Study of Foreign Ions. About 120 times larger amounts
of uranium(VI), 60 times of thorium(IV), 140 times of sodi-
um and potassium, 35 times of rare earths(”,
tungsten(VI), yttrium(III), ammonium; 18 times of zin-
c(II), 12 times of strontium(II), 10 times of copper(I1) and
zirconium(IV), 5 times of lead(II), 3 times of antimo-
ny(V), tantalum(\’), iron(III), equal amounts of titani-
um(1V) and 50 times of nitrate, 40 times of chloride and
phosphate, 140 times of sulfate did not show significant in-
terference in the estimation of 37.6 pm of niobium in 25 ml
by the recommended procedure.
However, in the presence of large amounts (greater than
10 times) of zirconium and (i), in presence of 4 to 5 times of
tantalum(V), iron(III), aluminum(III), fluoride ion; (ii) in
the presence of comparable amounts of nickel(II), tin(I1)
and (IV), beryllium(I1); and (iii) in the presence of small
amounts of molybdenum( VI), this method gives errors in
the range of 10%.
The interference of 9 times of iron(II1) could be sup-
pressed completely using ascorbic acid, but for molyb-
denum(V1) using citric acid as a masking agents, niobium
could be estimated with 3% of error by this method.
In the following synthetic mixtures in 25 ml of solution,
niobium(V) was estimated successfully from (1) niobium
(40 pg) with zirconium(1V) (91 pg) and uranium(V1) 3808
pg); (2) niobium(V) (37.16 pg) with zirconium(1V) (4.56 pg)
and uranium (VI) (190 pg); (3) Nb205 (44.59 pg); TazO5
(8.05 pg); Ti02 (1.62 pg); Si02 (0.22 pg); SnOz (1.48 pg);
Fez03 (14.77 pg); CaO (0.94 pg); MgO (0.03 pg); A1203 (0.79
pg); Tho2 (1.21 pg); Crz03 (0.81 pg); YzO3 (13.93 pg); MnO
(1.60 pg); PbO (2.04 pg); UOz (6.03 pg); NazO (0.29 pg);
K20 (0.21 pg) (relative standard deviation of 7 samples is
2.6%); (4) Niobium(V) (92.9 pg) tantalum(V) (90.8 pg); ti-
tanium(1V) (11.0 pg); tin(I1) (50 wg); antimony(II1) (50 wg),
and tungsten(V1) (800 pg).
Niobium estimation by this method in the presence of
other metals as given above will be useful for the analysis of
fuel elements or other types of alloys where uranium(V1)
concentration is high.
(1) I. D. Ali-zade and 0. A. Gamid-zade, Zh, Anal. Khim., 29, 735-9 (1974).
(2) A. K. Babko and V. V. Lukachine, Ukr. Khem. Zh., 27, 682-7 (1961).
(3) V. P. Madhava Menon, N. Mahadevan, K. Srinivasulu, and Ch. Venka-
teswarulu, J. Sci. hd. Res. (Hardwar, India), 218, 20-23 (1962).
(4) V. Patrovsky, Collect. Trav. Chim. Tchec., 23 1774 (1958).
(5) S. V. Elinson, L. T. Pobedina, and A. T. Rezova, Zavod. Lab., 37, 391-4
RECEIVED for review August 21,1975. Accepted December
1, 1975. Financial assistance from the Council of Scientific
and Industrial Research, India, to one of the authors
(MSR) as an award of Junior Research Fellowship is grate-
Colorimetric Assay for Aromatic Amines
Esther Rinde and Walter Troll”
New York University Medical Center, Department of Environmental Medicine, 550 First Avenue, New York, N. Y. 100 16
Aromatic amines can be detected in the nanomole range
with the reagent Fluram, with which they form stable yel-
low derivatives. Fluram In glacial acetic acid reacts only
with aromatic amlnes. The reaction is complete in 10 min-
utes and can be performed on thin-layer (TLC) chromato-
grams making possible the specific measurement of aro-
matic amines. The yellow product can be quantitatively
eluted from the TLC plates. Fluram is colorless, and the
blanks are zero.
The reagent Fluorescamine (Fluram) was introduced by
Udenfriend (1) for the quantitation of aliphatic amines in a
sensitive fluorescent assay. Aromatic amines, like the ali-
phatics, form fluorescent products with Fluram which are
unstable, but they also form stable yellow derivatives.
A number of aromatic amines have been found to be car-
cinogens (2); hence, it is important to have sensitive and
specific assays for their detection in the environment and
in body fluids. Using selective extraction procedures (3),
the excretion of aromatic amine in the urine of exposed in-
dividuals can be measured, and is a good criterion for de-
termining whether exposure has occured. The stability of
the product formed when Fluram reacts with aromatic
amines makes it possible to perform the assay on TLC from
which it can be quantitatively eluted. Moreover by the use
of TLC in conjunction with the assay, it is also possible to
differentiate one aromatic amine from another (4). One of
the colorimetric assays we had previously used was a modi-
fication of the Satake method (5) using the reagent trini-
trobenzene sulfonic acid (TNBS) which reacts with aro-
matic amines at pH 5. With TNBS, it is necessary to ex-
tract the product formed into organic solvent; otherwise
the yellow color of the reagent interferes. Fluram has the
advantage of being colorless (blanks are zero), eliminating
the need for extraction, and has a sensitivity in the nano-
Apparatus and Reagents. All solvents were Reagent Grade.
Aniline, purchased from Eastman Organic Chemicals was redis-
tilled 2X. Purified samples of the other amines tested were sup-
plied by Allied Chemical Company: benzidine, 2-naphthylamine,
dichlorobenzidine, 0-tolidine. Fluram was purchased from Fisher
Scientific (Catalog No. 43023). Monoacetyl benzidine was synthe-
sized from benzidine (6).
All of the aromatic amines were dissolved in acetone to give a
solution (dichlorobenzidine, because of its limited solubility
had to be prepared by first dissolving it in glacial acetic acid (glac.
HAC), then diluting with acetone to this concentration). Fluram
was used as a 1 mg/ml solution in glacial acetic acid which is stable
for weeks at room temperature.
For the aliquoting of microliter quantities, the “Drummond
Dialamatic Microdispenser” (Drummond Sci. Glass P6295) was
used. Thin-layer chromatography (TLC) plates were purchased
from Scientific Glass (Silica Gel 0.25 mm thick. Art 5762/0001).
The solvent system for development of the plates was chloro-
form (90) glac. HAC (5) methanol (5). Spraying was accomplished
with the “Lab Reagent Sprayer” (Analtech Catalogue No. A-100).
Procedure. Fluram Reaction in Solution. Aliquots of 2 to 50 ~1
were transferred to tubes and the acetone was evaporated with a
stream of nitrogen. Fifty gl of the Fluram solution was then added
ANALYTICAL CHEMISTRY, VOL. 48, NO. 3, MARCH 1976
v B E N Z I D I N E
MONO-ACETYL BENZ I D I N E
0 T O L I D I N E
A DICHLORO B E N Z I D I N E
A 2 - N A P H T H Y L A M I N E
3 6 4 0
Figure 1. Standard curves for the reaction of Fluram with aromatic amines in solution
followed by a 30-sec vortex mix. After 10 min, the reaction was
stopped by addition of 0.5 ml of methanol.
The optical density of the yellow product formed with each
amine was measured at the wavelength of maximum absorption, as
determined in a Gilford Spectrophotometer (0.5-ml quartz cu-
vettes with a 1-cm light path were used).
Measurement of Benzidine on TLC. Varying amounts of the
benzidine stock were applied to TLC plates. As soon as the plates
were dry after development, they were sprayed with the Fluram
solution, and again allowed to dry. The yellow spots were scraped
off the plates and the benzidine-Fluram product was eluted with
0.5 ml methanol, by vortexing for 1 min, then centrifuging for 10
min at 12 000 RPM in a Serval1 Refrigerated Centrifuge. This al-
lowed for almost complete recovery of the methanol, free from the
silica particles. Optical density was measured at 415 mF.
RESULTS AND DISCUSSION
The assay was linear over a wide range of amine concen-
tration (Figure 1). Benzidine gave the largest yield at every
concentration, monoacetyl benzidine and tolidine gave
values of about l/z, and dichlorobenzidine and 2-naphthyl-
amine about l / 4 that of benzidine.
The advantage of having the Fluram in glacial acetic acid
is that it will react only with aromatic amines, the pK of al-
iphatic amines being near 9. The reagent is stable as is the
yellow product formed (with the exception of the dichloro-
benzidine product which fades in 5 min). The benzidine
and monoacetyl benzidine products have absorbance maxi-
ma at 415 mp; tolidine, 2-napthylamine, and dichloroben-
zidine yellows read maximally at 390 mp. These aromatic
amines are easily separated using the solvent system given
in Experimental section. In a 10-cm run, we obtained a
2-cm separation between benzidine and 2-naphthylamine
and a 1-cm separation between benzidine and tolidine.
Assay of benzidine on a TLC plate was also linear in the
range measured and the color yield identical to that ob-
tained by reaction in solution (Figure 2). Precision data for
the 5 amines tested and for the TLC results are given in
The method is applicable to the monitoring of aromatic
amines in urine or in water. The quantitative recovery of
the yellow product (formed in minutes) from TLC makes
possible the specific measurement of aromatic amine. In
previously published work (3), we described the finding of
monoacetyl benzidine (MAB) in the urine of Rhesus mon-
L c c
Flgure 2. Data obtained for the reaction of Fluram with benzidine on
a TLC plate
ANALYTICAL CHEMISTRY, VOL. 48, NO. 3, MARCH 1976
Table I. Precision Data
No. of test
Aromatic amine, nmol
111. Mono acetyl benzidine
VI. Benzidine on TLC
a As defined in Anal. Chem., 41, 2139 (1969). b Benzidine, 6 nmol,
Sum of squares
from mean, D2
(D2/N - 1)
tolidine 6, 10, 16 nmol: variance was less than
key fed benzidine or benzidine azo dye and the measure-
ment of benzidine + MAB excretion using TNBS. With
this new procedure, it will be possible to measure benzidine
or its MAB metabolite separately.
We thank the Allied Chemical Corporation for supply-
ing materials used in this study.
(1) S. Udenfriend, S. Stein, P. Bohlen, W. Dairman, and W. Leimgruber,
Science. 178. 171 (1972).
(2) W. C. Hueper, "Occupational-And-Environmental Cancers of the Urinary
System", Yale University Press, New Haven, Conn., 1969, pp 118-180.
(3) E. Rinde and W. Troll, J. Nat. Cancer SOC., 55 (1). 181 (1975).
(4) E. Rinde and W. Troll, Roc. Am. Assoc. Cancer Res., 16, 79. San Diego,
Calif., May 1975, Abstract No. 314.
(5) K. Satake, T. Okuyama, M. Ohashi, and T. Shimoda, J. Biochem., 47 654
(6) S. Laham. J. P. Farant. and M. Potvin, Occup. Health Rev. 21, 14 (1970).
RECEIVED for review September 29, 1975. Accepted De-
cember l, 1975. Supported by the National Bladder Cancer
Project, Public Health Service Grant (2.415315 from the
National Cancer Institute; Public Health Service Core
Grant ES00260 from the National Institute for Environ-
mental Health Sciences; and the Allied Chemical Corpora-
ANALYTICAL CHEMISTRY, VOL. 48, NO. 3, MARCH 1976