Detection and determination of aromatic amines as products of reductive splitting from selected azo dyes

Article abstract of DOI:10.1006/eesa.2002.2191

The current environment-friendly regulations concerning textile products ban the marketing of textiles dyed with azo dyes capable of reductively splitting carcinogenic aromatic amines. The study analyzes seven azo dyes whose chemical structure determines various quantities of splitting aromatic amines, such as benzidine. For tests, seven commercially available azo dyes with aromatic amines in their structure were selected. These included two hazardous dyes: Acid Red 85 and Direct Blue 6, both capable of reductively splitting carcinogenic benzidine. Of the remaining five azo dyes, three - Ponceau SS, Sudan II, and Disperse Yellow 7 - are capable of splitting p-phenylenediamine and aniline, while Mordant Orange 1 and Disperse Orange 3 can split only p-phenylenediamine. For Acid Red 85 and Direct Blue 6, the quantity of benzidine split from them was analyzed, depending on the conditions of the reduction process (e.g., in the HPLC method, 104 g/kg of dye for reduction in NaOH, and 41 g/kg of dye for reduction in acetate buffer). The spectrophotometric method proved useful for preliminary analysis of amine content in examined samples. Spectrophotometric analysis may be used to determine the total content of amines counted as aniline. A full qualitative and quantitative analysis of amines released from azo dyes is possible using high-performance liquid chromatography (HPLC).

Full text of DOI:10.1006/eesa.2002.2191

Ecotoxicology and Environmental Safety 53, 42}47 (2002)
Environmental Research, Section B
doi:10.1006/eesa.2002.2191
Detection and Determination of Aromatic Amines as Products
of Reductive Splitting from Selected Azo Dyes
A. Pielesz,*ꢁꢀ I. Baranowska,- A. Rybak-, and A. W"ochowicz*
*University of Bielsko~Biala, Faculty of Textile Engineering and Environmental Protection, Textile Institute, Bielsko~Biala,
Willowa 2, Poland; and -Departament of Analytical and General Chemistry, Silesian Technical University, M. Strzody 9, Gliwice, Poland
Received May 15, 2001; published online June 28, 2002
draw up lists of dangerous azo dyes and amines [3, 22, 23].
The result is a systematic removal from the market of those
dyes that are suspected to have carcinogenic or mutagenic
in#uence on humans. The main hazard criterion for a dye is
its ability to split reductively aromatic amines in contact
with sweat, saliva, or gastric juice [12]. The process of the
The current environment-friendly regulations concerning tex-
tile products ban the marketing of textiles dyed with azo dyes
capable of reductively splitting carcinogenic aromatic amines.
The study analyzes seven azo dyes whose chemical structure
determines various quantities of splitting aromatic amines, such
as benzidine. For tests, seven commercially available azo dyes
with aromatic amines in their structure were selected. These
included two hazardous dyes: Acid Red 85 and Direct Blue 6,
both capable of reductively splitting carcinogenic benzidine. Of
the remaining 5ve azo dyes, three0Ponceau SS, Sudan II, and
reduction of azo dyes with the cleavage of aromatic R}NH
amines is one of the means of degradation of those dyes.
ꢂ
Figure 1 illustrates how amines are split from an azo dye.
Other methods of splitting amines from dyestu!s are photo-
Disperse Yellow 70are capable of splitting p-phenylenediamine degradation and biodegradation by means of hydroxyla-
and aniline, while Mordant Orange 1 and Disperse Orange 3 can
split only p-phenylenediamine. For Acid Red 85 and Direct Blue
6, the quantity of benzidine split from them was analyzed,
depending on the conditions of the reduction process (e.g., in the
HPLC method, 104 g/kg of dye for reduction in NaOH, and
41 g/kg of dye for reduction in acetate bu4er). The spectro-
photometric method proved useful for preliminary analysis of
amine content in examined samples. Spectrophotometric analy-
sis may be used to determine the total content of amines counted
as aniline. A full qualitative and quantitative analysis of amines
released from azo dyes is possible using high-performance liquid
chromatography (HPLC). ꢀ 2002 Elsevier Science (USA)
tion, oxidation, or hydrolysis. However, in humans, the
biological reduction of an azo dye is responsible for the
possible presence of toxic amines in organism. Aromatic
amines are speci"ed [12, 23, 20] in groups III A1 and III A2
of the Maximale Arbeintsplatz Konzentration (MAK) list as
well as in International Agency for Research on Cancer
(IARC) and Ecological and Toxicological Association of the
Dyestu!s Manufacturing (ETAD) lists (e.g., benzidine,
o-toluidine, 4-aminodiphenyl). The ability to form any of the
listed carcinogenic amines through the cleavage of one or
more azo groups is the reason for classi"cation as a banned
azo dye.
Key Words: azo dye; aromatic amines; spectrophotometric
method; HPLC.
Legally, the issue of toxic chemicals (including aromatic
amines that split from dyestu!s or dyed textiles) has been
regulated by the European Union directive [6]. However,
according to the European Union Law, each member state
may de"ne its own regulations to protect people's health.
Thus, "rst Germany [5] and then other European Union
countries approved bans on importing and marketing tex-
tiles dyed with dyes capable of reductively splitting carcino-
genic amines. In Poland also [5], e!orts are being made to
pass a relevant act complying with the European Union
directives. The danger to the user of a dyed product is
caused not only by the hazardous dye itself, but also by the
aromatic amine that splits o!.
INTRODUCTION
Publications in the literature [1, 3, 4, 11] have set forth
evidence that azo dyes can pose threats to public health. It is
claimed that textile chemicals not only act as allergens but
also exhibit toxic and even mutagenic or carcinogenic prop-
erties [2]. The sources of mutagenic activity are the dyes-
tu!s and the amines contained in their chemical structures.
The literature [3, 4] quotes opinions that the dyestu!s
themselves provide only a minor part of the mutagenic
potential, and for many years work has been conducted to
Regarding textile products, European Union law pro-
vides a list of hazardous substances and also speci"es their
maximum acceptable quantities and obligatory detection
ꢀ To whom correspondence should be addressed. E-mail: apielesz@
pb.bielsko.pl.
42
0147-6513/02$35.00
ꢀ 2002 Elsevier Science (USA)
All rights reserved.

DETERMINATION OF AROMATIC AMINES IN AZO DYES
43
EXPERIMENTAL
nR
mR
In Fig. 2the structural formulas of the dyes selected for
N
N
analysis are found. Acid Red 85 and Direct Blue 6 are
industrial dyes no longer produced; they were supplied by
Instytut Barwnikow i Produktow Organicznych from
Zgierz, Poland. Ponceau SS, Sudan III, Disperse Yellow 7,
Mordant Orange 1, and Disperse Orange 3 were supplied
by Sigma}Aldrich, Poland. The following reagents and
solutions were used for analyses: 2% solution of J-naph-
thol (p.a. Fluka, Sigma}Aldrich) in ethanol; 1 M sodium
reduction
nR
mR
nitrate NaNO (p.a. Fluka, Sigma-Aldrich); sodium hydrox-
ꢂ
H2N
NH2
ide NaOH (p.a. Fluka, Sigma-Aldrich); solutions 1 M and
+
30%; 1 M HCl (p.a. Fluka, Sigma-Aldrich); sodium dithion-
ite Na S O (p.a. Fluka, Sigma-Aldrich); methylene chlor-
ide absolute (Merck, Poland).
ꢂ ꢂ ꢃ
FIG. 1. Reductive splitting of an aromatic amine from an azo dye.
Two methods of chemical reduction were adopted. More
drastic conditions of reduction [21] involved 0.05 g of dye
(i.e., the quantity "tting to standard of 3% dyeing or 4 g of
methods. It is assumed that the legally allowed con- dyed cotton fabric with 3% dyeing). The sample was boiled
centrations of aromatic amines in textiles should be deter- in 1 M NaOH (30 ml) for 1 h in a round-bottom #ask under
mined by means of the TLC, GC/MS, or HPLC methods a re#ux condenser. After 30 min 0.8 g of sodium dithionite
[23, 7, 13, 14], and these are currently the routine proced- was added, and reduction was continued. The entire process
ures. It should be stressed, though, that those regulations lasted 1 h. For less drastic conditions [1], the reduction was
are binding only for the European Union countries, and the carried out a 703C at pH 6, using 17 ml of citrate bu!er with
limits do not have to be observed in Asian countries [19] or 0.8 g of sodium dithionite per 1 g of "ber. Other conditions
in East European markets.
were identical to those of the previous method.
It is therefore necessary to seek alternative, inexpensive
In each case, the extract obtained was transferred to
methods of chemical analysis of carcinogenic amines Extrelut 20 columns (Merck) to which 17 ml of methylene
in dyestu!s or dyed textiles. The aim is to be able to chloride was subsequently poured. Next, the samples were
quickly detect those products that do not conform to evaporated to dryness, and the remains were dissolved in
the Chart of Substance Safety Characteristics. The Chart volumetric #asks in 100 ml of 0.1 M HCl. For evaporating
is an attachment that producers supply to the buyers of the samples, a 350P vacuum evaporator (EQUIMED, Po-
their dyestu!s. It contains key product information (includ- land) was used. The temperature was set at 403C.
ing trade name, physical and chemical properties, toxicolo-
gical data) describing the product as a nonhazardous absorption spectrophotometer (Carl-Zeiss Jena Germany),
substance. using the method of aromatic amine determination de-
Spectrophotometric analysis was carried out in a Specol
Regardless of the method of determining the content of veloped by Maslowska and Swat [10], with slight modi"ca-
aromatic amines in samples, an area that needs to be ad- tion. In the original method, aromatic amines are isolated
dressed is that of structure}activity relations. These rela- using distillation with steam. In the current study, the
tionships reveal how the reduction conditions a!ect the chemical reduction was carried out as described above.
quantity of amines splitting o! a dyestu!. It is worthwhile to The spectrophotometric method of determining trace
determine whether this quantity is somehow dependent on amounts of aromatic amines [10] utilizes the color
the dye structure, i.e., the presence and position of speci"c reaction of diazotization of amines with J-naphthol. It
functional groups in the dye. Usually [11, 8, 15, 16, 9], only allows the determination of total aromatic amine content as
genotoxic, mutagenic, or carcinogenic properties of azo dyes aniline.
are examined, which result from the presence of certain
The working solution of J-naphthol was prepared im-
functional groups within the chemical structure of the mediately before carrying out the diazotization reaction.
dye. Here, a di!erent approach is suggested: combining A 100-ml #ask was "lled with 5 ml of 2% J-naphthol
the analytical objective, i.e., determining the amine solution in ethanol and 50 ml of 1 M NaOH solution. Stock
content in a sample, with a structural interpretation, i.e., solutions were prepared, using in each case the amine that
explaining how the chemical structure of a dyestu! and the split o! the dye analyzed, e.g., benzidine for Acid Red 85 and
reduction conditions a!ect the quantity of a released aro- p-phenylenediamine for Disperse Orange 3. Each solution
matic amine.
was made in the following way: 0.1000 g of amine was

44
PIELESZ ET AL.
Acid Red 85
HO
O
S
N=N
O
N=N
H
C
3
O
Na
3OS
SO
3Na
Direct Blue 6
H2N
NH2
OH
OH
N=N
SO3Na
N=N
Na
SO3
NaO3S
NaO3S
Ponceau SS
HO
SO3Na
N=N
N=N
SO
3Na
Sudan III
HO
N=N
N=N
Disperse Yellow 7
Mordant Orange 1
N=N
N=N
OH
CH3
COOH
OH
O2N
N=N
Disperse Orange 3
N
N
NH2
O2
N
FIG. 2. Structural formulas of the dyes selected for analysis.
weighed and transferred to a 100-ml volumetric #ask, to
For further examination, a working solution was pre-
which 10 ml of 1 M HCl was added together with distilled pared with a concentration of 100 ꢀg/ml (by diluting 10 ml
water. If an amine did not dissolve in the acid, hot water was of stock solution in a 100-ml volumetric #ask). Next, 0.1, 0.2,
added.
0.3, 0.4, , 1 ml of the amine solution were poured in
2

DETERMINATION OF AROMATIC AMINES IN AZO DYES
45
TABLE 1
Content of Aromatic Amines Determined by Spectrophotometric and HPLC Methods
Content of aromatic amines (g/kg dye)
No.
Dye?
Splitting amine
Benzidine
Spectrophotometry
HPLC
1.
2.
3.
4.
5.
6.
7.
Acid Red 85
Direct Blue 6
Ponceau SS
Sudan III
Disperse Yellow 7
Mordant Orange 1
Disperse Orange 3
128$0.11
23.2$0.07
160$0.07
110.8$0.10
256$0.07
94.7$0.10
143.6$0.10
104
24.6
129.1; 31.5
73.7; 15.1
165.9; 89.2
132.8
Benzidine
p-Phenylenediamine, aniline
p-Phenylenediamine, aniline
p-Phenylenediamine, aniline
p-PD
p-PD
0.088
?Dyes were analyzed under more drastic conditions of reduction process.
a series of test tubes with glass stoppers. Each sample was group is observed for water-soluble dyes [4], i.e., those with
supplemented with 1 ml of 1 M HCl and 0.2ml of 1 M groups such as }SO Na and }COOH in their structure.
ꢅ
NaNO . After 3 min of shaking, the test tubes were "lled
Those criteria were re#ected in the analyses of Ponceau
ꢂ
with 5 ml of freshly prepared working solution of ꢀ-naph- SS, Sudan III, and Acid Red 85, which have an }OH group
thol and "lled to 10 ml with distilled water, taking into in an ortho- position in relation to the azo bond and thus
consideration the amount of amine solution they already are easily reduced to one or more amines. The splitting of
contained. A zero sample was prepared in the same way, substantial amounts of amines for Disperse Orange 3 and
without the amine stock solution and "lled with 9 ml of Mordant Orange 1 is caused by the presence of a nitro
distilled water.
group in the dye structure. Nitro groups are reduced [18]
Absorbance was measured with the spectrophotometer at much more easily than azo groups. Introducing an }OH
wavelength ꢁ"520 nm. For study samples, 0.5 or 1 ml of group in a para- position in relation to the azo bond makes
each solution (product of dyestu! reduction) in 0.1 M HCl the dye less alkali-resistant [18], which, indirectly, may also
was used, with the remaining components added as de- facilitate reduction. This e!ect was observed for Disperse
scribed above.
Yellow 7. Only small amounts of amines split o! Direct Blue
Chromatography was performed using a Merck}Hitachi 6. This weak reactivity was caused by steric hindrances to
L4500A HPLC chromatograph equipped with a diode- the reduction process [11].
array detector, an L 7840 #uorescence detector, and column
packing of the chemically bonded alkyl stationary phases from the presence of speci"c functional groups in the azo
(Li Chrosorb C (7ꢀm) (250;4 mm)). dyes examined, are quantitatively justi"ed by the data in
The structural conditions described above, stemming
ꢀꢄ
From 100 ml of the samples dissolved in 0.1 M HCl, 3 ml Table 1. It lists the amounts, obtained by means of two
of each solution was taken to be vaporized, with the dry measuring methods, of aromatic amines released by the azo
remains dissolved in 1 ml of methanol. Twenty microliters dyes during their reduction, under drastic conditions of
of the solution was injected into the column. The meth- a strongly alkalic environment and at boiling point. The
anol:water (55:45 v/v) mixture was used as the mobile phase table proves that spectrophotometry allows determination
[14], at a #ow rate of 1.2ml/min.
of the total amount of amine (recounted as total aniline
content) that splits o! a dye during its chemical reduction.
For three dyes*Ponceau SS, Sudan III, and Disperse
Yellow 7*on the basis of the well-known procedure of dye
reduction (see Fig. 1), which involves the breaking o! of azo
bonds and recreation of amines, it is justi"ed to say that two
amines (p-phenylenediamine and aniline) split o! rather
than one. In this case, determining the amount of each
amine is possible only by using HPLC. Then comparable
amounts of released amines are expected, regardless of the
measuring method applied. This is con"rmed by Table 1.
For example, when Ponceau SS is reduced, 160 g of amine
splits o! every 1 kg of dye (in the spectrophotometric
method). For the same dye analyzed using HPLC, 129.1 g
RESULTS AND DISCUSSION
It is well known [17] that the degree to which a dye azo
group is reduced depends on the electron density around
the }N"N} bond. Electron-withdrawing groups such as
}NH and }OH decrease the electron density around the
ꢂ
}N"N} bond and facilitate the reduction of the azo group
with a simultaneous release of an aromatic amine. Also,
placing an electrodonor substituent (ED) in an ortho- posi-
tion in relation to the azo group causes a reduction through
the creation of hydrogen bonds on the nitrogen azo group.
A similar e!ect in a simpler reduction within the }N"N}

46
PIELESZ ET AL.
TABLE 2
Content of Aromatic Amines Determined by Spectrophotometric and HPLC Methods
Content of aromatic amines (g/kg dye)
No.
Dye and method of chemical reduction
Splitting amine
Spectrophotometry
HPLC
1.
2.
3.
4.
5.
Acid Red 85 in NaOH
Acid Red 85 in pH 6 bu!er
Direct Blue 6 in NaOH
Direct Blue 6 in pH 6 bu!er
Direct Blue 6 fabric in NaOH
Benzidine
Benzidine
Benzidine
Benzidine
Benzidine
128$0.11
39.2$0.10
23.2$0.07
17.5$0.09
10.7$0.07
104
41
24.6
21.9
6
of p-phenylenediamine and 31.5 g of aniline splits o! every color reaction waste. Also, it is probably more di$cult to
1 kg of dye, which gives about 160 g of amine per 1 kg of reduce a dye bound with a "ber than a dye that is an
dye.
Depending on the method of chemical reduction, di!erent
independent competent in a chemical reaction environment.
Comparing both analysis methods (spectrophotometric
amounts of splitting aromatic amines were observed and HPLC), one can see that similar amounts of amines are
(Table 2). It seems that a strongly alkaline environment is split o! during dye reduction, regardless of which technique
most conducive to dye degradation. For example, Acid Red has been employed. The absolute amount of amine produc-
85, when reduced in 1 M NaOH, splits o! 128 g of benzidine ed during the reduction depends on the structure of the dye
per 1 kg of dye analyzed spectrophotometrically, compared analyzed, that is, on the type of functional groups the
with 104 g of benzidine for the HPLC method.
dye contains, either facilitating or hindering the reduction
Less amine is released during a chemical reduction in process.
acetate bu!er. The smallest amount of amine splits o!
The spectrophotometric method, being simpler and less
during the reduction of a dye that is bound on a cotton expensive, may be used for preliminary analysis and detect-
fabric dyed with an azo dye. This indirectly explains the ing, generally, aromatic amines as impurities or dye reduc-
process of dye depletion from a dyeing bath. Normally, the tion products. When interpreting the results of the
fabric is dyed with 3% dyeing, but only a part of the dye spectrophotometric analysis, whether the dye is the product
binds to the fabric, the rest being retained in the solution as of a coupling reaction between an amine and ꢀ-naphthol
aniline
1400
1200
1000
1,4 - phenyldiamine
800
600
400
200
0
0
1
2
3
4
5
6
7
8
9
10
Retention Time (min)
FIG. 3. Chromatogram of separation of aromatic amines: aniline (t "5.34 min) and 1,4-phenylenediamine (t "8.61 min).
ꢀ
ꢀ

DETERMINATION OF AROMATIC AMINES IN AZO DYES
47
TABLE 3
5. Dritte Verordung zur Anderung der Bedarfsgegenstandenverordnung
vom 16 Dezember 1994, Bundesgesetzblatt, Jahrgang, Teil INRF
(1994).
Statistical Test in the Chromatography of Amines (n ؍ 3)
6. European Union Directive No. 76/769/EEC.
Amine
Added (ꢀg)
Xꢂ
SD
Recovery (%)
7. Idaka, E., and Ogawa, T. (1978). Degradation of azo compounds by
Benzidine
Aniline
p-Phenylenediamine
10
10
10
9.73
10.30
9.85
0.58
0.87
0.70
97.2
102.9
98.5
Aeromonas hydrophilia var.24B. JSDC 3, 91}94.
8. Kalopissis, G. (1991). Structure}activity relationships of aromatic
amines in the Ames Salmonella typhimurium assay. Mutat. Res. 246,
45}66.
9. Kalopissis, G. (1992). Structure}activity relationships of aromatic dia-
mines in the Ames Salmonella typhimurium assay. Part II. Mutat. Res.
269, 9}26.
should be taken into account. The ꢁ
ꢁꢂꢃ
should be selected
individually for each determined aromatic amine, as the
dyes produced in the diazotization reaction have diversi"ed
coloring. Also, it should be stressed that in the spectro-
photometric method the accuracy of amine determination is
limited by the adopted procedure of recounting amine as
aniline.
10. Maslowska, J., and Swat, B. (1990). A method for determination of
amounts of aromatic amines in dyed cotton textiles. Ann. Natl. Hygiene
Instit. 41(5, 6), 285}289.
11. Peters, A.T., and Freeman, H. S. (1996). Genotoxicity of azo dyes: Bases
and implications. In Physico-Chemical Principles of Color Chemistry,
pp. 254}290. Blackie Academic & Professional, an imprint of Chap-
man & Hall, London.
The chromatographic method earlier developed by the
authors [14] was used to identify and determine benzidine 12. Pielesz, A. (1999). The process of the reduction of azo dyes used in
dyeing textiles on the basis of infrared spectroscopy analysis. J. Molec.
Struct. 511, 512, 337}344.
(t "5.79 min) in the reduction products of Acid Red 85 and
ꢀ
Direct Blue 6, as well as p-phenylenediamine (t "8.61 min)
ꢀ
13. Pielesz, A., Swierczek, S., W"ochowicz, A., and Baranowska, I. (1999).
and aniline (t "5.34 min) in the reduction products of
ꢀ
Adsorption and partition TLC separation of MAK-type aromatic
amines, reduction products of azo dye. J. Planar Chromatogr. 12(5/6),
215}220.
other dyes examined. The statistical results for the amine
chromatography are found in Table 3.
Figure 3 is a typical chromatogram. Peaks 1 and 3 do not
represent amines (this was veri"ed through analyzing stan-
dards), but probably some other dye degradation products.
The chromatographic method can be used to analyze the
reduction products of dyes that release amines of the MAK
type.
14. Pielesz, A., Baranowska, I., Swierczek, S., and W"ochowicz, A. (1999).
Separation of aromatic amines of MAK group, which are the reduction
products of azo dyes by partition HPLC chromatography. Chem. Anal.
44, 495}504.
15. Rosenkranz, H. S., and Klopman, G. (1989). Structural basis of the
mutagenicity of phenylazoaniline dyes. Mutat. Res. 221, 217}234.
16. Shahun, M. M. (1989). Evaluation of the mutagenicity of azo dyes in
Salmonella typhimurium: A study of structure}activity relationships.
Mutagenesis 4(2), 115}125.
CONCLUSION
17. Shargel, L., Banijamali, A. R., and Kuttab, H. (1984). Relationship
between azo dye structure and rat hepatic azoreductase activity.
J. Pharm. Eur. ;niontical Sci. 73(2), 161}164.
18. Stiepanow, B. I. (1980). In The Essentials of Chemistry and Technology of
Organic Dyes. Wydawnictwo Naukowo-Techniczne, Warsaw.
By analyzing the test results provided in Table 1 and 2, it
is possible to compare the spectrophotometric and HPLC
methods. It is possible to determine the presence or absence
of a carcinogenic aromatic amine by employing the less
expensive and simpler spectrophotometric method.
19. Technical Support Sta! (AG). (1996). High performance thin layer
chromatography: Your prudent choice for analyses of banned carcino-
genic dyes, dyestu! intermediates and textiles. Colourage 9, 19}26.
20. Vierte Verordrung zur Anderung der Bedarfsgegenstandenverordnung,
Bundesgesetzblatt, Jahrgang, Teil I. (1995). Vol. 20, p. 7.
REFERENCES
21. Wilken, B. M., Muller, J., Pieler, H., and Hocker, H. (1995). In
Mutagenity Testing of Textiles. Proceedings of the Ninth International
Textile Research Conference Biella, Italy, June 28}July 5, Conference
paper, pp. 262}272.
1. Achwal, W. B. (1997). Problems during analysis of textiles as per
Eco-Standards and the Consumer Articles Ordinance (Part-I). Colour-
age 5, 29}31.
2. Anliker, R. (1979). Proceedings, ecotoxicological assessment of dyes
with particular reference to ETAD's activities. JSDC 9, 317}326.
22. Zimnicki, J. (1999). In Looking for Dangerous Dye. Proceedings of IV
International Conference of Science and Technology, Ecology and
Textile Industry, Ekotextil'99 |odz-Arturowek, May 18}19, Confer-
ence paper, pp. XXII/1}12.
3. Brzezinski, St. (1998). Polish system of ecological certi"cation of textile
products, Part III. ¹extile Rev. 11, 28}35.
23. Zweite Verordung zur Anderung, Bundesgesetzblatt, Jahrgang, Teil I.
(1994). Vol. 15, p. 7.
4. Chung, K. T. (1983). The signi"cance of azo-reduction in the mutagen-
esis and carcinogenesis of azo dyes. Mutat. Res. 114, 269}281.