TLC Sensitivity of
Six Modifications of Dragendorff's Reagent
LINDA B. RUBIA and RALPH GOMEZ"
Abstract 0 The relative sensitivities of six modifications of Dragen-
dorff's reagent were measured on TLC plates by spectrodensitometry,
using compounds containing specific functional groups. A correlation
between the structures of the compounds reactive to Dragendorff's re-
agents and the sensitivities of the reagents was made. Explanations for
the variations in sensitivities between different modifications of Dra-
gendorff's reagent are given.
Keyphrases Dragendorff s reagent-several modifications, effect on
relative TLC sensitivities of compounds containing various functional
groups 0 TLC-effect of modifications of Dragendorffs reagent on rel-
ative sensitivities of compounds containing various functional groups
Structure-activity relationships-compounds containing various
functional groups, effect of modification of Dragendorff's reagent on
relative TLC sensitivities
Dragendorff s reagent is a commonly used spray reagent
for the detection of nitrogen-containing compounds in the
TLC analysis of pharmaceutical products (1-3). Its sen-
sitivity of detection of different compounds, relative ease
of preparation, and stability are some reasons for its ex-
tensive use. Dragendorffs reagent has been known for over
a century (4) and has been used as a TLC reagent for over
25 years.
Many modifications have been suggested to increase its
sensitivity, specificity, and stability (5-8). The purpose of
this investigation was to determine whether any modifi-
cation offered advantages over the others and whether one
could be used as a general spray reagent for the detection
of nitrogen-containing active pharmaceuticals.
The sensitivities of six modifications of Dragendorff s
reagent were measured on TLC plates by spectrodensito-
metry, using compounds containing specific basic func-
tional groups. The sensitivities of the modifications were
determined, and their reactivity to the functional groups
was studied. The reactions that explain the differences in
sensitivities between the reagents are discussed.
EXPERIMENTAL
Equipment-A spectrodensitometer' equipped with a monochro-
mator was used. The wavelength was 500 nm, and the signal was fed into
a properly attenuated recorder.
Materials and Reagents-Plates-Silica gel G was slurried with
water and coated on 20 X 20-cm glass plates using a mechanically oper-
at.ed TLC coating apparatusz. The plates were dried a t 40" overnight;
the average layer thickness was 0.25 mm.
Solutions-The following solutions were prepared: 1% (w/v) aqueous
sodium nitrite, 40% (w/v) aqueous potassium iodide, 1.7% (w/v) bismuth
nitrate in 20% (v/v) acetic acid, and 10% (w/v) aqueous sulfuric acid.
Acid-Modified Dragendorff's Reagent (Reagent A)-Potassium iodide
solution, 100 ml, was added to 100 ml of bismuth nitrate solution, and
the mixture was diluted to 1 liter with 10% (w/v) sulfuric acid.
Peroxide-Acid- Modi fied Dragendorff's Reagent (Reagent B )-To
300 ml of Reagent A, 0.6 ml of 30% hydrogen peroxide was added.
Iodine-Acid-Modified Dragendorff's Reagent (Reagent C)-To 300
ml of Reagent A, 6 g of iodine crystals was added. Then the mixture was
shaken for 30 min.
Schoeffel model SD 3000.
Camag model 21-602.
n H
I I1
OH &H, \
Ill N
OH L
V
0
H O - Q - G H , C H C O O H II
N'(CHJ,Br- TSH, CH36NH,
VI VII vn1
XI XI1
NHZ
I
& NO,
XI11 XIV
Acid-Modified Dragendorff's Reagent Oversprayed with Sodium
Nitrite (Reagent D)-After spraying the TLC plate with Reagent A, the
plate was lightly sprayed with sodium nitrite solution.
Peroxide-Acid-Modified Dragendorff's Reagent Oversprayed with
Sodium Nitrite (Reagent El-After spraying the TLC plate with Re-
agent B, the plate was lightly sprayed with sodium nitrite solution.
Iodine-Acid-Modified Dragendorff's Reagent Oversprayed with
Sodium Nitrite (Reagent 8')-After spraying the TLC plate with Reagent
C, the plate was lightly sprayed with sodium nitrite solution.
Iodine Solution-To 10 ml of 40% potassium iodide solution, 2 g of
iodine crystals was added. The mixture was diluted to 100 ml with 10%
(w/v) sulfuric acid and mixed well.
Samples and Standard-The compounds studied were: 4-amino-
benzoic acid (I), 4-(N-methylamino)henzoic acid (11),4-(N,N-dimeth-
y1amino)benzoic acid (1111, edrophonium chloride (IV), dodecylamine
(V), tetramethylammonium bromide (VI), levodopa (VII), acetamide
(VIII), trimethylsulfonium iodide (IX), henzyltriphenylphosphonium
chloride (X), diphenylguanidine (XI), 4-nitroaniline (XII), 2-nitroaniline
1656 /Journal of Pharmaceutical Sciences
Table I-Relative ResDonse to Draaendor f f s Reagents
Com- Reagent Reagent Reagent Reagent Reagent Reagent
pound A B C D E F
I 0.2 0.7 0.9 1 .o 0.8 1.0
I1 0.2 0.9 1.6 1.2 1.2 1.6
111 0.4 1.4 2.8 2.4 2.5 2.9
IV 0.9 2.8 3.8 3.9 3.9 4.4
V 0.6 1.0 1.2 1.4 1.3 2.6
VI 1.7 3.7 4.2 3.8 3.6 4.4
VII 0.0 0.0 0.0 0.0 0.0 0.0
VIII 0.0 0.0 0.0 0.0 0.0 0.0
IX 0.9 2.7 3.4 3.5 4.3 4.i
X 1.4 3.5 4.7 4.5 4 8 5.1
XI 0.5 2.1 2.9 3.1 4.1 4.9
XI1 0.2 0.9 1.2 0.9 1.6 1.8
XI11 0.4 0.5 0.7 0.4 0.8 0.5
XIV 0.1 0.2 0.7 0.2 0.6 0.8
(XIII), and 3-nitroaniline (XIV). Solutions of the compounds were pre-
pared in either chloroform or chloroform-methanol (92:8) at a concen-
tration of 0.6 mg/ml. The standard was 1% (w/v) methanolic thymol blue
(thymolsulfonphthalein sodium salt).
Procedure-A 20 X 20-cm TLC plate was scored into 20 parallel 1-cm
lanes. With a syringe, alternate lanes were spotted with 5 pl of sample
solution. The spots were thoroughly dried, and the plates were sprayed
evenly with 10 ml of Dragendorff‘s reagent. Then the plates were im-
mediately covered with a 20 X 20-cm piece of clear glass and scanned. Five
determinations on five separate plates were performed for each sample.
On each plate, 2 pl of thymol blue solution was spotted. The areas of all
spots were measured relative to the area of the thymol blue spot.
RESULTS AND DISCUSSION
The relative sensitivities of 14 compounds sprayed with six modifica-
tions of Dragendorff’s reagent are summarized in Table I. In most cases,
Reagent F proved to be the most sensitive, followed by Reagent C. Re-
agents A and B had the lowest sensitivities, while Reagents D and E had
intermediate sensitivities.
An explanation for the observed order of sensitivities may be found
in the preparation of the reagents since the relative sensitivity of each
reagent can be roughly correlated to the amount of available iodine. To
illustrate this correlation, IV-VI and IX-XI were spotted on TLC plates,
sprayed with a solution of iodine of an equal concentration, and prepared
in the same solvent as Dragendorfrs reagent. The relative responses for
IV, V, VI, IX, X, and XI were 4.0,1.7,4.1,3.3,5.3, and 3.9, respectively.
These results show that the relative areas were similar to those obtained
for Reagents C and F and greater than those obtained for Reagents A and
B, which do not contain added iodine, and indicate that the high sensi-
tivity of Reagents C and F is due largely to their iodine content.
The contribution of sodium nitrite to the total sensitivity of the reagent
can be seen from the data in Table I. Comparison of the data for Reagents
A and D shows that the latter was more sensitive. Similar results are seen
when Reagents B and E and C and F are compared. The difference be-
tween reagents in each pair is that Reagents E and Fare oversprayed with
sodium nitrite solution while B and C are not. This increase in sensitivity,
therefore, is attributed to sodium nitrite overspraying. As Fike (9) pro-
posed, sodium nitrite releases iodine from the excess iodide in the re-
agents, which enhances the sensitivity.
A possible reaction that explains the role of sodium nitrite in the re-
agent is shown in Scheme I (10).
2N02- + 21- + 4H+ - 12 + 2NO + 2H20
Scheme I
The contribution of hydrogen peroxide to the sensitivity of the reagent
can be seen by inspecting Table I, which shows that Reagent B was more
reactive than Reagent A. The only difference in the two is that Reagent
B has added hydrogen peroxide. Hydrogen peroxide reacts with excess
iodide in the reagent to liberate iodine, similar to sodium nitrite.
A possible reaction is shown in Scheme I1 (10).
HzO2 + 2H+ + 21- - 12 + 2HzO
Scheme I t
Reagent 3 cannot be expected to be as sensitive as the other reagents
containing iodine or oversprayed with sodium nitrite, since the amount
of hydrogen peroxide added is small and, therefore, the amount of iodine
that can form and be available for reaction is small.
The data show that addition of crystalline iodine or formation of iodine
by reaction with sodium nitrite or hydrogen peroxide increased the
sensitivity of each reagent over its noniodine-containing counterpart.
Table I shows that IV, VI, IX, and X were the most reactive with
DragendorfPs reagents, while 111 and XI were close behind in reactivity.
These reactive compounds are quaternary nitrogen and phosphorus salts
and tertiary sulfur salts. Compounds 111 and XI are tertiary amines, which
can be easily converted to quaternary amine salts by reaction with the
excess sulfuric acid in the reagents. Presumably, reaction with the reagent
then proceeds by the same mechanism as with the salts.
Compounds I, V, XII, XIII, and XIV, all primary amines, had a rela-
tively low reactivity with the reagents, although V, the aliphatic primary
amine, had a higher sensitivity than the other four compounds.
Compounds VII and VIII, an amino acid and amide, respectively, did
not react with any of the six modifications of the reagent.
For a homologous series, 1-111, reactivity increased from primary to
secondary to tertiary amine; with ortho-, para- and meta-substituted
nitroanilines, reactivity was generally poor and the order of reactivity
varied with the reagent.
The role of light and air on the stability of plates sprayed with Reagent
C was investigated. Duplicate plates were spotted with 3 pg of IV, sprayed
with Reagent C, and immediately covered with a 20 X 20-cm piece of plain
glass. The sides were sealed with tape. One plate was left exposed to room
light, and the second was kept in the dark. Both plates were periodically
scanned on the densitometer. After 5 hr, the relative areas of the spots
were fairly constant for both plates.
The experiment was repeated. This time the plates were left uncovered.
After several hours, the spots had faded a great deal. However, the de-
crease was approximately the same for the plate exposed to light and the
plate kept in the dark. This result indicates that the spot on the plate is
not light sensitive but appears to be sensitive to air.
Since Dragendorffs reagent is used extensively in the pharmaceutical
industry for the detection of nitrogen-containing drug substances on TLC
plates, these results should aid in improving the detection limits of these
active drug substances and their degradation products on TLC plates.
REFERENCES
(I) “Thin-Layer Chromatography,” 2nd ed., E. Stahl, Ed., Academic,
(2) J. G. Kirchner, “Thin-Layer Chromatography,” Interscience, New
(3) K. Randerath, “Thin-Layer Chromatography,” Academic, New
(4) G. Dragendorff, Russ. J . Pharm., 5, 85 (1866). fb id . , 6, 663
(5) H. M. Bregoff, E. Roberts, and C. C. Deliveche, J . Riol. Chem.,
(6) R. Munier and M. Machebouf, Bull. SOC. Chem. B i d , 31,1144
(7) R. Munier, Bull. Soc. Chem. Riol., 35,1225 (1953).
(8) H. Thies and F. W. Reuther, Naturwissenschaften, 41, 230
(9) W. W. Fike, Anal. Chem., 38,1697 (1966).
New York, N.Y., 1969, p. 873.
York, N.Y., 1967, p. 160.
York, N.Y., 1 9 6 4 , ~ . 74.
(1867).
205,565 (1953).
(1949). Ibid., 33,846 (1951).
(1954).
(10) G. H. Ayres, “Quantitative Chemical Analysis,” Harper, New
York, N.Y., 1 9 5 8 , ~ . 471.
ACKNOWLEDGMENTS AND ADDRESSES
Received January 3, 1977, from the Research Products Section,
Quality Control Department, Hoffmann-La Roche Inc., Nutley, N J
071 10.
Accepted for publication January 31, 1977.
The authors acknowledge the assistance of the Scientific Literature
Department of Hoffmann-La Roche Inc.
T o whom inquiries should be directed.
Vol. 66, No. 11, November 1977 I1657