Ferric ammonium citrate

Toxicity to microorganisms

Description of key information

Iron (III): NOEC 500 mg Fe/l.

Citrate: Toxicity threshold (TT) value for bacteria (Pseudomonas putida) of >10000 mg/l

Ammonium: NOEC > 50 mg/l

Ammonium sulfate: NOEC >94 mg/l for specific nitrifying bacteria

Additional information

No studies are available for the toxicity to microorganims of reaction mass of ammonium iron (III) citrate and ammonium sulfate. However, ammonium iron (III) citrate and ammonium sulfate are both expected to dissociate under environmentally relevant conditions (see Section 1.3). Therefore, it is appropriate to read-across data relevant to the toxicity to microorganisms of iron, citrate, ammonium and sulfate. Further details may be found in an expert report (Peter Fisk Associates, 2012) attached in Section 13 of the IUCLID 6 dossier.

Iron (III)

Of the components of the registered substance, ferric iron is the component we must consider as of concern in the activated sludge process. Since iron is a required co-factor for bacterial biochemistry its effects on the growth and flocculation of sludge bacteria have been extensively studied, as has its redox chemistry in the treatment system.

Apart from a direct toxic effect on microorganisms, contaminants in wastewater systems may also disrupt the activated sludge process by interfering with the flocculation process by which the bacterial mat forms clumps and separates from the aqueous stream and consequently leads to problems of excess dissolved solids in the effluent from the treatment plant. It is in this process where the effect of bi- and trivalent metal cations including ferrous and ferric iron are likely to have the most effect and which has been studied most extensively.

Iron is generally considered as advantageous for wastewater treatment and indeed iron salts are frequently used in activated sludge treatment as coagulants or to improve reactor performance (Philips and Verstraete, 2002).According to experience, under normal conditions, these applications of these substances cause no disturbances in the normal operation of the biological degradation, which is clear evidence of an absence of inhibitory effects within the normal loading range.

 

In an Internet-published article Zhang and Hynninen (2012) measured the effect of 0 (control), 5, 10, 20, 30, 50, or 80 mg/l ferric iron (introduced as ferric chloride) concentrations on the settle-ability, dissolved and total chemical oxygen demand (COD), sludge volume index (SVI), mixed liquor suspended solids (MLSS), ash content, surface charge and dissolved oxygen content of sludge reactors. A summary of their conclusions is as follows:

 

1.        Iron introduced to activated sludge systems improved the settle-ability of the sludge.

2.        Iron decreased the SVI, and reduces the sludge bulking problems.

3.       Iron improved the treatment efficiency in concentrations up to 30 mg Fe/l, but overdosing up to 80 mg/l decreased the COD reduction.

4.        Iron caused the surface charge of the activated sludge particles to be slightly less negative and hence improved settling, but increasing amounts of Fe produced sludge with poorer de-water-ability.

5.        Iron increased the rate of metabolism of the sludge bacteria.

 

In spite of this generally positive effect, the presence of iron has been shown to have a subtler effect on the ecology of microbial communities in sludge treatment plants. Philips and Verstraete (2002), showed that addition of iron to bacterial communities in sludge at 2 mM (112 mg/l) of ferric iron resulted in breaking of the sludge flocs, resulting in massive growth of sessile-protozoa and lowering of the numbers of nitrifiers. Ferrous iron at the same dosage resulted in large coherent flocs. However, these effects are seen with much higher doses than the Zhang and Hynninen study. 

 

Both ferric and ferrous iron salts are commonly used for chemical phosphorus removal in the activated sludge process. Oikonomidiset al. (2010) compared Fe(III) and Fe(II) salts regarding their effect on the physical and biological properties of activated sludge and found that ferrous iron was a more effective flocculent than ferric, which they attributed to stronger ionic bonds being formed by ferrous iron with the floc surface prior to its oxidation to ferric iron. However, this conclusion is in contrast to other studies such as Murthy (1998) and da Silva Henriques (2006). Both of these salts were dosed at 25 mg/l, however, as was also seen by Zhang and Hynninen (2012), there was a reversal of this beneficial effect at higher iron concentrations.

 

The effect of ferric iron on protein and polysaccharide flocculation was extensively studied by Murthy (1998) who concluded that ferric iron was capable of coagulating much of the released biopolymers (biofilm) from microorganisms and suggested a mechanism through the adsorption of organic biocolloids onto iron-hydroxy mineral surfaces in the flocs. He also suggested that iron may play an important role in maintaining floc structure and retaining proteins within the floc and reducing the effluent COD. The concentrations of ferric chloride used were 0, 10, 20, 100, 200 and 400 mg/L, and a near linear decrease in supernatant turbidity (increase in flocculation effect) and decrease in COD was seen across the whole concentration range (c.f. Zhang and Hynninen (2012)).

 

The effect of iron on flocculation in activated sludge was also reviewed by da Silva Henriques (2006), who in addition covered the reduction of ferric to ferrous iron under anoxic conditions, as under these conditions iron-reducing bacteria may predominate  She concluded that, independent of the mechanism of iron reduction in the floc matrix, the process is always followed by a deterioration of floc structure and strength, and the reduction of the cation charge from +3 to +2 is alone sufficient for deflocculation to be observed (like Fe (III), Fe (II) is mainly associated with the floc matrix which leads to a deterioration of floc quality in the presence of ferrous iron). See also Nielsen (1996) and Caccavo et al. (1996) and c.f. Oikonomidis et al. (2010).

A number of results for iron salts are available for microbial inhibition, in several species and with varying reliability. In some cases the acidic pH (4-5) was noted by the reporting authors. The observed effects may be associated with the intrinsic acidity of Fe(III) salts.

In acute toxicity studies in microorganisms, an IC₅₀ result is available for ferrous sulfate of 40 mg Fe/l in a reliable Microtox study using Vibrio fischeri. The pH was not adjusted as part of this assay and the effects may well be attributable, at least in part, to acidification.

A short term study specifically investigating effects on nitrifying microorganisms in WWTP sludge concluded an EC₅₀ of 95 mg/l. The study is reliable (with some methodological shortcomings) and pH control was good. However it is apparent that under neutral conditions the majority of iron present will be undissolved; it is not clear how account is taken of this in the study. The IC₅₀that is concluded is well below the loading which is routinely used in real wastewater treatment plants, and so it would be misleading to base the PNEC on this result.

Inhibition of cellular respiration in activated sludge biomass was observed, with an IC₅₀of ca. 500 mg FeCl₃/l (equivalent to ca. 170 mg Fe(III) /l) under conditions where the pH was not adjusted to neutralise pH effects (Broglio, 1987). This was a non-standard shortened test and some details of test conduct and conditions are not reported, so the result is of non-assignable reliability. The author concluded that effects on aquatic microorganisms appear to be related to the pH of the test medium, which decreases significantly as more iron is added; pH fell to ca. 5.7 at the 500 mg/l loading from a baseline pH of ca. 7.4.

Results in the range of 100-183 mg/l have been reported in studies of uncertain reliability usingPseudomonas putida,P. fluorescensandPhotobacterium phosphoreum.

Several IC₅₀values (1-10 mg/l) have been measured for Fe(Cl)₃using 14 microorganism species, in studies of uncertain reliability. Available data on test conditions is minimal but the acidic pH (4-5) was noted.

In conclusion, iron generally has a positive effect on the bacterial colonies in activated sludge used for wastewater treatment and concentrations up to 30 mg/l of ferric iron increase their growth and metabolic rate and decrease the chemical oxygen demand of the effluent. Iron also has a beneficial effect on bacterial floc production and leads to a lower suspended solid concentration in the effluent. There is some evidence that concentrations of iron higher than 30 mg/l lead to an increase in the chemical oxygen demand and a deterioration of floc quality; however, other studies have shown a positive beneficial increase in these parameters at iron concentrations of up to 400 mg/l.The typical loading rate in waste water treatment plants varies between ca. 1 – 100 g Fe/m³waste water. The reported typical overall average usage (15 g Fe/m³waste water) is reported to result in concentrations of 200-500 mg Fe/l in activated sludge (Kronos 2000). Higher levels have been reported but this may be a seasonal or occasional effect; loading rates of iron vary significantly for different applications at different stages of the wastewater treatment processing, and the maximum concentration to which activated sludges are regularly exposed could easily be much higher than this. Overall there is substantial evidence to assume a NOEC of 500 mg Fe/l.

Citrate

In accordance with Column 2 of REACH Annex VIII, the activated sludge respiration inhibition study does not need to be conducted for citrate as the substance is readily biodegradable and the applied test concentrations are in the range that can be expected in the influent to a sewage treatment plant.

However, for the purpose of setting a PNEC the Bringmann and Kuhn (1980) study for citric acid withPseudomonas putidahas been selected since no other study on the test substance is available.

Under conditions of defined pH, addition of citric acid to solution is equivalent to addition of the citrate ion.

Ammonium and sulfate

Typical levels of free ammonia in municipal WWTPs range from ca. 12-50 mg/l in the WWTP collection system (Lehr 2005). A NOEC of >50 mg/l, can therefore be predicted for ammonia toxicity to microorganisms. Ammonium sulfate toxicity to sewage treatment microorganisms has been studied (Suwaet al., 1994) (OECD 2004). 10 strains of Nitrobacter spp. (ammonium oxidising bacteria) were studied. These were isolated from 34 different sludges including sludges from primary sewage treatment plants and sludges from night soil treatment plants. Both sensitive (no growth at ca. 5000 mg/l ammonium sulfate) and insensitive (growth at ca. 5000 mg/l ammonium sulfate) strains were found. These results indicate that a NOEC for specific nitrifying bacteria will be greater than ca. 100 mg/l.