Aluminium chloride

Endpoint summary

Description of key information

Oral data:

A number of oral repeated dose studies in animals were available for aluminium chloride. As these studies were of different qualities, they were evaluated in a weight-of-evidence approach.

In a non-guideline study, Bilkei-Gorzo et al. (1993) reported about the effects of subchronic treatment with aluminium on rat brain choline acetyltranserase activity. Male and female rats were treated for 90 days with water-soluble, insoluble or chelated aluminium compounds. The daily treatments given were as follows: controls, NaCl (100 mg/kg body weight) plus citric acid (30 mg/kg); AlCl3 (30 or 100 mg/kg); Al(OH)3 (100 mg/kg) plus citric acid (30 mg/kg); Al(OH)3 (300 mg/kg). The learning ability was determined in the labyrinth test at day 90, and the choline-acetyltransferase, acetylcholinesterase activity and aluminium content of the brains were measured. Soluble and chelated aluminium compounds worsened the learning ability, and the aluminium content of the brain was elevated. Acetylcholine esterase activity increased and choline-acetyltransferase activity decreased, but the changes were statistically significant in the groups treated with 100 mg/kg AlCl3 and with Al(OH)3 + citric acid, only.

The study has several weaknesses: The examined parameters were very selective (labyrinth test, AChE and CAT activity, Al content in brain) and cannot be evaluated properly without the context of other critical endpoints, such as those recommended in OECD TG 408. For instance, clinical signs and body weight were not monitored, and a comprehensive functional observation battery was not performed. Any findings observed by the authors may therefore be unspecific and secondary to other effects. Most importantly, however, the reported effects are not in line with the high-quality study by ToxTest, 2010, which was conducted according to OECD TG 426 (see IUCLID chapter 7.8.2). In this study, no effects on memory and learning were observed in rats treated with much higher Al concentrations (up to 300 mg Al/kg bw) and for much longer. The treatment-related deficits in learning reported by Bilkei-Gorzo are therefore implausible and seriously question the validity of the entire study. Thus, the study should be disregarded.

Sahin et al. (1995) reported about the impairment of motor coordination in mice after ingestion of aluminium chloride in a non-guideline study. Male animals were offered a concentration of about 22 mg per day continuously in the drinking water (study duration: 100 days). The data showed that Al produced an impairment of performance in the motor coordination task (rota-rod treadmill) of mice depending on the duration of treatment. There was no difference between the percentage of animals staying on the rod on the 38th and 90th days of the control group (p > 0.05). The decrease observed in the performance of the group treated with Al on the 90th day compared with the performance of the same animals on the 38th day was found to be statistically significant (p < 0.05).

The study is not considered reliable: Only one test substance concentration was applied, which makes evaluation of a dose-response relationship impossible. Furthermore, the test substance was applied via drinking water, but water intake was not monitored, so the actually ingested dose is unknown. Rota-rod performance was the only evaluated parameter. No other parameters were reported, such as clinical signs or body weight. This is all the more critical as body weight can influence the performance in the motor coordination test on the rota-rod treadmill. Similar to the study by Bilkei-Gorzo (1993), the findings by Sahin et al. are not consistent with the results of the OECD TG 426 study by ToxTest, 2010. Although Sahin tested mice and ToxTest tested rats, no impairment of the rota-rod performance was observed in the ToxTest study, which indicates that the Sahin study is unreliable as a whole. It should therefore be disregarded.

Cho and Joshi (1988)determined the effects of aluminium on hexokinase and glucose-6-phosphatase in the brain in a non-guideline study. Male rats were continuously offered 100 uM aluminium chloride via the drinking water over a study period of 1 year. The concentrations of Al in the homogenates, as computed on the original brain for the control and Al fed group were 40 ng and 80 ng/g wet weight, respectively. The activity of acetylcholine esterase was the same in both groups but that of hexokinase and G6PDH in the Al-fed group was about 73% and 80%, respectively, of the control. Dialysis restored the G6PDH but increased the hexokinase of the control group 2-fold and that of Al-fed group 2.7-fold. Thus, at this elevated level it was same in both groups. The contribution of Al from the undialysed homogenates during assay was too low to account for the inhibition. It is therefore suggested that a dialyzable inhibitor for hexokinase is normally present in the brain and that Al feeding increases its concentration to further inhibit the utilization of glucose.

The study by Cho and Joshi has several deficiencies: The information given in the report is generally very limited, only 5 control animals and 9 treated animals were examined, and the parameters assessed were limited to body/brain weights and enzyme activity. A comprehensive assessment of mortality, clinical signs, water/food consumption, pathology etc. was not conducted. The significance of the study to evaluate specific target organ toxicity upon repeated exposure to aluminium chloride is therefore regarded with reservations. Due to the above-mentioned deficiencies, the study is assigned a reliability rating of 3 and is consequently disregarded.

Newer data from oral repeated dose studies are meanwhile available for basic aluminium chloride (CAS 1327-41-9) and aluminium citrate (CAS 31142-56-0). The data from these supporting substances allow a read-across approach to cover the information requirement for anhydrous aluminium chloride more reliably. As described in chapter 7.1 of the IUCLID, the bioavailabilities of the following aluminium salts was shown to be comparable in an in vivo study conducted with radioactively labeled test compounds: Aluminium citrate, aluminium chloride, aluminium nitrate, aluminium sulphate and aluminium hydroxide. Furthermore, anhydrous and basic aluminium chloride can be expected to form the same dissociation products in aquaeous solution and to have the same systemic effects in the body. Therefore, read-across from aluminium citrate and basic aluminium chloride was considered appropriate to cover the endpoint of oral repeated dose toxicity for anhydrous aluminium chloride in accordance with section 1.5 in REACH Annex XI.

In a GLP-compliant combined repeated dose / reproductive screening study (OECD 422), administration of basic aluminium chloride by oral gavage to male and female rats at dose levels of 20, 200 or 1000 mg/kg/d (equivalent to 3.6, 18 and 90 mg Al/kg bw/d) was studied. Males were exposed for 28 days, females between 37 and 53 days. Overall, no significant signs of toxicity were observed, except for local effects on the stomach in high dose males: Foci on the glandular mucosa of the stomach, partly with thickening of the glandular mucosa or limiting ridge, were found in several male animals at 1000 mg/kg bw. No toxic effects were observed in females at any dose. Therefore, the overall NOAEL for female rats was established to be 1000 mg/kg bw/day(equivalent to 90 mg/kg bw/d). For males the NOAEL for local effects was established to be 200 mg/kg bw/day (equivalent to 18 mg/kg bw/d) and for systemic toxicity 1000 mg/kg bw/day (equivalent to 90 mg/kg bw/d). No abnormalities were detected with regard to the parameters of the functional observation battery.

In acombined developmental neurotoxicity and chronic study (according to GLP and OECD guidelines 426 and 452)(see chapter 7.8.2 of the IUCLID),rat pups previously exposed to aluminium citrate during the in utero and weaning period via the dams were further treated with aluminium citrate until 12 months of age via the drinking water. Due to the study design, it is difficult to differentiate between developmental or direct toxicity after weaning, which, however, does not affect the formal reliability of the study. The dose levels (for dams and corresponding offspring) were 30 mg Al/kg bw/day, 100 mg Al/kg bw/day, and 300 mg Al/kg bw/day. Except for diarrhea in the high dose dams, no treatment-related effect was observed in the dams.The most notable treatment-related effect observed in the offspring was renal pathology – hydronephrosis, ureteral dilation, obstruction and presence of calculi - most prominently in the male pups. Higher mortality and significant morbidity were observed in the male pups in the high dose group leading to euthanization of this group atca. study day 89. Clinical observations that showed a relationship with treatment, either directly or secondary to renal failure, were poor coat, weight loss, and haematuria. Diarrhoea was also observed. These signs were found only in the high dose Al-citrate treatment group. The results show a clear, consistent reduction of post-weaning body weight in the high dose Al-citrate group in both male and female pups. Overall, dosing of animals with aluminium citrate led to higher fluid consumption than in the control animals. In the offspring, neuromuscular effects (based on the reduction fore- and hindlimb grip strength and foot splay) were detected at both the high (300 mg Al per kg) and mid (100 mg Al per kg) dosage level. However, these effects are attributable to the lower body weights of the individual rats in this test. The individual data are not available, and neither are historical control data. Therefore, the neuromuscular effects cannot unequivocally ascribed to treatment with Al3 +. These effects formed the basis for the assignment of 100 mg Al/kg Body weight as the conservative LOAEL for this study. Concentrations of Al in blood, femur, liver and some CNS tissues were significantly higher than controls. However, there was no evidence of an effect on learning and memory in this study.

In conclusion, the two guideline studies conducted with basic aluminium chloride and aluminium citrate are considered to provide the most reliable data to evaluate the oral repeated dose toxicity of anhydrous aluminium chloride. General toxicity was observed at relatively high dose levels (90 mg Al/kg bw in the OECD 422, and 300 mg Al/kg bw in the OECD 426/452). The above-mentioned literature publications indicate Al-related changes in the brain on the molecular level and impairment of memory and learning ability in rats and mice; however, no effects on learning and memory were found in comprehensive functional observation batteries in recent guideline- and GLP-compliant studies.

Inhalation data:

Since no reliable repeated dose data for the inhalation route was available for anhydrous aluminium chloride (likely due to the hygroscopic properties of the substance), a 90-day repeated dose inhalation study has been conducted.

The registrant addressed the question of whether an experimental atmosphere of dust aerosol can be generated under standard laboratory conditions (i.e. at standard air humidity) in a basic technical experiment (please see attached document in IUCLID section 7.5.2) and could demonstrate that generation of a stable dust atmosphere of anhydrous aluminium chloride is in fact not possible.

The reactivity of the substance and the results from the technical trial mentioned above indicate that assessment of anhydrous aluminium chloride is not representative for exposure in the workplace. When present as dust in the atmosphere, e.g. during filling processes, the substance will likely start reacting with moisture in the air immediately. Thus, when aluminium is detected in the atmosphere, it will most probably not be present in the form of anhydrous aluminium chloride but in the form of the reaction products with water in the air.

No reliable information on the behavior of aluminium chloride dusts in moist air could be found.

To prove the hypothesis made above and to gain a better understanding of the products and kinetics of the reaction of the solid test substance in air and thus of the material to which workplace exposure could potentially take place, the registrant has performed a series of experiments at BASF's in-house analytics competence center (study report, BASF SE, 2017, report no. 16Y46195) with the aim of answering the following questions:

How fast is the hydrolysis reaction of AlCl3 anhydrous dusts in air?

Which Al species is the primary reaction product of that hydrolysis?

A detailed description of the experiments and the results can be found in IUCLID section 7.5.2 (attachment to the repeated-dose inhalation study; BASF SE, 2019).

The question of hydrolysis kinetics was addressed by means of ATR (attenuated total reflectance)-FTIR (Fourier transform infrared spectroscopy) measurements. The experiments show that that the hydrolysis of solid aluminium chloride anhydrous takes place rapidly when in contact with air and becomes faster with increasing surface of contact between the test material and the surrounding air. The spectra that were measured first (within seconds after loading the samples) already showed bands that can be assigned to the hydrolysis products, so it was not possible to spectrally determine the true start of the reaction. In conclusion, the ATR-FTIR experiments in combination with the technical aerosolization trial in IUCLID section 7.5.2 confirm that aluminium chloride anhydrous is not the representative aluminium species to evaluate worker exposure by inhalation. Atmospheric aluminium concentrations that are measured in the workplace (see CSR chapters 9 and 10 for workplace measurements conducted by BASF SE) cannot be attributed to aluminium chloride anhydrous but to its hydrolysis products that form extremely fast. Consequently, the potential health hazards of the hydrolysis products, not of aluminium chloride anhydrous, are of potential relevance for the worker and have been addressed in a repeated-dose inhalation study.

To characterize the hydrolysis product(s), several different methods were used:

Elemental analysis showed that the composition (chlorine and aluminium contents) of aluminium chloride anhydrous that had previously been exposed to air to let it react (hydrolyze) to exhaustion is very similar to that of aluminium chloride hexahydrate (AlCl3x6H2O, CAS no. 7784-13-6), indicating that the hexahydrate is the main reaction product.

ATR-FTIR measurements indicated that the hydrolysis products of aluminium chloride anhydrous in air and in water, respectively, are very similar. The spectra datasets correspond to reference spectra of aluminium chloride hexahydrate, which is in line with the results from elemental analysis and points to the hexahydrate as the primary hydrolysis product, both in water and in air.

By means of XRD measurements, aluminium chloride hexahydrate was identified as the main component of the hydrolysis products of aluminium chloride anhydrous in water and in air, respectively. The results indicated that in the sample that had reacted in contact with air, slow secondary hydrolysis reactions continue to take place over a long period of time.

Although spectra obtained by solid27Al-NMR were highly similar for the educt (aluminium chloride anhydrous), its hydrolysis products in air and in water, and the reference substance aluminium chloride hexahydrate (octahedrally coordinated aluminium representing the dominant resonance peak), they were in line with the other characterization experiments in that the spectra of the hydrolysis products showed the highest correspondence to that of the hexahydrate.

Finally, thermal decomposition patterns and volatile reaction products were compared by means of TGA-FTIR and revealed a high degree of analogy between the hydrolysis products in air and water, respectively, and the hexahydrate.

In conclusion, the analytics experiments showed that:

hydrolysis of anhydrous aluminium chloride takes place extremely fast when in contact with air.

aluminium chloride hexahydrate is the main hydrolysis product formed during the reactions of anhydrous aluminium chloride in air or in water.

Consequently, aluminium chloride anhydrous is not a relevant test substance for a 90-day repeated-dose inhalation study. Therefore, the requested 90-day inhalation study has been performed with the reaction product aluminium chloride hexahydrate to evaluate the hazard potential associated with handling anhydrous aluminium chloride as a dust.

Another technical trial was performed at BASF's inhalation lab (no report generated, oral communication with laboratory) to explore if commercially available aluminium chloride hexahydrate is better suited for generation of a solid aerosol atmosphere. The results showed that aluminium chloride hexahydrate is also hygroscopic and formed liquid droplets when in contact with air. It was therefore not possible to load the brush generator's cartridge or to generate a solid aerosol atmosphere. However, a 10% aqueous solution was prepared which was then sprayed into the exposure chamber using compressed air. The water component of the aqueous solution seemed to rapidly evaporate during this process, because in the chamber, a stable and reproducible atmosphere of solid hexahydrate at a concentration of 333 mg/m³ was found. Particle sizes were reproducible and as expected (MMAD 3.9 µm, GSD 1.6). Concentrations were determined gravimetrically after storing the filter in a drying cabinet at 50°C for 30 minutes or at 40°C for 45 minutes with subsequent cooling in an exsiccator for 10 minutes.

This technical trial shows that a stable solid aerosol atmosphere of aluminium chloride hexahydrate (the main hydrolysis product of anhydrous aluminium chloride) can be generated from an aqueous solution using the procedure above.

Therefore, it has been proposed to test aluminium chloride hexahydrate in a repeated-dose inhalation toxicity study instead of anhydrous aluminium chloride, which is technically not feasible.

In decision SEV-D-2114385103 -55 -01/F, ECHA has requested a 90-d inhalation study using the registered substance "in the form adapted to the experimental design (hydratation degree to be specified)", which is interpreted by the Registrant as agreement to the procedure described above (spraying of an aqueous solution).

In the final study, Aluminium chloride anhydrous ground was dissolved in ultrapure water, where it formed a stable AlCl3 hexahydrate and after appropriate dilution, sprayed as aqueous solution. The stability of the so formed aluminium chloride hexahydrate in aqueous solution was examined by Al-NMR analyses. All cascade impactor measurements of the particle size resulted in MMADs between 0.90 and 1.39 μm with GSDs<3. Thus, the aerosols were highly respirable for rats and a very high proportion of the aerosol particles reached the lungs.

Results were as follows: Inhalation exposure of 0.1, 0.5 and 2 mg/m³ Aluminium chloride anhydrous ground for 90 days (65 exposures) did not cause any clinical signs of toxicity, or any changes in body weight, food consumption and clinical chemistry parameters. Treatment-related changes of lavage parameters and histological changes in lungs and nasal cavity (only at 2 mg/m³) were observed in males and females exposed to 0.5 and 2 mg/m³. In female animals slight changes of lavage parameters were still observed at 0.1 mg/m³. Based on changes of lavage parameters, a No Observed Adverse Effect Concentration (NOAEC) for local effects could not be established for female animals under the current study conditions. For male animals, the NOAEC for local effects was 0.1 mg/m³. Haematological examination revealed increased absolute and relative neutrophile counts and decreased lymphocyte count, which most likely attributed to the local inflammation. Based on these changes in the haematology the NOAEC for systemic toxicity has been set at 0.1 mg/m³.

Justification for classification or non-classification

Taking into account the whole of the available information, Aluminium chloride anhydrous has to be labelled with EUH 071 according to Regulation (EC) No. 1272/2008 (CLP).