1 stp = Sewage Treatment Plant
n.a. = negligible amount.

Impurities of the dyes as well as decomposition by reductive cleavage of the azo dyes may result in transformation of the azo dyes to the degradation products/metabolites (aromatic amines), of which some are potentially carcinogenic. Estimation of the decomposition of azo dyes in the environment may be derived from knowledge of the structural and molecular composition of the azo dyes and of a stoichiometric equation.

The environmental exposure routes of the aromatic amines are essentially the same as the ones described for the dyes.

Degradation

An important natural abiotic degradation mechanism is photolysis and hydrolysis as a function of pH in the range of pH 4-9 (ETAD, 1992a).

The evidence of the role of hydrolysis in degradation of azo dyes is not conclusive. Hydrolysis is by Baughman and Perenich (1988b) not considered to be important. If the dye is not broken during rigors of biological waste treatment, it is unlikely to degrade rapidly in the less severe conditions of the environment. This is supported by Clarke and Anliker (1980), who states that the reductive cleavage of the azo-bond is the major degradation pathway for azo dyes.

For the reactive dyes the abiotic half-life due to hydrolysis is approximately 2 days (IUCLID).

Photo-reduction of azo dyes to hydrazines and amines is possible, but it is likely to be very slow, except in oxygen-poor water. The stability of the dyes to visible and UV-light is very high, and therefore only slow degradation has been shown (Clarke & Anliker, 1980).

The photo-stability of azo dyestuffs is high in pure water but in the pre- sence of natural humic materials, the photo decomposition is strongly accelerated, probably through oxidation by single oxygen or oxy-radicals (Brown & Anliker, 1988).

Shu et al. (1994) demonstrated photo-oxidation (UV/H₂O_2- -photo chemical reactor) of two non-biodegradable azo dyes in waste water (Acid Red 1* and Acid Yellow 23*). It was observed that the decomposition of both azo dyes was pseudo-first order reactions with respect to azo dye concentrations. The reaction rates were dependent on the pH, the initial dye concentration and the hydrogen peroxide dosage, e.g. with high concentrations of H₂O₂ (18.95 mM) the half-life was 6 minutes for Acid Red 1* (20 ppm).

Other advanced oxidation processes include Fenton’s reagent and TiO₂ photo-oxidation (Shu et al., 1994). A feasibility study by Dieckmann et al. (1994) indicates that azo dyes (Solvent red 1 and 4-hydroxyazoben-zene) can be degraded via sensitised photocatalysis on a surface of TiO₂.

Shu and Huang (1995) investigated 8 acidic azo dyes for degradation of UV/Ozone. They found that the degradation rate were of the first order with respect to both azo dyes and ozone concentrations. UV-light did not significantly enhance the degradation ability. The half-lives were in the range of 1.2 to 2.6 minutes.

It is assumed that the main abiotic removal mechanism for dyestuffs in wastewater treatment plants is adsorption of sludge. However, other effects like sedimentation, precipitation or flocculation may also play a role (Pagga & Taeger, 1994).

Anionic dyes may be expected to react with Ca, Mg etc. to form highly insoluble salts (i.e. pigments) and thereby reduce the concentration, which is available for other reactions or biological effects (Baughman & Perenich, 1988b).

Other physico-chemical processes are flocculation, flotation, membrane filtration, electrokinetic coagulation, electrochemical destruction, ion-exchange, chemical oxidation and different sorption techniques. A review of the different treatment technologies and techniques and their efficiency towards degradation of xenobiotics has been given by Matsumoto et al. (1995). However as Banat et al. (1996) conclude, not one specific treatment process seems to be able to handle decolourisation of all textile waste waters. Generally, a customised process, which probably involves a combination of different methods, will be more applicable. Ozonation has achieved the greatest practical importance for removal of colours, but also precipitation and flocculation procedures have given good results. Decolourisation with reductive agents such as hydrosulphite is a workable proposition (Clarke & Anliker, 1980).

Some of the aromatic amines may be susceptible to photolysis, e.g.

4-methyl-m-phenylamine (HSDB, 1998).

Hydrolysis is, generally, not an important route of degradation of the aromatic amines (HSDB, 1998).

Even though the dyes have absorption maxima in the range of visible and UV-light, photo-reduction does not play a dominant role in the environmental fate of dyes, although its contribution to the total mineralisation of widely dispersed trace amounts may be underestimated. Furthermore, hydrolysis seems not to be an important degradation pathway either, except for reactive dyes, which are hydrolysed rapidly in aqueous solution.

For the metabolites, photolysis may be of some importance, whereas hydrolysis not seems to be an important degradation route.

Razo-Flores et al. (1997a) estimate that due to the recalcitrance of azo dyes in aerobic environments, the azo dyes eventually end up in anaerobic sediments, shallow aquifers and in groundwater.

Extensive tests indicate that dyes are generally adsorbed to the extent of 40-80% by the biomass and are thus partly removed from the water phase in sewage treatment plants. They are, however, not biodegraded at this stage to any significant extent (Clarke & Anliker, 1980).

Dyes to be useful must possess a high degree of chemical and photolytic stability which implies that removal of dyes from effluents is difficult. Stability against microbial attack is also a required feature of azo dyes (Pagga & Brown, 1986). Subsequently, they are less amenable to biodegradation (Banat et al., 1996). It is thus unlikely that they, in general, will give positive results in short-term tests (e.g. OECD) for aerobic biodegradability (Brown & Anliker, 1988).

Furthermore, the electron-withdrawal character of azo-groups generates electron deficiency and thus makes the compounds less susceptible to oxidative catabolism. As a consequence, many of these chemicals tend to persist under aerobic environmental conditions (Knackmuss, 1996).

Biodegradation of azo dyes can occur in both aerobic and anaerobic environments. In both cases, the initial step in the biodegradation is the reductive cleavage of the azo-bond. Under aerobic conditions the initial step of cleavage of the azo-bond is typically followed by hydroxylation and ring opening of the aromatic intermediates (Zissi & Lyberatos, 1996).

Permeability through the cell wall has often been found to be the rate-limiting step in the reduction process. The microbial reduction of azo dyestuffs are either by reduction of living cells or by cellular extracts (Brown & Anliker, 1988).

Anaerobic and aerobic metabolic activities are a prerequisite for the complete biodegradation of recalcitrant aromatic pollutants, which contain electron-withdrawal substituents, such as azo dyes. Therefore, the recalcitrant nature of azo dyes can be overcome by utilising anaerobic-aerobic co-cultures (Field et al., 1995). This is supported by Clarke and Anliker (1980), who furthermore state that physical and chemical treatment is required as well. With the possible exception of basic dyes, the biological treatment processes (activated sludge) have in most cases proved to be insufficient for removal of dyestuffs from waste waters (Clarke & Anliker, 1980).

Bacteria - anaerobic

Brown and Laboureur (1983b) investigated the primary biodegradation of 13 azo dyes in an anaerobic sludge inoculum. The dyes were selected as commercially significant and represented both monoazo, diazo and polyazo dyes. The monoazo dyes were Mordant Blue 13, Mordant Black 9, Basic Red 18, Acid Yellow 151 and the diazos Direct Red 7, Acid Red 114*, Direct Blue 15*, Direct Yellow 12, Reactive Black 5 and Acid Blue 113*. All of these were substantially biodegraded (75-94%), whereas the polyazos Direct Black 19 and Direct Black 22 were only decolourised between 51-61% in a time period of 0 to 42 days.

Later results by Brown and Hamburger (1987) have confirmed that azo dyes are likely to undergo primary biodegradation in an anaerobic environment. The decolourisation was more than 90% in the time range of 0 to 56 days. The dyes tested were Acid Orange 7*, Acid Yellow 25*, Acid Yellow 36*, Acid Yellow 151, Acid Red 114*, Acid Black 24, Direct Red 7, Direct Blue 14*, Direct Blue 15*, Direct Yellow 12, Direct Yellow 50*, Mordant Black 9 and Mordant Black 11. This was also confirmed by Boethling et al. (1989) for Direct Red 28*.

Shaul et al. (1991) also found evidence of biodegradation of Acid Orange 7*, Acid Orange 8 and Acid Red 88. In 24 hours, 81 to 86% were degraded. The presence of sulfo groups on the aromatic component of some azo dyes seemed to inhibit the biodegradability significantly.

Direct dyes (Direct Red 28*, Direct Blue 1* and Direct Blue 14*) are degraded with more than 90% in anaerobic sediment-water systems with half-lives ranging from 2 to 16 days. The degradation is inhibited when the dyes are strongly bound to the sediment (Weber, 1991)

In sediments, Yen et al. (1991) showed that the degradation of two disperse azo dyes (Disperse Red 1 and Disperse Red 5) had half-lives within hours when the concentrations were kept below 10 ppm in the sediment. The reduction of nitro groups to amino groups and/or cleavage of the azo groups to give nitroanlines were found to be major pathways.

Zissi and Lyberatos (1996) demonstrated that Bacillus subtilis is, at least partly, able to degrade the disperse azo dye p-aminobenzene under anoxic conditions growing in a batch-reactor. The results proved that Bacillus subtilis co-metabolises p-aminobenzene under denitrifying conditions in the presence of glucose as a carbon source, producing aniline and p-phenyldiamine, as the N=N double bond is broken.

Other authors have reported degradation of disperse dyes with half-lives in order of minutes as well, e.g. Disperse Blue 79* (Weber & Adams, 1995; Freeman et al., 1996) and Disperse Red 1 with a half-life of less than 8 hours (Baughman & Weber, 1994).

The non-ionic dye Solvent Red 1 has been reported to have a half-life of 2.2 to 4 days (Baughman & Weber, 1994).

The reduction of benzidine azo dyes to free benzidine by soil bacteria has been reported for four aminobenzene azo dyes. The soil bacteria are Pseudomonas cepacia and Pseudomonas sp.. The initial reaction was azo reduction and cleavage, followed by acetylation and aromatic ring hydroxylation. The azo dyes were reduced with 42 to 91% at an aqueous concentration of 5 to 30 ppm during 24 hours of incubation. Similarly, a Plesiomonas bacterial species isolated from textile waste water has shown to degrade 5 different azo dyes under anaerobic conditions. Mixtures of sewage and soil bacteria (e.g. Pseudomonas aeuginosa) may also effectively degrade azo dyes. The dyes undergo azo-bond cleavage followed by carboxylation, hydroxylation and acetylation metabolism of the initial aromatic amine azo-reduction metabolites (Brown & DeVito, 1993).

Examples of removal of dyes in use in Denmark under anaerobic conditions are summarised in Table 5.5 below.

Fjernelse af azofarver anvendt i Danmark under anaerobe forhold.

Bacteria - aerobic

Like dyes in general, the hydrolysed dyes are practically not biodegraded in the short retention time of the aerobic treatment processes. Most dyes are degraded under anaerobic conditions. Such conditions are met in the anaerobic digestion process at sewage treatment plants, and in sediments and soils (ETAD, 1991).

Pagga & Brown (1986) tested 87 dyes in a short-term aerobic biodegradation based on the OECD Guideline for a static test method with activated sludge. They found no significant biodegradation, but substantial colour removal was observed which was attributed to the elimination of the dyes by adsorption. The tested dyes represented all the ionic characters and chemical types.

A study by Zhang et al. (1995) revealed that Acid Orange 7* and Acid Orange 8 can be degraded aerobically in a rotating drum biofilm reactor. The more complex Acid Orange 10* and Acid Red 14*, however, were not aerobically degraded. However, the authors demonstrated that cleavage of the azo bond occurred easily under anaerobic/anoxic biofilm conditions.

Knackmuss (1996) suggests that a total biodegradation of azo dyes may be accomplished by bacteria, harbouring a highly efficient uptake and an azo reductase system which are used in a two-step anaerobic/aerobic process, at least with regards to biodegradation of sulphonated naphthalenes.

Fungi

Microbial degradation of lignin-containing pulp and paper waste water has been demonstrated by several authors, especially with the white-rot Basidiomycete fungus: Phanerochaete chrysoporium. The mechanism of colour removal involves lignin peroxidase and Mn-dependent peroxidase or laccase enzymes. The degradation of azo dyes is apparently dependent on the availability of nitrogen. If there is a high concentration of N the degradation rate decreases. Banat et al. (1996) have reviewed the literature and found out that azo dyes may be degraded by the fungus between 23 and 90% in a time span of 3 to 21 days with different concentrations. A wide variety of dyes has been tested, among them Acid Red 114*, Acid Red 88, Direct Blue 15*, Disperse Yellow 3, Disperse Orange 3 and Solvent Yellow 14* (Spadaro et al., 1992). In addition, other anionic dyes, like Reactive Orange 96, Reactive Yellow 5 and Reactive Black 5, have been demonstrated to be biodegraded by the white-rot fungus Phanerochaete chrysosporium, too (Heinfling et al., 1997).

The actinomycete strains, mainly streptomycetes, isolated from soil samples have been demonstrated to decolourise effluents containing different types of reactive dyes. In a study carried out by Zhou and Zimmermann (1993) it was concluded that the decolourisation of Reactive Red 147 was due to adsorption rather than biodegradation. Banat et al. (1996) has reviewed the studies of other fungal biodegradation of azo dyes and several (7 in total) other species have shown to decolourise but mainly by way of adsorption.

There are conflicting evidence of the influence of the substituents on the aromatic ring with regards to the effect on biodegradability by Phanerochaete chrysoporium. Paszczynski et al. (1992) found that substitution with sulfo groups on the aromatic component of some azo dyes not seemed to affect the biodegradability of the anionic azo dyes significantly. Pasti-Grigsby et al. (1992), however, found that significant degradation of the azobenzene derivative dyes and naphtol-derivative dyes (e.g. Acid yellow 9 and Acid Orange 12 (anionic)) occurred solely when the hydroxy group was in a specific position relative to the azo linkage. Spadaro et al. (1992) showed that when the aromatic rings of the neutral dyes (Solvent Yellow 14*, Disperse Orange 3 and Disperse Yellow 3) had substituted hydroxyl, amino, acetamido or nitro groups, the mineralisation was greater than by those with unsubstituted rings.

Algae

Decolourising with algal cultures has been found by Jinqi and Houtian (1992). The reduction of algae resembles that of the bacteria. The azo reductase of the algae Chlorella and Oscillatoria is responsible for degrading azo dyes into aromatic amines. The aromatic amine is then subject to further degradation by the algae. As for bacteria, azo compounds with a hydroxy or an amino group are most likely to be readily degraded than those with a methyl, methoxy, sulfo or a nitro group.

Mineralisation

FitzGerald and Bishop (1995) found an almost total decolourisation in the first stage of an anaerobic/aerobic treatment of sulphonated azo dyes (Acid Orange 10*, Acid Red 14* and Acid Red 189). Analyses of the intermediates at the first and second stages (aerobic) showed virtually no concentration of intermediates, which may indicate a total anaerobic mineralisation. In contrast, Seshadri et al. (1994) found that the aromatic amines remained undegraded in an anaerobic fluidised bed reactor.

Razo-Flores et al. (1997b) have demonstrated that Mordant Orange 1 may be completely degraded (mineralised) in a continuous upward-flow anaerobic sludge bed reactor in the presence of co-substrates.

Razo-Flores et al. (1997a) have further demonstrated that the azo dye, azodisalicylate is completely biodegradable in the absence of oxygen. The dye is mineralised in an adapted methanogenic consortium to CH₄ and NH₃ in both batch assays and continuous bioreactors.

Degradation of metabolites

Free aromatic amines are generally susceptible to environmental degradation (Brown & DeVito, 1993). Zerbinati et al. (1997) have found that naphthalenesulfonates can undergo oxidative degradation under physico-chemical conditions similar to those occurring in a river. However, other studies have shown that, e.g. benzidine is bound with the humic acid fraction of the soil (Weber, 1991).

Brown and Laboureur (1983a) showed in aerobic biodegradation tests that the four aromatic amines: aniline, p-anisidine, p-phenetidine and

o-toluidine are ready biodegradable and that both o-anisidine and

3,3´-dichlorobenzidine are inherent biodegradable in accordance with the OECD test guidelines. Brown and Hamburger (1987) confirmed these results for the lipophilic aromatic primary amines, but depending on their precise structure, some sulphonated aromatic amines may not be degradable.

Under aerobic conditions another type of recalcitrance can be recognised, namely, the tendency of certain compounds, susceptible to free radical reactions, to undergo oxidative coupling. These coupling reactions can result in the formation of recalcitrant humic-like polymers or in irreversible covalent binding of the pollutant into the soil humus. Aromatic amines and nitroaromatics are susceptible to these polymerisation reactions. Formation of azo compounds by oxidative coupling has been demonstrated in aerobic enrichment cultures from the aromatic amines (Field et al., 1995)

The metabolites of aromatic primary amines are not rapidly degraded under anaerobic conditions (Brown & Hamburger, 1987). Electron donating amino groups are expected to pose a serious problem to further reductive biotransformations by anaerobes. However, there is evidence for anaerobic aniline biodegradation by sulphate reducing bacteria and in mixed cultures under denitrifying conditions. Aniline degradation by a methanogenic consortium has also been claimed. Aromatic amines with carboxy, hydroxy and methoxy substituents are potentially mineralisable under methanogenic conditions (Field et al., 1995). Another example is o-toluidine, which is not degraded under anaerobic conditions (HSDB, 1998).

Summary

Various microbial species, i.e. fungi, bacteria and algae may be able to biodegrade azo dyes in an anaerobic environment. Total mineralisation or further degradation of the metabolites may predominantly take place in an aerobic environment.

The universal degradation route seems to be initial reductive cleavage of the azo bond followed by e.g. acetylation, carboxylation and aromatic ring hydrolysation.

The rate limiting step for bacterial degradation is the uptake across cell membranes for intracellular reduction, whereas some fungi may degrade the dyes extracellularly.

The substituents and the substitutional pattern may also significantly influence the biodegradability. The reported effects are contradicting, but the ionic azo dyes with hydroxy or amino groups are most likely to be readily biodegraded, compared to those with methyl, methoxy, sulfo or nitro groups. For the non-ionic dyes (disperse, solvent and mordant) an enhanced biodegradation is observed with hydroxyl, amino, acetamido or nitro groups compared to unsubstituted rings.

It is difficult to generalise about degradation rates and the degree of removal for specific azo dyes or for the different chemical classes based on the findings in the literature, because the experimental conditions vary.

However, biodegradation of azo dyes varies, in general, from hours to several months or more depending on, among other things, the physico-chemical properties of the dyes. The molecular size of the azo dyes, especially solvent and disperse dyes, may reduce the rate and probability of biodegradation. This is due to limited uptake possibilities, and the substituents may also influence the degradation rate.

The metabolites are primarily biodegraded under aerobic conditions. Some of the metabolites are ready biodegradable, and some of the sulphonated aromatic amines may not be degradable.

Distribution

Data concerning the volatilisation of azo dyes from aqueous surfaces are not available. With respect to volatilisation it is prerequisite to distinguish between the ionic dyes and non-ionic dyes, because ionic compounds are generally non-volatile. (Brown & Hamburger, 1987). Therefore, volatilisation will not be important for acid, direct, basic and reactive dyes. In principle, the solvent and disperse dyes have the potential to be volatile, but as they are large, complex molecules they can be expected to have low vapour pressures. Another reason for volatilisation to be unlikely for the uncharged dyes is that the escaping tendency or fugacity, which drives volatilisation, is also the driving force for both sorption and bioconcentration (Baughman & Perenich, 1988b).

Baughman and Perenich (1988a) calculated Henry’s law constants from solubility and vapour pressure. The values show that the disperse dyes will be entirely vapour-phase controlled in the environment in their rate of volatilisation from water and that this process is extremely slow. The vapour pressures lie in the range of 2 ´ 10^-14 to 1 ´ 10^-6 mmHg and the solubility in the range of 2 ´ 10^-9 to 4.5 ´ 10^-6 mol/l. The Henry law constant is on average 10^-10 atm ´ m³/mol for disperse dyes.

In general, the metabolites, i.e. the 22 potentially carcinogenic aromatic amines, show moderate to low volatilisation with Henry’s law constants in the range of 4.7 ´ 10^-11 (3,3´-dimethoxybenzidine) to 2.0 ´ 10^-6 atm ´ m³/mol (o-toluidine).

Due to the chemical characteristics of the azo dyes, volatilisation from surfaces of either water or soil (wet or dry) is considered to be insignificant for both ionic and neutral (non-ionic dyes). This applies for the metabolites, too.

Adsorption

Because of dyestuffs inherent high affinity to substrates, they are adsorbed onto the sludge during sewage treatment and are thus removed from the final treated effluent (Anliker, 1986). But due to the chemical composition of some of the dyes, they may pass the sewage treatment unaffected and thus end up in the aquatic environment. Extensive testing indicates that dyestuffs are generally adsorbed to the extent of 40-80% by the biomass and are thus partially removed in sewage treatment plants (Clarke & Anliker, 1980). However, due to their relatively low affinity to substrates, the removal of the hydrolysed dyes (e.g. Reactive dyes) by adsorption onto the sewage sludge is only in the range of 0-30% (ETAD, 1991).

In the practical concentration range of 10 to 50 mg dye/l, there is an almost linear relationship between the concentration in solution and the amount adsorbed. The adsorptive capacity of activated sludge for dyes investigated was, in neutral media, in the range 0.01 to 4% of dyestuff on dry weight sludge (Clarke & Anliker, 1980).

The chemical properties and substitutional pattern of the chemical structure of the dyes and the composition of the waste water influences the degree of adsorption. The adsorption depends on the pH, salinity and the concentration and nature of organic contents.

Based on the properties of sediments, cation exchange is anticipated to be extensive and rapid for the basic dyes. A similar situation should exist for the anionic acid and direct dye, but the equilibrium constants would probably be much smaller (Baughman & Perenich, 1988b).

Shaul et al. (1991) investigated the partitioning of water-soluble azo dyes in the activated sludge process. A total of 18 dyes were tested and categorised according to their behaviour in the tests (Table 5.6). For Group 1 it was concluded that the high degree of sulphonation enhanced their water solubility and limited their ability to adsorb onto the biomass. Although the dyes in Group 2 were highly sulphonated, their greater molecular size was thought to account for their greater degree of adsorption.

Weber (1991) has demonstrated that the sorption of several weakly basic benzidine-based dyes (Direct Red 28* (disulphonated) and Direct Blue 14* (tetrasulfonated)) strongly depend on the pH and the nature and concentration of inorganic salt in solution in an anaerobic sediment-water system. Sorption is strongly favoured with decreasing pH and increasing salt concentration. The sorption was enhanced especially for Direct Red 28*, which was less substituted.

Pagga and Taeger (1994) have found that the colour elimination of acid and disperse azo dyes (Acid Orange 7*, Acid red 88, Disperse Orange 29 and Disperse Yellow 5) depends on the hardness of the water. A high concentration of calcium ions favours adsorption as well as flocculation or precipitation processes or a better settling of the sludge and less turbidity.

In a study by Yen et al. (1990), it has been shown that newer disperse dyes show a higher degree of partitioning into the sediment than older disperse dyes based on calculated sediment concentrations.

The metabolites adsorb, except for 4-methyl-m-phenylenediamine, mode- rately to strongly onto sediments and soil. 4-methyl-m-phenylenediamine

does not adsorb to any significant degree (HSDB, 1998).

The removal of various dyes from different classes has been studied and the removal pattern may be summarised as shown in Table 5.7.

Extensive adsorption onto soil and sediment has been demonstrated in several experiments. It is concluded that adsorption is the major route of removal of dyes in the environment. Adsorption is an important removal pathway for the metabolites, as well.

A high degree of solubility and sulphonation reduces adsorption, whereas increasing molecular size, hardness of the water and salinity favour sorption. This applies for a decreasing pH, as well.

Bioaccumulation

The obtained data on bioaccumulation are primarily derived from fish-

tests.

The uptake rates are influenced by the partition coefficient (log K_ow) (Erickson & McKim, 1990). Other factors may be of primary importance for the uptake as well, e.g. diffusional resistance, molecular size, respiratory volume and gill perfusion (Niimi et al., 1989).

The elimination rates for hydrophobic chemicals are low. For hydrophobic chemicals it has often been shown that uptake and clearance between fish and water is a first-order exchange process (Van Hoogen & Opperhuizen, 1988).

Anliker et al. (1981) have presented, estimated and experimentally assessed the log K_ow and have experimentally assessed the log BCF (bio- concentration factor) in fish (MITI standard) for 50 azo dyes, representing both the ionic forms and the neutral forms. The average values for the different dyes are presented in Table 5.8 below.

Fordelingskoefficienter og den modsvarende målte bioakkumulationsfaktor for 50 azofarvestoffer fordelt på 5 kemiske (tekniske) grupper.

The survey shows, with a few exceptions, that the very hydrophilic (ionic) dyes have a log BCF of - 1 to 1, although from the log K_ow lower log BCFs may have been predicted. This is explained by the adherence of dyes to the outside of the fish or to the intestine. None of the dyestuffs bearing at least one charged group has showed a log BCF larger than 1. It has been demonstrated that disperse dyes do not bioaccumulate in fish even though their log K_ow values were larger than 3. The molecular weight was relatively high, between 450 to 550 g/mol, making the transport across membranes difficult.

These findings have been confirmed in other studies. The partition coefficients of 21 reactive dyes were very low (log K_ow < 0) and none of these dyes have showed any tendency of bioaccumulation in the flow-through tests (MITI-standard) in the carp. (ETAD, 1991). In Carrassius sp., which was exposed to 2 mg/l and 0.2 mg/l for 42 days, the BCFs were less than 1.1 and less than 11, respectively (IUCLID).

ICI has carried out eight-weeks accumulation studies on the Carp (Cypri- nus carpio) (MITI standard). The results indicate that neither the 30 water-soluble² nor the 12 disperse dyes² with exposure levels up to 10 mg/l were accumulated. For the soluble dyes the accumulation factor was below the detection limit. The low accumulation of the soluble dyes may be expected, and the low accumulation factors found for the disperse dyes may be due to their relatively high molecular weight (typically 300-500) or because their absolute fat solubility is relatively low (Brown, 1987).

Anliker and Moser (1987) have investigated the melting point, the log K_ow, solubility in water and n-octanol and the log BCF in fish for 8 disperse dyes (nitroazobenzene and phenylazopyridone types):

molar weight:	360 to 546 g/mol
melting point:	117 to 225 0C
solubility in water:	n.d.
solubility in n-octanol:	81 to 2,430 mg/l
log Kow:	2.5 to 5.4 (majority above 3)
log BCF (exp.):	0.3 to 1.76

They found that the high K_ow suggested strong bioaccumulation tendencies, but the bioaccumulation was below 100. It was hypothesised that this behaviour may be due to their pronounced aggregation tendency, making transport across membranes difficult. The findings of Opperhuizen et al. (1985) support this. Their results indicated that for extremely hydrophobic chemicals with an effective cross section over 9.5 Å, a lack of uptake into biota (fish) can be expected, as membrane permeation seems practically impossible.

A study of 75 disperse dyes, even highly lipophilic ones, by Anliker (1986) and a later study by Anliker et al. (1988) on 23 disperse dyes³, including highly lipophilic ones, confirmed the above mentioned observations.

Similar results have been reported for the BCFs of chlorinated aromatic amines in guppies (Poecilia reticulata). The experimentally demonstrated values of BCF are significantly smaller that the calculated values of BCF (Wolf et al., 1992).

The azo compound (not dye) 3,3´,4,4´-tetrachloroazobenzene (TCAB), a common contaminant from 3,4 dichloroaniline based herbicides and of agricultural soils, has been tested (short-term) on the aquatic snail Indoplanorbis exustus by Allison and Morita (1995a). They found that even at detrital exposures of 2,500 ppm, the maximum level only reached 287 ppb (whole body basis). The authors above (1995b) also came to the same result in the Japanese Medaka (Oryzias latipes). The fish inhabit still waters and paddy fields and their drains. The study showed that TCAB is bioadsorbed and to some extent bioaccumulated in the fish. The contaminant was administered through the food.

Apart from the above stated findings, only a small amount of data was found in the literature and databases on the log K_ow and the log BCF for specific azo dyes included in the present survey. The results of the literature study are presented in Table 5.9.

Fordelingskoefficient og biokoncentrationsfaktor for azofarvestoffer anvendt i Danmark.

Compared to the findings of Anliker et al. (1981) shown in Table 5.8, the log BCF for Acid Red 114* is above the range reported, whereas Acid Yellow 23* is in agreement. The remaining dyes are incomparable, as they are based on estimates rather than actual experimentally measured values of BCF. The solvent dyes were not included in the Anliker et al. (1981) study.

However, the calculated log BCF value for Disperse Blue 79* is in agreement with the findings of Anliker et al. (1981), whereas Direct Black 38* is twofold higher than reported by the above authors.

Metabolites

Both measured and estimated log BCFs for the cleavage products of the dyes, i.e. the 22 potentially carcinogenic aromatic amines, are according to ECDIN and HSDB (1998) below 3. The highest estimated log BCF has been found for 4-o-tolylazo-o-toluidine (2.75) (HSDB, 1998). The majority lies in the range of 1.5 to 2.0. The lowest values (<1.47) have been reported for o-anisidine (0.85), o-toluidine (1.2) and 4,4´-methylenedi-aniline (1.1). The log BCF values for the aromatic amines indicate that there is a risk of biomagnification for a great majority of the metabolites.

Summary

For the compounds with log BCFs larger than 3, there is a high risk of bioaccumulation, whereas for compounds with log BCFs between 1.47 and 3, the risks of biomagnification and secondary poisoning are important. For compounds with log BCF values below 1.47, there is no immediate concern with regard to bioaccumulation (Franke et al., 1994).

When looking at the values of the dyes included in the present survey, it is indicated that Acid Red 114 may bioaccumulate in fish, whereas the remaining ionic dyes do not seem to have any significant bioaccumulation potential. However, the estimated log BCFs for the non-ionic dyes, i.e. disperse and solvent, indicate a potential risk of bioaccumulation.

The estimated values for log BCF are generally to high which several authors have found. Therefore, the evidence of the risk bioaccumulation of the non-ionic azo dyes must be further validated, taking the potential barrier of uptake into account, as a result of the high molecular size of these compounds.

Generally, the cleavage products of the azo dyes, i.e. the aromatic amines, have a potential for bioaccumulation, too.

Aquatic compartment

Monitoring data

Only a few monitoring studies of environmental levels of dyes have been found, and data from Denmark have not been obtained.

In a study conducted by the US EPA, effluents from 25 textile industries were measured. The average TOC was measured to 276 mg/l (range 55 to 1,120 mg/l). The dyestuff itself has not been measured, but it is estimated that the dye contributes between 2 and 10% of the TOC and COD indicating worst case levels of 5.5 to 112 mg/l. However, the typical dye concentration lies in the range of 10 to 50 mg/l. Decolourised effluents contain less that 1 mg/l dye, and the TOC contribution of dyestuff following the primary and biological treatment stages is normally considerably less than 0.5 mg/l. In the same study, effluents from a tannery were measured, and the raw effluent contained 22 to 56 ppm dyestuff (Clarke & Anliker, 1980).

The following Table 5.10 and Table 5.11 summarise additional data regarding environmental monitoring of dyes in water and sediment.

Moniteringsdata for farvestofskoncentrationer i vand.

Moniteringsdata for farvestofskoncentrationer i sediment.

Estimation of PEC for the aquatic compartment

In most industrialised countries only about 20% or less of the release from processes will reach open water due to effective adsorption in the primary and the biologic treatment stages (Clarke & Anliker, 1980).

However, in the present calculation of PEC _{effluent, stp}, two scenarios will be presented.

The estimation of PEC _{effluent, stp} is based on the following assumptions:

The processing industries do not treat waste water in agreement with TGD (1996).

Between 40 and 80% of the azo dyes are adsorbed in the sewage treatment plant (STP) (Clarke & Anliker, 1980). Resulting in a worst case scenario of an adsorption of 40% (60% release to the effluent) and a best case scenario of an adsorption of 80% (20% release to the effluent).

Adsorption is the only removal route of azo dyes in the STP, i.e. there is no abiotic or biotic degradation.

Furthermore, a standard STP scenario, in compliance with TGD (1996), is used. According to this standard the values presented in Table 5.12 are standard characteristics of a STP:

Standardkarakteristika for et rensningsanlæg.

Ref.: TGD (1996).

The calculation of PEC _{influent, stp} is simplified and based on the equation below:

PEC _{influent, stp} = Release _wastewater /Waste _{inhab ´} Capacity _{stp ´} 365

The calculation of PEC _{effluent, stp} is simplified and based on the assumptions mentioned above. In addition, the PEC _{effluent, stp} for the processing industry is corrected for the number of sites present in Denmark , i.e. 40 sites for textile colouring and 1 site for leather dyeing. For the use, the number of inhabitants in Denmark (approximately 5 millions) is normalised to the capacity_stp.

PEC _{effluent, stp} = PEC _{influent, stp} ´ (1- adsorption factor/ (number of sites) or inhabitants in Denmark (Table 5.13).

PEC _{surface water} = PEC _{effluent, stp ´} dilution factor

According to the TGD (1996), the dilution factor is 10.

Table 5.13

Estimeret PEC_{udløb, stp} og PEC_{overfladevand} for azofarvestoffer.

The PEC_sediment is calculated from:

PEC_sediment = PEC_{surface water}  adsorption factor

In Table 5.14, the PEC_sediment is presented.

Estimeret PEC_sediment for azofarvestoffer.

Concerning the concentration of azo dyes in the sludge, the estimation is based on that the production of sludge amounts to 170,000 tonnes dw/year in Denmark (Miljøstyrelsen, 1996b). The "worst case" of adsorbed azo dyes onto the sludge is 80% and the "best case" is 40%. The calculated concentration in sludge is based on the following equation:

PEC _sludge = (Release ´ adsorption factor ´ 10⁶/ Sludge rate)/(number of sites) or inhabitants.

Sludge rate = 170.000 tonnes/year

Estimeret PEC_slam for azofarvestoffer.

The estimated PEC_{surface water} for processing and use is in the range of 0.04 to 1.44 mg/l. According to the monitoring studies (Table 5.10) a range of 0.012 to 0.523 mg/l for treated sewage effluent has been found. If comparing the two, the estimated PEC_{effluent, stp} is approximately 3 times higher. Compared to the concentrations found in river water, the estimated PEC_{surface water} is 2 to 6 times higher which may be due to the dilution effect. The estimated PEC_sediment is, on the other hand, below the range of the monitored data (Table 5.11), namely 0.002 to 0.11 mg/kg dw.

Due to the lack of monitoring data of environmental concentrations of azo dyes in Denmark, it is not possible to validate the estimated PECs based on Danish data. The basic assumption, however, that the processing industries do not carry out waste water treatment prior to outlet (PEC_{influent, stp}) is unlikely, because most of these companies, if not all, are encompassed by a special section of the Danish Environmental Protection Law (chapter 5). Hence, their emissions are restricted and must be approved by the authorities. Subsequently, the companies are obliged to have some degree of waste water treatment prior to the outlet to the municipal STP.

Assuming that 40 to 80% of the dyes are removed from waste water before the outlet from the industry and likewise in the STP, this indicates that the actual PEC_{effluent, stp} for the processing and use phase is more likely to be in the range of 0.024 to 0.864 mg/l and the PEC_{surface water} in the range of 0.002 to 0.086 mg/l. These concentrations are within the same range, and for PEC_{surface water} approximately 4 times higher compared to the findings in the aforementioned monitoring studies. This indicates at least for the best case scenario, that the estimated PECs may be realistic.

If it is estimated that the PEC_{surface water} is too high, then the PEC_sediment has to be reduced in the same order of magnitude. Resulting in a concentration of 0.001 to 0.090 mg/kg from processing and use which is within the low range of the monitoring studies (Table 5.11).

If it is assumed that the companies carry out waste water treatment, the PEC_{sludge, stp} may also be reduced 2 to 5 times, depending on the degree of adsorption (40-80%) at the companies, and this results in a range of 1.18 to 5.62 mg/kg for processing and use.

Atmosphere

No data have been obtained concerning monitoring of azo dyes in the atmosphere or bound to particulate matter.

It has not been attempted to calculate the atmospheric PEC, but it is estimated that the PEC is very low, because volatilisation is highly unlikely for the azo dyes from both moist and dry surfaces. Furthermore, the release from the processing industry and from incineration is considered to be very low (approximately equal to 0).

Terrestrial compartment

Monitoring data

There are no direct route by which agricultural soils may become contaminated with synthetic dyes. In principle, it is possible that the disposal of sludge from sewage treatment plants, which receive dye-house effluents, may provide an indirect route of exposure. Although, there appear to be no reported data of the levels of dyestuffs on agricultural soils, estimates based on the principles elaborated by the OECD, would indicate a worst case level of 1 mg/kg (w/w of dry soil) (Brown & Anliker, 1988).

Furthermore, deposition of particulate matter may be a potential pathway for the terrestrial environment, but as stated above it is considered to be an unlikely pathway.

A practical demonstration has showed that sewage sludge contaminated with dyes, when held under simulated landfill conditions, does not release dyes into the leachate. The amine metabolites, which may be expected to be produced from these dyes, cannot be found in the leachate or interstitial water either (Brown & Anliker, 1988)

However, no data have been obtained on terrestrial monitoring of azo dyes.

Estimation of PEC for the terrestrial compartment

The sources of environmental releases of azo dyes in the terrestrial environment are waste disposal in landfills and sludge applied to agricultural soil.

It is estimated that the total amount of sludge per year in Denmark is 170,000 tonnes of dry weight. About 114,000 tonnes (67%) are used in agriculture and 20,000 tonnes (12%) are deposited in landfills. The rest (21%) is incinerated (Miljøstyrelsen, 1996b).

It is not known how many hectares of agricultural soil which are fertilised with sludge in Denmark. But according to the TGD (1996), the following characteristics of soil and soil uses are accepted:

Standard miljøkarakteristika for jord.

Ref.: TGD (1996).

In section 2.3.4 of the TGD (1996), the standard environmental characte- ristics are defined, and on this basis it may be calculated that the density of the soil is 1.7 t/m ³. By application of the depth of soil of 0.2 m in accordance with the TGD (1996), it is estimated that the weight of soil per square meter is equal to 0.34 tonnes.

Subsequently, assuming that in a worst case scenario 80% of the azo dyes are adsorbed onto the sludge, and that in a best case scenario 40% are adsorbed onto the sludge, then the amount of azo dyes on the agricultural fields can be estimated from the following equation:

PEC_{agri sludge} = (release ´ adsorption factor ´ fraction to agriculture)/ (sludge amount/application rate) ´ soil weight.

Estimeret PEC_{agri slam} for azofarvestoffer.

The allocation of sludge to landfill disposal amounts to 20,000 tonnes (dw)/year. The contribution of sludge adsorbed azo dyes to the total amount of azo dyes in landfills may be calculated on the basis of the equation shown below:

Sludge amount to landfills = release ´ adsorption factor ´ fraction to landfills.

In a worst case scenario, the total contribution (processing + use) may be 13.5 tonnes per year, and in a best case scenario 6.70 tonnes per year, which is approximately 6% and 3%, respectively of the total amount of dyes deposited in landfills.

Thus, the total release of azo dyes to landfills may be estimated to approximately 240 tonnes per year in worst case and 233 tonnes per year in best case.

Assuming that the processing industry carries out waste water treatment, the PEC_{agri sludge} is reduced to the range of 0.2 to 0.3 mg/kg. The contribution from the use phase is unchanged with 0.25 to 0.5 mg/kg soil. Due to the lack of monitoring data, it is not possible to validate the calculated PECs. However, these concentrations are, compared to a worst case level of 1 mg/kg (w/w of dry soil) reported by Brown and Anliker (1988), lower.

The fate of products containing dyes released to landfills is uncertain, but there may be a potential release of dyes to soil from this compartment.

Ecotoxicity

Reactive Black 5 (diazo) has a low toxic potential in aquatic organisms (fish LC₅₀ 100-500 mg/l; bacteria EC₅₀ > 2,000 mg/l) as well as the hydrolysed dye (fish LC₅₀ > 500 mg/l; Daphnia magna EC₅₀ (48h) > 128 mg/l) (Hunger & Jung, 1991, IUCLID). Very little information is available on the aquatic toxicity of the hydrolysed reactive dyes, but their loss of ability to react with various groups of vital organic materials, such as proteins and DNA, reduces the potential hazard considerably (ETAD, 1991).

Spencer (1984) has examined the effect of Aquashade (a mixture of Acid Blue 9 and Acid Yellow 23*) on the oxygen consumption of the crayfish Orconectes propinquus and has not found any effect at a concentration of 1 mg/l at an exposure of five days.

A survey of available fish toxicity data on over 3,000 commercially available organic dyes by ETAD member companies indicated that about 98% have a LC₅₀ greater than 1 mg/l, a concentration at which colouring of a river normally would be observable. The remaining 2% were acute toxic (LC₅₀ < 1 mg/l). The latter, consisted of 27 different chemical structures including four Acid dyes, sixteen Basic dyes of which 10 were of the triphenylmethane (not azo) type. In only one case, the LC₅₀ was as low as 0.01 mg/l (Clarke & Anliker, 1980). The LC₅₀ for 59% was more than 100 mg/l (Anliker, 1986), indicating that 41% of the organic dyes are potentially toxic or toxic at levels in the range of 1 to 100 mg/l.

Many acid dyes, including azo dyes, exhibit high toxicity to fish but do not significantly inhibit algal growth (Clarke & Anliker, 1980).

Zhang et al. (1995) showed that azo dyes competitively inhibit COD utilisation or respiratory rates of biofilms at concentrations of 10 mg/l of Acid Orange 14. However, the inhibition effect was much less significant in biofilms, compared to a suspended activated sludge system. Furthermore, the results indicated that the aerobically non-biodegradable dyes, Acid Orange 10* and 14, were more toxic compared to biodegradable dyes such as Acid Orange 7* and 8.

Brown et al. (1981) reported the results of a study of possible inhibitory effects of dyes on aerobic waste water bacteria measured as respiratory rate. They tested both acid, direct, disperse, reactive, basic, vat, solvent and mordant dyes. The study indicated that 18 out of 202 dyes showed an IC₅₀ less than 100 mg/l, including three dyes with an IC₅₀ between 1-10 mg/l. These 18 dyes were all basic dyes. Unfortunately it was a mixture of chemical classes of dyestuffs, including azo dyes, so it is not possible to relate the results directly to specific azo compounds.

ICI found no adverse effects on the carp (Cyprinus carpio) exposed to less than 10 mg/l of 30 water soluble (ionic) and 12 disperse dyes for 8 weeks (Brown, 1987).

Dyes in the aquatic environment were reported to affect microbial populations and their activities. Azo dyes such as Basic Brown 4, Direct Brown 95*, Direct Black 80, Mordant Black 11, Acid Black 52, Direct Red 81* and Direct Yellow 106 were inhibitory to microbial oxidation processes in both activated sludge and stream water. The inhibition by the basic dyes were stronger than the inhibition by acid dyes when the pH was above the isoelectric point of the micro-organism. The inhibition was weakened by introduction of the functional groups methyl, nitro, sulpho or acid to the azo dye or by replacement of the benzene ring with a naphthalene ring. However, introduction of chlorine or bromine strengthened the observed inhibition (Chung & Stevens, 1993). The IC₅₀ was not stated.

In an ADMI (American Dye Manufacturers Institute) study, the toxic effects of 56 selected dyes to the green alga Selenastrum capricornutum were examined. The growth of the algae was assessed after 7 and 14 days in the presence of 1 and 10 mg/l of dyes. 15 dyes (27%) strongly inhibited growth at a test concentration of 1 mg/l after 7 days of incubation (Brown & Anliker, 1988).

The following short term test results are available from a study by ETAD and presented on a seminar in 1992 (ETAD, 1992b). ETAD carried out an investigation of 47 dyes of different chemical dye classes. Even though the specific amount of azo dyes in the investigation is not stated, the results are shown in Table 5.18, Table 5.19, Table 5.20, Table 5.21, Table 5.22 and Table 5.23 below, in order to gain insight to the toxicity of the different chemical (technical) groups of dyes.

It is indicated that the bacteria are less susceptible to the different classes of dyes compared to other test organisms. Among the tested dyes, the bacteria were only susceptible to basic dyes at concentrations below 100 mg/l, which is in agreement with the findings of Brown et al. (1981) and Chung and Stevens (1993).

From the tables it is indicated that the zebra fish is susceptible to (in declining order) basic dyes > acid dyes > disperse dyes at a level less than 100 mg/l. For the other chemical classes, hydrolysed reactive, direct and mordant dyes, the LC₅₀ is above 100 mg/l. The susceptibility to acid and basic dyes for fish is in agreement with the findings of Clarke and Anliker, 1980.

The susceptibility of Daphnia resembles that of the zebra fish, but the order is different, basic > disperse > acid. The remaining chemical classes all show a LC₅₀ above 100 mg/l. The study confirms the findings reported by Hunger and Jung (1991) and IUCLID that the reactive dyes and hydrolysed reactive dyes have a low toxic potential in aquatic organisms.

The alga is apparently the most susceptible organism, because it for all the tested dyes showed a susceptibility to the dye below 100 mg/l. The susceptibility was in the following declining order Mordant > Basic/acid/-disperse > direct > hydrolysed reactive dyes.

Based on literature and database studies, it was possible to obtain results of the LC₅₀ for some of the azo dyes in use in Denmark, but in general there are only a few data available on effects through the normal sources (AQUIRE, IUCLID, HSDB, MITI, etc.). Furthermore, only data on various fish species were obtained, and it was not possible to obtain data on the basic, mordant and the disperse dyes which are used in Denmark.

In Table 5.24, the lowest effect concentrations for azo dyes used in Denmark are presented. The data indicate that various fish species are susceptible to acid and direct dyes at a level between 1 to 10 mg/l. The susceptibility regarding solvent dyes is in one instance below 1 mg/l. It is not known, if some of these azo dyes were included in the study by ETAD (1992b).

1 AQUIRE.

2 MITI.

In addition to the figures shown in Table 5.24, one fish species (Oryzias latipes), exposed 48 hours to Acid Yellow 36, had a LC₅₀ of 68 mg/l. For the remaining dyes, amongst them 5 acid, 6 direct and 2 solvent dyes, the LC₅₀ was above and well above 100 mg/l. Apparently, the different almost exclusively fish species show very variable susceptibility. For further details, see Appendix 2.

Azo compounds

At exposure levels of 2,500 ppm of the azo compound 3,3´, 4,4´-tetrachloroazobenzene on diet, the mortality of the Japanese Medaka (Oryzias latipes) was significantly higher compared to the control group (Allison & Morita, 1995b). On the other hand detrital exposure levels of 2,500 ppm of the same compound did not appear to cause any harmful effects towards the aquatic snail (Indoplanorbis exustus) (Allison & Morita, 1995a).

Metabolites

Couch and Harshbarger (1985) presented an overview of the effects of carcinogenic agents on aquatic animals in experimental studies. They found reports on the effects of aminoazotoluene on fish, adult guppy and adult Medaka at dietary exposure levels of 120 mg/l and 600 mg/l, respectively. Further neoplasm in the liver was induced within 12 weeks and 24 weeks, respectively. The argument of the authors was that in the environment the susceptibility to xenobiotics may differ among different species. Subsequently, proliferation and cellular disorder are neoplasms, which may be caused by xenobiotics, viruses or an interaction of both.

Hermens et al. (1990) investigated the influence on enzyme induction (MFO P450) on the acute toxicity (96-hr LC₅₀) of 4-chloroaniline (p-chloroaniline) to the rainbow trout (Salmo gairdneri). The 95-hr LC₅₀ was from 11.0 to 14.0 mg/l, and the results showed no significant difference between prior induced trout (50 mg/kg Aroclor 1254) and not induced trout, suggesting that metabolic activation does not necessarily play a role in the acute toxicity of aromatic amines to fish.

Metabolic activation of aromatic amines has been shown in the phyla: Mollusca, Crustacea and Echinodermata, e.g. Mytilus edulis, Mytilus galloprovincialis, Carcinus maenas, Asterias rubens, resulting in mutagenicity to Salmonella typhimurium (Marsh et al., 1992).

Dumpert (1987) showed that p-chloroaniline has a lethal effect on the embryo of Xenopus laevis at a concentration of 100 mg/l. Its development is inhibited (teratogenic) at concentrations of 1 and 10 mg/l, respectively.

In a study of bacterial growth kinetics to in vitro toxicity assessment of substituted phenols and anilines, Nendza and Seydel (1990) demonstrated that these compounds were inhibitory, and that the toxic action was probably caused by damage to the bacterial cells. This was documented by decrease in growth rate and in change of the Na⁺/K⁺ ratio with an increase in the Na⁺ and a decrease in the K⁺ concentrations. Furthermore, the authors found a good agreement between growth kinetics of E.coli and fish tests (guppy and zebra fish) for phenols - a linear relationship between log 1/LD_{50 guppy} and log 1/I_{50 E.coli}.

p-aminoazobenzene (10.23 mg/l) was by Zissi and Lyberatos (1996) found to result in a decrease of 15% in the specific growth rate of Bacillus subtilis.

The lowest effect concentrations found for the restricted aromatic amines are presented in 5.25 below. No data were obtained on the naphthalene based amines.

Table 5.25

The lowest effect concentrations for some of the metabolites.

Laveste effektkoncentrationer for visse nedbrydningsprodukter.

Name	Organism	Effect	Conc., mg/l
4-chloroaniline	Daphnia magna	EC50, 24 h	0.061
	Lepomis macrochirus	LC50, 96 h	2.01
	Pimephales promelas	LC50, 96 h	121
	Salmo gairdneri	LC50, 96 h	141
	Xenopus laevis	Teratogen	11
4- aminobenzene	Oryzias latipes, juv	LC50, 24 h	1.72
	Oryzias latipes, juv	LC50, 48h	0.72
o-anisidine	Daphnia magna	EC50, 48 h	6.83
	Poecilia reticulata	EC50, 336 h	18
benzidine	Limoria lignorum, adult	LC100, 18 h	>0.052
	Oryzias latipes, juv	LC50, 24 h	16.02
	Oryzias latipes, juv	LC50, 48 h	10.52
1 Dumpert (1987). 2 ECDIN. 3 Federal Ministry of the Environment, Youth and Family, Austria 1997.

As shown in Table 5.25, it has been reported that benzidine and 4-aminobenzene are acute toxic (LC₅₀ < 1 mg/l) to some crustaceans and juvenile fish. The EC₅₀ for Daphnia magna is as low as 0.06 mg/l for

4-chloroaniline. In general, the LC₅₀ of 4-chloroaniline for various fish species is in the range of 12 to 46 mg/l which indicates potential toxicity. o-anisidine has a LC₅₀ (336 h) of 165 ppm towards adult Poecilia reticulata. o-toluidine has a LC₅₀ (336 h) of 81 ppm towards juvenile Poecilia reticulata. For further details, see Appendix 3.

Summary

No specific data were obtained on basic, reactive, mordant and disperse dyes for any of the dyes encompassed in the present survey.

But it may be concluded that some of the acid and basic dyes are acute toxic to toxic to aquatic organisms (fish, crustaceans, algae and bacteria), which also applies for at least some of the direct dyes, e.g. Direct Blue 14*. Reactive dyes (Reactive Black 5) generally have very high effect concentration levels (>100 mg/l) and are not considered to be toxic to aquatic organisms.

Furthermore, it is indicated that the non-ionic (disperse, mordant and solvent) dyes are toxic and potentially toxic. Solvent dyes may even be acute toxic to aquatic organisms. The mordant dyes may, according to the present findings, not exhibit any toxicity at levels below 100 mg/l.

Algae are generally susceptible to dyes, but the inhibitory effect is thought to be related to light inhibition at high dye concentrations, rather than a direct inhibitory effect of the dyes. According to ETAD (1994), this effect may account for up to 50% of the inhibition observed.

The effects of the substitutional pattern of the dyes are inconclusive, but it has been suggested that introduction of the functional groups; methyl, nitro, sulpho or acid, weakens the inhibition of bacteria, whereas introduction of chlorine and bromine strengthens the inhibition.

In general, it should be noted that toxicity data of chronic low-level exposures for most of the commercial dyes and their derivatives are lacking.

It is indicated, in general, that the effects of the metabolites to aquatic organisms, except for algae, are at levels where potential toxicity is recognised (LC₅₀ < 100 mg/l). This applies for all of the three groups: anilines, benzidines and toluidines. No data were obtained for the naphthalenes.

Anilines and benzidines are both acute toxic and toxic depending on the specific species. The anilines seem to be more toxic to Oryzias latipes juv than benzidine. The findings of the toluidines indicate potential toxicity for various aquatic organisms.

PNEC - Dyes

Applying an assessment factor of 100 on the EC₅₀ from respiration inhibition test (Table 5.20), the following PNEC is derived in accordance with TGD Part II, section 3.4:

PNEC_stp is in the range of 10 : g/l to 100 : g/l.

It should be noted, however, that it is not known if the observed effect is caused by azo dyes or other dye types. But the significance of possible inhibitory effects of azo dyes to the bacteria in the sewage treatment plant is of great importance, therefore, the estimate of PNEC_stp is included.

Short term data from each of the three trophic levels (alga, fish, daphnia) of the base set are available. Hence, according to TGD Part II, section 3.3.1 an assessment factor of 1,000 is applied at the lowest L(E)C₅₀. However, as stated above it is not known, if the observed effect is caused by azo dyes in the case of algae and daphnia, cf. Table 5.18 and Table 5.19. But the lowest observed effect is observed for fish (Table 5.24) with a LC₅₀ of 0.7 mg/l for Oryzias latipes, arriving at a PNEC of:

PNEC_{aquatic organisms} = 0.7 : g/l.

Metabolites

No data were obtained on bacterial inhibition of the metabolites.

Data of two trophic levels were obtained for the metabolites. The lowest observed effect is found in daphnia (Table 5.25) with a LC₅₀ of 0.06 mg/l for Daphnia magna, arriving at a PNEC of :

PNEC_{aquatic organisms} = 0.06 : g/l.

Atmosphere

No data were obtained on atmospheric exposure.

Terrestrial compartment

ETAD has organised a study of the possible effects of dyes on plant germination and growth. Four dyes were used and among them an acid dye of the azo type (C.I. 13155). All four dyes were incorporated into a seed compost at concentrations of 1, 10, 100 and 1,000 mg/l and germination and growth of three plant species (sorghum, sunflower, and soya) were assessed. No effects were observed on seed germination. With respect to the growth rate, there was no observed effects at a concentration of 100 mg/l. At a level of 1,000 mg/l, however, there was a variable growth depending on the dye and the species of the particular plants. After a growth period of 21days, the plant foliage was analysed. At the 1,000 mg/kg soil level the dyestuffs, among them the azo (C.I. 13155), were just detectable in the plant foliage (max. 2 mg/kg) (Brown & Anliker, 1988).

Chung et al. (1997) found out that growth of the soil living nitrogen-fixing bacterium Azotobacter vinelandii is inhibited by p-phenylene- diamine and 2,5-diaminotoluene, which are derivatives after azo reduc-tion of e.g. Basic Brown (C.I. 21010) and Direct Black 80. The nitroge- nous activity was also significantly inhibited at a concentration of 50 : g/ml. p-phenylene- diamine was found to be inhibitory to the growth of other common aquatic and soil bacteria.

PNEC - terrestrial

According to the TGD Part II section 3.6.2.2, an assessment factor of 1,000 should be applied for L(E)C₅₀ short-termed toxicity tests for soil. Brown and Anliker (1988) have reported effects at a level of 1,000 mg/kg for plants, indicating a PNEC of :

PNEC_soil = 1 mg/kg.

Risk characterisation

The PEC/PNEC ratios which can be derived with the available data are shown in Table 5.26.

Table 5.26

PEC/PNEC ratios for the aquatic and terrestrial compartments.

PEC/PNEC forhold for vand- og jordmiljø.

Compartment	Site		PEC (mg/l)			PEC/PNEC
			Worst	Best	Worst		Best
STP	Sludge	Processing
		Textile	8.24	4.11	82.4		41.1
		Leather	4.70	2.45	47.0		24.5
		Use
		Textile	0.68	0.34	6.8		3.4
Aquatic	STP effluent	Processing
		Textile	1.44	0.48	2057		685
		Leather	0.82	0.27	1171		386
		Use
		Textile	0.12	0.04	171		57
	Sediment	Processing
		Textile	0.11	0.02	1.1		0.2
		Leather	0.06	0.01	0.6		0.1
		Use
		Textile	0.008	0.002	0.1		0.02
	Surf. water	Processing
		Textile	0.14	0.048	200		69
		Leather	0.08	0.027	114		39
		Use
		Textile	0.012	0.004	17		8
Terrestrial	Agr. sludge	Processing
		Textile	0.484	0.242	0.5		0.2
		Leather	0.007	0.003	<0.01		<0.01
		Use
		Textile	0.498	0.249	0.5		0.3

For substances with a PEC/PNEC ratio of less than 1 there is, according to TGD, no need for further testing and risk reduction measures beyond those which are already being applied. A ratio higher than 1, however, indicates a need for further information and/or testing or even a need for limiting the environmental risks.

It is indicated that there is a need for further testing and information with regards to the concentration of dye in the aquatic compartment, except for the sediment, because the PEC/PNEC is higher than 1. Whereas the PEC/PNEC ratios for the terrestrial compartment indicate, that there is no need for further testing and/or information.

Furthermore, it is indicated that there is a need of further information with regards to the concentration of dyes in the STP, because the PEC/PNEC is well above 1.

With reference to the assumptions and recalculation of the PECs, it is indicated that the PEC/PNEC ratios presented in Table 5.26 are to high.

Recalculation of the PEC/PNEC ratios indicates:

PEC/PNECsludge, stp	3.4 to 5	(>1)
PEC/PNECeffluent, stp	34 to 1,234	(>>1)
PEC/PNECsurface water	3 to 123	(>1)
PEC/PNECsediment	0.01 to 0.9	(<1)
PEC/PNECagri sludge	0.5 to 0.8	(<1)

Summary

The survey indicates that there is a need for further information and testing in order to assess the environmental risk in the STP, the STP effluent and surface water, whereas the releases associated with sludge application in agricultural soil not seem to present any immediate concern.

[Front page] [Contents] [Previous] [Next] [top]