adsorption at lower pH values Electrostatic attractionalone is not sufficient to explain the extent of removal observed, however Through surface interactionmodeling, Theis and Richter [34] demonstrated that Ni(CN)2 −

4 must bond at specific surface sites

on goethite, in surface complexation reactions that involve high free energies of interaction Theyproposed the formation of a goethite–cyanide–metal surface complex via a surface complexationreaction:

≡Fe2(OH)2 +

2 + Ni(CN)2 −

4 + 2H+= ≡Fe2–(CN)2–Ni–(CN)2+ 2H2O (5.27)

quantify the amount of cyanide associated with them (seeChapter 7)

given inFigure 5.6 (Note: Equilibrium constants from Sehmel [5] were not used in calculating

Figure 5.7 The adsorption of nickel observed in the system with CN

of nickel.Figure 5.8shows data for similar sets of batch adsorption experiments performed withThe stability constants given inTable 5.3provide an indication of the propensity of weak acid

3 4 5 6 7 8 9 10

60 40 20 80

60 40 20

0.01 M ), and (b) 10−3M cyanide (NiT= 10−4.77M, I = 0.01 M) (Source: Reprinted with permission from

Theis, T.L and Richer, R.O., Particulates in water, 189, 73, 1980 Copyright 1980 American Chemical Society.)

where≡Fe2(OH)2 +

2 is a surface hydroxyl site on the surface of goethite in aqueous suspension, and

≡Fe2–(CN)2–Ni–(CN)2is the surface species formed by adsorption of Ni(CN)2 −

4 on the goethite.The uptake of H+ shown in the reaction occurs commonly in adsorption of inorganic anions onoxides, and is related to the commonly observed pH dependence for anion adsorption: maximumadsorption at lower pH and decrease in adsorption with increasing pH [13] Formation of a metal–ligand–metal ternary surface complex as shown in the reaction of Equation (5.27) has been proposedfor other metal–ligand systems [35–38]

Theis and West [8] studied the adsorption of cadmium, copper, and zinc divalent cations and theirmetal–cyanide complexes on goethite in aqueous suspensions Some typical results for adsorption of

Cd2+, Cu2 +, and Zn2 +

were conducted with total metal concentration of approximately 2in the absence of cyanide are presented inFigure 5.9× 10−5 All of the experimentsM and with 0.6 g/L

3 4 5 6 6 7 8 9 10 11 0

10−4.77M, I = 0.01 M, SiO2= 29.41 g/L (
) CNT= 10−5M, () CNT= 10−4M, () CNT= 10−3M,

(•) CNT= 0 (Source: Reprinted with permission from Theis, T.L and Richer, R.O., Particulates in water,

189, 73, 1980 Copyright 1980 American Chemical Society.)

cation adsorption on metal oxides: an increase from 0 to 100% adsorbed with increasing pH Batchadsorption experiments conducted with free cyanide showed no adsorption of the free cyanide ongoethite for any pH from 3 to 11 (data not shown) Experiments with free cyanide added in excess of

At lower pH values, adsorption of the cadmium, copper, and zinc was unaffected by the free cyanide

as may be seen by comparison with Figure 5.9 Above pH 6.5 to 7.0, however, adsorption of the metalswas inhibited by the presence of the cyanide At the higher pH values, metal–cyanide complexesdominate the speciation of the metals (e.g., see the aqueous phase speciation diagram for zinc inhave no affinity for the goethite surface at neutral to alkaline pH values

The examples presented in this section demonstrate that some weak metal–cyanide complexescan adsorb on soils and soil components under some conditions, but the extent of adsorption dependsstrongly on the particular metal–cyanide species, mineral adsorbent, and solution conditions Solu-tion pH is an especially important governing parameter, as is the case for adsorption of all ions

on oxidic minerals [13] The data presented also demonstrate that the presence of free cyanide in

FeOOH(s) The pH adsorption edge plots shown inFigure 5.9exhibit the typical characteristics for

the metal concentrations were also performed Results for the three metals are given inFigure 5.10

Figure 5.5) The data in Figure 5.10 indicate that the cadmium–, copper–, and zinc–cyanide species

3 4 5 6 7 8 9 10 11 0

() CNT= 10−3M, (•) CNT= 0 (Source: Reprinted with permission from Theis, T.L and Richer, R.O.,

Particulates in water, 189, 73, 1980 Copyright 1980 American Chemical Society.)

a systems with metals, leading to the formation of metal–cyanide complexes, can result in enhanced

or reduced adsorption of the metals The metal–cyanide complexes may interact with the surface to

a greater or lesser extent than the metals alone An interrelated, complex group of factors governsmetal–cyanide species adsorption, and it is diffcult to form generalizations

5.2.1.4 Oxidation

Weak metal–cyanide complexes generally are readily oxidized by oxidizing agents such as chlorine orozone The more strongly bonded complexes in the WAD category, such as nickel, silver, and mercurycyanide complexes, oxidize more slowly [15] The more weakly-bonded complexes, including those

of cadmium, copper, and zinc, decompose rapidly in the presence of oxidizing agents

most common approach used to treat waters bearing free cyanide A number of weak metal–cyanidecomplexes are also readily oxidized in this process In order to identify the fraction of measuredtotal cyanide, which includes metal–cyanide complexes and free cyanide, that is treatable by alkaline

As discussed in Section 5.1.5, and in more detail in Chapter 20, alkaline chlorination is the

10−4.72M, () ZnT = 10−4.62M, (
) CdT = 10−4.68M (Source: Theis, T.L and West, M.L., Environ Technol Lett., 7, 309, 1986.)

10−4.71M, CNT= 10−4.00M (Source: Theis, T.L and West, M.L., Environ Technol Lett., 7, 309, 1986.)

chlorination, an analytical measurement known as “cyanide amenable to chlorination” [15] has longbeen employed The method involves measurement of total cyanide on samples with and withouttreatment by chlorination, with the difference giving the amount of cyanide in the sample amenableexistence of the method speaks to the facile oxidation of a number of weak metal–cyanide complexes

In some cases, the presence of weak metal–cyanide complexes can enhance the rate of free cyanidedecomposition through catalysis by the metal This has been demonstrated for copper cyanide com-plexes [20] Gurol and Holden [20] studied the effect of copper(I) on the removal of free cyanide by

to chlorination (Chapter 7) While the CATC method has limitations, as discussed in Chapter 7, the

ozone in alkaline solution They performed experiments in solutions at pH 11.5 in systems with anexcess of free cyanide over copper(I), giving Cu(CN)3 −

4 as the dominant copper species They foundthat the presence of copper increased the rate of free cyanide oxidation significantly Comparison ofinitial rates of cyanide disappearance for systems with and without copper indicated a fivefold higherrate for the system with copper cyanide species Further investigations revealed that the observedenhancement was likely due a very fast, independent oxidation–reduction reaction between Cu(I)and free cyanide The following reaction sequence was proposed:

An equilibrium between the copper(I) cyanogen species and Cu(CN)3 −

4 is rapidly established(Equation [5.31]) In the last step, Equation (5.32), cyanogen goes through a disproportionationreaction to yield free cyanide and cyanate The net reaction from the above sequence is thus asfollows:

2Cu(CN)3 −

4 + 3CN−+ 3O3→ 2Cu(CN)3 −

4 + 3CNO−+ 3O2 (5.33)Thus, the oxidation of 3 mol free cyanide requires 2 mol of ozone and produces 3 mol of cyanate, aswould be expected, but the rate of the reaction is much accelerated due to the presence of the Cu(I)

5.2.2.1 Formation

The cyanide anion can form strong complexes with a number of transition heavy metals, the mostnotable among them are cobalt, platinum, gold, palladium, and iron Most of these metals fall inGroups IB, IIB, and VIII of the periodic table As iron is by far the most abundant of these elements inthe environment and in process waters, iron–cyanide complexes are the strong metal–cyanide com-plexes of greatest interest Gold–cyanide complexes are of great commercial interest, as the strength

of the gold–cyanide bond is exploited in hydrometallurgical gold mining for aqueous extraction ofarranged in tetrahedral or octahedral forms with strong bonding energy existing between the heavymetal atom and the cyanide ligand [1] Because they can only dissociate under strongly acidic

pH conditions (pH < 2), they are referred to as strong acid dissociable complexes, or simply as

strongly-complexed cyanide [15]

As some of these species are formed very slowly [1,33], it is difficult to determine the equilibriumformation constants Formation data determined by direct thermodynamic methods are available onlyfor complexes of gold(I), and palladium(II) [1] For other metals, like iron, electron transfer betweencomplex ions of the element in the+2 and +3 oxidation states is rapid enough such that the ratio of

the formation constants can be determined from measurement of redox potentials This ratio can then

be combined with standard enthalpy and entropy change measurements for the formation reaction

of interest

gold from ores (seeChapters 4and26) The metal–cyanide bonds in these complexes are mostly

Source: Data from Sehmel, G.A., Cyanide and antimony thermodynamic

data-base for the aqueous species and solids for the EPA-MINTEQ geochemical code,

PNL-6835, Pacific Northwest Laboratory, Richland, WA, 1989, (Table 5).

Table 5.4 lists the equilibrium constants for the reversible formation of iron–cyanide complexes,which are of primary interest with respect to cyanide in the environment The constants reported arefrom the compilation by Sehmel [5], which was selected for the reasons discussed in Section 5.2.1.1.Among all the iron–cyanide complexes, the most commonly occurring are ferrocyanide, Fe(CN)4 −

6 ,where iron is the+2 oxidation state, and ferricyanide, Fe(CN)3 −

6 , where iron is in the+3 oxidation

state Another iron–cyanide complex only recently identified and not presented in Table 5.4, with

a chemical formula, Fe(CN)5NHCH4−

3 , has been found to dominate groundwater at many formermanufactured gas plant sites [39]