Diiron tris(sulphate)

Endpoint summary

Description of key information

Acute toxicity, oral: H302: Harmful if swallowed, Category 4, OECD TG 423; study Choi 2004a
Acute toxicity, inhalation: No adverse effect observed in limit study, EPA OPP 81-3; US EPA 1993/Robbins 1991
Acute toxicity, dermal: Non-toxic, OECD TG 402; study Choi 2004b

Key value for chemical safety assessment

Acute toxicity: via oral route

Endpoint conclusion

Endpoint conclusion:

adverse effect observed

Dose descriptor:

LD50

Value:

500 mg/kg bw

Quality of whole database:

The studies with different iron species were conducted between 1961-2004. Some of them are pre GLP and pre-guideline studies.

Acute toxicity: via inhalation route

Endpoint conclusion

Endpoint conclusion:

no adverse effect observed

Dose descriptor:

discriminating conc.

Value:

300 mg/m³ air

Quality of whole database:

Three entries with following endpoints: A discriminating dose tested in conformity guideline standards, a derived TLV and an inappropriate 8 h aerosol test with unknown resulting exposure concentration.

Acute toxicity: via dermal route

Endpoint conclusion

Endpoint conclusion:

no adverse effect observed

Dose descriptor:

discriminating dose

Value:

881 mg/kg bw

Additional information

This endpoint is covered by the category approach for soluble iron salts (please see the section Toxicokinetics, metabolism and distribution for the category justification/report format).

- oral:

• animal data:

A plethora of acute oral studies conducted with the different members of the iron salt category are available. The available reliable studies for the different metal salts have been compiled (see the below Table). There is marked variability in the values quoted for acute toxicity for all salts. This is probably a reflection of the variety of protocols used and the duration over which observations were recorded. In addition when looking at species differences it is noted that mice seem to be a bit more susceptible to iron toxicity than rats. Nevertheless all substances clearly have LD50 values > 300 mg/kg bw. Study Choi (2004a) using FeCl2 as test item is chosen as the representative key study as it is, of the reliable studies, the one reporting the most critical LD50 value. This is in line with the assumption that the ferrous ion (Fe(II)) has a higher oral bioavailability than the ferric ion (Fe(III)). Choi (2004a) followed closely the protocol of OECD TG 423 and GLP using gavage in rats and therefore is deemed fully reliable. Hypoactivity and piloerection were reported at both 300 and 2000 mg/kg bw, prone position, reddish change and oedema on ears, fore-legs and hind-legs were only seen in the higher dose. At 2000 mg/kg bw, nasal discharge (reddish or clear) was observed externally in all animals. Haemorrhage on lymphatic nodes, stomach and intestine in all animals and haemorrhage on the thymus of one animal was observed. At 300 mg/kg bw, haemorrhage on lymphatic nodes and intestine were observed in the one animal (of totally six) that died during the study. At 200 mg/kg bw all 3 animals died. This distribution of deaths in the two dose groups is indicative of a LD50 of 500 mg/kg bw according to the OECD TG 423.

Table: Overview on acute toxicity measurements

Iron salt species analysed	Identifier of the study	Klimisch score and relevance	Animal species	LD50/LC50 based on tested substance	LD50/LC50 based on Fe
Oral, LD 50 in mg/kg bw
FeCl3	Hosking 1970	K2 KS	mouse, female	1278 (871 - 1830, 95 % CV)	440 (300 -630, 95 % CV)
FeCl2	Choi 2004a	K1 KS	rat	500 (300 – 2000, toxic classes tested)	220 (132 – 881, toxic classes tested)
Fe2(SO4)3	ICI 1991	K4 to date SS	rat	500 – 2000 (females), > 2000 males	139 – 558 (females), > 558 males
FeSO4	Weaver 1961	K2 SS	mouse	1025 (802 - 1311, 95 % CV)	377 (295 - 482, 95 % CV)
FeSO4	Weaver 1961	K2 SS	rat	2625 (2323 -2966, 95 % CV)	965 (854 -1090, 95 % CV)
FeSO4 x 7 H2O	Bayer AG 1985	K2 SS	rat	3200 (2900 -3700, 95 % CV)	643 (583 -743, 95 % CV)
FeSO4	Boccio 1998	K2 SS	mouse	670 (females) 680 (males)	246 (females), 250 (males)
FeSO4 x 7 H2O	MHLW 2002	K1 KS	rat	> 2000	> 4021
FeSO4	Parent 2000	K2 SS	rat	3200	1177
Inhalation, LC50 in mg/m³
FeCl3	ACGIH 1999	K4 SS	unknown species	Substance is irritating
FeCl3	BASF AG 1977	K4 SS	rat	LC0 is larger than the saturation concentration of a mist generated from a 40 % (w/w) solution when exposed for 8 h
Fe2(SO4)3	US EPA 1993	K4 SS	rat	> 1.1	> 0.15
Dermal, LD50 in mg/kg bw
FeCl2	Choi 2004b	K1 KS	rats	> 2000	> 881

KS: key study; SS: supporting study; K1: reliable without restriction, Klimisch 1; K2: reliable with restriction, Klimisch 2; K3: not reliable, Klimisch 3; K4: not assignable, Klimisch 4

• human data:

The Expert Group on Vitamins and Minerals (EVM 2003) considered that the following acute toxic doses apply:

- for infants (under the age of six years) 20 mg/kg for gastrointestinal irritation, with systemic effects not occurring below 60 mg/kg bw

- for children, 200-300 mg/kg bw

- for adults 1400 mg/kg bw acute oral toxicity in humans. It is not stated in the source material whether the toxicity values are mg salt/kg bw, or mg Fe/kg bw. It is not always stated which salt was used.

McElhatton (1991) reports that in a follow up study of 49 pregnancies where iron overdose was a factor there was no evidence of a correlation between serum iron concentrations and birth weight. No causal relationship between iron overdose or Desferoxamine (DFO) treatment and malformations was observed. There was no evidence to suggest that DFO caused toxicity in the mother or baby.

A summary entry of ACGIH (1980) & Ellenhorn (1988) describes the estimated fatal dose in humans for soluble ferric salts to be 30 g. Toxic doses of iron overwhelm the normal gastrointestinal regulatory mechanism (suggesting that saturable transport systems are not involved at high concentrations); this results in massive iron absorption. Major toxicity occurs when serum iron levels exceed the iron-binding capacity of transferrin. Free circulating iron damages systematic blood vessels. The release of serotonin and histamine potentiate the vascular damage caused by free serum iron. In severe iron overdose, the coagulative necrosis with platelet aggregation appears similar to the damage caused by corrosive agents.

According to Gilman 1980, the studies on human subjects have established that physical intolerance to iron does occur. With a dose of 200 mg or iron per day divided into three equal portions, symptoms occurred in approximately 25 % of individuals, compared to an incidence of 13 % among those receiving placebos; this increased to 42 % when the dosage of iron was doubled. Nausea and upper abdominal pain were increasingly common manifestations at high dosage. Constipation and diarrhoea were not more prevalent at higher dosage, nor was heartburn. Available evidence suggests that the normal individual is able to control absorption of iron despite high intake, and it is only individuals with underlying disorders that augment the absorption of iron might develop haemochromatosis.

In IUCLID 2000 (summary of summaries) it is explained that ferrous sulphate is widely used in iron pills for the treatment of anaemia. There is therefore a large amount of human data on the use and misuse of the compound. The lowest lethal dose in humans by oral ingestion is estimated at being in the range of 40 - 1600 mg/kg. Young children appear to be more susceptible. Other estimates are considerably higher. The main cause of death is haemorrhagic gastritis with oedema, (also demonstrated in animal experiments on treatments for iron poisoning). Liver damage has also been reported after gross ingestion of ferrous sulphate. Ferrous sulphate has been used for many years and has been used to replace some irritant agents. However, ferrous sulphate is regarded as a skin and eye irritant for the purposes of handling.

Hoppe (1955) reports that death has occurred from the oral ingestion of ferrous sulphate at doses ranging from 40 to 1600 mg/kg.

According to Aisen (1990) acute symptoms of iron toxicity are characterized by vomiting, diarrhoea, mild lethargy, upper abdominal pain, pallor, and hyperglycaemia with more severe clinical findings including cyanosis, stupor, acidosis, haematemesis, shock, and coma.

In addition two case studies on oral iron poisoning are available. Ling (1988) describes the following case: A 25 year old woman fatally ingested 200 mL ferric chloride solution (pH 1), equivalent to 230 mg/kg bw of elemental iron. She had hypoxemia and severe metabolic acidosis with respiratory alkalosis initially. Three hours after her ingestion she presented with drowsy consciousness, tachycardia, tachypnea and projected vomiting. Laboratory studies showed leukocytosis, elevated glucose, aspartate aminotransferase, amylase, lactate dehydrogenase, and total bilirubin, coagulation defect and haemolysis. Aspiration and vision loss were also reported. Four hours after ingestion cardiopulmonary arrest suddenly occurred after severe vomiting and the woman died. Toxicological studies showed marked elevation of serum iron (2440 µg/dL).

Ellenhorn (1998) describes a case where a pregnant mother died of iron poisoning (serum iron, 1700 mg/dL) after deferoxamine was withheld.

The available human data is of low reliability or from a secondary source (EVM 2003). Therefore for the derivation of classification and labelling is based on the available animal data. Nevertheless it should be emphasized that the human data is not in contradiction to the animal data.

- dermal:

For the dermal exposure route only one reliable study is available. Choi (2004b) followed closely the protocol of OECD TG 402 and GLP using 24 h semiocclusive exposure of the test item FeCl2 in corn oil to rats (dose group: 2000 mg/kg bw). For Fe(III)-salts no studies are available. Nevertheless FeCl2 can be regarded as a worst case as the general assumption is that the ferrous ion (Fe(II)) has a higher bioavailability than the ferric ion (Fe(III)). The higher corrosivity of Fe(III)-salts is covered under the respective chapter on irritation/corrosion. Nevertheless it has to be taken into account that upon the availability of water and oxygen FeCl2 is oxidised and hydrolysed and hydrochloric acid is freed according to the following (idealised) equation:

4 FeCl2 + O2 + 10 H2O -> 4 Fe(OH)3 + 8 HCl

Therefore during an incubation period of 24 h FeCl2 can exhibit a certain corrosive effect on the local skin tissue. Accordingly it can be regarded as a good surrogate for the other iron salt in acute dermal toxicity.

In Choi (2004b) there were no unscheduled deaths and bodyweight gains were normal during the study. A yellowish-brown change was observed at the application site in all test animals, and was considered to relate to the colour of the test article. Two males and 4 females had a reddish nasal discharge on Day 2, which was considered a distress symptom caused by pressure of taping on the thorax area.

During macroscopic examination, scarring was observed at the application site on one male and one female. This was considered to relate to the application of the test article and can most probably be assigned to the corrosive action of the test substance. Internally, no abnormalities were observed.

- inhalation:

According to Annex VIII as specified under section 8.5, column 2, no testing is required via inhalation route since high reliability studies are already in place via the oral and dermal route.

A reliability score 4 study (BASF 1977) in which a saturated atmosphere of aerosol generated using a 40 % aqueous solution of ferric chloride did not cause any fatalities in rats following eight hours of exposure. In the study with rats (US EPA 1993/Robins 1991) using iron (III) sulphate, no mortality occurred and an LC50 greater than the highest tested concentration of 1.10 mg/L was evidenced.

- other routes:

A number of studies are available for the different iron salts that were applied via i.v., i.p. or subcutaneous injection. None of these studies are fully reliable and as they are not of relevance for the exposure to chemicals they are disregarded here.

Justification for selection of acute toxicity – oral endpoint
Reliable study according to OECD TG 423, reporting the most critical LD50 value

Justification for selection of acute toxicity – inhalation endpoint
Only one guideline limit study at 300 mg Fe /m³ with no mortality and no gross pathology effects observed.

Justification for selection of acute toxicity – dermal endpoint
Only one available study, performed according to OECD TG 402, no effects at 881 mg Fe/kg bw

Justification for classification or non-classification

Oral toxicity

Based on the above stated assessment of the acute oral toxicity of the iron salts, the LD50 values from reliable studies range between 300 and 2000 mg/kg bw. Accordingly the salts belonging to the iron salt category need to be classified as “R22 Harmful if swallowed” according to Commission Directive 2001/59/EC (28th ATP of Council Directive 67/548/EEC) and as Category 4, “Warning - H302: Harmful if swallowed” according to CLP (5th ATP of Regulation (EC) No 1272/2008 of the European Parliament and of the Council) as implementation of UN-GHS in the EU.

Table: Acute toxicity label elements for category 4 (CLP, 5th ATP, Annex I, Table 3.1.3)

Element		Type
GHS Pictogram		GHS07 exclamation mark
Signal Word		Warning
Hazard Statement (Oral)		H302: Harmful if swallowed
Precautionary Statements	Prevention (oral)	P264, P270
	Response (oral)	P301 + P312, P330
	Storage (oral)	none
	Disposal (oral)	P501

 

Dermal toxicity

Based on the above stated assessment of the acute dermal toxicity of FeCl2 which is used as surrogate for all iron salts of this category (absence of toxicity up to 2000 mg/kg). Thus the salts belonging to the iron salt category do not need to be classified according to Council Directive 2001/59/EC (28th ATP of Directive 67/548/EEC) and according to CLP (5th ATP of Regulation (EC) No 1272/2008 of the European Parliament and of the Council) as implementation of UN-GHS in the EU.

 

Inhalation toxicity

Based on the above stated assessment of the acute inhalation toxicity of Fe2(SO4)3 which is used as surrogate for all iron salts of this category, the LC50 result from available study is >>1.1 mg/L (US EPA 1993/Robbins 1991). The absence of 50 % mortality up to 5 mg/L is the upper level of relevance for classification. Therefore this study result might be used for a precautionary classification as “R20 Harmful by inhalation” according to Council Directive 2001/59/EC (28th ATP of Directive 67/548/EEC) and as Category 4, “Warning - H332: Harmful if inhaled” according to CLP (5th ATP of Regulation (EC) No 1272/2008 of the European Parliament and of the Council) as implementation of UN-GHS in the EU. According to CLP (5th ATP) 3.1.2.3.2., “Inhaled particles between 1 and 4 microns mean mass aerodynamic diameter (MMAD) will deposit in all regions of the rat respiratory tract. This particle size range corresponds to a maximum dose of about 2 mg/L.” Based on the measured MMAD of 2.75 µm in the study, the used gravimetric concentration of >1.1 mg/L in can be therefore regarded as the limit concentration for testing the substance. Together with the fact, that neither mortality nor signs of toxicity were noted according to the short report, the determined LC50 of >1.1 mg/L would hence not justify a classification as Acute Tox. Cat. 4 after acute inhalation exposure.