Registration Dossier

Data platform availability banner - registered substances factsheets

Please be aware that this old REACH registration data factsheet is no longer maintained; it remains frozen as of 19th May 2023.

The new ECHA CHEM database has been released by ECHA, and it now contains all REACH registration data. There are more details on the transition of ECHA's published data to ECHA CHEM here.

Diss Factsheets

Environmental fate & pathways

Additional information on environmental fate and behaviour

Administrative data

Endpoint:
additional information on environmental fate and behaviour
Type of information:
other: experimental results on similar substance
Adequacy of study:
weight of evidence
Reliability:
2 (reliable with restrictions)
Rationale for reliability incl. deficiencies:
other: Literature data; no GLP.

Data source

Reference
Reference Type:
publication
Title:
Transformation of Dyes and Related Compounds in Anoxic Sediment: Kinetics and Products
Author:
George L. Baughman and Eric J. Weber
Year:
1994
Bibliographic source:
Environ. Sci. Technol. 1994, 28, 267-276

Materials and methods

Principles of method if other than guideline:
The reactions of several azo, anthraquinone, and quinoline dyes were studied in settled sediments not following an official method, nor guideline.
GLP compliance:
not specified

Test material

Constituent 1
Reference substance name:
Several azo dyes
IUPAC Name:
Several azo dyes
Details on test material:
CHEMICALS
1-aminoanthraquinone (97%) from Aldrich Chemical Co. (Milwaukee, WI)
1-chloroanthraquinone (98%) from Aldrich Chemical Co. (Milwaukee, WI)
Diazald from Aldrich Chemical Co. (Milwaukee, WI)
1-hydroxyanthraquinone (98+ %) from TCI America (Portland, OR)
sodium dithionite (purified) from Fisher Scientific (Atlanta, GA)
Aluminum powder from EM Science (Gibbstown, NJ)
Disperse Red 1 (DR1) (CI 11110)supplied courtesy of Atlantic Industries (Nutley, NJ)
Disperse Red 9 (DR9) (CI 60505)supplied courtesy of Atlantic Industries (Nutley, NJ)
Disperse Violet 1 (DV1) (CI 61100)supplied courtesy of Atlantic Industries (Nutley, NJ)
Disperse Red 11 (DR11) (CI 62015)supplied courtesy of Atlantic Industries (Nutley, NJ)
Solvent Red 1 (SR1) (CI 12150) supplied courtesy of Atlantic Industries (Nutley, NJ)
Solvent Yellow 33 (SY33) (CI 47000) supplied courtesy of Atlantic Industries (Nutley, NJ)
Disperse Blue 3 (DB3) (CI 61505) used as component of Sublaprint Blue 70013, supplied courtesy of Keystone Aniline Corp. (Chicago, IL)
Disperse Blue 14 (DB14) (61500) used as component of Sublaprint Blue 70013, supplied courtesy of Keystone Aniline Corp. (Chicago, IL)

Press cake samples of the dyes contained numerous components as shown by high-performance liquid chromatography (HPLC) and thin-layer chromatography
(TLC). Thus, all of the dyes were initially recrystallized from ethanovwater solutions. The recrystallized dyes were then flash-chromatographed on silica gel (32-63 pm, Universal Adsorbents). Due to the inability to obtain pure materials, DB3 and DB14 were used as the mixture isolated from Sublaprint Blue 70013. However, in the studies, each compound was resolved and measured separately.

Results and discussion

Any other information on results incl. tables

PRODUCT STUDIES

1-Chloroanthraquinone. In initial experiments without GPC sample cleanup, the detection of reaction products by GC-MS was not possible. Therefore, one of the first compounds examined was 1-CA on the premise that Cl would serve as a tracer in the mass spectra. The resulting mass chromatogram showed a Cl- containing component that eluted before 1-CA and which had an apparent molecular weight 14 less than the parent compound (MW 242). Based on the mass spectrum this component was presumed to be an anthrone. The reduction of 1-CA with sodium dithionite or Al resulted in reaction mixtures for which mass chromatograms showed two major components that had molecular weights of 228 and 244. The spectrum and retention time of the MW 228 component matched that of the component from sediment. The MW 244 component must be due to either the hydroanthraquinone or its tautomeric oxyanthrone. Selected ion mass scans showed that the other anthrone isomer coeluted with 1-CA and was thus hidden in the synthesis extracts. This product was not detected in sediment extracts. 1-CA is the only anthraquinone for which a hydroanthraquinone was detected, either from synthesis or in sediment. Dithionite reduction of 1-hydroxyanthraquinone resulted in product extracts that also indicated anthrone formation. In this case, the mass chromatogram showed that the isomers eluted on either side of the parent compound. However, 1-HA was not examined in sediment.

Disperse Red 9. By GC-MS analysis, the sediment reaction of DR 9 (mass 237) also was found to give two products (each with mass 223) eluting before and after the parent dye. Extracts of Al metal reductions had only two major products, and they matched the retention times and mass spectra of the components from sediment. The component eluting after the dye was significantly less stable than the other (ANT), and often it was not observed in sediment extracts, presumably because of the time required for sample preparation and analysis. The compound, ANT, eluting before the dye was synthesized and purified by flash chromatography in sufficient quantity for HRMS analysis and for NMR and melting point (mp) determination. The HRMS obtained by a solid probe gave a mass of 223.100, consistent with the formula C15H13NO. The spectrum also confirmed that the major fragment peak at mass 206.098 was due to the loss of OH. This suggests, but does not prove, that the earlier eluting product is the hydrogen-bonded isomer. The mp at 115-117 °C was higher than the 111-113°C reported for 1-methylamino (9,10H) anthraceneone. Isomer confirmation was obtained by GC/FTIR of the extracts. The FTIR spectrum of DR9 has carbonyl stretching peaks at 1686 (m) and 1643 (m) cm-1 corresponding to the free and intramolecularly hydrogen-bonded carbonyl groups, respectively. It also has an N-H stretching peak at 3308 (w) cml. The spectrum of the first eluting peak, ANT, has only one carbonyl stretching peak [1644 (s) cmll and an N-H stretching peak at 3307 (w) cm-1. The mass and infrared spectra combined provide proof that the first eluting product, ANT, is the hydrogen-bonded isomer, 1 -rnethylamino (9,10H)anthraceneone. The proton NMR spectrum of ANT shows a broad N-H resonance at 9.51 ppm with singlets at 4.23 and 2.95 ppm for the (2-10 methylene group and the methyl group of the amine, respectively. The other downfield signals are from the aromatic hydrogens and are between 6.5 and 8.5 ppm. All 15 carbon signals are present in the 13C spectrum, and DEPT 135 indicates the methyl group at 29.4 ppm, the methylene carbon at 33.2 ppm, and seven CHs between 107 and 134 ppm. The carbonyl carbon is located at 186.8 ppm. The complex pattern of aromatic hydrogens was resolved in a C-H correlation experiment into the expected four doublets and three triplets. A H-COSY experiment indicates coupling between the aromatic hydrogens but also shows some coupling between the protons of the methylene group and the aromatic protons. In sediment, there was no evidence of N-demethylation of DR9 (i.e., no 1-aminoanthraquinone (1-AA) was detected) even though the possibility was always examined. Similarly, the 4-hydroxy compound was never detected as more than a trace component. Further, the 2-hydroxyanthraquinones are too involatile for GC and, hence, would not have been detected from either 1-AA or DR9. These findings are quite different from results on metabolism in sheep, which have an anaerobic rumen. In that case, metabolites were 1-aminoanthraquinone, its 2- and 4-hydroxy derivatives and their glucuronides.

1 -Arninoanthraquinone. Results with 1-AA were similar to those with DR 9. GC-MS analysis showed 1-AA (mass 223) to react in sediments with the formation of one anthrone (mass 209). The other anthrone could not be detected. Reduction of 1-AA with aluminum resulted in the formation of both anthrones, one of which eluted on either side of the parent compound. Like the other anthrones, both isomers have the molecular ion as the base peak in the mass spectrum. Also, the second and third largest peaks (masses 180 and 152, respectively) are the same in both spectra. Other peaks were in the background. By analogy with the anthrones from DR9, the first eluting peak for 1-AA was assumed to be the hydrogen-bonded isomer. This was also confirmed by GC-FTIR. Data suggest that, for mono 1-substituted anthraquinones in anoxic sediments, the initial transformation step is the reduction to anthrones. Also in the case of DR9 and 1-AA, the hydrogen-bonded isomers are stable but the non-hydrogen-bonded isomers break down rather easily. The more highly substituted compounds behaved quite differently.

Disperse Violet 1. Upon chemical reduction, DV1 can be converted to the “leuco” form of the dye, 1,4 -diamino-2,3 -dihydroanthraquinon (DDA) (mass 240). DDA, however, is reported to be unstable toward hydrolysis, the products of which react with ferric ions to produce a fluorescent compound. Furthermore chemical reduction of DV1 (followed by aeration) yields quinizarin, MW 242. Quinizarin is probably the compound reacting with ferric ions to produce the fluorescent product; thus, quinizarin can also be anticipated to undergo metal complexation and reduction to leuco-quinizarin in sediments. Many different sediment experiments were conducted with DV1 in attempts to identify products, especially DDA. Unfortunately, only two products were identified with certainty-quinizarin and 1-hydroxyanthraquinone. Both compounds were identified by comparison of GCMS retention times and spectra with those of purchased material. Also, quinizarin was identified by HPLC with fluorescence detection. Both the sample peak and quinizarin had exitation and emission maxima at 490 and at 565-570 nm, respectively. In all cases, the product peaks were very small. Other products tentatively identified by GC-MS were leuco-quinizarin and 1-aminoanthraquinone. Many anthraquinones and related compounds were difficult to identify with certainty at trace levels by GC-MS because the molecular ion is the base peak in the mass spectrum and other peaks are often of low intensity. The reaction of DV1 with dithionite resulted in two major products. These were identified as quinizarin and leuco-quinizarin as expected. Identification was based on GC-MS, HPLC, and TLC for quinizarin. GC-MS of the reaction mixture shows quinizarin (mass 240) and the leuco compound (mass 242). HRMS of these components resulted in molecular weights of 240.043 and 242.059. These results strongly suggest that Disperse Violet 1 is transformed in sediment to products, perhaps including DDA, that are converted to compounds containing OH groups in place of NH2. As noted above, such products are likely to reoxidize during workup. Also, the complexation by metals in sediment could adversely affect extraction (recovery) of quinizarin and other 1-hydroxy compounds. The addition of large amounts of EDTA, however, did not result in increased recovery of quinizarin.

Disperse Red 11. Experiments with DR 11 showed the presence of a major product that was not cleanly separable from the parent dye by HPLC, was not amenable to GC-MS analysis, could not be extracted from aqueous alkali, and had a diode array spectrum (UV-vis) very similar to that of DR11. The electron impact mass spectrum is typical of an anthraquinone with MW 254 and was tentatively ascribed to 1,4-diamino-2-hydroxyanthra quinone (HA), the O-demethylation product. To verify the identity of this reaction product, the methyl group was cleaved from the parent dye with cold concentrated sulfuric acid to give a red compound that could not be extracted from basic aqueous solutions. The HPLC/TLC retention characteristics, diode array and mass spectra matched that of the sediment transformation product. HRMS confirmed the expected fragmentation as m/e 254, molecular ion; m/e 226, M-CO; and m/e 197, M-NH2C2OH. This compound has not been reported previously in the chemical literature.

HA was also examined by NMR (lac and lH). The spectra show the presence, for the same ring, of an alcohol proton at 11.5 ppm and an aromatic proton at 6.6 ppm. Two additional sets of aromatic protons are located at 8.2 ppm (C-5 and C-8 protons) and at 7.7 ppm (C6 and C7 protons).

Only 13 carbon signals are present in the 13C spectrum since there is a high intensity peak at 125.7 ppm due to two carbon resonances. The carbon bearing the hydroxyl group resonates at 153.4 ppm. The DEPT 45 spectrum has five signals due to the five aromatic CHs. The parent compound, containing a methoxy group at C-2, shows only the aromatic protons and the methyl signal at 3.1 ppm. There are no other resonances downfield of 8.0 ppm. These NMR results unequivocally established HA as the 1,4 -diamino-2-hydroxy compound. The FTIR spectrum is also consistent with this conclusion. No other products were verified, either from DR11 or from its daughter product, HA, though many attempts were made. It is probable, however, that products are formed from either the dye or HA as a result of the replacement of one or more of the amino groups with protons or a hydroxy group. For example, quinizarin was identified tentatively in several product runs but at concentrations too low for confirmation. The above facts are consistent with the observation that the reaction between dithionite (or aluminum) and the dye results in the formation of 1-amino-2-methoxyanthraquinone as the major product. Similar reduction of HA resulted in the analogous 2-hydroxy compound. As noted earlier, anthraquinones having a hydroxyl group in Ifthe above conclusion is correct, other products may still not have been detected because they are not amenable to GC-MS analysis, they cannot be extracted because they are good complexing agents, or their chromatographic peaks were hidden behind those of the sediment background components.

Solvent Red 1. Products of the azo dye, SR 1, were expected to be o-anisidine and 1-amino-2-naphthol resulting from reductive cleavage of the azo bond. However, these compounds could not be identified by HPLC because they coeluted. After GPC cleanup, each product was detected by GC-MS with both spectra and retention times identical to those of purchased materials. Several other azo dyes have been studied in sediment systems. In all cases, the primary products are amines resulting from nitro group reduction and/or cleavage of the azo bond. This conclusion is also consistent with environmental data on DB79 in that one of its cleavage products recently has been reported in river water. Thus, we can confidently conclude that reductive cleavage of azo groups is general in anoxic environments.

Solvent Yellow 33. Although the kinetics of this quinoline dye were studied in three different sediments, the dye was so unreactive that a product was only identified from Herrick Lake sediment that had been standing with dye for almost 2 years. Detection of the product was prompted on discovering that reduction of SY33 (mass 273) with aluminum resulted in the formation of a minor, stable product having a molecular weight of 275. Selected ion GC-MS analysis of the sediment extract after GPC cleanup showed the presence of a very small amount of a compound having the same retention time and mass spectrum. GC-HRMS of the product gave a molecular weight of 275.098 corresponding to C18H13NO2. GC-FTIR analysis of the synthetic product shows a strong carbonyl frequency at 1802 cm-1 (reasonable for a five-membered ring) and a weak aliphatic CH at 2932 cm-1. The parent dye has no aliphatic CH and its carbonyl frequency is 1815 cm-1 . The spectrum of neither compound has an OH band. Many attempts were made to reduce SY33 in sufficient quantity for NMR. However, the yield was always low, and the compound decomposed on flash chromatography. HPLC, with diode array detection, of a small amount of purified material showed that the product has no absorption maximum in the visible region.

KINETIC RATE

Constants

The half-lives of the dyes vary from 0.1 to 140 days or about 1000-fold. However, except for SY33, only the two anthraquinone dyes, DB3 and DB14, have half-lives as long as a few weeks. Thus, SY33 is the only dye that might persist in anaerobic sediments. The stability of SY33 is consistent with results for two other quinoline dyes, Disperse Yellow 54 and Disperse Yellow 64, which also were found to be very stable in sediment. It is possible that the slowness of reaction is due to metal complexation and/or existence of the dye as a zwitterion. The latter, and possibly the former, may also lead to sorption of the dyes by cation exchange.

As expected, there is considerable scatter in the rate constants. Nevertheless, it is surprising that the overall averages of the rate constants for dye loss have coefficients of variation that are less than 100%, except for DR11. During the summer, DR11 kinetics in Herrick Lake sediment were very erratic and, sometimes, unusually fast. This may have been due to an algal bloom since the sediment extracts at those times were quite green. A possible, but improbable, source of variability in the rate constants is the presence of solid dye in the sediment. Systematic studies of variation in transformation rate constant with initial dye concentration were not conducted because of analytical limitations and because of the need to conduct studies below the calculated dye saturation

limit. However, exceeding the saturation limit would decrease the apparent initial rate. Moreover, the extrapolated initial concentration, Di, would be greater than that actually prepared. Precisely the opposite situation was nearly always observed and can be accounted for by a faster reaction that terminates quickly. A fast initial reaction also has been observed in the sediment transformation of halogenated compounds. Little is known about temperature effects on sediment transformations, so approximate activation energies were determined for the disappearance of DR9 and DR11.

The activation energies to be in the range of 50-80 kJ/mol. Although such limited data are of little mechanisticvalue, they can be used to estimate the effect of temperature differences. The data are not adequate for assessing temperature effects on product distribution.

Kinetic model

To permit a more detailed kinetic analysis of transformation of DR9 and DR11 formation and decay of ANT and HA were determined simultaneously by HPLC.

Unfortunately, experiments for following product disappearance were so lengthy that only a few could be performed. For those experiments, rate constants kl, k2, and k3 were extracted individually by curve fitting with the NLIN procedure of SAS.

Application of the model to DR9 and DR11 gave comparable results by regression and by curve fitting. The lines were computed using the average

rate constants obtained by regression. Curves for other runs gave similar fits. This clearly suggests the formation of undetected transformation

products of the dye. Undoubtedly, the unstable, non-hydrogen bonded anthrone is a major component of the other reaction products that could not be detected. The reverse situation is found for DR11, which suggests that the formation of the product HA is the dominant, but probably not the only, reaction. Taken with our results and the known biochemistry of aromatic O-demethylation, it is probable that the reaction, like azo bond cleavage, is generally facile in anoxic sediments. Another product that might be expected from transformation of DR11 is suggested by the foregoing in combination with our following results for chemical reduction. The major products, isolated after dithionite or aluminum treatment of DR11 or HA, result from replacement of the 4-amino group with a proton, yielding mass 253 and 239, respectively. Unfortunately, the mass 239 product does not go through the GC and could only be seen after methylation. Further, these products suggest that the reactions may be under thermodynamic rather than kinetic control. Also, it is reasonable to expect that reduction products that do not survive the synthesis workup would not survive isolation from sediment. This leads to the tentative reaction scheme for DR11. In our experiments, neither product was seen either before or after methylation. However, under these conditions observed, it is probable that neither product would be seen even if special studies were designed for that purpose.

Factors Influencing Sediment Kinetics

A number of factors might be expected to significantly influence kinetics in sediment systems. They include organic carbon content, presence of solvent, role of microorganisms, sediment handling practices, and particle size distribution. Although these factors were not examined systematically in this study, it seems important to relate our observations because their role is virtually unknown.

The organic carbon content of the different sediments used for this work has about a 5-fold range. However, the precision of the analyses is only 50%. Since it is well known that organic content varies inversely with particle size, the definition of the influence of these parameters on sediment activity will require very detailed and careful experiments. Little information is available to guide the handling of sediment or to suggest how stirring and sieving (aeration) might affect activity. We consistently found that stirring and mixing in air (required for sediment preparation) affected the kinetics erratically. For example, experiments with DR5 and SR1 (identical experiments except for air stirring) showed the effect to be as large as a factor of 6.

The effect, as anticipated, was most pronounced for the fastest reactions. Thus, time constraints due to sample processing and analysis clearly indicate that half-lives of less than several hours to 1 day must be considered upper limits. The possible role of microorganisms was examined in

several ways. Phenol (1 9%) and NaNO2 (0.01 and 0.03 M) were added in some kinetic experiments to determine the effect of these biological inhibitors. For DR5 and NaN02, the higher concentration clearly slowed the reaction.

However, replicates with phenol reacted slightly faster than those without phenol. Phenol in the presence of DR11 had only a minor effect on the rate of dye loss compared to controls, but no product (HA) was formed. Heat sterilization of the sediment before the dye addition also dramatically reduced the rate of dye loss. In other experiments, the sediment was respiked with dye upon termination of the kinetic run. In each case, the respikes had a lower rate but only by about a factor of 2. This is consistent with the fact that the reactions usually deviate from first order after a few half-lives. The above results are also consistent with reactions that are not directly metabolic but that are biochemical in nature.

The effect of solvent (added with the dye) on the rate constant was examined for DR9. Data recorded show that there is no perceivable difference in rate constants over a 10-fold range in ACN concentrations below 1.5%.

Similar results have been found for other compounds. This does not guarantee, however, that reactions of other compounds will be insensitive to the solvent.

CONCLUSION

In summary, it can be stated with reasonable certainty that, in most anoxic sediment environments, uncharged azo and simple anthraquinone dyes will not persist. However; the quinoline dyes are much more stable than the azo and anthraquinone dyes under anoxic conditions. Solvent Yellow 33 is the least stable of three quinoline dyes that have been studied and it has a half-life of approximately 6 months.

Disperse Red 9 and 1-aminoanthraquinone are transformed partially to hydrogen-bonded anthrones that are much more stable than the isomer that cannot hydrogen bond. Disperese Red 11 undergoes rapid 0-demethylation as its major first step, giving 1,4-diamino-2-hydroxyanthraquinone.

The products from both dyes are much more stable than the parents and have half-lives of a few months in anoxic sediments. The anthraquinone dyes containing amino or alkylamino groups in the 1- and 4-positions also are lost from anoxic sediment with half-lives less than a few weeks. The transformation pathways for these dyes are not known, but are more complex than for the other anthraquinones examined.

Applicant's summary and conclusion

Conclusions:
Considering the conclusion that in most anoxic sediment environments, uncharged azo and simple anthraquinone dyes will not persist, and given the polyazo moieties of DBk22, it is likely that DBk22 will follow the same fate when adsorbed to sediment.
Executive summary:

The reactions of several azo, anthraquinone, and quinoline dyes were studied in settled sediments.

Result

Several 1-substituted anthraquinones were lost from sediment with halflives less than 10 days. For monosubstituted 1-amino and 1-methylamino (Disperse Red 9) compounds, the most stable product is the intramolecularly hydrogen-bonded anthrone. The 1,4-diaminoanthraquinone (Disperse Violet

1) and 1,4-diamino-2-methoxyanthraquinone (Disperse Red 11) were lost without formation of detectable products except for a demethylation product of the latter. Both the anthrone from Disperse Red 9 and the demethylation product of Disperse Red 11 reacted with halflives of a few months, but other major products were not detected. An azo dye (Solvent Red 1) and a quinoline dye (Solvent Yellow 33) were transformed with half-lives of a few days and months, respectively. The azo dye reacted by reductive cleavage of the azo bond.

Conclusion

Considering the results achieved and the conclusion that in most anoxic sediment environments, uncharged azo and simple anthraquinone dyes will not persist, and given the polyazo moieties of DBk22, it is likely that DBk22 will follow the same fate when adsorbed to sediment.