in copyright - non-commercial use permitted rights ...34246/... · especially the modernnon-aqueous...
TRANSCRIPT
Research Collection
Doctoral Thesis
The quantitative determination of the barbituric acid derivatives
Author(s): Adel, Mohamed Sherif
Publication Date: 1959
Permanent Link: https://doi.org/10.3929/ethz-a-000113912
Rights / License: In Copyright - Non-Commercial Use Permitted
This page was generated automatically upon download from the ETH Zurich Research Collection. For moreinformation please consult the Terms of use.
ETH Library
Prom. Nr. 2963
The Quantitative Determination
Of The Barbituric Acid Derivatives
THESIS
presented to
the Swiss Federal Institute of Technology Ziirich
for the Degree of Doctor of Natural Sciences
by
MOHAMED SHERIF ADEL
B. Pharm.; M. Pharm.
Citizen of Egypt
Accepted on the Recommendation of Prof. Dr. J. Biichi and Prof. Dr. H. Fliick
Juris-Verlag Zurich
1959
Leer - Vide - Empty
TO THEM
WHOSE PRESENCE IS HOME
Leer - Vide - Empty
- 5 -
Acknowledgment
I wish to express my sincere gratitude to Professor Dr. J. Biichi, the Direc¬
tor of the Pharmacy Institute of the Swiss Federal Institute of Technology, Zurich,
Switzerland, under whose auspices this work was initiated and matured and whose
help and encouragement have accompanied this work from the beginning to its
completion.
I wish also to express my appreciation and thanks to the Swiss Federal Institute
of Technology for the kind assistance extended to me during the last period of this
work.
To the supervisor of the Pharmacy Institute, Mr. R. Schwegler, my hearty
thanks for his valuable and continuous help in supplying the chemicals and apparatus.
To Dr. X. Perlia, Coworker with Professor Dr. J. Biichi, my best thanks
for his valuable discussions during some parts of the experimental work and his
guidance in measurements carried out with the Beckman Spectrophotometer in the
Hygiene Institute.
I wish to thank Dr. A.M . Shams-el-Din ,Coworker with Professor Dr. G.
Triimpler, for discussing some of the results of the potentiometric titration curves.
I am indebted to Professor Dr. A. Abdel Rahman,Professor of Pharmaceu¬
tics, Faculty of Pharmacy, Cairo University, for his permission to this study vacation.
Leer - Vide - Empty
- 7 -
Table of Contents Page
I INTRODUCTION 9
II GENERAL PART 10
A) A survey of the official barbituric acid derivatives and their
methods of assays in official literature 11
B) The plan of the work 13
HI EXPERIMENTAL PART 15
A) The identification and the purity of the used samples of barbiturates 15
B) Physical properties of the barbituric acid derivatives 16
1. Melting points 162. Sublimation 163. Solubility 18
4. Dissociation constants and dissociation exponents (pK values) 20
5. Ultra violet spectrum 36
C) Chemical properties of the barbituric acid derivatives and their
use in the quantitative estimation 48
1. Acidity 48
2. Reaction with silver 48
3. Reaction with mercury 50
4. Halogenation of the unsaturated radicals 51
5. Colour reactions 51
a) Production of complex salts with cobalt 51
b) Production of complex salts with copper 53
c) Production of colour with selenious acid 54
6. Condensation of barbiturates with p-nitro benzyl chloride 54
7. Reaction with different reagents 55
D) Methods of the quantitative estimation of the barbituric acid
derivatives 56
A survey and discussion of the literature
Application of the method of estimation and the results obtained
Criticism and summary of the results
1. Acidimetric methods 56
a) Titrations in aqueous solvents 56
b) Titrations in non aqueous solvents 63
ot) Vespe and Fritz method 67
p) Heiz's method 76
y) Chatten's method 83
6) Ryan, Yanowsky and Piter method 85
2. Argentometric method 92
- 8 -
a) Budde's method 94
b) Bodin's method 98
c) Mangouri and Milad method 104
d) Chavanne and Marie method 106
3. Mercurimetric method - Pedley's method 111
4. Bromometric method 115
5. Colorimetric method - Nuppenau's method 116
6. Kjeldahl's method 122
IVa) SUMMARY AND CONCLUSIONS 127
b) PROPOSAL OF THE RECOMMENDED METHODS 131
c) ZUSAMMENFASSUNG 132
V) BIBLIOGRAPHY 138
- 9 -
I INTRODUCTION
(B)Since the discovery by EmilF ischer and von Mering in 1903 that Veronal
,
a barbituric acid derivative, had a profound hypnotic and sedative action, enormous
interest had arised in preparing and studying similar compounds. Several hundred of
these barbituric acid derivatives have been prepared, many of them were studied for
their pharmacological action and quite a good number is in common use in therapy.
Although the progress in the synthetic preparation and medical investigation
of these group of compounds readily followed the discovery of Fischer, it was up
till rather recently that their analytical estimation has attracted the attention. Budde
in 1934 was the first to open the line of research making use of the solubility of the
silver salts of the barbituric acid derivatives in sodium carbonate solution. Other
estimation methods then followed. Some depended on the acid character, while others
on the determination of the nitrogen content of the compound. Colorimetric and Spec-
trophotometric methods also found application.
It is remarkable that although the barbituric acid derivatives occupy a prominent
position in most of the Pharmacopoeiae, yet no reliable definite method for their
estimation is recommended. In the present study, the recommended methods of
assaying barbituric acid derivatives are compared with one another with the aim of
establishing the most reliable one and thus helping in filling the vacant gap present
in many of the Pharmacopoeiae.
- 10 -
II GENERAL PART
A) A survey of the official barbituric acid derivatives and
their methods of assays in different pharmacopoeiae
The following official literature will be dealt with:
a) The Swiss Pharmacopoeia 1933 and its supplements (Ph.Helv. V; Suppl. I,
II &m)
b) The British Pharmacopoeia 1958 (Brit. Ph. 1958).
c) The United States Pharmacopoeia 1955 (USP XV)
d) The Danish Pharmacopoeia 1948 and its addendum (Ph.Dan. IX & Add. 1954)
e) The French Pharmacopoeia 1949 (Codex Gall. 7 & Suppl. 1954).
f) The German Pharmacopoeia 1926 (DAB 6)
g) The Netherlands Pharmacopoeia 1958 (Nederl. Ph. VI)
h) The Egyptian Pharmacopoeia 1953 (E.P. 1953).
The following table 1 summarises the official barbituric acid derivatives and
their methods of determination in the mentioned official literature.
N.B. In all the following text we are going to adopt the official names of the Swiss
Pharmacopoeia 1933 and its addenda given to our barbiturates, omitting the
latin ending um i. e.
Allobarbital (5 : 5-diallyl barbituric acid)
Cyclobarbital (5-ethyl 5-cyclohexenyl barbituric acid)
Hexobarbital (5-methyl 5-cyclohexenyl 3-methyl barbituric acid)
Methylphenobarbital (5-ethyl 5-pheno-3-methyl barbituric acid)
Pentobarbital sodium (the sodium salt of 5-ethyl 5-methylbutyl barbituric acid)
- 11 -
Table 1
Official barbituric acid derivatives and their methods of assays in different Pharmacopoeiae
Official Barbituric
Acid Derivatives
Ph Helv V &
Suppl i,n&in1933/1955
i)
Brit Pb 1998
J)
E P 1953
3)
USP XV 1955
*)
Ph Dan IX ft
Add 1954
5)
Codex Gall 7
ft Suppl 1954
«)
DAB 6 1S26
n
Nederl Ph
VI 1958
8)
1-Allobarbltalum
(5.5-diaHyl)
No method ... ... ... Bromometric ...... Bromometric
2-Barbitalum
(5.5-diethyl)
No method No method ... Kjeldahl No method No method Acidlmetric
S-Barbitalum solublle
(sodium salt of
5:5-diethyl)
Alkallmetrlc Gravimetric Gravimetric a) Alkallmetrlc
b) Gravimetric
... No method Alkallmetrlc
4-Phenobarbitalum
(5-phmyl 5-ethyl)
No method Acidlmetric
(non-aqueous)
No method No method Kjeldahl Acidlmetric No method Acidlmetric
S-Phenobarbltalum solublle
(sodium nit of 9-phenyl-
5-ethyl)
Alkallmetrlc Gravimetric Gravimetric Gravimetric a) Alkallmetrlc
b) Gravimetric
No method Alkalimetric
6-CyclobarbltaIum calcium
(calcium salt of 5-ethyl-
5-cyClohei-1'-eny1)
Gravimetric Acidimetrtc
(non-aqueous)
a) Alkallmetrlc
b) Bromometric
...
7-Hexobarbitalum
(S-cyclohei-l'-enyl 1:5-
dlmethyl)
No method No method ... Bromometric a) Addimetric
b) Gravimetric
No method
8-Henobtrbitthini solublle
(sodium salt of 9-cyclobex-
l'-enyl 1:5 dimethyl)
Gravimetric ... Gravimetric ... Alkalimetric ... Alkalimetric
9-Metbylphenobarbttalum
(5-ethyl-5 -phenyl-1 - me¬
thyl)
No method No method ... Kjeldahl ... Acidimetrtc
10-Pentobarbttalum solublle
(sodium salt of 5-ethyl-
5 -(V -methyl butyl))
Gravimetric Gravimetric Gravimetric ... ... Alkalimetric
11 -Allypropymalum
(5 -allyl -5-tsopropyI)
... — ... ... Bromometric ...— ...
12-Amobarbital
(5-ethyl-5-lsoamyl)
... ... ... No method Kjeldahl ......
13-Amobarbital sodium
(sodium salt of S-ethyl-
S-lsoamyl)
Gravimetric
14 -Amylobarbitone
(S-ethyl-S-tS'-methyl butyl))
... Acidimetrtc
(non-aqueous)
... ... ...... —
15 -Butobarbitone
(5-n-butyl-S-ethyl)
... Acidlmetric
(non-aqueous)
... Acidlmetric
lfl-Quinalbarbitone sodium
(sodium salt of 5-»llyl-5-
(l'-methyl butyl))
Gravimetric Gravimetric
- 12 -
Summary of the methods of estimation of barbiturates in
different Pharmacopoeiae
The Swiss Pharmacopoeia 1933 and its addenda, the British Pharmacopoeia
1953, the Egyptian Pharmacopoeia 1953, the United States Pharmacopoeia 1955 (XV),
the Netherlands Pharmacopoeia 1926 and the German Pharmacopoeia 1926 did not
describe any method for the assays of the barbituric acid derivatives.
The Danish Pharmacopoeia 1949 prescribed the Kjeldahl method for the
assay of several barbituric acid derivatives. It also described the Bromometric
method for the estimation of the barbituric acid derivatives possessing unsaturated
radicals. The French Pharmacopoeia 1949 does not recommend any method for
assaying Barbital (5:5 -diethyl). For Phenobarbital (5-ethyl 5-phenyl) it recommends
the use of acetone as a solvent and titration with potassium hydroxide in methanol
using thymol blue as indicator till a clear blue colour is obtained.
Recently, the British Pharmacopoeia in September 1958, recommended the
titration of the barbituric acid derivatives in non-aqueous media using lithium
methoxide as a titrant, dimethylformamide as a solvent and quinaldine red as indi¬
cator. The new Netherlands Pharmacopoeia 1958 recommended the estimation of the
barbituric acid derivatives by the acidimetric method. The acids being dissolved in
neutralized alcohol, are titrated against standard alkali using thymolphthalein as
indicator.
From Table 1, it is clear that with the exception of the Danish Pharmacopoeia
1948 and its addenda the methods of the quantitative determination of barbiturates
are incomplete. This calls forth a comparative study of the methods of estimation,
especially the modern non-aqueous titrations, argentometric and mercurimetric
methods.
After finishing our experimental research, the new British Pharmacopoeia
1958 and the Netherlands Pharmacopoeia 1958 appeared and described the acidimetric
titrations of barbiturates. The first pharmacopoeia recommended the non-aqueous
titrations, the second used aqueous medium. These methods as will be seen in our
experimental part, are of successful applications in the assaying of the barbituric
acid derivatives especially those in non-aqueous media.
- 13 -
B) The plan of the work
O R3
/c\ /c-°Rg C NH
A
Barbituric acid derivatives of the above general formula (I) could be generally
classified into:
1) DialkyI derivatives of barbituric acid
a) Saturated dialkyl derivatives.
b) Unsaturated dialkyl derivatives.
R, may or may not be R...
2) Aromatic and Alicyclic derivatives
R, is either aromatic or alicyclic, while R, is alkyl.
3) Nitrogen substituted derivatives
Rj is either aromatic or alicyclic, R2 is alkyl and R, is alkyl.
Representative examples of each of these mentioned groups were chosen for
our experimental work.
For the first group, Pentobarbitalum solubile (Ph.Helv. V) was chosen as a
representative of the saturated dialkyl derivatives:
Rl = " C2H5' R2 = " C3H7"CH "
» R3 = " H•
CH3
Chemically it is the sodium salt of 5-ethyl 5-1-methyl butyl barbituric acid. It is
also found in the market under the trade name Pentobarbital^ (Lilly) and Nembutal^
(Abbott).
As a pepresentative for group 1) b), Allobarbitalum (Ph.Helv. V) was chosen,
where
R-j = Rn = - CHn = CH - CHq- i Rq = - H
Chemically it is 5:5-diallyl barbituric acid. It is also found in the market under the
- 14 -
tradenames, Curral® (Roche) and Dial® (Ciba).
As a representative of the second group, Cyclobarbitalum served as a typical
representative, where
h=(\->
R2Rl =
\_y~'' R2 = -C2H5"' R3 = -H-
Chemically it is 5-ethyl 5-cyclohex-l'-enyl barbituric acid. It has several trade
names e.g. Cyclobarbitone® (Burroughs & Wellcome), Cyclosedal® (Burroughs
& Wellcome), Phanodorm® (Bayer), Phanodorn® (Winthrop-Stearns).
For the nitrogen substituted derivatives, two representatives were used i. e.
Hexobarbitalum (Ph.Helv. V) and Methylphenobarbitalum (Ph.Helv. V).
Hexobarbitalum has the following radicals:
Rl =
\ /~» R2 = R3 = CH3 "
•
Chemically it is 5-methyl 5-cyclohex-l'-enyl 3-methyl barbituric acid. Several
other names are met with e.g. Cyclonal® (May & Baker), Evipal® (Winthrop-
Stearns), Evipan® (Bayer), Hexanastab® (Boots), Noctivan (Theraplix),
Priv6nal® (Th6raplix), Tobinal® (Siegfried).
Methylphenobarbital has the following radicals:
Rl "
\ 7~ ' R2 ~ C2H5 "
' R3 = CH3
Chemically it is 5-ethyl 5-phenyl 3-methyl barbituric acid. Trade names are present
in the market e.g. Isonal® (Roussel), Mebaral® (Winthrop-Stearns), Phemitone®
(Boots), Prominal® (Bayer).
Thus the five examples studied in this research work represent the group of
the barbituric acid derivatives, besides they are of common use in therapy. The
different methods of estimation of barbituric acid derivatives are applied to these
representatives, in the present investigation.
- 15 -
III EXPERIMENTAL PART
A) Identification and Purity of the used Barbiturates
The identification and purity tests described by the Swiss Pharmacopoeia 1933
and its addenda for our representatives of the barbituric acid derivatives were
carried out. All our tested samples were conforming with the requirements of this
pharmacopoeia.
As a confirmatory measure for our results of the quantitative determination,
and for the purity of the samples used, we sent our dried samples (in vacuum at
0.01 mm. mercury for 48 hours at 60 C) to analysis for the carbon-, hydrogen- and
nitrogen content. This was carried out in the Microanalytical Laboratory of Organic
Chemistry ETH, Zurich, Switzerland (Director W. Manser).
Table 2
Confirmatory, Carbon, Hydrogen and Nitrogen content determination
of the used barbiturates
Barbituric acid
derivative
% of Nitrogen
calc. found
%oi
error
%ofC
calc. found
%of
error
%ofH
calc. found
%of
error
Allobarbital
Cyclobarbital
Hexobarbital
Methylpheno-barbital
Pentobarbital-
sodium
13.46 13.58
13.57
11.86 11.80
11.86 11.83
11.38 11.63
11.28 11.20
+0.89
+0.81
-0.50
-0.25
+2.2
-0.71
57.68 57.58
57.72
61.00 60.19
61.00 61.06
63.40 63.54
53.22 49.25
+0.89
+0.069
+1.33
+0.097
+0. 022
-6.26
5.81 5.83
5.93
6.83 6.27
6.83 6.78
5.73 5.76
6.90 6.80
+0.342
+2.08
+8.2
+0.73
-0.524
+1.45
- 16 -
B) Physical properties of the barbituric acid derivatives
The barbituric acid derivatives and their salts are white, crystalline, odour¬
less powders of a bitter taste.
The following properties will be dealt with:
1. Melting points
2. Sublimation
3. Solubility
4. Dissociation constants and dissociation exponents (pK value)
5. Ultra violet spectrum
1. Melting Points
The following (uncorrected) melting points were taken from official literature;
our (uncorrected) melting points were recorded by the microscope method as shown
in the following table:
Table 3
Melting points of Barbiturates
Barbituric acid
derivative
Ph. Helv. V.
& Suppl.^
B.P.C.
1954 9)
U.S. P. XV
4)Ph. Danica
19485)
Our deter¬
minations
AUobarbital
Cyclobarbital
Hexobarbital
Methylpheno-barbital
Pentobarbital
170-172° C
167-171° C
142-145° C
173-177° C
125-128° C
172-174° C
173-176°C
145-147° C
178-181°C
127-130° C 127-131° C
172-175° C
171-175° C
143-147° C
177-180° C
172° C
173° C
142° C
179° C
128° C
2. Sublimation
All solids have some tendency to pass directly into the vapour state. At a given
temperature each solid has a definite, though generally small vapour pressure; the
latter increases with the rise of temperature. Sublimation is the term applied to the
process of transforming a solid to vapour without intermediate passage through the
- 17 -
liquid state. The objects of sublimation are:
1) to purify volatile solids from admixed and fixed impurities.
2) to provide a convenient means of collecting volatile solids resulting from
chemical reactions at either high or low temperatures.10)
P e r 1 i a gave a full description of the isolation of barbituric acid derivatives
by microsublimation. He discussed the factors which have influence on the formation
and shape of sublimate e.g. atmospheric pressure, temperature, duration of subli¬
mation etc. He also studied the possibilities of identification of the sublimate by its
crystal form, micromelting point determination, refractive index of the melted
sublimate and microchemical reactions.
The possibility of occurence of sublimation during the excessive drying of the
barbituric acid derivatives obtained in the gravimetric estimation of the salts of
barbituric acid derivatives should be taken in consideration. The following directions
are found in different pharmacopoeiae in regards to the drying of the residues of
barbituric acid derivatives.
The Swiss pharmacopoeia 1933 directs that the residues of barbituric acid
derivatives obtained in the gravimetric estimation of the salts should be dried at
103 - 105 C for 30 minutes e. g. in cases of calcium salt of cyclobarbital and sodium
salt of pentobarbital.
The British Pharmacopoeia 1958 directs the drying of the residue as follows:
dry the residue to constant weight at 105 C. e.g. in cases of Barbitone sodium
Hexobarbitone sodium etc.
The United States Pharmacopoeia 1955 directs the drying of the residue as
follows: dry the residue for 2 hours at 105° C. e.g. in cases of Pentobarbital sodium
and Secobarbital sodium. In case of Amobarbital sodium it is only directed to dry the
residue for 30 minutes at 105 C.
We examined the effect of the time of drying on the residue of Pentobarbital
obtained in the assay of Pentobarbital sodium as directed by the Swiss Pharmacopoeia
1933. The following procedure was adopted:
About 0.50 g. dried pentobarbital sodium accurately weighed was introduced in a
separating funnel of 100 ml. capacity and dissolved in 20 ml. of water. After the
addition of 10 ml. dilute hydrochloric acid the contents of the separating funnel were
snaked well with 25 ml. of chloroform. Extraction of the precipitated pentobarbitalwas repeated twice with 25 ml. portions of chloroform. The chloroform extracts
were collected in a tared Erlenmeyer flask of 200 ml. capacity. The chloroform was
distilled and the remaining residue was dried for 30 minutes at 103 - 105° C. and
weighed.
10) X. Perlia, "Beitrage zum Nachweis therapeutisch wichtiger Barbiturate" Diss.
ETH. Zurich 1953.
- 18 -
It was found that 30 minutes drying of the residue gave rather high results so
drying was continued for 1, 3 and 6 hours. The assays were carried out four times.
The following results were obtained which represent the mean values:
Table 4
The Effect of the time of drying on the Pentobarbital residue
for 1 hour
Drying
for 3 hours for 6 hours
Mean per¬
centage of
Pentobarbital
found
98.87% 98.61% 98.38$
In our opinion, the drying of the residues of the barbituric acid derivatives could
be carried out at 105 C. till constant weight is achieved, since no loss occurs by
sublimation under the mentioned conditions; In case of excessive heating e.g. more
than 8 hours some crystals as sublimate appear in the upper part of the Erlenmeyer
flasks showing the liability of some loss, which might occur under this severe condi¬
tion of drying.
3. Solubility
Barbituric acid derivatives are slightly soluble or even insoluble in water,
practically insoluble in petroleum ether. They are variably soluble in ether, alcohol,
chloroform, benzene and acetone. They are soluble in alkalies and alkali carbonates.
The solubility of the barbituric acid derivatives dealt with in this thesis is
shown in the following table:
- 19 -
Table 5
The solubility of barbiturates
(one part of barbiturate in x parts of solvents)
Barbituric
acid
Water
cold boiling
Alcohol Ether Chloroform Acetone Literature
Allobarbital 300 50 20 20 - v. s. Ph.Helv.V. X)
800 50 11 21 100 - Ph. Dan.5'
700 - 15 20 - - B.P.C.9)
Cyclobarbital v.sl.s. 100 5 20 -_ Perha 10>
sl.s. 100 v. s. - -- B.P.C. 9)
Hexobarbital v.sl.s.
3000 -
sp.s.
f.s.
in anhyd.alcohol
sp.s.
f.s.
v. s.
f.s. -
Ph.Helv.V.1'Suppl. II
B.P.C.9'
3500 250 40 50 6 - Ph. Dan.5'
Methylpheno-barbital
v.sl.s.
insol. 500
f.s.
200
sl.s.
150 60
- Ph.Helv.V.1^Suppl. II
Ph. Dan.5'
Pentobarbital sl.s. - f.s. f.s. f.s. f.s. Perha10)
The following letters were used in the previous table:
v. s. = very soluble = soluble in less than 1 part of solvent,
f.s. = freely soluble = soluble in from 1-10 parts of solvent,
s. = soluble = soluble in from 10 - 30 parts of solvent,
sp. s. = sparingly soluble = soluble in from 30 - 100 parts of solvent,
sl.s. = slightly soluble = soluble in from 100 - 1000 parts of solvent,
v. si. s. = very slightly soluble = soluble in from 1000 - 10 000 parts of solvent,
insol. = insoluble = soluble in more than 10000 parts of solvent.
- 20 -
4. Dissociation constants and dissociation exponents (pK values)
of barbituric acid derivatives
a) Literature
After Biggs' and according to Erlenmeyer et al. ' who have shown
that replacement of all four hydrogen atoms by deuterium occurs in a solution of
barbituric acid (n) in heavy water: it follows that enolization must occur in at least
two of the 1:2-,1:6- and 5:6 positions or the equivalent 2:3-, 3:4-, and 4:5-
positions and may occur in all three (III).
HN -C =0 N C-OH
I \ « I_
I IO = C 2 5 CH2
-
HO - C CH
I 3 4 I••
«
HN- - C = O N C-OH-i,
n m
Biggs' had found by potentiometric titration that, as might be expected, 5:5-
diethyl-l:3 dimethyl barbituric acid was devoid of acid properties and possessed no
absorption at wavelengths of 241 mu or longer i. e. there was no absorption in the
region where less fully substituted barbituric acid absorbed freely. A trisubstituted
acid, 5-cyclohexenyl l:5-dimethyl barbituric acid in which enolization can occur only
in the 2:3 or 3:4 - positions, shows only weak acidity (pK 8. 37) but if there is a3.
hydrogen atom in the 1 position, then with increased possibility of enolization the
acidity increased. Thus seven 5:5-disubstituted acids were found to have pK values
ranging from 7. 73 to 7. 99 (the 5-benzyl-5-ethyl acid having pK 7.45) i. e. they are
about 3 times stronger. In l:3-dimethylbarbituric acid enolization can occur in the
4:5 and 5:6 positions and this leads to acidity of a much higher order (pK 4. 68).
Barbituric acid, the 1-methyl acid and the 1: 3-dimethyl acid have spectra of the same
type (markedly different from that of the 5:5 disubstituted acids) and it is therefore
probable that they are structurally similar and enolization which might be expected
in the first two of these acids at 2:3 and 3:4- positions does not occur to any marked
extent here. The decreasing acidity of barbituric acid (pK 4.04), the 1-methyl acid
(pKa 4. 35) and the 1:3 dimethyl acid (pKa 4. 68) may therefore be ascribed to the
introduction of methyl groups. Finally, if enolization is reduced by substitution in
the 5 position, as in the 5-isopropyl acid (pK 4. 94), then, as would be expected the
acidity is also decreased.
11) A.I. Biggs, J.Chem.Soc. 2485 (1956).12) H. Erlenmeyer et al. Helv.Chim.Acta., 19, 354(1936).
- 21 -
The difference in substitution at C 5 with different groups have a little effect
on the dissociation constants. Comparison of the dissociation constants given by
Wood ' between:
.7barbituric acid 1051 x 10
5-ethyl barbituric acid 383 x 10"7_7
5:5-dimethyl barbituric acid 0. 73 x 10
_75:5-diethyl barbituric acid 0.37 x 10
showed that the strength of the barbituric acid derivative depends on whether one
hydrogen atom at C 5 or the two are substituted. When the two hydrogen atoms are
substituted a very weak acid is obtained. In case of diethyl barbituric acid when one
ethyl is substituted with a phenyl group the acid character is strengthened. On the
other hand substitution with cyclohexenyl group (Cyclobarbital) gives a weaker acid
characters. Through the introduction of bromine containing radicals the dissociation
constants are slightly increased. The methylation of Nitrogen atom of barbituric acid
gives derivatives with a lower dissociation constants.
Poethke and Horn14\ Biggs11' C lowes, Keltch and Krahl15'
gave the following dissociation constants (K) and dissociation exponents (pK ) for
some barbituric acid derivatives, as shown in the following Table 6.
13) J.K.Wood, J.Chem.Soc. Transactions, 89, n, 1831 (1906).14) W. Poethke and D.Horn, Archiv der Pharmazie, 287, 487 (1954).
15) G.H.A. Clowes, A.K. Keltch and M.E. Krahl, J.Pharmacol. & Exper.Therapy, 68, 312 (1940).
- 22 -
Table 6
Dissociation constants and dissociation exponents of barbiturates
Barbituric
acid derivative
Method (a)
at 25°
K PKa
Method (b)
at 18°
K pK*
a
at 20°15
Biggs11) Krahl
AUobarbital
Barbital
Cyclobarbital
Hexobarbital
Methylphenol-barbital
Phenobarbital
Phentobarbital
1.55 xlO"8 7.81
4.47 xlO~8 7.35
0.91xl0"8 8.04
2.24xl0"8 7.65
5.01xl0"8 7.30
1.23xl0"8 7.91
3.09xl0"8 7.51
0.83xl0"8 8.08
1.74xl0-8 7.76
6.17xl0"8 7.21
7.77 7.62
7.97 7.91
7.36
8.37 8.24
7.45 7.41
8.02
Both methods (a) and (b) are recorded by Poethke and Horn '
Method (a) = Determination of pK was carried out colorimetric with Zeiss Pulfrich
Photometer at 25° C.
Method (b) = Determination of pK was carried out potentiometric with Ionometer
after Lautenschlager with a quinohydrone electrode at 18° C.
The comparison between the given pK values in Table 6 shows a big difference
which could not only be due to the difference in temperature at which the pK values
were determined. This stimulated us to determine all the pK values of the official
barbiturates in the Swiss Pharmacopoeia and its addenda.
b) Determination of the dissociation constants and dissociation Exponents (pKa values)
We carried out the determination of the dissociation constants of the barbituric
acid derivatives, official in the Swiss Pharmacopoeia, in the Analytical Chemistry
Institute of the Swiss Federal Institute of Technology, Zurich. I wish here to express
my thanks to Prof. Dr. G. Schwarzenbach for his permission to carry out this work.
My thanks are also extended to Dr. G. Anderegg for his supervision and guidance
during the measurements and calculations.
The dissociation constants were calculated from the pH measurements.
- 23 -
Apparatus and Procedure: A special closed measuring cell was used. This
was adjusted to constant temperature (20° C) by a thermostat. Carbon dioxide free-
nitrogen was bubbled all the time of experiment in the measuring cell. The barbituric
acid derivative was dissolved in specially distilled water (carbon dioxide free, distilled
from an all glass pyrex apparatus). This solution was introduced in the measuringcell and titrated against 0.1 M sodium hydroxide using glass and calomel electrodes
attached to a Beckman's pH meter. A magnetic stirrer was used to ensure mixing.
The following steps were carried out to determine the dissociation constants:
1. Determination of the factor of 0.1 M sodium hydroxide,
2. Caliberation of the Beckman's pH meter,
3. Titration of the barbituric acid derivatives,
4. Calculation of the dissociation constants.
1. Determination of the factor of 0.1 M sodium hydroxide
The determination of the factor of sodium, hydroxide was done as follows:
5.00 ml. 0.1 M hydrochloric acid (factor = 1.00) and 10 ml. of 1 M potassiumchloride were introduced in 100 ml. measuring flask and completed to volume with
distilled water. This solution was introduced in the measuring cell and titrated
gradually against 0.1 M sodium hydroxide under an atmosphere of nitrogen as shown
in the following example:
ml- ml.
0.1 M NaOH pH 0.1 M NaOH pH
5.51
5.67
5.86
6.23
7.43
10.05
10.70
11.10
This process was repeated till concordant results were obtained, and was carried
out everytime before the titration of barbituric acid derivatives.
5Factor of 0.1 M sodium hydroxide = = 1.035.
4.83
2. Caliberation of the Beckman's pH meter
This was done by using 1.00 ml. of 0.1 M hydrochloric acid (factor = 1.00) and
10 ml. 1 M potassium chloride which were introduced in a measuring flask of 100 ml.
capacity and completed to 100 ml. with distilled water. The solution was transferred
to the measuring cell and titrated against 0.1 M sodium hydroxide (factor = 1.035)under an atmosphere of nitrogen, measuring the pH after each addition. The caliberation
of the apparatus was carried out everytime before the titration of the barbituric acid.
An example of the caliberation of the apparatus is as follows:
3.72 4.72
1.00 3.83 4.76
2.00 3.92 4.78
3.00 4.17 4.80
4.00 4.51 4. 82 end point 4.83
4.20 4.62 4.84
4.40 4.80 4.87
4.60 5.05 4.90
4.70 5.33
- 24 -
ml.
MNaOH PH c.pH D
_ 4.40 3.017 1.38
0.10 4.45 3.066 1.38
0.20 4.52 3.122 1.40
0.30 4.58 3.186 1.39
0.40 4.66 3.261 1.40
0.50 4.76 3.352 1.41
0.60 4.88 3.467 1.41
0.70 5.05
0.80 5.33
0.8i 5.40
pH = measured pH value after each addition of alkali.
c. pH = corrected pH value; calculated from the end point found in caliberation of
pH meter and the normality factor of 0.1 M sodium hydroxide as follows:
0.1 M NaOH [h+J = hydrogen ion concentration c. pH = - log[rI+J0.93 x 1.035 x 10"3 = 0.962 x 10"3 3.017
0.1 0.83 x 1.035 xlO"3 = 0. 859 x 10"3 3.066
0.2 0.73 x 1.035 x 10"3 = 0.755xl0"3 3.122
D = the difference between the corrected and the measured pH values. The mean of
the difference (D) in this case was equal to 1.39. This was used to correct the
measurements of pH values in the titration of barbituric acid by substraction from
the measured pH values.
3. Titration of the barbituric acid derivatives
ex) Barbituric acid derivatives (free acids)
-4About 1x10 of the molecular weight of the dried barbituric acid derivative
(3 hours 103 - 105°) was accurately weighed and introduced in 100 ml. measuringflask and dissolved in about 80 ml. distilled water by the aid of a hot current of water
or in a water bath (i.e. about 1 x 10-3 mol. weight per liter). 10 ml. 1 M potassiumchloride were added after cooling the previous solution and completed to 100 ml. with
distilled water. This solution was transferred to the measuring cell and titrated
gradually with 0.1 M sodium hydroxide (factor = 1.035).
An example of titration of allobarbital using 0.02147 g. gave the following data:
ml.
0.1 M-NaOH
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
ml.
0.1 M NaOH PH
0.84 5.50
0.86 5.65
0.88 5.83
0.90 6.24
0. 92 end point 0. 93 7.18
0.94 9.12
0.96 10.41
0.98 10.741.00 10.97
pH c. pH/pH
7.30 5.91
8.19 6.80
8.505 7.11
8.72 7.33
8.91 7.52
9.09 7.70
9.26 7.87
9.46 8.07
9.70 8.31
10.01 8.62
10.69 9.30
- 25 -
From the corrected pH values obtained in the previous titration a curve is drawn
plotting the corrected pH values against the amount of 0.1 M sodium hydroxide
(factor = 1.035) added, e.g. see Curve 1 (Allobarbital).
4. Calculation of the dissociation constant
The dissociation of a weak acid can be represented by the equation:
H A_
- H+ + A"
And according to the law of mass action
[h+] [a"3K =
[HA]
If to the solution is added a small amount of a strong base, say sodium hydroxide,
a part of the acid is neutralized.
Let C represent the analytical concentration of the acid per liter.
cs = [h a] = [h a] + [a- ]total
By measuring the pH value of the solution [h J is known and |_OH~J can be
calculated from this equation:
Kw20° = 10"13-96 = [h+] [OH"] = O.llxlO"13
,-13
T^TT-1 0.11 xwhere
[OH J=
10
[h+]
Since the amount of sodium hydroxide added is equivalent to
[NaOH] = [a" ] + [OH-] - [h+ ]
[a"] = [NaOH] - [oh"] + [h+]
J [NaOH]ml. NaOH x factor of NaOH
volume of titrated solution
[ha] = [ha] - [A" ]total
Substituting the values of [a"] , [H+ ] and [h a] in the following table, the
dissociation exponent (pK values) can be determined.
- 26 -
Table 7
4) Calculation of pK -value for Allobarbital
ml.
0.1 M-
NaOHpH [«1 [oh-] [saOH] M [ha]-c,-[a-]
-logKMean pK
0.20 7.11 0. 776 X 10"7 0.141 X 10"6 0.207xl0"3 0.207 X10-3 0.826 xlO-3 -0.601 7.711
0.30 7.33 0.468 X10-7 0.234x10"* 0.310xlO"3 0.310 xlO"3 0.729 x 10"3 -0.372 7.702
0.40 7.52 0.302 xlO"1 0.364 xlO'8 0.414 xlO-3 0.413 xlO"3 0.619 xlO-3 -0.177 7.697
0.50 7.70 0.200 X 10"' 0.550 x 10'6 0.517xl0-3 0.517 xlO-3 0.516xl0-3 0.000 7.7007.69
0.60 7.87 0.170X10-7 0.647 XlO-6 0.621 xlO-3 0.620 xlO-3 0.412 XlO'3 0.177 7.693
0.70 8.07 0.085 XlO"7 1.290 xlO"6 0. 724 x 10"3 0.723 x 10"3 0.309 x 10"3 +0.380 7.690
[NaOIi] = ml. 0.1 M sodium hydroxide (factor = 1,035) added multiplied by the
molarity.
C = the analytical concentration of the barbituric acid per liter. This was
calculated by dividing the weight of the barbituric acid used over the molecular weight
of the barbituric acid which in this case was equal to 1.033 x 10.
R) Salts of barbituric acid derivatives
In case of sodium salts of barbituric acid derivatives e. g. Pentobarbital sodium
the following method was used in titration of the salt:
-4About 4 x 10 of the molecular weight of the salt was accurately weighed and
introduced in a measuring flask of 100 ml. capacity. This was dissolved in distilled
water and completed to volume. 25.00 ml. of this solution were introduced in a
measuring flask of 100 ml. capacity, 10 ml. potassium chloride were added and
diluted to about 90 ml. with distilled water. 2.00 ml. of 0.1 M hydrochloric acid
(factor = 1.00) were added and the volume completed to 100 ml. with distilled water.
This solution was transferred to the measuring cell and titrated gradually against0.1 M sodium hydroxide (factor = 0.984), with 0.0248 g Pentobarbital sodium. We
obtained the following data:
27 -
ml. PH c- pH0.1 M - NaOH (pH - 2.
. 5.38 2.990.10 5.42 3.03
0.20 5.48 3.09
0.30 5.56 3.17
0.40 5.61 3.22
0.50 5.69 3.30
0.60 5.79 3.400.70 5.91 3.52
0.80 6.08 4.69
0.82 6.13 3.74
0.84 6.18 3.79
0.86 6.23 3.84
0.88 6.30 3.91
0.90 6.37 3.98
0.92 6.46 4.07
0.94 6.57 4.18
0.96 6.72 4.33
39)ml. pH c-pH
0.1 M - NaOH (pH - 2. 39)
0.98 6.93 4.54
1.00 7.29 4.901.02 8.12 5.73
Vr1.04 8.81 6-42
1.06 9.07 6.68 P°6.87
1-1.08 9.26
1.10 9.39 7.00
1.20 9.80 7.41
1.30 10.05 7.66
1.40 10.25 7.86
1.50 10.42 8.03
1.60 10.60 8.21
1.70 10.80 8.411.80 11.01 8.62
1.90 11.30 8.91
2.00 11.77 9.38
The end point [1.017 ml. 0.1 M NaOH (factor = 0.984)] found in the previous
titration was substracted from the added ml. of 0.1 M sodium hydroxide in order to
calculate the dissociation exponent as shown in the following table 8 under ml. The
previously described method of calculation was also here applied.
Table 8
Calculation of the pK -value for Pentobarbital sodium
ml.
0.1 M-
NaOH
PH [»i [OH"] [NaOH] [A"] [ha]-cs-[a-][ha]
-log K
-pKamean pK
0.283 7.66 0.2188xl0~7 0.5025 x 10"6 0.2780 xlO"3 0.2774 x 10"3 0.7236 xlO"3 - 0.416 8.076
0.383 7.86 0.1381 xlO-7 0.7960 x 10"6 0.3770xl0"3 0.3762 XlO"3 0.6248 XlO"3 - 0.220 8.080
0.483 8.03 0.0934 xlO-7 1.1780 x 10"6 0.4750 xlO"3 0.4750 xlO"3 0.5272 x10"3 - 0.047 8.0778.09
0.583 8.21 0.0616 xlO-7 1.7860 x 10"6 0.5730 xlO*3 0.5712xl0"3 0.4298 XlO"3 + 0.124 8.086
0.683 8.41 0.0389x 10"7 2.8300 x 10"6 0.6720 xlO-3 0.6691 x10"3 0.3319xl0"3 + 0.305 8.105
0.783 8.62 0.0240 x 10"7 4.5800 x 10"6 0.7700 xlO"3 0.7654 x 10"3 0.2356 xlO"3 + 0.512 8.108
[NaOH] = ml. 0.1 M sodium hydroxide added multiplied by the molarity.
C = the analytical concentration of the barbituric acid per liter. This was cal¬
culated by dividing the weight of the barbituric acid used over the molecular
weight of the barbituric acid, which in this case was equal to 1.001 x 10.
- 28 -
X) Water insoluble barbituric acid derivatives
In case of water insoluble barbituric acid derivatives e. g. Methylphenobarbital,
the procedure was modified as follows:
About 100 mg. of Methylphenobarbital were introduced in a measuring flask of
100 ml. capacity, 10 ml. of potassium chloride were added and the contents completedwith distilled water to volume (a slight excess of water not more than 1 ml. was added
to compensate for the solution lost by filtration). The solution was shaked occasionally,left for about 14 hours and then filtered through acid free filter paper in another mea¬
suring flask of 100 ml. capacity. This solution was transferred to the measuring cell
and titrated gradually with 0.1 M sodium hydroxide (factor = 0. 984) using smaller
quantities of sodium hydroxide".
We obtained the following data:
ml.
0.1 M - PH corrected pHNaOH (pH -2.35)
- 8.86 6.31
0.02 9.387 7.03
0.025 9.50 7.15
0.050 9.83 7.48
0.075 10.11 7.76
0.100 10.37 8.02
0.125 10.66 8.310.150 11.10 8.75
0.175 11.52 9.17
0.200 11.80 9.45
0.225 11.99 9.64
0.250 12.17 9.82
0.275 12.27 9.92
0.300 12.38 10.030.325 12.46 10.110.350 12.52 10.17
0.400 12.64 10.29
0.500 12.80 10.45
0.600 12.92 10.57
The calculation of the dissociation exponent (pK value) for Methylphenobarbital
was carried out in another way since the concentration of the titrated acid was not
known.
- 29 -
In case of a very weak acid e. g. Methylphenobarbital (more than pK 7) X H which
represents a weak acid, dissociates in water according to the following equilibrium
X H= H+ + X" (1)
The analytical concentration of this acid could be represented by
Mtotal
= tXH] +M <2>
_
[x-] fe»]
[XH]
If to this acid solution sodium hydroxide solution is added, so that partial neutralization
occurs, we get the following relation:
The dissociation constant i.e. K,. =
,. I; (3)
Let Z = — and Y=[x](
|[NaOH] - [OH] |
[NaOH] = [X-] + [OH^] (4)
[X-] = [NaOH] - [OH-] (5)
Also from (2) [xh] = [x]total - [x-]
i.e. [XH] = [x]tQtal - {[NaOH] - [OH-]} (6)
substituting the values of [x-] and [x h] in equation (3) we get
K =
fc+l ' 1 [NaOH] - [OH~])(7)
W total -{[Na0Hl " [OH"]}Reversing this equation we get:
M._
CX]total-{[NaOH3 -^OH^lK {[NaOH] - [OH^}
10-pK {[NaOH] - [OH]}
—-f^T
r il+1 =° (10)
10-pK J [NaOH] - [OH]j
(8)
-dK'
J total
10 pK
then Z 10"pH -.X +1=0 (11)
- 30 -
Equation (11) is a linear equation having the values of Z and Y as unknown. A method
to solve this equation could be obtained by determining the pH values following the
addition of successive known amounts of sodium hydroxide solution. For each pH
value a straight line was drawn in a Caeserian system (when plotted graphically)
having and <[NaOH] - [OH][ as the Z and Yaxes respectively. The point,10-PH l '
where all these lines intersect, represents the pK and [x] . . . of this weak acid.
In equation (11) the following conditions could be assumed:
When Z = 0 then Y = I [NaOH] - [OH"]Jandwhen Y = 0 then Z =
10-PH-13.96
Knowing that [oh"J [h+] = io"13,96 at 20° C. then [OH-] = -j? —
[H+]The following example illustrates the application of these equations in the determination
of both the dissociation constant K and the solubility of Methylphenobarbital. Data for
the measurements are given on page 28. From these data Y and Z are calculated after
addition of 0.05 ml. 0.1 M sodium hydroxide (molarity = 0.492 x IO-4)
[NaOH] = 0.05 x 0. 984 x 10"3
[OH'] =1£^
__ 10-6-48=0.331xl0-610-7.48
Y = [NaOH] - [OH] = 0.492xl0-4 - 0.331 x 10"6 = 0.489 xlO-4
Z = - _J_ = -_J_
= . i.10+7-48 = -3.02 xlO710-pH 1Q-7.48
after the addition of 0.075 ml. 0.1 M. NaOH, Z and Y are calculated as follows:
[NaOH] = 0.075 x 0. 984 x 10"3 = 0. 738 x IO"4
[OH"] .12^= l0-6-20
= 0.63 xlO'610-7.76
Y = [NaOH] - [OH] = 0.738 x IO-4 - 0.631 x 10-6 = 0.732 x 10"4J-7.76
Z =1
= -l.io7-76 = -5.75 xlO-710
after the addition of 0.05 ml. 0.1 M. NaOH, Z and Y are calculated as Follows:
[NaOH] = 0.10 x 0. 984 x 10-3 = 0. 984 x IO-4
- 31 -
Y = [NaOH]-[oH"J = 0.984 x 10-4 - 0.115 xlO"5 = 0.975 x 10"4
Z = - =-1.108-02 = -1.047 xlO810-8.02
after the addition of 0.125 ml. 0.1 M NaOH, Z and Y are calculated as follows:
[NaOHJ = 0.125x0.984 x 10"3 = 1.23xl0"4
[OH'] = IP'13'96 = 10"5-65 = 0.224 xlO-5
10-8.31
Y = [NaOH] - [OH'J = 1.23 xlO-4 - 0.224 xlO-5 = 1.210 xlO'4
Z = - = -1.108-31 = -2.042 x 10810-8.31
Z and Y are plotted in Curve 8' on millimeter paper and from this figure their values
1were determined where the four lines intersect. Z was found to be equal to 6.75 x 10
_4and Y was found to be equal to 1. 60 x 10
.
since Z1
i.. e.1
= 6, 75 x 107
10-PK 10 -PK
10pK = 6..75 x:107 = io7- 829
£K = 7.,829 = 7^83
since Y = [x]t = 1.60 x 10-4 x Molecular Weight
_4i.e. Tthe solubility of methylphenobarbital = 1.60x10 x 246.13 = 0.0394 g./liter.
- 32 -
Results of the determination of the dissociation exponents (pK values) at 20 C of
the official barbituric acid derivatives of the Swiss Pharmacopoeia 1933
and its addenda
The previously described procedure and method of caluclation by which the
dissociation exponents (pK„) of Allobarbital was determined, were applied here fora
Barbital, Cyclobarbital, Hexobarbital, Methylphenobarbital, Phenobarbital and
Pentobarbital sodium. The determined dissociation exponents are given in the follo¬
wing table:
Table 9
Our determined dissociation constants (K) and dissociation exponents (pK ) of
the barbituric acid derivatives
Barbituric acid
derivative
Experiment 1
K PKa
Experiment 2
K PKaExperiment 3
K pKaCurve
Allobarbital
Barbital
Cyclobarbital
Hexobarbital
Methylpheno¬barbital
Phenobarbital
Pentobarbital
(sodium)
2.042 xlO"8 7.69
1.202x10"® 7.92
2.090x10"® 7.68
0.514x10"® 8.29
1.480x10"® 7.83
3.803x10"® 7.42
0.760x10"® 8.12
1.95xl0-8 7.71
1.26x10"® 7.90
2.14x10"® 7.67
0.525x10"® 8.28
1.480 xlO"8 7.83
3.803x10"® 7.42
0.814x10"® 8.09
0.525x10"® 8.28
1
2
3
4
7, 8&8'
5
6
- 33 -
277-
2 0 4 0 6 0 8 1
Curve 1
pKa value for Allobarbltal (0 02147 g )
12 ml 0 1H NaOH
>
7t
» -4
ir
?
S/
7
T_'
/
It^
5
02 04 06 08 1 1 2 ml 0 1M NaOH
Curve 2
pK^ value for Barbital (0 0168 % )
iJ
/'
y
r
'
//
y'
/
11
02 04 06 08 12 mLO 1H NaOH
Curve 3
pKg value for CyclobarbiUl ( 0 0236 g )
0-/- -Jr
Tr^
z.j/_
/_7
^/
/
^tLr
el32 04 06 08 1 1 2 ml 0 1M NaOH
pKa value for Hexobarbital (0 0236 g )
- 34 -
0.2 0.4 0.6 0.8 1ml. O.lMNaOH
pK value for Phenobarbltal (0.0327 g.)
pH 101 1 1 1 1 1 1
mp
2 ml 0. lMNaOH
Curve 6
pK„ value for Pentobarbital Sod. (0.0248 g.)
1ft I
Solubility of Methylphenobarbltal = 0.00394 %
0.1 0.2 0.3 0 4
Curve 7
pKa value lor Methylphenobarbltal (0.00394 g.)
8 ml. 0 1 H NaOH.175 .2!
ml. 0.1 M NaOH
Curve S
pKa value for Methylphenobarbital
6xL07
iter
g./l
0394
0.
=
246.13
x1.6xl0"4
=
Mol.Wt.
x1.6x10
=Solubility
7.83
=7.829
=pK
i.e.
107-
829
=6.75xl07
=10
pK
10"PK
6.75xl07
-
1
1.6x10
=Y
and
6.75x10
=Z
-4
7
cuurve
10'
+1}
-lxl
|07
ilO"
0.1
'
/
:clOH
0.3.
/
clO"'
1.6
Y
- 36 -
5. Ultra violet spectrum
Absorption spectrophotometry in analysis of barbituric acid derivatives
a) Literature
Elvidge summarised the application of absorption spectrophotometry in
pharmaceutical analysis in the following points:
1. As applied to the determination of the purity of any one sample, spectros¬
copic data are to be regarded as secondary and not as primary standards. Considerable
impurities may be present in certain circumstances without being detected spectros-
copically. Taken in conjunction with other physical and chemical data, however, they
do afford a good measure of the degree of purity or normality of a sample.
2. Spectroscopic methods in general are capable of a high degree of accuracy
and once the value for a compound has been established, it may be relied upon to
give good results, comparable with other methods of analysis and superior in
sensitivity and to allow the accurate determination of much smaller quantities
ranging from 1/10 to 1/1000 of the amount possible by ordinary analytical methods
used in pharmacy. They are particularly suitable to routine work and are of great
use in those cases, where only small amounts of sample are available.
3. In qualitative work their value is very restricted and can only be used in the
sense of offering confirmatory evidence of results obtained by other methods.
4. In practically every case where an absorbing compound is concerned some
suitable spectroscopic feature can be found which is adaptable to quantitative work.
5. Solvent influence is important and requires investigation as widely as possible
in any particular case.
Elvidge stated that the strength of alkali had an influence on the extinction
values of barbiturates. He also pointed out that it is difficult by absorption methods
to distinguist between Barbital, Phenobarbital and Allobarbital. Hexobarital differed
from the remainder by failing to show selective absorption in alkaline solution.
17)E lvidge discussed the influence of solvent on the extinction coeffiecient of
barbiturates. In general acids themselves show only general absorption with no peak.
The introduction of sodium to form a soluble salt produces a marked peak and this
becomes more pronounced and displaced slightly towards longer wave length with
increase of caustic soda e.g. Barbital, Phenobarbital and Allobarbital.
According to E lvidge ,the use of absorption spectroscopy in qualitative
16) W.F.Elvidge, Quart. J. Pharm. Pharmacol., 14, 134(1941).17) W.F. Elvidge, Quart. J. Pharm. Pharmacol., T7, 219(1940).
- 37 -
analysis is severely restricted by the fact that compounds of similar chemical
constitution yield similar absorption curves. In a homologous series the shape of
the curves will remain practically the same although there may be slight displace¬
ment of the peak absorption and generally the extinction values increase with
increasing the molecular weight. From a practical point of view such differences
are difficult to detect and no great relibability can be placed on them for work of
this nature.
For the determination and differentiation of barbiturates, Goldbaum '
developed a method which is based on the change in the absorption spectra of
barbiturates in strong alkali and in solution at pH 10. 5. When the optical densities
in pH 10. 5 solution are substracted from those in strong alkali (0.45 N NaOH),
differences appear that are highly characteristic of all barbiturates except the
n-methyl- and thio-derivatives. By comparing the differences at various wavelengths
with that at 260 m u, ratios are obtained that can differentiate among many of the
commonly used barbiturates.
19)Stuckey examined the ultra-violet absorption spectra of 1-methylbarbituric
acid, 1: 3-dimethylbarbituric acid and barbituric acid,in acid, alkaline solution and
at varying dilutions in water. From the similarities shown in the spectra of all three
compounds and in particular the almost identical value of £ at ca 2600 A in
alkaline solution, he suggested that barbituric acid in aqueous solution undergoes
only one enolization involving the active methylene group in position 5.
Stuckey'examined the ultra-violet absorption spectra of a number of 5 : 5-
disubstituted barbituric acid derivatives. He found that in general all these substituted
compounds show a peak in alkaline solution ca 2500 A. This peak absorption can be
used for estimating small quantities of these compounds. He gave the following data:
Table 10
Barbituric acid derivatives Cmax.
Wavelength
inO.lN NaOH in A
1 -Allobarbital 7200 2520
2-Cyclobarbital 8400 2540
3-Methylphenobarbital 9000 2460
18) L.R. Goldbaum, Anal.Chem., 24, 1604(1952).19) R.E. Stuckey, Quart. J. Pharm. Pharmacol., 14, 217(1941).20) R.E. Stuckey, Quart. J. Pharm. Pharmacol., 15, 370, (1942).
- 38 -
He concluded that there is no general relation between the substituent group
and the wavelength or the molecular extinction coefficient. Further, there is no
method of distinguishing between individual barbituric acid derivatives, the maximum
of absorption of each being ca 2500 A, and the molecular extinction coefficient being
generally within the limits 7000 and 9000.
b) Determination of the ultra-violet absorption spectra of barbituric acid derivatives
Experimental: The extinction coefficient was measured with either,a) Beckman D U spectrophotometer equipped with ultra-violet accessories; or,
b) PMQ II Carl Zeiss spectrophotometer; using quartz 1 cm. cells.„i\
Reagents: a buffer solution having a pH of 9. 5 was prepared after Mattson'
as follows:
1-Boric acid and potassium chloride, 0.1 M (dissolve 6.185 g. of boric acid and
7.455 g. potassium chloride in water and dilute to 1000 ml. with water).2-Sodium hydroxide, 0.1 N prepared carbonate free and standardized volumetrically.The buffer solution having a pH of 9.5 was prepared by adding 34.43 ml. of 0.1 N
sodium hydroxide to 50 ml. of boric acid potassium chloride solution and diluting to
100 ml. with water.
The following data are given as an example of spectrophotometric measurements
of Allobarbital in buffer solution at pH 9.45, using Beckman's DU spectrophotometerand 1 cm. cells
21) L.N. Mattson, J. Amer. Pharm.Ass. (Sci.Ed.), 43, 22(1954).
- 39 -
Table 11
Spectrophotometry measurements of Allobarbital
Sol. No.
mg.
1
0.14
2
0.15
3
0.16
4
0.17
5
0.18
6
0.19
7
0.20
m u
244
242
240
E
0.229
0.237
0.240
E
0.245
0.255
0.257
E
0.261
0.275
0.276
E
0.275
0.290
0.294
E
0.291
0.305
0.310
E
0.310
0.322
0.327
E
0.321
0.336
0.340
238
236
0.238
0.225
0.252
0.240
0.272
0.255
0.288
0.274
0.302
0.286
0.320
0.305
0.332
0.316
m u
244
242
240
8480
8780
9250
Molecuh
8490
8830
8910
ir Extincl
8480
8930
8990
ion Coefi
8420
8880
9010
icient (6
8400
8820
8980
)
8490
8820
8980
8320
8730
8820
Mean
8440
8820
8990
238
236
8800
8320
8720
8320
8850
8290
8810
8400
8730
8280
8780
8350
8630
8210
8760
8310
Sol. No. = solution number.
mg. = mg. of dried Allobarbital present in 25 ml. of buffer solution.
E = the measured extinction coefficient.
m u = wavelength in millimicron at which measurements were taken.
£ = Molocular extinction coeffient.
The molecular extinction coefficient (£ ) were calculated by dividing the extinction
coefficients (E) over the molecular concentration per liter (Molarity) of each dilution.
The mean values of the molecular extinction coefficients were plotted against
the wavelengths as shown in Curve 14, (pH 9.5, buffer).
- 40 -
The procedure and measurements carried out in case of Allobarbital were
applied to Cyclobarbital, Hexobarbital, Methylphenobarbital and Pentobarbital sodium
at pH 9. 5; using the same buffer solution. In case of Allobarbital and Methylpheno¬
barbital Beckman's DU spectrophotometer was used. PMQ Carl Zeiss spectrophoto¬
meter was used in case of Cyclobarbital, Hexobarbital and Pentobarbital sodium. The
measured maximum extinctions of the previously mentioned barbituric acid derivatives
are summarised in the following table:
Table 12
Spectrophotometric measurements of the examined barbiturates
Allobarbital mg.
E 240 mu.
0.14
0.240
0.15
0.257
0.16
0.276
0.17
0.294
0.18
0.310
0.19
0.327
0.20
0.340
Curves
9
Cyclobarbital mg.
E 238 mu.
0.16
0.282
0.18
0.315
0.20
0.351
0.21
0.365
0.23
0.400
0.25
0.432
0.27
0.47310
Hexobarbital mg.
E 244 mu.
0.16
0.216
0.18
0.241
0.20
0.263
0.21
0.276
0.23
0.301
0.24
0.313
0.25
0.32511
Methylpheno¬
barbital
mg.
E 244 mu
0.16
0.235
0.17
0.235
0.18
0.250
0.19
0.264
0.20
0.275
0.21
0.286
0.22
0.30012
Pentobarbital
sodium
mg.
E 240 mu
0.15
0.235
0.17
0.254
0.19
0.272
0.20
0.295
0.22
0.325
0.24
0.354
0.26
0.38113
mg. = mg. of the dried barbituric acid derivative present in 25 ml. of buffer
solution.
E mu. = Maximum extinction coefficient at the specified wavelength in milli¬
micron.
Curves were drawn from the previously given data, plotting the weights of barbi¬
turic acid derivatives against the extinction coefficients e. g. Curve 9 - 13. All these
curves gave straight lines. These curves can be used in the quantitative estimation of
very small amounts of barbituric acid derivatives.
The mean values of the molecular extinction coefficients were calculated as
shown in the case of Allobarbital Table 11, and were plotted in curves 14 - 18.
The effect of the strength of alkali on the extinction values was studied by deter¬
mining the absorption spectra of barbituric acid derivatives in strong alkaline solutions.
- 41 -
In strong alkaline solutions e. g. 0.1 N sodium hydroxide, the barbituric acid deriva¬
tives showed decomposition (pH about 12. 5) and the measurements of the extinctions
should be carried out using only freshly made solutions e. g. in cases of Cyclobarbital,
Hexobarbital and Methylphenobarbital. In cases of Allobarbital and Pentobarbital sodium
the decomposition was not very rapid, compared to the previously mentioned barbi¬
turic acid derivatives. In general, freshsly prepared solutions should be used in
measurements of the extinction coefficients in alkaline solutions.
In all the following measurements, PMQ II Carl Zeiss spectrophotometer was
used. 0.1 N sodium hydroxide was used as a solvent in all cases except in case of
Allobarbital where 1 N sodium hydroxide was used. The obtained values of the molecular
extinction coefficients are shown in Curves 14 - 19. We compared these measurements
of our examined barbiturates in water solution (pH about 6.2), in buffer solution (pH
about 9.5) and in 0.1 N sodium hydroxide (pH about 12).
Table 13
Summary of the maximum and minimum molecular extinction coefficients of the examined
barbituric acid derivatives in different aqueous solvents
1. Allobarbital
a) in water solution pH 5.55
b) in buffer solution pH 9.5
c) in 1 N sodium hydroxide pH about 13
no peakmax. 8990
max. 6490
max. 13350
min. 3640
mu
240
256
222
234, 236
Curve
14
2. Cyclobarbital
a) in buffer solution pH 9.5
b) in 0.1 N sodium hydroxide pH 12.05
max. 10300
max. 7000
min. 5088
238
250, 252
234
15
3. Hexobarbital
a) in water solution pH 5.55
b) in buffer solution pH 9.5
c) in 0.1 N sodium hydroxide pH 12.05
max. 7760
max. 7790
max. 7340
min. 3770
220
244
244
224
16
4. Methylphenobarbital
a) in buffer solution pH 9.5
b) in 0.1 N' sodium hydroxide pH 12.05max. 8570
max. 9490
max. 11320
min. 5780
244
246
216, 218
228
17
5. Pentobarbital sodium
a) in buffer solution pH 9.45
b) in 0.1 N sodium hydroxide pH 12.05
max. 9180
max. 7310
max. 11910
min. 5400
240
242
218
230
18
6 = molecular extinction coefficient,
max. = maximum molecular extinction coefficient,
min. = minimum molecular extinction coefficient,
mu = wavelength in milli micron.
- 42 -
0.13 0.15 0.17 0.19 0.21
Curve 9
Spectrophotometric measurements of Allobarbital at 240 mu.
at pH 9.45 in buffer solution (Beckman's Spectrophotometer)
0.23 mg. Allobarbital
//
/1
0.39
0.35
//
0.31
//
0.16 0.18 0.20 0.22 0.24 0.26 mg. Cyclobarbital
Curve 10
Spectrophotometric measurements of Cyclobarbital at 238 mu.
in buffer solution of pH 9.5 (PMQ II C. Zeiss Spectrophotometer)
- 43 -
l Jl
0 29
/
0 23
0 21
0 28
0 26
/
0 22) '
0 18 0 18 0 20 0 22 0 24 0 26
Curve 11
Spectrophotometric measurements of Herobarbital at 244 mu In buffer
solution erf pH 9 5 (PMQ D Car] Zeiss Spectrophotometer)
mg Heiobarhital0 18 0 18 0 20 ] 26 mg Methylphenobarbital
SpectrophoWmetric measurements of Methylphenobarbital at 244 mu
in buffer solution of pH 9 45 (Beckman s Spectrophotometer)
/
/'
//
/
/
0 16 0 IB 0 20 0 22 0 24 0 26 mg Pentobarbital
sodium
Curve 13
Spectrometrlc measurements of Pentobarbital sodium at 240 mu in buffer
solution of pH 9 45 (PMQ H Carl Zeiss Spectrophotometer)
hydroxide)
sodium
N(0.1
12.05
ph
at
Measurements
(buf
fer)
9.5
pH
at
Measurements
Cyclobarbitai
for
curves
Spectrophotometrte
15
Curve
hydroxide)
sodium
(N
14
about
taken
Measurements
(buf
fer)
9.45
pH
at
taken
Measurements
(wat
er)
5.55
pH
at
taken
Measurements
Allobarbital
for
curves
Spectrophotometrte
X
0
\\\
\\t\
\H
\\\
\/
\
j
iT
.
\'
\\
\
\Ky
\
i
\/
v
\\
\(i
i
\\
\\
r\
/1
\/
\1\
/I
1-N
1X
111
\
1
j1
1\i
i
ft
14000
(buf
fer)
9.5
pH
at
taken
Measurements
hydroxide)
sodium
N(0.1
12.05
pH
at
taken
Measurements
Methylphenobarbital
for
curves
Spectrophotometry
17
Curve
hydroide)
sodium
N(0.1
12.05
pH
at
taken
Measurements
(buf
fer)
9.5
pH
at
taken
Measurements
(water)
5.55
pH
at
taken
Measurements
Hexobarbital
for
curves
metrie
ctrophoto
Spe
16
Curve
m;i
290
nui
290
x^
\
\
\
\\\
2000
\\
VVV
\v
1
//
if
o
{)
M.E.C.
V\\\
.
11I
i
/i
/i
/i
VI1
VV\1)
8000
ii
iI
A
*
1
\
\\S
«
i
<'\
\\\\\\
i//\
u
\V
///
/h11
\
^yo
\,/
-'"
\
1t
- 46
A1
i
*
i
i
f< r"N\
1 t
1 /1 I
V1
\\
\ \
\ \
I\
\\
\
\
X
\
\
\\\
\
\ I
210 230 250 270
Curve 18
Spectrophotometrie curves of Pentobarbital sodium
290
m j]
Measurements taken at pH 12.05 (0.1 N sodium hydroxide)
- 47 -
Summary and conclusions of the spectrophotometric analysis of
barbituric acid derivatives
1. The absorption spectra of AUobarbital, Cyclobarbital, Hexobarbital, Methyl-
phenobarbital and Pentobarbital sodium in different aqueous solvents were studied.
2. The influence of the solvent on the extinction coefficients of barbiturates was
examined. In general, acids themselves showed in water only general absorption with
no peak e. g. AUobarbital Curve 14. In case of Hexobarbital a very slight peak was
obtained (Curve 16). The introduction of sodium ion to form a soluble enolic salt,
produced a marked peak and this became more pronounced e. g. in case of Methyl -
phenobarbital and displaced towards longer wavelengths with increase of sodium
hydroxide e. g. in case of AUobarbital, Cyclobarbital, Methylphenobarbital and Pento¬
barbital sodium. (Curves 14, 15, 17, 18). In case of Hexobarbital no displacement
of the peak of the wavelength occured (Curve 16). The peaks in the other cases i.e. of
AUobarbital, Cyclobarbital, Hexobarbital and Pentobarbital sodium were less pronounced
with the increase of sodium hydroxide than the peaks where the buffer were used.
3. From the previously given curves the quantitative estimation of AUobarbital,
Cyclobarbital, Hexobarbital, Methylphenobarbital and Pentobarbital sodium could be
carried out. Best results being obtained in buffer solution (pH about 9.5).
4. It is difficult to distinguish by absorption methods between individual barbituric
acid derivatives e. g. AUobarbital, Cyclobarbital, Hexobarbital, Methylphenobarbital
and Pentobarbital sodium. The peak absorption of each in buffer being from 238 - 244 mp.
and the maximum molecular extinction coefficients being generally within the limits
8500 - 10000. The peak absorption of each in 0.1 N sodium hydroxide being from
242 - 252 mu. and the maximum molecular extinction coefficients being with in the
limits 7000 - 9490.
- 48 -
C) Chemical properties of the. barbituric acid derivatives
and their use in the quantitative estimation
1. Acidity
Barbituric acid derivatives can occur in two tautomeric forms; a keto form (IV)
and an enol form (V).
? tf
XC<6 ^C-O— XC<6 Vc-OH
/ \4 3>C-° / C<4 3S*C°H
R, C Nlf Rf X N^
1 II xII
O O
IV V
The form (V) possesses the acid properties through the hydrogen atom at C 2. This
hydrogen can be replaced by metals to obtain the corresponding salts of barbituric
acid derivatives e.g. sodium, potassium and calcium salts. These salts react alka¬
line in aqueous solution.
The acidimetric quantitative estimation method either in aqueous or non aqueous
solvents depend on the acidic properties of the enolic form (V). In aqueous alcoholic
solvent, barbituric acid derivatives could be titrated against standard alkali using
thymolphthalein as indicator. In non aqueous solvents e. g. dimethylformamide,
pyridine and chloroform, barbituric acid derivatives are estimated using either sodium
or lithium methoxide as titrants and thymol blue as indicator.
2. Reaction with silver
Barbituric acid derivatives dissolved in alkaline solutions react with silver
nitrate giving salts which are generally insoluble in water. This property was used
to estimate barbituric acid derivatives argentometrically. Schulek and Rozsa '
used borax solution as a solvent for barbituric acid derivatives and titrated with
silver nitrate till a reddish colour predominated. One molecule of barbituric acid
derivative corresponds to 2 equivalents of silver. The structure of the disilver
22) E. Schulek and P. Rozsa, Z.Anal.Chem., 40, 415(1938).
- 49 -
compound formed in this method was given the following structure (VI) by23)
Danielson ':
H C_N
C C-O-Ag
V ft—N^o
VI
Budde ' dissolved the barbituric acid derivatives in sodium carbonate
solution pH and titrated with silver nitrate to the appearance of a permanent turbidity.
One molecule of barbituric acid derivative is equivalent to one molecule of silver.
The sodium silver salt formed according to this method was discussed by23)
Danielson ' and the following possible structures were given:
O NaV Ag
R< C ——*kK„/ \„
~ TCL C-O-Na
R/ C —2 II
-*'O
f
Rj C N.
and ^^ _; C-O-Ag
2 IIO
vii vin
Danielson ' used potassium metaborate as a solvent for the barbituric acid
derivatives and titrated with silver nitrate; potassium chromate being used as indi¬
cator and as a comparison solution. The compounds formed contain 1 atom of silver,
1 of potassium or sodium and 2 molecules of barbituric acid. This was confirmed by
analysis, and was given the following structure (IX):
OII ?
C9H. .C NH .NH C. C9H.
25\ / \ / \„/2 5
c' ^ C-O-Ag : K-O-C^ CQC2H5/ C~N N \' C2H5
O
IX
23) B. Danielson, Svensk farm. T., 55, 125(1951).24) H. Budde, Dtsch. Apoth. Ztg., 49, 2"9~5 (1934).
- 50 -
25)According to Stainier et al.
, ordinary barbituric acid derivatives could
fix 2 atoms of silver and the disilver compound obtained is slightly dissociated and
slightly soluble in dilute nitric acid.
e. g. Phenobarbital sodium + 2 Ag NOg Phenobarbital Ag2 + HNO, + NaNO,
The filtrate of the previous solution is acidic due to the liberation of nitric acid. In
case of the barbituric acid substituted at the nitrogen atom the monosilver compound
is the only possibility, and the filtrate is neutral.
e. g. Methylphenobarbital sodium + Ag NO, Methylphenobarbital Ag + Na NO,
The thiobarbiturates can fix 4 atoms of silver with the formation of silver sulphide
and disilver compound of barbituric acid derivative.
3. Reaction with mercury
2fi ^P e d 1 e y
' used mercuric perchlorate solution to precipitate the barbituric
acid derivatives dissolved in boiling water. The excess of mercuric perchlorate
being titrated against ammonium thiocyanate. Analysis of the precipitate obtained
with a barbiturate containing an unsubstituted imido-nitrogen indicated a monomole-
cular compound with mercury e. g. Barbital (CgH, ,0,N2) gave a precipitate of the
following formula (CgH^OgNg) Hg containing 7.26% of Nitrogen and 51.43% of
mercury conforming with the calculated percentages of nitrogen and mercury which
were 7.33% of nitrogen and 52. 3% of mercury.
In case of barbiturates contaning a substituted imido nitrogen one molecule of
mercury combines with 2 molecules of the barbiturate giving the following formula (X).
Ri
°
c —
R, R-I3N—
OII
-c Ri\ / \ / \ /
*
>c xc -O-Hg-O -C <
II0
-*' ^N-II0
^R2
25) C. Stainier, Ch. Lapifere etS. de Ti&ge Robinet, Ann. pharm. franc., 14,26) E. Pedley, J. Pharm. Pharmacol., 2, 39 (1950). 384 (1956)7"
- 51 -
4. Halogenation of the unsaturated radicals
The unsaturated radicals e. g. allyl or cyclohexenyl could be estimated by
bromination whereas the double bonds are saturated with two atoms of bromine and
the excess bromine is titrated e. g. the allyl radical CHo = CH - CH, - fixes two
atoms of bromine to give CH„Br - CH Br - CH, -
.
This method of bromination of the unsaturated radicals is used by the Danish
Pharmacopoeia 1948 to estimate Allobarbital, Hexobarbital and Cyclobarbital.25)
According to Stainier,the New and Nonofficial Remedies (1955)
described the same procedure of the Danish Pharmacopoeia 1948 for the estimation
of Hexobarbital. But in case of Apobarbital sodium,the NNR (1955) it described the
direct bromometric titration using a solution of bromine bromate in acid medium
(in presence of hydrochloric acid) and methyl red indicator till the disappearance of
its colour.
The reduction of the colour of potassium permanganate by the unsaturated
radicals in alkaline solution, carrying a blank, is used as a qualitative test to detect
their presence.
5. Colour Reactions
a) Production of complex salts with cobalt
The reaction depends on the formation of a bluish violet complex of barbituric
acid derivatives and cobalt salt in presence of a base in non aqueous solvent.
10} 21)
According to Perlia and after Zwikker ' three cobalt complexes of
diethyl barbituric acid are possible to be produced namely red, brown and blue
complexes. The red cobalt complex exihibit a bright red violet colour, and is very
sensitive to water. It is named diamine cobalt dibarbital and has the following
formula (XI):
O OII II
C9H- X N^^ ^-S-N C r,H.
^c c=o7>cp~rTo=c' c'
C2H5 X<jf—NH/ NH3 'NH3 NH—C^ C2H5o o
XI
27) J.J.L. Zwikker, Pharm.Wbl., 68_, 975 (1931).
- 52 -
The brown cobalt complex e.e. hexamincobaltdibarbital hydroxide is a brown
glistening crystalline powder having the odour of ammonia and possesses the following
composition (XII):
m
Co (NH3)6
xn
,OH
•(Barbital)2
The previous compound (XII) liberates 1 molecule of ammonia and the odour
of ammonia is lost giving pentaminhydroxycobalt dibarbital of the following
composition (XIII):
in
Co (NH3)5
xni
OH
^(Barbital),
The blue cobalt complex has the following composition:
25% Dichlorodibarbital-cobalt-potassium (XIV)
f Cl2 (Barbital)2 Co 1 K2
XIV
75% Chlorohydroxydibarbital-cobalt-potassium (XV)
Tci (OH) (Barbital)2 Co"| K2
XV
When carrying the cobalt colour reaction, all the test materials should be
completely water free. The reacting materials i. e. the barbituric acid derivatives,
cobalt salt and the base should be present in optimal proportions to each other. The
stability and the intensity of the colour depends mainly on the type of the base used.
When an excess of an inorganic base is used, the basic cobalt salt is precipitated.28)
Parr i ' used for the detection of barbital, a cobalt salt and ammonia in
27)alcoholic solution where he obtained a violet colour. Zwikker carried out the
cobalt reaction using cobalt chloride, anhydrous methanol to dissolve the barbituric
28) W. Parr i, Boll. chim. farm., 36, 401(1924).
- 53 -
acid derivative and a saturated solution of barium oxide in methanol as a base. Other
bases e.g. potassium hydroxide, sodium hydroxide, piperidine, isopropylamine and
isobutylamine were used by different authors. According to Perlia ' and after
29}Robles ', the stability of the colour reaction of cobalt could be increased using
the following procedure: Dissolve the barbituric acid derivative in chloroform;
3 drops of 5% cobalt nitrate in water, 1 drop of 0.1 N potassium cyanide and 3 drops
of pyridine are added. After standing a cherry red colour develps which is stable
for 30 days.
To detect the presence of barbituric acid derivatives, several authors used
strips of filter paper dipped in 1% alcohilic cobalt nitrate solution and dried. By
means of a capillary pipette few drops of the dissolved substance to be tested, in
ether or alcohol, and few drops of 5 to 10% aquoues ammonia were added to the
strips of filter paper. In presence of barbituric acid derivatives the violet colour
was produced.
The quantitative determination of barbiturates by the cobalt amine reaction was30)
applied by different authors e. g. Baggesgaar d-Ras mussen and Jerslev'
and Nuppenau '. The method recommended by Nuppenau will be discussed
and applied later.
b) Production of complex salts with copper
27)
Z wikker' used a mixture of pyridine and 10% copper sulphate and water
to purify the crude barbiturate which precipitate in toxicological examinations. He
suggested that the resulting red violet copper barbiturate pyridine complex was of
the composition Cu (pyridine)0(Barbiturate)„.32)
Flo tow found that the copper barbiturate pyridine complexes were soluble
and stable in chloroform and that 0.1 mg. of Barbital could be detected when extracting
with chloroform.33}
G o m a h r and Kresbach 'used the Zwikker reaction in the qualitiative
detection of barbiturates and other pharmaceuticals of similar structure. The reaction
was carried out by dissolving few milligrams of the tested substances in a mixture
of pyridine and chloroform: and copper sulphate solution was added. After shaking
all barbiturates gave violet colour to the chloroform layer. In case of thiobarbiturates
a green colour was obtained.
29) G. Robles, Bol.Soc.Quin, Peru, 15, 70(1949).30) H. Bagesgaard-Rasmussen and B.Jerslev, Dansk T. Farm, 25, 29 (1951).31) H. Nuppenau, Dansk T.Farm., 28, 194(1954).32)E.Flotow, Pharm. Zhalle, 88, 158" (1949).33)H.Gomahr and H. Kresbach, Scientia pharm., 1_9, 154(1951).
- 54 -
It should be noted that barbiturates and thiobarbiturates could not be differentiated
with the cobalt colour reaction; whereas by the copper sulphate pyridine chloroform
reaction,thiobarbiturates could be differentiated from barbiturates.
c) Production of colour with Selenious acid
34)Turf itt described the following colour reaction of barbiturate with selenious
acid as a qualitative reaction. By heating a mixture of barbiturate and selenious acid
in presence of concentrated sulphuric acid a green colour developed. After cooling
and transference of the mixture to a porcelain dish, a red turbidity developed on the
addition of few drops of alcohol.
6. Condensation of barbiturates with p-nitro benzyl chloride
After Perlia the condensation of barbiturates with p-nitro benzyl chloride
could occur not only through the 2 imido hydrogen atoms in position 1 and 3 but also
through the 2 methylene hydrogen atoms in position 5. A tetra derivative could be
produced by the condensation of barbituric acid with p-nitro benzyl chloride having
the following formula:
O OR
H C-NH^ R .C-Ax>C .C = ° ^-^ JZ C = 0 + 2HC1
v/ XC-NH>/ R = -CH2-C6H4-N02 R/ N—n/O OR
XIV XV
In case of the mono alkyl barbituric acid, a tri-substitution product is possible;
in case of the dialkyl barbituric acid a di-substitution product is obtained; while in
case of the dialkyl N-methylated barbituric acid the mono-substitution product is the
only possibility.
The p-nitro benzyl derivatives of barbiturates are white crystalline powders
soluble in chloroform and slightly soluble in water and alcohol.
Perlia gave a valuable table of the melting points of the p-nitro benzyl deri¬
vatives of barbiturates determined by different authors. The main use of the previous
reaction is in the identification of barbiturates by determining the melting points of
34) G.E. Turfitt, Quart. J. Pharm. Pharmacol., 21_, 1 (1948).
- 55 -
the p-nitro benzyl barbiturate derivative. Together with the determination of the
melting point of the barbiturate itself a more reliable identification possibility is
achieved.
7. Reaction with different reagents
25)
Stainier et al. 'examined the reactions of barbituric acid derivatives with
different reagents i. e. Millon's, Denige's, Mercuric acetate, Parri's, Zwikker's,
and Xanthydrol reagents. They found that all ordinary barbituric acid derivatives
gave positive reactions with these reagents. Barbituric acid derivatives substituted
at the nitrogen atom behaved differently e.g. Methylphenobarbital which gave no
precipitate with Millon's and Xanthydrol reagents; with mercuric acetate and Denige's
reagents it gave weak positive reactions. Hexobarbital gave negative reactions with
all the previous reagents except with Parri's reagent. These reactions are not only
specific for barbituric acid derivatives but also other substances e. g. sulphonamides
and purine derivatives gave positive reactions with the previous reagents.
- 56 -
D) Methods of the quantitative estimation of barbituric
acid derivatives
1. Acidimetric method
a) Titration in aqueous media
Survey of the literature
The acidimetric method depends on the principle that the commonly used
disubstituted barbituric acid derivatives, due the lactam-lactim tautomerism, react
in the lactim form as weak acids (see page 48 Form. IV & V) and so they can be
titrated with alkalies.
351Palme found that the use of phenolphthalein as indicator for the titration
of diethylbarbituric acid was unsatisfactory. By using thymolphthalein or a mixture
of thymolphthalein and alizarin yellow as indicator, 0.1 N alcoholic potassium hydroxide351
as a titrant, together with a comparison solution, Palme ' obtained a degree of
accuracy of 0.5%. Babitsch used thymolphthalein as indicator in the titration
of phenyl ethyl and diethyl barbituric acid derivatives, in alcoholic solution with 0.1 N37)
sodium hydroxide. Gervay ' described the estimation of phenyl ethyl and diethyl
barbituric acid derivatives in 33% methanol using thymolphthalein as indicator for
38)phenyl ethylbarbituric acid and alizarin yellow for diethyl barbituric acid. Morin
stated that although the barbituric acid derivatives form mono alkali salts, they
cannot be determined by titration withaqeous alkalies on account of the extent to
which the salts hydrolyse. He used acetone as a solvent for the barbituric acid
derivatives to repress the hydrolysis, 0.1 N potassium hydroxide in methanol as a
titrant and thymol blue as indicator.
To abtain accurate results in the titrations of weak acids, the knowledge of
their dissociation constants is of a great importance in the choice of the indicators
14)to be used. Poethke and Horn studied the dissociation constants of several
barbituric acid derivatives. They found that accurate results could be obtained in the
acidimetric method using carbonate free 0.1 N alkali, carbon dioxide free alcohol
(50%), thymolphthalein as indicator and a copper dichromate solution as a comparison
35) H. Palme, Pharmaz. Ztg., 75, 1347 (1930).36) S. Babitsch, Chem. Zbl. 313B (1936) n.
37) V. Gervay, Pharmaz. Zentralhalle, 83, 494(1942).38) Ch. Morin, Chem. Zbl. II, 3948 (193577 Quart.J. Pharm. & Pharmacol. 8, 691
(1935).
- 57 -
solution. They described the following procedure:
About 1 mmol. of the barbituric acid derivative (accurately weighed) is dissolvedin 24 ml. ethanol (carbon dioxide free, best done by distillation over sodium). To
this solution 1 ml. 0.1 % alcoholic thymolphthalein solution and 15 ml. carbon dioxidefree water are added. Titrate with carbonate free 0.1 N alkali till the and point is
reached as compared with a comparison solution. Carry out a blank using the same
reagents to find the correction factor. The correction factor is estimated byPoethke and Horn14) in their assays to be 0.03 ml. of alkali to be substractedfrom the number of ml. of the alkali used. The comparison solution is an ammoniasolution containing 2 mg. copper (7. 85 mg. CUSO4. 5 H2O) and 0.1 mg. chromium
(0.285 mg. K2Cr207).The results of the assays carried out by Poethke and Horn
'adopting the
previous method showed deviation of ± 0.6 %.
In regard to the acidimetric titration, the found differences between the deter¬
mined dissociation constants of the barbituric acid derivatives are of no important
significance. According to Poethke and Horn 'they are between 0.83 x 10" -
_0 _Q6.17 x 10 ; and according to our results (page 32 ) they are between 0.51 x 10" -
-83. 80 x 10
. The previous authors proved that acids of this strength could be titrated
with 0.1 N alkali using a comparison solution with satisfactory accuracy. In regard
to the choice of the indicators, they used the following equation to calculate the pH
values at the equivalence point. This equation was also referred to by Hamilton
^ q-39)
and Simpson'
pH = 1/2 pW + 1/2 pA + 1/2 log. C
where pH = pH of the solution at the equivalence point.
pW = Dissociation exponent of water.
pA = Dissociation exponent of acid.
C = Concentration at equivalence point.
For phenyl ethyl barbituric acid the dissociation exponent is 7. 3 and the con¬
centration = 0.02 at the equivalence point and by substituting these values in the
previous equation gives:
pH = 6.99 + 3.91 - 0.85 = 10.05
These found two pH values at the equivalence point of 9. 79 and 10.05 occur in
the change interval of the indicator thymolphthalein (9.3 - 10.5). So thymolphthalein
is suitable to be used. Due to the slight solubility of barbituric acid derivatives in
water, the titration is carried out in alcohol water mixture.
39) L-F. Hamilton & S.G.Simpson "Calculations of Analytical Chemistry" Inter¬
national Chemical Series 5th. Edition Mc Graw-Hill Book Company p. 176, (1954).
- 58 -
40) 14)
Horsch applied the method proposed by Poethke and Horn 'to
diethyl barbituric acid and he obtained more or less concordant results. Since thymol-40)
phthalein is a sensitive indicator to carbon dioxide of the atmosphere, Horsch
recommended to titrate rapidly and with the least shaking near the end point,
preferably using a magnetic stirrer. He also drew the attention to the fact that
recently well boiled distilled water reacted acidic, and consequently he recommended
to carry out a blank to estimate the used alkali to neutralize the mixture of solvents
used (alcohol and water).25)
Stainier et al. ' stated that the tests and standards of the NNR recommended
the titration using thymolphthalein for allylisopropyl; butyl n ethyl; diallyl; and ethyl
cyclohexenyl, barbituric acid derivatives. The assays being carried out by
dissolving the barbituric acid derivative in 25 ml. alcohol and adding 25 ml. of
water. The French Codex 1949 described for the assay of Phenobarbital an analogous25)
method using acetone as a solvent. Stainier et al. stated that the acidimetric method
without special precautions gave a precision of 1 -1.5%.
The National Formulary (1955) described the estimation of Allobarbital
using alcohol as a solvent, thymolphthalein as indicator and 0.1 N alkali as a titrant.
titrant.
Discussion of the literature
Most of the authors in the given survey of the literature recommended the use
of thymolphthalein as indicator, 0.1 N alkali as a titrant and a mixture of alcohol
and water as a solvent for the barbituric acid derivatives. Several precautions should
be taken e. g. the use of carbonate free alkali, carbon dioxide free alcohol and water,
the use of a comparison solution and a blank, and the least shaking to avoid the access
of carbon dioxide of the atmosphere to the titrated solution.
Application of the method of estimation and the results obtained
The National Formulary (1955) used the following method for the assay of
Allobarbital acid:
Dissolve about 500 mg. of Allobarbital acid, previously dried at 105° C for
4 hours and accurately weighed, in 40 ml. of neutralized alcohol. Dilute with 25 ml.
of water previously bioled to remove carbon dioxide, add thymolphthalein test solution
and titrate with 0.1 N sodium hydroxide. Each ml. of 0.1 N sodium hydroxide is
equivalent to 20.82 mg. of cioHi2N2°3 (diallylbarbituric acid).
40) W. Horsch, Pharmazie, 12_, 122, 212(1957).
- 59 -
The United States Pharmacopoeia (1955) describes neutralized alcohol to be
made by neutralizing alcohol with 0.02 N or 0.1 N sodium hydroxide using phenol -
phthalein test solution until a faint pink colour is obtained. In the previous assay it
is clear that thymolphthalein test solution should be used for neutralization of alcohol.
To determine the normality factor of sodium hydroxide we used the following
method:
Dry about 0. 5 g. of potassium hydrogen phthalate at 105° C for 3 hours (if the
salt is in the form of large crystals, they should be crushed before drying). Dissolve
it in 75 ml. of carbon dioxide-free water, add 2 drops of thymolphthalein test solution
and titrate with 0.1 N sodium hydroxide to the production of a permanent blue colour.
Table 14
Results of assays of Allobarbital as directed by N.F. 1955
Assay No. 1 2 3 4
I 0.985 490.3 23.54 99.96 %
II 0.986 491.8 23.61 99. 95 %
m&iv 0.985 500 23.98 99. 86 %
1 = Normality factor of 0.1 N sodium hydroxide
2 = Weight of Allobarbital in mg.
3 = ml. of 0.1 N sodium hydroxide used for Allobarbital
4 = Percentage of found Allobarbital
The same percentages of Allobarbital (99.86%) were found when using different
percentages of alcohol as solubilizing agent (33%, 40%, 56% V/V alcohol).
- 60 -
It is obvious from Curve 19 that the pH values at the visual end points were
shifted to higher pH values with the increase of alcohol percentages.
The previous results agree with the statement of Kolthoff and Laitinen41'
in regard to the influence of the medium on the indicator. They stated that when
organic liquids e. g. ethyl alcohol, methyl alcohol or acetone with a lower dielectric
constant than water were added to an aqueous solution the equilibrium conditions
were changed. The addition of alcohol to an aqueous solution decreases the ionization
constants of weak acids and bases. Consequently indicator acids will become more
sensitive to hydrogen ions in the presence of organic solvents and their colour change
interval will be shifted to higher pH values in mixtures of water and alcohol.
The same procedure adopted in the assays of Allobarbital was applied, using
dried Cyclobarbital which is readily soluble in alcohol.
Table 15
Results of the assays of Cyclobarbital
AssayNo.
Weight of
Cyclobarbitalin mg.
ml. of 0.1 N
NaOH
Factor of
0.1 N NaOH
Percent of
Cyclobarbitalfound
1
2
3
500
500
400
21.5
21.5
17.25
0.989
0.989
0.989
100.5 %
100.5 %
100.8 %
The same procedure adopted in the assays of Allobarbital was applied, using
dried Hexobarbital instead of Allobarbital. Hexobarbital is not readily soluble in the
mixture of alcohol and water (40 % V /V ) used. A very small amount remained
insoluble in the assay No. 3, increasing in amount in the assays 2 and 1.
Table 16
Results of the assays of Hexobarbital
AssayNo.
Weight of
Hexobarbital
in mg.
ml. of 0.1 N
Na OH used
Factor of
of 1 N Na OH
Percent of
Hexobarbital
found
1
2
3
500
200
100
21.0
8.42
4.2
0.989
0.989
0.989
98.06 %
98.37 %
98.06 %
41) I.M. Kolthoff and H.A. Laitinen, "pH and Electro Titrations".
2nd Edition (1942) John Wiley and Sons page 30.
61
11
10
t1
'•/
*
s*
,'fl'' '
'' 1''' 1
111
II11f 1
t
/
t
11
1
,.-'"'
-"
s
20
y
nl.H20/
s
/
1
/
/
'/t
//1
1
11
1I
6 10 14 18 22 24 26 30 ml
0.1 N NaOH
Curve 19
The effect of using different concentrations of alcohol, in the acidimetric method, on
the visual end points (°)
about 56% alcohol in the titrated solution, the visual end point being at
pH 10.40
about 38£ alcohol in the titrated mixture, (20 ml, of boiled and cooled
distilled water were added ofter the addition of 20 ml. 0.1 N sod. hydroxide)
about 33% alcohol in the titrated solution, the visual end point being at
pH 9. 95
about 40% alcohol in the titrated solution, the visual end point being at
pH 10.15
- 62 -
Results of the assays of Methylphenobarbital
The same procedure adopted in the assays of Allobarbital was applied, using
dried Methylphenobarbital. 500 mg. of dried Methylphenobarbital was used. A large
amount remained insoluble in the mixture of alcohol and water used. The end point
was very unstable and after vigorous shaking disappeared.
Percent of found Methylphenobarbital: 92.49%, 103,5%.
Methylphenobarbital could not be assayed by this method, due to its insolubility
in the alcohol used.
Criticism and summary of the results
1. Our experience shows that the previous method can be used for the assay of
barbituric acid derivatives, which are fairly soluble in alcohol e. g. Allobarbital,
Cyclobarbital. In case of Hexobarbital smaller quantities than 500 mg. should be used.
The previous method could not be applied to Methylphenobarbital which is sparingly
soluble in alcohol.
2. A further trouble is encountered with the slightly soluble barbituric acid
derivatives in the alcohol water mixture used, in relation to the requirement of vigorous
shaking which is contraindicated in this procedure to guard against the access of carbon
dioxide of the atmosphere. To minimize the access of carbon dioxide, a card with a
hole to cover the Erlenmeyer flask, and a magnetic stirrer should be used.
3. The addition of alcohol to an aqueous solution decreases the ionization constant
of weak acids and bases. Consequently indicator acids will become more sensitive to
hydrogen ions in the presence of organic solvents and their colour change interval will
be shifted to higher pH vakues in mixtures of water and alcohol. This was practically
confirmed by using different percentages of alcohol. So the use of stronger alcoholic
solutions to effect complete solubility of the barbituric acid derivatives is undesirable
as it may shift the pH beyond the interval change of Thymolphthalein.
4. Alcohol has also an action on the colour of thymolphthalein, the blue colour of
the indicator in alkaline medium is greenish in presence of alcohol. So the use of a
comparison solution and a blank as proposed by Poethke and Horn'gives more
accurate results.
- 63 -
b) Titrations in non-aqueous solvents
A survey and discussion of the literature
42)Pif er et al. concluded that the field of non-aqueous titrations offers a
variety of possibilities for the assay of pharmaceutical compounds, many of which
cannot be determined readily by other methods. Non-aqueous titrations present many
opportunities for further theoretical and practical investigations. It appears that the
time is not too distant when non-aqueous titrations will be used frequently as titrations
in water for the assay of pharmaceuticals.
The previous authors attributed the importance of titrations in non-aqueous
solvents to the following qualifications:
Specifity. Depending upon which part of a compound is the physiologically active
moiety, it is often possible to titrate that part by proper selection of solvent and
titrant.
Solubility. The variety of organic solvents available for the titration in water
free medium permits the choice of the most desirable for solubilizing the sample.
Simplicity. Most of the titrations involved are of an acid-base nature, which
can be performed either visually using a variety of indicators, or potentiometrically
with various electrode combinations. These methods are advantageous for routine
control, since a minimum of equipment is required, the determinations can be carried
out rapidly, and the solvents and the indicators used are available at moderate cost.
Sensitivity. Organic solvents permit far less ionization than an aqueous medium.
By proper selection of solvents therefore, many determinations can be conducted with
very small quantities of sample without loss of sensitivity at the end points.
Selectivity. Recently developed techniques permit the differentiation between
various acidic or basic functional groups within the same molecule or in mixtures -
42) C.W. Pif er,E.G. Wollisch and M. Schmall, J. Amer. Pharm. Ass. ,Sci.
Edit. 42, 509(1953).
- 64 -
a procedure which usually cannot be carried out in aqueous solution.
Accuracy and precision. The accuracy and precision obtainable is comparable
in most cases to those of conventional titrations in water.
Practical advantages of the non-aqueous method over the aqueous gravimetric42)
official method was finely shown after P if e r et al. table which states the following
Table 17
Comparison between official and non-aqueous procedures
(after Pif er et al.)
Operation AqueousUSP-method
Non-aqueous(Schmall extractor)
1 Grind tablet mass X X
2 Weigh sample X X
3 Solution in base X X
4 Liberation by acid X X
5 Extraction X
(minimum of
six hand
extractions)
X
(automatic)
6 Prepare tared
vessel X
7 Evaporate solvent X X
8 Ether addition
and evaporation X
9 Dry residue X
(2 hours)
10 Weigh residue X
11 Titrate residue X
x indicates that this particular operation must be performed in the analysis
and mostly wereThe following abstracts were taken after P if e r et al.'
conforming with our practical experience.
Basic titrants: Pif er et al. ' stated that the alcoholates often produce
gelatinous precipitates with organic solvents which may obscure the visual end point
or coat the electrodes, causing a sluggish inflection. These authors therefore
recommended the use of lithium methoxide in benzene-methanol since they rarely
have encountered this undesirable condition.
- 65 -
Standardization of basic titrants: All basic titrants including potassium, lithium
methoxide, sodium aminoethoxide and sodium triphenylmethane are best standardized
against benzoic acid.
Solvents: Solvents can be conveniently classified into 4 groups, namely: acidic,
basic, relatively neutral and mixed solvents. The proper selection of the solvent
system may contribute to the success of many a titration. The dielectric constant of
the solvent has a direct effect upon the dissociation of a compound. According to
43)Hammet any specified acid forms the most acidic solution, i.e. the solution
which is the most effective proton donor, in that solvent in which it is least ionized.
The same concept applies to basic compounds. In order to obtain a solution in which
ionization is depressed, it is often necessary to dissolve the sample in a minimum
quantity of acid or basic solvent to effect solubilization, followed by an excess of a
miscible organic solvent of low dielectric constant, e.g. benzene to acidic or basic
solvents, or p-dioxane to water. This technique increases the sharpness of the visual
and potentiometric end points and is particularly useful when applied to titrants of
0.02 N or lower normality. It has been proved that the ionization of a compound is
not a requirement for a titration since those carried out in aprotic solvents are far44)
more sensitive. Markunas and Riddick ' have set forth some general
requirements for solvents such as
a) they must be commercially available in reasonably pure form at a moderate
cost;
b) they must dissolve the substance being titrated;
c) the products of the titration should be free from gelatinous type precipitates;
and
d) the solvents should not enter into the reactions.
Interferences in titrations of acids
1) Weakly acidic substances;
2) water which should be restricted to a minimum;
3) carbon dioxide which can be held to a minimum by titrating in a closed system
or under nitrogen;
4) these precautions should be specially observed when titrating with weaker
than 0.1 N solutions;
5) esters (ethyl acetate);
6) some halogen containing compounds;
7) interferences from titrants, which form gelatinous precipitates.
43) L.P. Hammet, Phys.Org.Chem., McGraw Hill Book Co., New York 1940, p. 261.
44) P.C. Markunas and J.A. Riddick, Anal.Chem. <23, 357 (1951).
- 66 -
Choice of indicators. Thymol blue is the most commonly used of the proposed
indicators, i.e. o-nitroaniline and azoviolet (p-nitro-benzol-azo resorcinol). Thymol
blue changes from yellow to green and further to blue on the addition of alkalies.
Effect of temperature on titrations. Usually the titrants contain methanol and
benzene which have high expansion coefficients and it is of great importance to determine
the factor of the titrant before assaying the acids.
Effect of addition of water. Due to the amphoteric characters of water it should
be mentioned that the presence of water in basic medium destroys the titration.
Fritz ' stated that the water content over 1% leads to more consumption of the
titrant and simultaneously obscure the end point. In case of the use of dimethylfor-
mamide as a solvent, the presence of water leads to hydrolysis of the solvent giving
formic acid. Hydrolysis also is liable to occur when heating dimethylformamide to
help solution in case of sparingly soluble substances; so it should never be heated.46)
It is recommended by H e i z 'to dry pyridine or the pyridine bases with potassium47)
hydroxide and followed by distillation. It is stated by B e r g e r' that less than
0.1 % of water content was obtained by the previous method. He further noted that
the dark colour which the solution acquires on standing could be diminished by
guarding against light. He added that the precipitates occurring in titrations using
pyridine as a solvent are bigger in amount and more gelatinous than those occurring
in case of dimethylformamide.
Effect of carbon dioxide of atmosphere on titrations. The basic solvents easily
absorb carbon dioxide of atmosphere. It is of great importance to neutralize the
medium before proceeding to estimate the acids and to guard against access of air.
This could be achieved by bubbling dried carbon dioxide-free nitrogen or using a
specially made closed system.
45) J.S. Fritz, Anal.Chem. 23, 589(1951).46) R. Heiz, Dansk Tidskr.Farm. 26_, 69(1952).47) J. Berger, Dansk Tidskr. Farm. 27, 53(1953).
- 67 -
Application and results of the different methods of non-aqueous titrations
48)
et) Vespe and Fritz method '
These authors worked with 12 pure barbituric acid derivatives; they concluded
that barbiturates can be titrated as acids in non-aqueous solutions using a visual end
point. Secondly titration of phenobarbital extracted from tablets gives results in
agreement with the official gravimetric method and obviates the use of tared flasks.
Thirdly barbiturates and sulpha drugs can be accurately determined by direct titration
of powdered tablet samples. They also discussed the limitations and possible further
applications of their method.
Their exact procedure was applied on our representatives of barbituric acid
derivatives, using both the visual end point and the potentiometric method using an
antimony electrode as an indicator electrode and a glass electrode as a reference
electrode. The exact procedure as directed by Vespe and Fritz is as follows:
The size of the sample should be such that about 0. 3 - 0. 8 milliequivalent of acidic
substance is present. The sample is dissolved in 20 - 30 ml. of dimethylformamide,3 drops of thymol blue added and titrated with 0.1 N sodium methoxide to a clear blue
colour. A microburette (10 ml.) which can be read accurately to 0.01 ml. is used
instead of the usual 50 ml. burette. For best results the solution should be protectedfrom carbon dioxide of the atmosphere during titration. This is conveniently done bycarrying the titration in a small flask or beaker covered by a cardboard with a hole
to admit the burette tip. The use of a magnetic stirrer adds to the convenience of the
titration. The titrant is standardized by titration against benzoic acid. Dimethylfor¬mamide contains always some acidic impurities for which a correction must be made.
In all the titrations reported a blank of 0.02 ml. of 0.1 N sodium methoxide was'
required for approximately 25 ml. dimethylformamide.
All barbiturates are readily soluble in dimethylformamide and give very sharp
visual end points.48)
The following reagents were used by Vespe and Fritz ':
a) Benzene, purified grade
b) Benzoic acid, primary standard grade
c) Dimethylformamide, technical (du Pont)
d) Methanol, absolute as purchased commercially
e) Sodium methoxide, 0.1 to 0. 2 N. Dissolve about 3 g of freshly cut sodium in
50 ml. of methanol, protecting the solution from carbon dioxide while the sodium
is reacting; cool if the reaction is violent. Add 100 ml. of methanol and 750 ml.
of benzene and store the reagent in borosilicate glassware. The solution should
be essentially colourless.
f) Thymol blue solution. Dissolve 0.3 g. of thymol blue in 100 ml. of methanol.
48) V. Vespe and J.S. Fritz, J.Amer.Pharm.Assoc. 4_1, 197 (1952).
- 68 -
In our assays we used the following reagents:
a) Benzene; dried with metallic sodium and distilled
b) Benzoic acid, analytical reagent (British Drug House, London)
c) Dimethylformamide, technical (Fluka Buchs, Switzerland)
d) Methanol treated with sodium and distilled
e) Sodium methoxide 0.1 N was prepared after Vespe and Fritz; and kept
in an automatically filled burette.
f) The same solution of thymol blue used by Vespe and Fritz was used in
our assays.
A magnetic stirrer was used to insure mixing. A card with 4 holes to cover the
beaker and allow the passage of the 2 electrodes, the tip of the burette and the bubbled
dried, carbon dioxide-free nitrogen was used.
The following results were obtained applying the visual procedure of Vespe48)
and Fritz 'to the assay our Barbiturate:
40 ml. of dimethylformamide was introduced in an Erlenmeyer flask. 3 dropsof thymol blue was added. Nitrogen was bubbled in the empty flask and during the
experiment. About 200 mg. of benzoic acid is accurately weighed and dissolved in the
neutralized dimethylformamide and titrated against sodium methoxide. The obtained
value from the above titration gives the normality factor of sodium methoxide, which
was determined everytime before each titration of barbiturates (see Curve 20).An accurately weighed amount of Allobarbital, about 200 mg., was put in another
40 ml. of dimethylformamide (previously neutralized with sodium methoxide to a paleblue colour of thymol blue). Then titrated with sodium methoxide. A magnetic stirrer
was used and a cardboard with a hole to cover the Erlenmeyer flask and allow the
passage of the nozzle of the burette.48)
We compared the visual titration method of Vespe and Fritz ' with the
following potentiometric method:
The accurately weighed amounts of benzoic acid and barbituric acid derivatives
were dissolved in 40 ml. dimethylformamide and titrated against sodium methoxide
using an antimony electrode as indicator electrode and a glass electrode as a reference
electrode and covering the 100 ml. beaker used with a 4 holed cardboard allowing the
passage of the 2 electrodes, the tip of burette and the bubbled dried, carbon dioxide-
free nitrogen gas. The apparatus used is Metrohm Type E 157*).
- 69 -
Table 18
Results of the visual and potentiometric titrations of our barbiturates
(V e s p e and Fritz -method)
Barbiturate visual titrations potentiometrictitration
Allobarbital
Cyclobarbital
Hexobarbital
Methylphenobarbital
1. 99.05%2. 99.38%3. 98.50%
4. 100,90 %5. 102.00 %6. 98.67%
1. 99.59%2. 99. 93 %3. 96.47%
1. 100.20%2. 99.84 %3. 99.41%
1. 99.90%2. 99.56%3. 98,54%
100.1 % Curve 21 & 21'
99.59% Curve 22 & 22'
99. 9 % Curve 23 & 23'
98.1 % Curve 24 & 24*
Criticism and summary of the results
48)
The method of Vespe and Fritz' recommended the use of the visual end
point titration for the estimation of barbiturates. We tried both the visual and
potentiometric titrations. The Metrohm pH-meter apparatus (E 157), an antimony
electrode as indicator electrode and a glass electrode as a reference electrode were
used.
Dimethylformamide is a good solvent for the titrations of both barbiturates and
benzoic acid used to standardize the sodium methoxide titrant. No precipitation was
encountered with in all assays of barbiturates. In case of benzoic acid precipitation
took place only in the very near of the end point and so did not distrub much the
potentiometric reading.
The potentiometric titration curves (Curve 21*) are characterized by a reasonable
potential jump in the vicinity of the end point. However, such a potential jump extends
over a certain volume of the titrant. A precise determination of the exact end point
could not be directly determined from these curves. This necessitates the analysis
of these curves. It is known that these usual S shape curves for potentiometric titration
have their inflection points at the corresponding end points of titration. Thus, if the
- 70 -
potential of the working electrode is differentiated with respect to the volume of the
titrant, and the values thus obtained are drawn against the mean value of the titrant,
one obtains the usual differential potentiometric titration curves. This procedure
was applied for the barbiturates used (see Curve 21'). The exact end points are
directly obtained from such curves, being the points of maxima.
In the case of Allobarbital a potential jump of 210 m. v./ml. titrant is observed.
Cyclobarbital, Hexobarbital and Methylphenobarbital are characterized by potential
jumps of 780, 1080 and 420 m.v./ml. respectively. These results suggests that
potentiometric titration for the above mentioned acids is sensitive and applicable
when present alone in solution. Inspection of the table of the ionization constants of
these acids (page 32 ) shows that it is inpractical to specifically determine these
acids when in mixture. This is because the ionization constants are very near to one
another.
- 71 -
/
/
//
/
<
1 t00t0^i
1 3 5 7 9 11 13 15 ml. 0.1 MCHjONa
Curve 20
The factor determination of titrant sod. methoxide using benzoic acid as directed
by Vespe & Fritz method as a demonstration curve for several potentiometric
titrations which should be always carried out before the assay of barbiturates.
72
iri.V.
260
240
220
200
180
160
140
120
100
80
60
40
20
t
/I
•"s
'
1 9 11 ml. 0.1 M CHjONaCurve 21
The potentiometric titration curve of AUobarbital by Vespe & Fritz as demon¬
stration from 3 titrations
AE
Ami.
220
180
l\
\\
\
\>\
140
100
60
20
7.5 8.5 mean ml.
Curve 21'
Differential potentiometric titration curve for
AUobarbital corresponding to Curve 21
- 73 -
560
520
480
440
400
360
//
320
/
5 7
Curve 22
11 ml. 0.1 M CH,ONa
The potentiometric titration curve of Cyclobarbital by Vespe & Fritz
as demonstration from 3 titrations
1
600
'
0
_—J"6.5. 7.5 mean ml.
Curve 22'
Differential potentiometric titration curve for
Cyclobarbital corresponding to curve 22
- 74 -
azu
740
660
A*—>
5801\
500
420
i4n
1 3 5 7 9 ml. 0.1 M CHjONa
Curve 23
The potentiometric titration curve fo Hexobarbital by Vespe &
Fritz as demonstration from 3 titrations
AE
Ami
600
a
6 6.5 7 7.5
mean ml.
Curve 23'
Differential potentiometric titration curve for
Hexobarbital corresponding to curve 23
- 75
m.T,
860
780
700
620
540
460
380
300
^/1
/
13 5 7
Curve 24
The potentlometric titration curve of Methylphenobarbital
by Vespe & Fritz as demonstration from 3 titrations
AE
6.5 7.5 mean ml.
Curve 24'
Differential potentlometric titration curve for
Methylphenobarbital corresponding to Curve 24
- 76 -
/3 ) H e i z 's Method
46)H e i z proved the possibility of titrating barbituric acid derivatives and related
substances in a mixture of pyridine bases as a solvent. The titration could be applied
successfully in all cases potentiometrically and in most cases visually using thymol
blue, phenolphthalein and in certain derivatives also thymolphthalein could be used.
He also reported that precipitation of gelatinous precipitates was seldom encountered
with and could be largely decreased by increasing the volume of solvents used. Access46)
of carbon dioxide of the atmosphere to the titrated solution should be avoided. H e i z'
gave a table containing 13 barbituric acid derivatives which he assayed. In his opinion
Cyclobarbital could not be titrated with any of the 3 indicators used.
Procedure: We adopted the same procedure and reagents used by H eiz with
the following very slight modifications:
Pyridine itself was used, dried and distilled over potassium hydroxide. Thymol
blue as indicator (0.2% in methanol) was the only indicator used. Metrohm E 157
apparatus and electrodes were used (Metrohm AG., Herisau, Switzerland). Dried,
carbon dioxide-free nitrogen gas was bubbled always during the experiments carried
for assaying the barbituric acid derivatives. A cardboard with 4 holes was used to
cover the beaker containing the tirated solutions, allowing the passage of the two
electrodes, the nozzle of the burette and the bubbled nitrogen. In all the following
assays 60 ml. of neutralized pyridine (with sodium methoxide using thymol blue as
indicator) were introduced in a beaker of 150 ml. capacity, benzoic acid or the barbi¬
turic acid dissolved and titrated to a blue colour of the indicator. Both visual and
potentiometric end points were recorded. We carried out the factor determination of
the titrant every time before the titration of the barbituric acid derivative as it varies
with atmospheric temperature.
- 77 -
Criticism and summary of the results of Heiz's method
For the standardization of the titrant (sodium methoxide), using benzoic acid,
a gelatinous precipitate occurs which masks to a certain extent the visual end point
and interferes to a great extent with the potential readings. Since it is always required
to standardize the sodium methoxide just before every estimation of barbiturates, it
seems clear that the use of pyridine as a solvent has no advantage over the other
solvents. This is due to the salting out of the benzoic acid in the form of a gelatinous
precipitate. This difficulty could be overcome when a large excess of pyridine as a
solvent is used. In this case the potential jumps at the end points observed in the
standardization of the sodium methoxide are large.
According to our results, Heiz's method is applicable using the visual end point
titration with sufficient success in the assays of Allobarbital and Methylphenobarbital.
In the case of Cyclobarbital and Hexobarbital the visual end points were not sharp.46)
According to H e i z ', Cyclobarbital could not be titrated by this method using thymol
blue, phenolphthalein or thymolphthalein. In case of Hexobarbital the visual assays
gave us rather low concordant percentages. In the contrary ,Heiz obtained high
percentages.
In the determination of Hexobarbital and Methylphenobarbital, the potential
jumps at the end points are very distinct amounting successively to 1870 and 1200
m. v./ml. (Curves 28' and 29'). Allobarbital does not show any potential jump at the
end point, and hence could not be determined potentiometrically in this solvent (Curve
26). Heiz 'however, reported the applicability of the potential titration to Allobar¬
bital. Cyclobarbital does show a potential jump which is, however, very faint (110
m.v./ml.) (Curve 27').
Pyridine being of a very bad smell and irritant to the respiratory system, it is
not popular to be in common use in analytical work.
- 78 -
m. v.
820
740
660
580
500
420
340
1 3 5 7 9 ml. 0.1M CHgONa
Curve 25
The factor determination of titrant (sod. methoxide) using
benzoic acid as directed by Heiz's method as demonstration
curve for several potentiometric titrations which should be
always carried out before assaying barbiturates.
t
- 79 -
m.v.
620
/
480 /y
/
440//
11 ml, 0.1 M
Potentiometric titration curve for AUobarbltal according to Heiz's
method as demonstration from 3 titrations
A
s/
//
\1
/J
yi
,
/
1 3 11 ml.
UM CH,ONa
Curve 27
The potentiometric titration curve of Cyclobarbital by Heiz
as demonstration from 3 titrations
AE
Ami.
80VV\
0
7.5 mean ml.
Differential potentiometric titration Curve
for Cyclobarbital corresponding to Curve 27
- 80 -
J/
730 /
[650
490
4101 3 5 7 9 ml. 0.1 M CHgONa
Curve 28
Potentiometric titration curve for Hexobarbital by Helz's method as
demonstration from 3 titrations
AE
Ami.
2000
1600
1200
800
4001
1
" 1
S-^
6.5 7.5 mean ml.
Curve 28'
Differential potentiometric titration curve
corresponding to curve 28 for Hexobarbital
- 81 -
-millivolt
850
770
690
610
530
450
//
//
/
9 ml. 0.1 MCHjONa
Curve 29
The potentiometric titration curve of Methylphenobarbital
by Heiz as demonstration from 3 titrations
AE
Ami.
1600
1200
800
400
•
6.5 7.5 mean ml.
Curve 29'
Differential potentiometric titration curve
corresponding to Curve 29 for Methylphenobarbital
- 82 -
Table 19
Results of the visual and potentiometric titration for our barbiturats
(Heiz's method)
Barbiturate visual titration Potentiometric titration
Allobarbital
Cyclobarbital
1. 99.3%
2. 99.3%
3. 99.9%
no visual endpoint
no potential jump (Curve 26)
faint potential jump (Curve 27 and 27')
Hexobarbital
1. 95.61%
2. 91.87%
3. 91.87%
good potential jump
93.11% (Curve 28 and 28*)
Methylpheno-
barbital 1. 98.24%
2. 101.5 %
3. 100.1 %
good potential jump
99. 94% (Curve 29 and 29')
- 83 -
X) Chatten's method
49)Chatten proposed a new non-aqueous technique which was rapid and could
be performed visually. He applied his procedure successfully to commercial samples
of phenobarbital tablets as well as those of phenobarbital and aminophylline. He
extended the use of potassium hydroxide in methanol as a titrant in non-aqueous
titrimetry to barbiturates.
Procedure of Chatten
Apparatus: 5 ml. or 10 ml. burette, graduated in 0.02 ml., electromagneticstirrer, I25~mL suction flask and a small Biichner funnel.
Reagents: 1) Chloroform A.C.S. grade; 2) anhydrous methanol A.C.S. grade;3) benzoic acid A.C.S. grade; 4) potassium hydroxide A.C.S. grade; 5) dimethyl-
formamide, Eastman grade white label; 6) potassium hydroxide 0.1 N in anhydrousmethanol and 7) thymol blue indicator as 0,5% in anhydrous methanol.
Stanjia£dj^ation of Jtitrantj Accurately weigh approximately 200 mg. of benzoic
acid and~dTs~soTve in Wml""of chloroform, and 1 ml. of methanol and 4 drops of thymolblue indicator and titrate to a violet colour. A blank with the solvent system used here
was approximately 0.10 ml. of titrant. The contents of the beaker or flask can be
conveniently protected from the atmosphere by using a piece of rubber dental dam or
cardboard with a hole sufficiently large to permit the burette tip to pass through.
procedures: a) To assay bulk barbiturates, accurately weigh a sample of 40 mg.to 50 mg".~inTo"a~r50 ml. beaker, dissolve in 50 ml. of chloroform by stirring electro-
magnetically, add 1 ml. of anhydrous methanol, 4 drops of thymol blue indicator and
cover the beaker. Titrate to a violet end point with 0.1 N potassium hydroxide in
methanol.
49)Chatten ' confirmed the precipitation of the barbiturate salts when the acids
were titrated in chloroform alone as reported by Swartz and Foss '. It was noted
by Chatten ' that if 1 ml. of anhydrous methanol was added to the chloroform
before beginning the titration precipitation did not occur. Since the titrant employed
in this investigation was potassium hydroxide in anhydrous methanol, the addition of
a further 2. 5 ml., the amount required for most titrations, provided sufficient methanol
to keep the barbiturate salt in solution. It was further noted that the precipitate which
was formed when 1 ml. of methanol was not added, frequently redissolved before the
titration was completed. This did not occur in every instance, however, and consequently
it was deemed advisable to add 1 ml. of methanol as a step in the standard procedure.
This resulted in a sharper end point than that obtained in chloroform alone. He obser¬
ved that the end point was more permanent in either the chloroform or the chloroform
methanol system than in dimethylformamide.
49) I.G. Chatten, J.Pharm. and Pharmacol. 8, 504 (1956).50) C.J. Swartz andN.E.Foss, J. Amer. Pharm.Ass., Sci.Ed., 44_, 217(1955).
- 84 -
The following procedure after Chatten 'was adopted in our assays of the
barbituric acid derivatives:
Standardization of the titrant: 50 ml. of chloroform + 1 ml. anhydrous methanol
+ 4 drops'of thynToT6ruVIn3f<ratbr"were neutralized with potassium hydroxide in
methanol to a pale violet colour. 200 mg. of benzoic acid were dissolved and titrated
to a violet colour with potassium hydroxide in methanol.
Assay_of barbitjur_ic_aci_d derivatiyesj 50ml. of chloroform, 1 ml. of anhydrousmethanol and 4 d6rps~6TfhYmol~biue"were"neutralized with potassium hydroxide to a
violet colour. In most of the assays carried 100 mg. of the barbituric acid derivative
were dissolved in the neutralized chloroform. The beaker was covered during titration
with a cardboard with a hole to permit the burette tip to pass through; a magneticstirrer was used to insure mixing of the titrated solutions.
Table 20
Results of the visual titration of our barbiturates
(Chatten's method)
Barbiturate Percentage found Mean
percentage
Allobarbital 98.80 99.93 99.93%
100.2 100.2 100.2 %
99.87%
Cyclobarbital 98.72 98.72 99.24% 98.89%
Hexobarbital 98.72 98.72 98.72% 98.71%
Methylpheno-
barbital99.91 99.45 99.91% 99.75%
Criticism and Summary of the results of Chatten's method
49)Chatten used either 5 or 10 ml. burette graduated in 0.02 ml. We used in
our assays 25 ml. burette graduated in 0.05 ml. The weight of the barbituric acid
derivatives used in each assay was increased from 50 mg. as recommended by Chatten
to 100 mg. This gave better results as it increased the volume of titrant used and hence
decreasing the percentage of error. Concordant results were obtained in all the 4
barbituric acid derivatives assayed. The end points in all the assays were sharp.
- 85 -
6) Ryan, Yanowsky and Pifer method
These authors were the first to propose the use of lithium methoxide in titrations
of barbiturates. They obtained good results and used comparatively small volumes of
DMF (dimethylformamide) as a solvent to dissolve the barbiturates. The following
experimental procedure was described by them:
Reagents: 0.1 N lithium methoxide. Dissolve 0.6 g of freshley cut lithium metal
in 150 ml. absolute methanol, cooling the flask during the addition. When the reaction
is complete add 850 ml. dry benzene. When cloudiness or precipitation occurs, add
a sufficient amount of absolute methanol to clarify the solution. It is desirable to keepthe volume of methanol at a minimum. Store the reagent in the reservoir of an auto¬
matic burette, protected from carbon dioxide and moisture.
DMF (dimethylformamide) du Pont, technical grade. Thymol blue indicator
1.0% in DMF.
Standardization of lithium methoxide: Accurately weigh out approximately 0.5 gof benzoic acid (National Bureau of Standards) and dissolve in 10 ml. DMF. Perform
the titration visually, using 3-5 drops of thymol blue as the indicator, the end pointbeing taken as a dark blue colour. A blank is run to correct for acidic impurities in
the DMF and usually is of the order of 0.1 ml. of titrant.
Titration fo free barbiturates: Weigh about 3 milliequivalents of the barbiturate
accurately and introduce in 250 ml. -Erlenmeyer flask and dissolve in 10 ml. DMF.
Add 5 drops of thymol blue and titrate with standard lithium methoxide to the appearanceof a dark blue colour. Excessive swirling of solution should be avoided. In order to
avoid the interference of atmospheric carbon dioxide a stream of nitrogen gas may be
directed over the surface of the solution during the titration. The authors have found
that a magnetic stirrer-gives sufficient agitation without carbon dioxide interference.
Procedure: Dried carbon dioxide-free nitrogen gas was bubbled in every
experiment in our assays. For standardization of lithium methoxide benzoic acid
(Analytical Reagent) was used. 0.50 g. benzoic acid consumed about 47 ml. of about
0.1 N lithium methoxide. Only 0. 250 g. of benzoic acid were used afterwards. 10 ml.
of DMF was first neutralized from acidic impurities till a dark blue colour is obtained
using thymol blue as indicatorj it consumed 4 drops of lithium methoxide (i.e. 0.075 -
0.10 ml.).200 mg. of barbituric acid derivative was DMF accurately weighed and dissolved
in the neutralized and titrated with lithium methoxide till a dark blue colour. Usuallythe end point is approached when a green colour predominates but should be continued
till a dark blue colour which is the real end point. No precipitates occured duringtitration of benzoic acid or the barbituric acids derivatives used. The end point in
titration of benzoic acid was a quick change from yellow to blue but in barbituric deri¬
vatives was gradual change from yellow green to dark blue.
51) J.C.Ryan, L.K. Yanowsky and C.W. Pifer,J. Amer. Pharm. Ass., Sci.
Edit. 43, 656 (1954).
- 86 -
Table 21
Results of the visual and potentiometric titrations of our barbiturates
adopting (Ryan et al. method)
Barbiturate visual
titrations
Potentiometric titrations
Allobarbital
98. 35%
98. 35%
98.55%
98.35%
Mean 98.40%
faint potential jump (Curves 31 and 31')
97. 75%
Cyclobarbital
99.30%
99.05%
99.05%
99.30%
Mean 99.17%
could not be assayed potentiometrically
(Curve 32)
Hexobarbital
98.70%
98.70%
98.70%
Mean 98.70%
slight potential jump (Curves 33 and 33')
94.15%
Methylpheno-
barbital 100.1 %
100. 35%
100. 35%
Mean 100.26%
slight potential jump (Curves 34 and 34')
100.3%
- 87 -
Discussion and summary of the results obtained by applying Ryan et al. method
Our results of assays show that non-aqueous titrations are applicable to the
analysis of barbiturates. The non-aqueous method offers two advantages, i.e. being
less tedious and less time consuming; besides it is equivalent to the official method
of the USP XV-method in accuracy and precisions.
The use of lithium methoxide as a titrant gave no difficulty in titrations. It is511
stated by Ryan et al. ' that with the exception of barbituric acid itself no precipitates
were encountered with in other barbituric acid derivatives and even the precipitate
obtained in case of barbituric acid was not gelatinous and of a tough texture which did
not interfere with the end point. In the contrary, the use of sodium methoxide gave
tedious gelatinous precipitates. Furthermore, it was noted by the same authors that
the use of sodium methoxide instead of lithium methoxide under the same conditions
gave a viscous gel which obscured the end point and gave non-reproducible results.
An additional economical advantage of this method is the small volume of DMF
(dimethylformamide) used as a solvent; in all other waterfree methods a large
consumption of solvents occurred. In addition this small volume used tend to favour
to a good extent a high ratio of benzene to solvent at the end point. In accordance with
42)
the experience of Pif er et al. ' the end points of visual titration are sharpened to
a slight extent when a high ratio of benzene to solvent is employed.
The change of the colour of thymol blue indicator is from yellow-green-blue;
this green stage adds to the convenience of its use. Titration to a dark blue colour is
the real end point and gave good reliable results.
Concordant results were obtained in our assays of Allobarbital, Cyclobarbital,
Hexobarbital and Methylphenobarbital titrated visually. In comparison with other water-
free medium methods it is highly recommended and gave good results visually. For
the standardization of lithium methoxide using benzoic acid, the potential jump at the
end point was very pronounced (Curve 30). The curve show a peculiar crescent-shaped
curve with a gradual decrease in potential towards the end point and then a pronounced
potential jump at the end points. The usual curves obtained in Vespe and Fritz method
and Heiz's method showed gradual increase in potential and then potential jump at the
end points.
Allobarbital shows a very faint change in potential at the end point i.e. 40 m. v./ml.
(see Curve 31 and 31'). On the other hand, Cyclobarbital does not show any potential
jump at the equivalence point, and so it cannot be estimated by this method potentio-
metrically (see Curve 32).
The determination of the end point potentiometrically, when using lithium
methoxide as a titrant, is nore or less convenient in the case of Hexobarbital and to a
- 88 -
lesser extent in the case of Methylphenobarbital. These titrations are characterized
with slight potential jump of 145 and 65 m.v./ml. at the equivalence points, for Hexo-
barbital and Methylphenobarbital respectively (see Curves 33' and 34').
- 89 -
-millivolt
450
// \ /
\/
18 22 ml. 0.1 M
lithium methoxide
-millivolt
//
//
i
/350
/
A/
-,(1
310
Curve 30
The factor determination of titrant (lithium methoxide) using
benzoic acid as directed by Ryan et al. method as demonstration
curve for several potentiometric titrations which should be
always carried out before assaying barbiturates.
7 9 11 13
ml. 0.1 M lithium methoxide
The potentiometric titration curve of Allobarbital by Ryan et al.
method as demonstration from 3 titration
AE
Ami.
10.5 11.5 mean ml.
Curve 31'
Differential potentiometric titration curve for
Allobarbital corresponding to curve
- 90 -
1
440
/
/420
400
380
360
i ... ,1
i 11 ml. 0.1 M
lithium methoxide
y7
4
/
9 11 ml. 0.1 M
lithium methoxlde
The potentiometric titration curve-of Cyclobarbital by Ryan et al.
method as demonstration from 3 titration
The potentiometric titration curve of Hexobarbltal by Ryan et al.
method as demonstration from 3 titration
140I
A
;20
j/ \^
9.5 mean ml.
Differential potentiometric titration curve for
Hexobarbltal corresponding to curve 33
- 91 -
-millivolt
500
480
460
440
420
400
380
360
340
13 5 7 9 11 ml.
0.1 M lithium methoxide
Curve 34
Potentiometnc titration for Methylphenobarbital as directed by Ryan et al. method as
a demonstration curve from 3 titrations
AE
Ami.80
60
40
20
8.5 9.5 mean ml.
Curve 34'
Differential potentiometnc titration curve for
Methylphenobarbital corresponding to curve 34
i
- 92 -
2. Argentometric methods
Survey and discussion of the literature
In general, the argentometric method depends on the fact that the barbiturates
ir. alkaline media form soluble silver barbiturate (see page 48, formula VI-K), which
are not dissociated. Excess of the silver ion, when sodium carbonate is used as a
solvent will cause the precipitation of silver carbonate. In case of sodium hydroxide
being used as a solvent, silver oxide precipitates.241
B u d d e' recommended the following method: Dissolve the barbituric acid
derivative in sodium carbonate solution and titrate with 0.1 N silver nitrate to the
appearance of a permanent turbidity. One molecule of the barbituric acid derivative
is equivalent to one equivalent of silver.53)
Mangouri and Mi lad ' showed that the percentage figure is dependent on
the ratio of the sodium carbonate to the weight of the barbituric acid derivative taken.
They stated that concordant results could be obtained by this method, if the weight of
the sodium carbonate added from the beginning is approximately five times the weight
of the barbituric acid derivative. On no account should the weight of sodium carbonate
considerably exceed the specified limit or much higher results exceeding 100 % may
be obtained.
54)Kalinowski recommended dissolving the barbituric acid derivative in N
sodium hydroxide and titrating with 0.1 N silver nitrate in presence of alcohol till a
permanent turbidity is produced. One molecule of barbituric acid derivative is
53)equivalent to one equivalent of silver. Mangouri and Milad
,on applying this
method noticed that the figures obtained were comparatively high and that the addition
of a slight excess of silver nitrate darkens the solution and thus masks the end point.
S c h u 1 e k and R o z s a dissolved a known weight of the barbituric acid
derivative in hot 5 % borax solution and titrated with 0.1 N silver nitrate till a reddish
colour predominates; one molecule of the barbituric acid derivative corresponds to
53)2 equivalents of silver. On trying this method, Mangouri and Milad 'obtained
the best results when the titration was carried out while the solution was simmering.
They pointed out the difficulty of observing the colour change at that high temperature.
They also showed that the end point varied according to the temperature of the solution,
thus the required colour change was found to take place earlier in a cold solution than
in a warm one and that it disappeared on warming; consequently the final results
depended on the temperature of the solution.
53) H.A. Mangouri and L. Milad, Quart.J.Pharm.& Pharmacol., 2CK 109(1947).54) K. Kalinowski, Chem.Zentr. I_, 2391 (1936).
- 93 -
53)
Mangouri and Mil ad ' dissolved the barbituric acid derivative in sodium
acetate solution and few drops of ammonia; the excess ammonia being expelled by
boiling. After cooling a measured excess of silver nitrate was added together with
about 0.1 g. of pure calcium carbonate and the solution was boiled to ensure curdling
of the precipitate. After cooling and filtering through Gooch crucible, the excess of
silver nitrate was estimated by ammonium thiocyanate using ferric ammonium sulphate
as indicator. One molecule of the barbituric acid derivative is equivalent to 2 equi¬
valents of silver in the case of acid derivatives or their sodium salts e. g. Barbital or
soluble Barbital; 4 equivalents of silver in the case of thiobarbituric acid derivatives
or their sodium salts; one equivalent of silver in the case of N-methylated derivatives
25)or their sodium salts. Stainier et al. applying the previous method, obtained
unsatisfacory results and they explained this by the possibility, that necesary conditions
for the total precipitation of the silver salt are difficult to realize.23)
Danielson 'proposed an argentometric method for the estimation of the
barbiturates using potassium metaborate solution as a solvent and titrating with silver
nitrate using potassium chromate as indicator and as a comparison solution. One
molecule of silver nitrate corresponds to 2 molecules of the barbituric acid derivative.
The method could be used for the assay of Allobarbital and Barbital. 5 Allyl 5 iso-
propylbarbituric acid can also be determined but the precision is not so good. Attempts
to apply the method to the following derivatives have failed for Cyclobarbital, Hexo-
barbital, 5-sioamyl-5-ethyl and Phenobarbital.55) 23)
Chavanne and Marie' modified the method of Danielson ', avoiding
the use of potassium metaborate which is difficult to obtain and used a mixture of
ethanolic potash and boric acid to dissolve the barbiturates.
56)Recently it was demonstrated by Vastagh and Szaboles that the argen-
tometric determination of Hexobarbital may be performed with considerable exactness
provided certain experimental conditions are strictly observed (among other the pH
must be kept constant at 8. 80). Thus Hexobarbital sodium may be determined as well
as the substance proper. Phenobarbital may also be titrated by argentometry both in
buffered solutions (pH 7.8) and in alkaline alcoholic solutions. The latter determination
is also possible in the presence of alkaloid bases, or chlorohydates, phenazones,
amidopyrine, caffeine and very small amounts of bromides.
55) P. Chavanne andH. Marie, Ann.farm.franc. 11_, 91 (1953).56) G. Vastagh andE. Szaboles, Arzneim. Forsch., j5 ,
355 (1958).
- 94 -
a) The application and results of Budde's Argentometric method
04)Budde described the following method for the barbituric acid derivatives
with the exception of the nitrogen methylated derivatives:
"About of the dried 0.2 - 0. 3 g. Barbiturate, accurately weighed, and dissolved
in 30 ml. water, 1 g. of exsiccated sodium carbonate is added and to the clear solution
0.1 N silver nitrate is added from burette till a clear turbidity occurs for a certain
time. One ml. 0.1 N silver nitrate is equivalent to 1/10*000 Mol. of the used barbi¬
turic acid derivative.
We applied the described method of Budde visually and potentiometrically
using the Metrohm apparatus E 157, (Herisau, Switzerland), silver electrode immersed
in the solution to be titrated and a calomel electrode immersed in a beaker of saturated
potassium chloride and was connected to the titration beaker by means of a salt bridge.
Solutions were mechanically stirred. The salt bridgewas made of potassium nitrate
in agar (30 g. KNO«, 3 g. agar and 100 ml. of water).
Gentle heating was used to effect the solubility of the barbituric acid derivative
in the sodium carbonate solution, then cooling and titrating with 0.1 N silver nitrate.
Table 22
Summery of the found percentages of barbiturates by Budde's Method
Barbiturate visual
titrations
Potentiometric
titrations
Cyclobarbital 1. 100.93%2. 100.93%3. 100.93%
1. 83.75%(Curve 35 and 35')
Hexobarbital 1. 53.12%2. 53.12%3. 53.12%
Inapplicable(Curve 36)
Methylpiieno-
barbital
1. 43.68%2. 43.68%3. 43.68%
Inapplicable(Curve 37)
Pentobarbital
sodium
1. 111.64%2. 111.64%
86.75%(Curve 38 and 38')
- 95 -
-millivolt
300
260
220
isa
14a
100-
60
'
7 9 11 ml. 0.1N AgNOj
Curve 35
Potentiometric titration for Cyclobarbital as directed by Budde's method as a demon¬
stration curve from 3 titration
AE
Ami.
100
80
60
40
20
\\1
\/o \
\\\
9 mean ml.
Curve 35'
Differential potentiometric titration curve for Cyclobarbital
corresponding to curve
- 96 -
-millivolt
y^
240
160
120
] I I 4 5 3 7 ml. 0.
Curve 36
Potentiometrlc titration for Hexobarbital as directed by Budde's method as a demon¬
stration curve from 3 titration
-millivolt
180
140
mn
6 7 ml. 0.1 N Ag NOg
Curve 37
Potentiometric titration for Methylphenobarbital as directed by Budde's method as a
demonstration curve from 3 titration
- 97 -
-millivolt
280
240
200
160
120
80
•^-"^
1 5 9 11 113 ml. 0.1 N Ag N03
Curve 38
Potentiometric titration for Pentobarbital sodium as directed by Budde's method as
a demonstration curve from 3 titrations
AE
~SmT.
60
50
40
30
20
10
v\ /
00
9 mean ml.
Curve 38'
Differential potentiometric titration curve for Pentobarbital
sodium corresponding to curve 38
- 98 -
Criticism and summary of the results of the application of Budde's method
The only case that could be determined by this method, as revealed from the
summary of the results, is Cyclobarbital. On the other hand, Pentobarbital sodium
gave an error of about + 11%. Hexobarbital and Methylphenobarbital are not deter¬
minable by the above suggested method. The error obtained in these two cases were
very high amounting respectively to about - 47% and - 57%.
The method of Budde was also carried out potentiometrically. The above 4
mentioned acids were assayed. The results of such experiments are presented in
Curves 35, 36, 37 and 38, and the corresponding differential potentiometric titration
curves 35' and38'. Pentobarbital and Cyclobarbital were the only two acids that gave
a measurable potential jump at the end point. The potential jump at the end point
amounts to 60 and 80 m. v./ml.respectively (Curve 35' & 38'). However, in the two
previous cases, negative error for the concentration was always obtained. The other two
acids, i.e., Hexobarbital and Methylphenobarbital do not show any potential jump
characteristic of an end point (Curves 36 and 37). The potential of the working electrode
changed linearly with the quantitty of silver nitrate solution added. This latter result
supports Budde's statement, that these acids are not determinable by his method.
b) Application and Results of Bodin's Argentometric methods
The potentiometric method described by Bo din 'is based upon the reaction
24)conditions used by B u d d e and consists of the quantitative reaction of 1 molecule
of silver ion with 1 molecule of barbiturate to form a slightly dissociated salt, which
is soluble in dilute sodium carbonate solution. At the end point excess silver ion
precipitates silver carbonate such that the end point potential is that of saturated
solution of silver carbonate in the presence of excess carbonate ion.
59)In a previous paper by B o d i n it was shown that the method as proposed by
Mattocks and Vo shall 'was found to be unsatisfactory in that the end point
potential of their dead-stop titration was not reproducible. The method described
here eliminates the uncertainity in the end point potential by determinning the potential
of a standard blank solution (saturated with silver carbonate for each sample just prior
to titration). This end point potential varies from time to time depending on the condi¬
tion of the electrodes, but is relatively constant to + 1 millivolt for the length of time
required to perform several titrations.
57) J.I. Bodin, J.Am. Pharm. Ass. Sci.Ed. 45, 185(1956).58) A.M. MattocksandE.C. Voshall, J. Am. Pharm. Ass. Sci.Ed. 49, 28 (1950).59)J.I.Bodin andA.Taub, J. Am. Pharm. Ass. Sci.Ed. 44, 296 (19"5l5).
- 99 -
Experimental
In our assays we used the following apparatus and reagents adopting the same
57)procedure of Bo din '.
a) Apparatus: Metrohm pH-meter E 157 (Herisau, Switzerland); silver and saturated
calomel electrode of the same manufacturers; saturated calomel electrode of the
same manufacturers; saturated potassium nitrate; agar potassium nitrate bridge.The Metrohm apparatus was used on the milivolt scale. The silver indicator
electrode was immersed in the solution to be titrated, the calomel electrode in a
beaker of saturated potassium chloride and was connected to the titration beaker
by means of a salt bridge. Solutions were mechanically stirred by a magneticstirrer.
b) Reagents: 0.1 N silver nitrate; anhydrous sodium carbonate, 3%, alcohol, 95%.c) Standard"blank: A solution having the same composition as the solvent in which
the barbituric acid derivative was dissolved was prepared. This solution consisted
of 10 ml. of 95 % alcohol, 50 ml. 3% sodium carbonate solution, 1 ml. 0.01 N
silver nitrate and distilled water to make 100 ml. The potential of the standard
blank solution was determined prior to the titration of each sample and it served
as the end point potential to which the sample was titrated. To correct for the blank,0,1 ml. was substracted from the volume of 0.1 N silver nitrate used for each
titration.
d) Procedure of titrations: About 500 mg. of the barbituric acid derivative, accuratelyweighed, were introduced in 250 ml. volumetric flask. Exactly 25 ml. of alcohol
95% and 25 ml. of distilled water were added. The flask was swirled till the
undissolved solid was evenly dispersed. 125 ml. 3% sodium carbonate solution
were added, the flask shaked thoroughly and diluted to mark with distilled water.
100 ml. aliquots were titrated to the standard blank potential.
Table 23
Summary of the found percentages of Allobarbital by Bodin's electro-
metric method
Percentage of sod.
Carbonate used
Percentage accordingto potential readingof standard
Percentagecalculated
potentiometri-cally
Percentageaccording to
turbidity (Visual)
1 1. 97.85%2. 96.29%3. 97.85%4. 97.85%
1. 91.5%(Curves 41
and 41')
1. 98.89%2. 96.29%3. 93.69%4. 96.29%
2 1. 105.14%2. 113.48%3. 123.87%4. 124.92%
1. 94.19%(Curves 40
and 40')
1. 98.89%2. 98.89%3. 98.89%4. 98.89%
3 1. 108.26%2. 105.66%3. 100.45%4. 109.3%
1. 97.1%(Curves 39
and 39')
- 100 -
Criticism and summary of the results of the application of Bodin's method
57)The method proposed by Bodin ' necessitates the performance of a blank
experiment. This experiment implies the addition of a 0.1 ml portion of 0.1 N silver
nitrate solution to the solution not containing the barbiturate. In performing an
experiment after it, the titration is conducted until this determined potential is
reached. However, such a potential is on the flat portion of the titration Curve, i. e.
after the exact end point is reached. This is clarified by Curve 39 (for Allobarbital
using 3% sodium carbonate) in which the predetermined potential was found to be
- 240 mv versus the saturated calomel electrode, corresponding to 10.50 ml. of
0.1 N silver nitrate solution. However, the plot of the differential potentiometric
titration curve, Curve 39! shows clearly that the end point is 9.4 ml. silver nitrate
solution. This latter value corresponds to the theoretical concentration of the acid
with an error amounting to - 3%. The concentration value of the acid as determined
by the method of Bodin ' leads to an error of + 8.25%, revealing the nonexactness
of the standard potentiometric end point.
On performing the same experiments, but using the visual turbidimetric method
for the determination of the end point, the same amount of error as shown above was
also noticed. This might be due to the presence in solution of traces of carbon dioxide
resulting from the dissolution of the barbiturate in the carbonate solution. Thus, on
titrating with silver nitrate solution there might exist the probability of formation of
some soluble silver bicarbonate which necessitates the addition of an excess of the
reagent before the end point is obtained.
The effect of the concentration of the sodium carbonate on the sharpness of the
end point was also studied. Three solutions having respectively the concentration of
1, 2, and 3% sodium carbonate were used.
The end point was determined either turbidimetrically or electrically. In the
determination of the latter both the standard potential method due to Bodin and that
from the differential potentiometric titration curves were calculated. From Curves 391,
40'and 41'it is shown, that the potential jump increases from 1 to 2 and 3% sodium
carbonate solution.
The 1% sodium carbonate solution was the best in detecting the end point by the
standard potential method. In all the solutions studied, the estimation of the end point
by the differential titration curves was low and is not to be recommended.
The turbidimetric determination of the end point is best obtained in the 2% sodium
carbonate solution.
- 101 -
-millivolt
280
240
200
160
120
80
//
/
//
/'
7 9 11 ml. 0.1 NAgNOg
Curve 39
Potentiometric titration for Allobarbital as directed by Bodin's method as a demon¬
stration curve from 3 titrations
AE 60
Ami.
50
40
30
20
10
o
11 mean ml.
Curve 39*
Differential potentiometric titration curve for Allobarbital corresponding to
curve 39
- 102 -
ZBU
|240
s
120
i^*^
80
5 7 9
Curve 40
11 13 ml. 0.1 N AgNOg
Potentiometric titration curve for Allobarbital by Bodin's method using 2% sodium
carbonate as an exemple from 3 titrations
A E
Ami
100
60
20II
5 7
Curve 40'
11 mean ml.
Differential potentiometric titration curve for Allobarbital
corresponding to curve 40
103
260
/
100
fiO
1 3 5 7 9 11 13 15 ml. O.lNAgNOj
Curve 41
Potentiometric titration curve for Allobarbital according to Bodinfs method using 1% sodium
carbonate solution
A E
Ami.
35
:3
5 7
Curve 41'
11 mean ml.
Differential potentiometric titration curve for Allobarbital
corresponding to curve 41
- 104 -
c) Application and results of Mangouri and Milad argentometric method
531Mangouri and Milad recommended the following method for the assay
of barbituric acid derivatives:
To a known weight of a barbituric acid derivative, suspended in about 20 ml. of water
add 10 ml. of sodium acetate solution (10% w/v) and dissolve it by the aid of gentle
heat, in a slight excess of dilute ammonia solution added in drops, then boil off the
excess of ammonia. Add a known excess of 0.1 N silver nitrate and about 0.1 g. of
pure calcium carbonate and boil the liquid for 2-3 minutes, cool and filter throughasbestos-packed Gooch crucible and wash the precipitate with 5 ml. quantities of
freshly boiled and cooled water till silver free. Acidify the combined filtrate and
washings with nitric acid and titrate the excess of silver nitrate with 0.1 N Ammonium
thiocyanate, using ferric ammonium sulphate as indicator. 1 mol. of the barbituric
acid derivative is equivalent to 2 equivalents of silver in the case of acid derivatives
or their sodium salts e. g. Barbital or soluble Barbital; 4 equivalents of silver in the
case of thiobarbituric acid derivatives or their sodium salts; 1 equivalent of silver
in the case of N-methylated derivatives or their sodium salts. The duration of boilingis generally 2-3 minutes, but it should be increased to 5 - 6 minutes in the case of
thiobarbituric acid derivatives; while in the case of sodium N-methylated derivatives
no boiling is necessary as just simple heating 60-70 C. is sufficient to bring about
the required curdling of the silver compound. When the 2 imido hydrogen atoms in
the nucleus of the barbituric acid derivative are replaced, the one by a metal and the
other by a radical, the addition of ammonia solution is not necessary.
Table 24
Summary of the found percentages of barbiturates by Mangouri and Milad method
BarbituratePercentage found
Mean
percentages
Allobarbital
Cyclobarbital
Hexobarbital
Methylpheno-barbital
Pentobarbital
sodium
95.38 98.29 95.74 98.20%
74.73 76.65 75.65%
97.59 89.64 92.41 91.22%
86.74 91.07%
90.46 94.32 78.52%
83.01 85.67 83.56%
96.9%
75.67%
91.46%
87.7%
84.08%
- 105 -
Criticism and summary of the results of the application of Mangouri
and Milad method
The exact procedure recommended by Mangouri and Milad 'was applied in this
study; using porcelain filtering crucibles (Berliner Hartporzellan fur hbchste An-
spriiche LM 3) instead of the Gooch crucibles used by the authors.
No concordant results were obtained in any of the assayed acids. Low percenta¬
ges were obtained in all the assayed acids. This might be attributed either to the partial
solubility of the precipitated silver salts, or to the incomplete precipitation of the
silver barbiturates under the conditions specified by the authors.
- 106 -
d) Application and results of Chavanne and Marie argentometric method
551 601Chavanne and Marie
'modified that method used by Daniel son
,
avoiding the use of potassium metaborate, which is difficult to obtain. They recommen¬
ded the following method:
0.6 g of the barbiturate is dissolved in 7 ml. of N ethanolic potash and the
solution is diluted with 33 ml. of water and treated with 0.45 g. of boric acid. After
warming to dissolve the boric acid and cooling, the mixture is titrated with 0.1 N
silver nitrate in presence of 1.5 ml. of 10% potassium chromate solution as indicator.
The end point is indicated when the solution assumes permanently a colour different
from that of the comparison solution containing 2 g. of precipitated calcium carbonate,1.5 ml. potassium chromate solution and 55 ml. of water. 1 mol. of silver nitrate
corresponds to 2 mol. of barbiturate. They applied their method to Allobarbital and
Barbital. It may also be used for Hexobarbital, but in this case there is no precipitateand the comparison solution does not contain calcium carbonate. For Phenobarbital
a different pH value is necessary and the amount of boric acid is therefore increased
to 1.2 g.
The potential change at the end point is small in presence of sodium carbonate241
as directed by Budde's ' method. A much greater voltage jump is obtained by using
potassium metaborate in place of carbonate, owing to the solubility of silver meta¬
borate being greater than that of carbonate. It is essential that the standard solution
of silver nitrate is slowly added all the time, not only near the equivalence point.
- 107 -
Table 25
Summary of the found percentages of barbiturates by Chavanne and Marie method
Barbiturate Visual Titration Potentiometric
titrations
Allobarbital
Cyclobarbital
Hexobarbital
Methylpheno¬
barbital
Pentobarbital
sodium
98.20 97.83 97.83 97.51%
97.35 98.20 97.10 97.10%
Mean 97.64%
93.65 94.23 93.85%
Mean 93.91%
96.16 96.16 96.16 96.16%
Mean 96.16%
104.58 104.16 104.58 104.58%
Mean 104.47%
112.78 114.6 114.6%
Mean 113.99%
Moderate potential
jump 103.6%
(Curve 43 and 42')
Inapplicable
(Curve 43)
Inapplicable
(Curve 44)
Inapplicable
(Curve 45)
Inapplicable
(Curve 46)
Criticism and summary of the results of Chavanne and Marie method
The method recommended by Chavanne and Marie for the estimation
of barbiturates was studied using both the visual end point titration and potentiometri-
cally. The results of the assays of Allobarbital visually gave more or less concordant
results and an error of - 2,4%. In case of Cyclobarbital, about -6% error was estimated
and the end point was rather difficult to detect due to the pale red colour developed
with the ethanolic potassium hydroxide. In case of Hexobarbital low concordant results
were obtained giving an error of about - 4%. In the assays of Methylphenobarbital
an error of + 4.5% was recorded. In the assays of Pentobarbital, about + 14% error
was obtained.
The potentiometric titration curve for Allobarbital, Curve 42, showed a moderate
jump in potential near the end point i.e. 400 millivolt/ml. (Curve42% However, the
percentage of found Allobarbital potentiometrically gave an error of + 3,6%. All the
other potentiometric titration curves namely curves 43, 44, 45 and 46 for Cyclobar¬
bital, Hexobarbital, Methylphenobarbital and Pentobarbital showed very faint potential
jumps near the end points. These mentioned four acids could not be estimated
potentiometrically by this method.
- 108 -
-350
-250
-150
-50
0
+50
+100
r*'
o
1 t0t^r
—
9 11 13 15 0.1 O.lNAgNO,
Curve 42
Potentiometrlc titration curve for Allobarbital by Chavanne and Marie method as an
example from 3 titrations
A E
Differential potentiometrlc titration curve for
Allobarbital corresponding to Curve 42
- 109 -
350-
26 ml. 0.1 N AgNOg
Potentiometric titration curve for Cyclobarbital according to Chavanne and Marie Method
as demonstration from 3 titration
volt3
millic e
i
150
0
1 9 11 13 15 ml. 0,1N AgNOgCurve 44
Potentiometric titration curve for Hexobarbital according to Chavanne and Marie method
as an example from 3 titrations
- 110 -
13 350
ai
150
0
7 9
Curve 45
11 13 15 ml. 0.1 N AgNO,
Potentiometric titration curve for Methylphenobarbital according to Chavanne and Marie method
as an example from 3 titrations
5 300
6
1>
—
0
15 ml. 0.1 N AgNOj1 3 5 7 9 11 13
Curve 46
Potentiometric titration curve for Pentobarbital sodium according to Chavanne and Marie Method
as an example from 3 titrations
- Ill -
3. Mercurimetric method
Survey and discussion of the literature
25)
Stainier et al. ' examined the identification reactions of the barbituric acid
derivatives with different reagents i.e. Millon's reagent (mercuric nitrate), Denige's
reagent (mercuric sulphate) and mercuric acetate. They found that all ordinary
barbituric acid derivatives gave positive precipitation reactions with these reagents.
The barbituric acid derivatives substituted at the nitrogen atom behaved differently
e. g. Methylphenobarbital gave no precipitate with Millon's reagent, with mercuric
acetate and Denige's reagent it gave weak positive reactions. Hexobarbital gave
negative reactions with all the previous reagents. These reactions are not only
specific for the barbituric acid derivatives, but also other substances e. g. sulphon-
amides and purine bases derivatives gave positive reactions with the previous reagents.
Pedley used mercuric perchlorate solution to precipitate the barbituric
acid derivatives dissolved in boiling water. The excess of mercuric perchlorate being
titrated against ammonium thiocyanate. As the barbituric acid derivatives themselves
are insoluble in water, preliminary experiments were performed by adding the
mercuric perchlorate solution to a boiling solution of the barbiturate in water. Owing
to the slight solubility of the precipitate in boiling water however low results were
obtained. This tendency was also emphasized when the volume of water used to dissolve
the barbiturate was increased. At high concentrations also low figures were obtained
probably due to complex formation, but this was only manifested in concentrations
over 8 millimols per liter. If the concentration was maintained between 2.5 and 5.25
millimolecule per liter concordant results were obtained. In the method finally
adopted, mixing of the two solutions was carried out at room temperature at a dilution
of about 0.2 g. in 150 ml. The effect of buffering the barbituric acid solution was also
investigated, as the sodium salts of the barbiturates are alkaline in reaction, and here
again it was found that a wide variation of pH was possible without appreciable effect
on the results obtained. The precipitate was found to be appreciably soluble in excess
of perchloric acid, however no precipitate is obtained in the presence of chloride ion
or mineral acids. In order to avoid a long washing of the precipitate, complete
filtration was avoided and an "aliquot part" method was used throughout. The results
of the estimation of Barbital, Phenobarbital and their sodium salts, carried out by
Pedley'
gave almost 100% results.
Brauniger and Borgwardt'applied the method of Ped 1 ey to the
61) H. Brauniger and G. Borgwardt, Pharmaz. Zhalle, 93,
266 (1954) and
93,299 (1954).
- 112 -
following barbituric acid derivatives i. e. Barbital, Cyclobarbital, Methylphenobarbital
and Phenobarbital, using different quantities of each to examine the effect of the
concentration of the barbiturate. They found that the results of the assays for concen¬
trations below 2.5 millimol per liter are very low. They also examined the effect
of using different percentages of methanol as a solvent. They also compared their
results obtained by applying Pedley's method with the results obtained by the acidi-
metric method using methanol as a solvent. They obtained better results by their
acidimetric method than those obtained by Pedley's method.
Application and results of Pedley's mercurimetric method
Pedley' recommended the following volumetric method for estimation of
barbituric acid derivatives:
Experimental: Mercuric Perchlorate Solution; An approximately 0.1 M solution was
prepared by boiling an excess (25 g.) of mercuric oxide with 28 g. of 60% perchloricacid in 200 ml. of water adjusting to 1 liter and filtering. This solution contains
approximately 36.6 g. of mercuric perchlorate and remains stable indefinitely.
Method adopted: Weigh out approximately 0. 2 g. of the barbiturate (or its equivalentin powdered tablets) and dissolve in 50 ml. of boiling distilled water. Boil for a few
minutes after solution has been achieved and add about 80 ml. of water and allow to
cool to room temperature. Transfer the solution to a 200 ml. measuring flask and
wash the container several times adding the washings to the contents of the flask,until the volume is about 150 ml. Now add slowly, with rotation of the contents of
the flask, 25 ml. of the solution of mercuric perchlorate and allow the mixture to
stand with frequent shaking for 15 minutes. Adjust the volume of the solution to 200 ml.
and then filter through a double fluted filter paper into a dry 100 ml. measuring flask.
Reject the first 50 ml. of the clear solution and then collect 100 ml. Transfer this
filtrate to a conical flask, washing the measuring flask with 20 ml. quantities of 10%nitric acid solution and adjust the volume of liquid to about 250 ml. Add 1 ml. of a
saturated solution of ferric alum and titrate with N/10 Ammonium thiocyanate. Repeatthis operation without the barbiturate. The difference in the 2 titrations representsthe number of ml. of mercuric perchlorate solution required for the barbiturate.
Each ml. of N/10 ammonium thiocyanate is equivalent to the molecular weightof the barbiturate/20*000, e.g. barbitone 0.00921 g.; barbitone sodium 0.0103 g. etc.
Methylphenobarbital on the other hand, though closely related to Hexobarbital
gives a practically insoluble mercury derivative, but owing to the almost complete
insolubility of the Methylphenobarbital itself the method had to be modified in the
following way:
Heat 0.4 - 0.5 g. of the Methylphenobarbital with 20 ml. of 0.1 N sodium hydroxideand 120 ml. of water to 60° C until dissolved. While maintaining at this temperatureadd 10 ml. of 10% acetic acid followed immediately by 25 ml. of mercuric perchloratesolution. Allow the mixture to cool to room temperature, transfer to a measuring flask
with washing, adjust to 200 ml., stand for 15 minutes with shaking and proceed as in
the original method. In this case 1 mol. of mercury combines with 2 mol. of the
- 113 -
barbiturate so that 1 ml. of a 0.1 N ammonium thiocyanate is equivalent to 0.02464 g.of Methylphenobarbital.
Procedure adopted was exactly as directed by the original method of P e d 1 e y .
The solubility of the precipitated mercury salt was examined in the present
study. The precipitate was washed with distilled water and suspended in dilute solution
of perehloric acid (about 1. 5 per liter of water) having the same pH value as that of
the filtrate and completed to 200 ml. in a measuring flask. The suspension was shaken
occasionally for 15 minutes and filtered through double fluted filter paper. 100 ml.
of the filtrate was acidified with nitric acid and titrated against 0. IN ammonium
thiocyanate. The quantity obtained (ca. 0.2-0.3 ml.) was used to correct the amount
of ml. of thiocyanate equivalent to the excess of mercuric perchlorate. By adopting
this correction for the solubility of the precipitated mercury salt of cyclobarbital
better results were obtained i.e. 100.5 %, 99.2 % and 98.5 %.
In the case of the precipitated mercury salt of Pentobarbital adopting the
previously described method of correction, the results obtained were nearly the same
as those without correction i. e. 92.3 % and 92. 3 %.
Table 26
Summary of the found percentages of barbiturates by Pedley's method
Barbiturate Assay number
1 2 3
Mean
percentage
Allobarbital
Cyclobarbital
Hexobarbital
Methylphenobarbital
Pentobarbital
sodium
117.11 117.7 118.52%
95.62 95.62 95.62%
51.94 63.18 63.18%
97.2 97.5 97.5 %
92.41 92.41 92.41%
117.77%
95.62%
59.4 %
97.4 %
92.41%
Criticism and summary of the results of the application of Pedley's method
Concordant results were obtained in the assays of Methylphenobarbital giving
an error of - 2.6 %. In case of Allobarbital high percentages were obtained and the
error amounts to + 17. 77 %. It may be attributed to the reduction of the mercuric
perchlorate to a slightly soluble mercury salt e.g. mercurous chloride. In case of
- 114 -
Cyclobarbital low concordant results were obtained with an error of - 4.4 %. For
Pentobarbital sodium also low results were found and the error amounting to -7.6 %.
In case of Hexobarbital very low variable results were obtained and this was explained
by Pedley to be due to the solubility of the precipitated mercury salt.
- 115 -
4. Bromometric method
Survey and discussion of the literature
The bromometric method depends on the fact that unsaturated radicals e. g.
allyl or cyclohexenyl can be estimated by bromination, whereas the double bonds
are saturated with two atoms of bromine and the excess of bromine is determined.
For example, the allyl radical (CH2 = CH-CH2 -) fixes two atoms of bromine to give
the dibromoderivative (CHjBr-CHBr-CHg -). This method of bromination of the
unsaturated radicals is used by the Danish Pharmacopoeia 1948 II to estimate
Allobarbital, Hexobarbital, and Cyclobarbital. It describes the following procedure:
0.15 g. of the dried barbituric acid derivative, accurately weighed, is dissolved
in 15 ml. of chloroform, introduced in a glass stoppered Erlenmeyer flask. 50 ml.
0.1 N potassium bromate, 5 g. potassium bromide and 10 ml. 2 N sulphuric acid
are added. Shake vigorously for one minute, keep in the dark for 45 minutes, shakingoccasionaly. Add 10 ml. of potassium iodide (1 + 9), shake vigorously and titrate
slowly with vigorous shaking ageinst 0.1 N sodium thiosulphate till the blue colour of
the starch solution used as indicator near the end point disappears. Carry out a blank
using the same reagents without using the barbituric acid derivative. The difference
between the amount of sodium thiosulphate used in the blank and that used in the
experiment corresponds to the barbituric acid derivative.
Schill ' described an identical procedure to that of the Danish Pharmacopoeia
1948,for the estimation of Allobarbital and Hexobarbital. He recommended the use
of 2 ml. of methyl alcohol with the 15 ml. of chloroform and 2 minutes shaking for
Allobarbital acid and 5-allyl 5-isopropyl barbituric acid. One minute shaking and
only chloroform as solvent for Hexobarbital and let stand for 60 minutes before
titration.
Table 27
Summary of the found percentages of unsaturated barbiturates by
the bromometric method
Barbiturate Percentage found Mean
percentage
Allobarbital
Cyclobarbital
Hexobarbital
99.85 100.33 99.29%100.14 99.48 100.42%99.72 99.15 99.72%
94.73 94.73 94.73%
99.70 99.70 99.49%
99.78%
94.73%
99.63%
52) G. Schill, Chem. Abstr. 6022-8, (1946).
- 116 -
Criticism and summary of the results
More or less concordant results were obtained in the assays of Allobarbital and
Hexobarbital. In case of Cyclobarbital, rather low concordant results were obtained.
It is probable that the time of bromination (45 minutes) is not sufficient in the latter
case. The bromometric method has the advantage of determining the unsaturated
barbituric acid derivatives when present in a mixture with other barbiturates.
5. Colorimetric method
Survey and discussion of the literature
Parri" 'was the first to describe the reaction of cobalt nitrate with the
barbituric acid derivatives in the presence of ammonia. The reaction was carried
out by treating a little amount of cobalt nitrate with few drops of ammonia just in
slight excess and adding the barbituric acid derivative. A rose violet colour develops
on cold. It was shown by Parri6 ' that excess of ammonia destroyed the colour.
The reaction of Parri could be used for the estimation of the barbiturates in certain25)
limited cases. As pointed out by Stainier et al. ' who adopted the conditions
given by L junber g for the estimation of Phenobarbital, satisfactory results
were obtained. They pointed out that the application of this method is limited from
5-10 mg. and every time a standard is made. Several authors have shown, that the
intensity of the reaction depends on the respective proportions of the cobalt salt and
the base for a certain quantity of the barbiturate. Zwikker ' studied the complex
cobalt salts produced with Barbital and proposed the formulae for them (see page 51 ).64)
Griffon and Le Breton used diethylamine in place of ammonia and applied65}
their reaction to certain toxicological analysis. Koppanyi et al. ' modified
Parri's reaction, using cobalt acetate and isopropylamine which gave advantages over
the original reaction.30)
Baggesgaard-Rasmussen and Jers 1 ev '
gave many important obser¬
vations on the reaction. They found that the results obtained by the different authors
are somewhat contradictory and often difficult to compare owing to the variations in
the performance of the reaction. They stated that most cobaltous salts are deliquescent
62) W. Parri, Boll. Chim.Farm., 36, 401 (1924).63) A.S. Ljunberg, Collect. PharmTSueeica (1952).64) H. Griffon and R. Le Breton, Ann. pharm. franc., 5, 393(1947).65) T. Koppanyi, W. Murphy and S. Krop, Arch. int. PKarm.Th6r., 46, 76(1933).
- 117 -
and not of a well definite composition. Anhydrous cobaltous acetate is stable when
kept in calcium chloride and potassium hydroxide desiccator; its solution (0.01 M)
in absolute methanol should not be kept for more than 3-4 weeks. Any primary
aliphatic amine gives the colour reaction but isobutylamine is preferred because its
boiling point compares well with that of the solvent. A molar solution in chloroform
is used which should not be kept for more than 5-10 days when used for the quanti¬
tative determination. The intensity of the colour increases with increasing the amine
concentration; with 100 moles of amine per mole cobalt which was chosen as the most
preferable, small variations in the amine concentration do not effect the optical
density to any measurable degree. Piperidine produces a colour which is as intense
as but less stable than the colour given by the primary aliphatic amines. The stability
of the colour is very poor in methanol solution but much better in solution containing
large amounts of chloroform. A solvent consisting of 12.5% (v/v) methanol in chloro¬
form is preferred.
The following observations all refer to the reaction used as a quantitative method:
The colour is not very stable, but in the same experimental series well reproducible
results are obtained when the measurements are made immediately after the prepa¬
ration of the solutions. The intensity of the colour varies form one day to another and
the same applies to the extinction given by solutions containing no barbiturates. For
that reason it is not possible to give definite standard curves, but a comparison curve
must be worked out the same day it is to be used. This is most conveniently done by
measuring the extinction value for three solutions, one with no barbiturate content
and two with concentrations of the barbiturate concerned corresponding to the first
straightlined part of the extinction concentration curve. The unknown sample must be
diluted to a concentration within this range; extrapolations are not permitted. The
slope of the straightlined parts of the extinction concentration curves varies consider¬
ably for different barbiturates. Serious inaccuracies in the quantitative determinations
consequently occur if the composition of the measured barbiturate is unknown. The
colour reaction is not specific for barbiturates. Apparently the reaction is character¬
istic for compounds with -CO-NH-CO- in a heterocyclic ring; the presence of active
hydrogen atoms in the ring seems to prevent the formation of the colour. Furthermore,
sulphonamides give similar reaction.31)
Nuppenau' studied the cobalt colour reaction, especially the influence of
the proportions of the reagents and the stability of the reaction towards hydration.
He used absolute methanol, anhydrous isopropylamine in absolute methanol, dried
cobalt acetate at 105 C and anhyrous chloroform. The measurements were taken at
- 118 -
a weve length of 565 m u. Under his experimental coditions the colour is stable for
3 hours and he got 9 straight lines for the 9 studied barbiturates for quantities between
0-5 mg., when plotting the concentration of the barbiturate against the extinction.
The method of Nuppenau will be applied to our barbiturates.
Application and results of Nuppenau Colorimetric method
Nuppenau' recommended the following method for the quantitative deter¬
mination of the barbiturates by the cobalt amine reaction:
Cobalt acetate reagent:. 0.125 g. dried cobalt acetate (dry the tetrahydrate for 2
hou?s~at"IS§°~c7 dissolved in anhydrous methanol to a volume of 100 ml. The salt
is very hygroscopic and increase in water content occurs on storage over calcium
chloride.
I_sopro_pj_lamine reagent^ 25 ml. anhyrous isopropylamine is measured off at 20 C
andmixe3~wfth"anEy3rous methanol to a volume of 100 ml.
Standard soluticmsof barbituric_aci_d derivatiyesj 0.50 g. of the barbituric acid deri-
vative~is~diss6lve"dInlTnhydYou¥7:hIorofoYm*fda volume of 1000 ml. The barbituric
acid derivatives are recrystallized twice from 50% ethanol and dried to constant
weight at 105° C.
4l*ydjpus_mj3thanol^ is prepared by distillation of methanol over magnesium turningsusfnglodfnVaVa'catalyst.
d£hydj°us_c^orjrform: is prepared by drying with anhydrous sodium sulphate.-The~reagents"and standard solutions are prepared and used at 20° C.
For measurements: the standard solutions are placed in 25 ml. volumetric flasks,51mir-cob~alt~acetati reagent and 5 ml. isopropylamine reagent and anhydrous chloro¬
form added to complete to 25 ml. The blank is prepared from 5 ml. cobalt acetate
reagent and anhydrous chloroform to 25 ml., as the blank is unstable when isopropyl¬amine is added.
31)The reagents recommended by Nuppenau were used in the present research
and his procedure was exactly adopted. Beckman quartz spectrophotmeter (DU with
UV accessories) and 1 cm. cuvette were used to measure the extinction. The
wavelength was adjusted at 565 m u. which corresponds to the maximum absorption
under the previously described conditions. The following are the measurements taken
for Allobarbital, Hexobarbital, Cyclobarbital and Methylphenobarbital.
Table 28
Measurements of extinction for Allobarbital
(Curve 47)
Allobarbital in mg.
directly after preparation
1
0.07
2
0.100
3
0.138
4
0.179
5
0.207
6
0.246
7
0.280
- 119 -
Table 29
Measurements of extinction for Cyclobarbital
(Curve 47)
Cyclobarbital in mg. 1 2 3 4 5
directly after pre¬
paration
after 9 minutes
after 20 minutes
after 60 minutes
0.060
0.060
0.059
0.060
0.104
0.101
0.104
0.104
0.146
0.150
0.150
0.150
0.191
0.194
0.194
0.194
0.239
0.240
0.235
0.240
Table 30
Measurements of extinction for Hexobarbital
(Curves 47 & 47')
Hexobarbital in mg. 1 2 3 4 5
directly after pre¬
paration
after 10 minutes
after 20 minutes
after 1 hour
0.051
0.050
0.050
0.050
0.082
0.083
0.083
0.076
0.105
0.107
0.107
0.105
0.141
0.141
0.140
0.140
0.170
0.168
0.169
0.169
After 3 hours the colour in 1 and 2 mg. experiments completely disappeared
changing to a pale qellow colour. The original colour of the previous solutions was
rose violet increasing in intensity in higher concentrations of Hexobarbital.
Table 31
Measurements of extinction for Methylphenobarbital
(Curve 47)
Methylphenobarbitalin mg.
1 2 3 4 5
directly after pre¬
paration
after 10 minutes
after 20 minutes
after 160 minutes
(2 & 2/3 hours)
0.055
0.047
0.052
0.048
0.078
0.076
0.076
0.078
0.108
0.107
0.106
0.107
0.139
0.136
0.137
0.138
0.168
0.169
0.167
0.168
- 120 -
All the solutions showed a rose violet colour increasing in intensity in higher
concentrations. In 1 mg experiment the colour disappeared completely after 3 hours.
- 121 -
E
0.28
0.24
0.12
0.08
0.04-
• • • • Methylphenobarbital
4 4 & a Hexobarbital
» » * x Cyclobarbital
o o o o Allobarbital
.
\/>
3 5 mg, barbiturate
per 25 ml. reaction mixture at 565 mu
Colorimetric estimation of barbiturates by
Nuppenau's method
E
0.20
0.16
0.12
0.08
_^.—
3 5 mg.
Hexobarbital
per 25 ml. reaction mixture at 565 m>i
Colorimetric estimation of Hexobarbital by
Nuppenau's method
- 122 -
Criticism and summary of the results of the application of Nuppenau's method
1. Nuppenau's method is applicable to the estimation of small amounts of the barbi¬
turates, ranging from 1 -5 mg. per 25 ml. of the reaction mixture.
2. Straightlined curves were obtained by ploting the extinction against the concentration
of the barbiturate e.g. Curve 47.
3. Absolute dehydration of the reagents is of utmost importance. It was shown by31}
Nuppenau' that more than 0.1% of humidity destroys rapidly the colour. In this
study it was also noticed that atmospheric humidity destroys the colour if it is not
guarded against.
4. The disadvantages of the Nuppenau's method are:
a) One should have a good idea of the concentration of the unknown barbiturate.
b) Dilutions should be made within the limited range of 1 - 5 mg. quantities per
25 ml. of the reaction mixture.
c) Fresh standard comparison solutions of the examined barbiturate should be
prepared and curves plotted each time before the estimation is carried out.
It is not possible to give definite standard curves, but a comparison curve
must be worked out the same day it is to be used under the same conditions.
d) The composition of the measured barbiturate should be known otherwise
serious inaccuracies in the quantitative determination occur.
6. The Kjeldahl method
Survey and discussion of the literature
One of the most time-honoured procedures for the determination of the nitrogen
content of organic substances is that due to Kjeldahl .It implies the treatment
of the substance with hot concentrated sulphuric acid whereupon the nitrogen is fixed
as ammonium sulphate. Excess alkali is then added and the ammonia expelled by
distillation is collected in a standard acid. The excess acid is titrated with standard
alkali. Since its invention, the method has undergone some modifications to increase
the severity of the reaction and reduce the digestion ime. Thus, Gunning pro¬
posed the use of potassium sulphate as a means of rainsing the boiling point of the
digestion medium. Folin and Wright used phosphoric sulphuric acid mixture
as a successful digestion medium.
66) J. Kjeldahl, Z.anal.Chem., 22, 366, (1883).
67) J.W. Gunning, Z.anal.Chem.7T8 ,188(1899).
68) O. Forlin and L.E. Wright, J.Biol.Che., 38,461 (1919).
- 123 -
It is often necessary to increase the severity of the reaction through the use of69)
an oxidising agent. Dowell et al. '
suggested the use of potassium permanganate70) 71)
while Koch and McMeekin ' used hydrogen peroxide. Mea-rs and Hussey '
reported the successful use of perchloric acid as an acid to digestion.
At frequent times researchs were directed to find catalysts to produce a further
increase in the velocity of the reaction. In 1885 Wilfarth '
reported on a number
of compounds that could be used as catalysts, mercury was stated to be the most
efficient. The digestion mixture of copper sulphate, potassium sulphate and mercuric
oxide in sulphuric acid has been given official status by the Association of Official
Agricultural Chemists '.74)
Winkler 'proposed collecting the ammonia from the steam distillation in
4% boric acid, rather than the standard acid called for in the original Kjeldahl method.
Boric acid being an extremely weak acid, does not cause a colour change with the
indicators used. However, ammonia is fixid by it and as such can be titrated directly
with acid. This technique obriates the need for more than one standard solution.
The method of Kjeldahl is not, however, applicable for all nitrogen containing
compounds. Thus certain alkaloids and other nitrogen containing organic compounds
will not yield all of their nitrogen to digestion with sulphuric acid.
Application and results obtained with Kjeldahl's method
The procedure used in this research work was that recommended by the United
States Pharmacopoeia XV with very slight modivications. The material was wrapped
in ashless filter paper (Schleicher and Schuell, Feldmeilen, Switzerland) to facilitate
the introduction of the material to the bottom of the Kjeldahl's flask. Another filter
paper of similar weight was used in the blank experiment. The distillation apparatus
used in the present work was that recommended by Parnas (Semi-micro Parnas
apparatus, Jena Glaswork Schott and Gen. Jena).
All rubber used in the apparatus was boiled for 10 minutes in approximately 1 N
sodium hydroxide and thoroughly washed with water before its first use. The steam
generatur of the distillation apparatus was filled with water to which has been added
a few drops of sulphuric acid. Fractions of glass tubes were used to prevent bumping.
69) C.T. Dowell, W.G. Friedemann and D.C. Cochrane, Ind.Eng.Chem., 13,358 (1921).
70) F.C.Koch andT.L. McMeekin, J. Am.Chem.Soc., 46, 2066 (1924).71) B. Mears and R.E. Hussey, J.Ind.Eng.Chem., 13_, T054 (1921).72) H. Wilfarth, Chem.Zentr., 56, 17, 113(1885).73) cited in Organic Analysis III, Interscience Publishers N.Y. page 137 (1956).74) Z.Winkler, Angew.Chem. 26, 231 (1913).
- 124 -
Following the procedure of the American Pharmacopoeia XV which recommends
the use of a boric acid solution to fix the liberated ammonia, Table 32, represents the
typical results thus obtained.
Table 32
Results of group of experiments 1
Barbiturate Percentage of Nitrogen Percentage of
calculated found error
Allobarbital 13.46 9.25 -31.3
Methylpheno¬barbital 11.38 10.00 -12.14
Pentobarbital
sodium 11.28 10.39 - 8.56
Table 32 shows clearly that in the 3 mentioned cases the estimation of the nitrogen
gives much lower results than that expected.
Table 33 represents some typical results for the determination of the nitrogen
content of Allobarbital, Methylphenobarbital and Pentobarbital sodium. In this case,
0.1 N sulphuric acid was used as absorbant for the liberated ammonia, the reaction
mixture in these cases was heated directly after the addition of the reactants.
Table 33
Results of group of experiments 2
Barbiturate Percentage of Nitrogen Percentage of
calculated found error
Allobarbital 13.46 13.75
12.85
11.55
+ 2.15- 4.68- 14.20
Methylenpheno-
barbital
11.38 10.94
10.79
10.62
- 2.87- 5.19- 6.69
Pentobarbital
sodium
11.28 12.91
11.59
10.46
11.85
+ 14.43
+ 2.72- 7.27
+ 5.05
- 125 -
Table 33 shows that when one heats the reaction mixture directly after the
addition of the reactants non concordant results are obtained.
The non concordance of the results (seen in Table 33) may be attributed to the
loss of a part of the nitrogen as its volatile products e. g. Nitrogen oxides, volatile
cyanide under the severe coditions of the reaction.
An interesting remark that was observed in the course of the present study is
the fact that when one leaves the reaction mixture to digest at the room temperature
overnight (14 to 16 hours) then heats for complete reaction; concordant results are
obtainable. In these cases, group of experiments 3, the heating time to effect complete
digestion was found to be rather shorter than if one heats directly (2-3 hours instead of
4-6 hours).
Table 34 illustrates some typical results for Allobarbital, Cyclobarbital, Hexo-
barbital, Methylphenobarbital and Pentobarbital.
Table 34
Results of group of experiments 3
Barbiturate Percentage of Nitrogen Percentage of
calculated found error
Allobarbital 13.46 13.43
13.52
13.44
-0.22
+ 0.44
+ 0.14
Cyclobarbital 11.86 11.62
11.58
11.73
11.58
- 2.02
-2.36- 1.09- 2.35
Hexobarbital 11.86 11.65
11.86
11.51
- 1.77
0.00- 2.95
Methylpheno¬barbital
11.38 11.06
11.32
11.02
11.26
-2.8- 0.53- 3.17- 1.05
Pentobarbital
sodium
11.28 11.32
11.32
11.29
+ 0.35
+ 0.35
+ 0.08
- 126 -
Criticism and summary of the results of the application of Kjeldahl's
method to barbiturates
1. Various conditions of the Kjeldahl reaction were discussed; the best results were
obtained when the reaction mixture was left to react on cold overnight and then heated
(see Table 34).
2. The percentage of error was as follows: (Table 34)
Allobarbital, it ranged from -0.22 to +0.44%
Cyclobarbital, it ranged from -1.09 to -2.35%
Hexobarbital, it ranged from - 1.77 to -2.95%
Methylphenobarbital, it ranged from - 0.53 to - 3.17%
Pentobarbital sodium, it ranged from +0.08 to - + 0.35%
3. The Kjeldahl's method is one of the most reliable methods for the estimation of the
barbiturates. Several assays should be carried out until concordant results are obtained.
- 127 -
Summary and Conclusions
In our experimental work we studied the quantitative determination of the barbi¬
turic acid derivatives by the following methods.
1. Acidimetric methods
a) Titrations in aqueous media
Due to the insolubility of the barbituric acid derivatives in water, alcohol water
mixtures were used as solvents. However, we found that not all the barbituric acid
derivatives are freely soluble in the described water-alcohol mixtures. Furthermore,
the presence of alcohol in the titration medium effects the variation of the colour and
the colour change of the indicator, especially that of thymolphthalein indicator. Accu¬
rate results could be obtained in the acidimetric method using carbonate free 0.1 N
alkali, carbon dioxide free alcohol and water mixture, thymolphthalein as indicator
and a copper dichromate solution as a comparison solution. A blank should also be
carried out to find the correction factor, using the same reagents without the acid.
Allobarbital, Cyclobarbital and Hexobarbital (the latter being used in smaller quantities),
could be successfully determined quantitatively by this method. Methylphenobarbital
could not be estimated by this method being sparingly soluble in the mixture of alcohol
and water used.
b) Non-aqueous titrations
Non-aqueous titrations were carried out using dimethylformamide, pyridine and
chloroform as solvents. Sodium methoxide, lithium methoxide and potassium hydroxide
in methanol were used as titrants. Thymol blue as indicator found application in all
the methods performed. The change of the colour of thymol blue indicator is from
yellow-green-blue; this green stage adds to the convenience of its use. Titration to a
clear blue colour is the usual real end point. Both visual and potentiometric titrations
were carried out. In the potentiometric titrations a combination of an antimony,
electrode as indicator electrode, a glass electrode as a reference electrode and the
Metrohm E 157 apparatus were used (Herisau, Schweiz).
The following methods were applied to our barbiturates:
o<) Vespe and Fritz method: In this method 'dimethylformamide as a solvent,
sodium methoxide in methanol and benzene as a titrant, and thymol blue as indicator.
Both the visual and potentiometric estimations gave reliable results in all the barbitu¬
rates examined i.e. Allobarbital, Cyclobarbital, Hexobarbital and Methylphenobarbital.
- 128 -
ft) Heiz's method:
461
H e i z' used pyridin as solvent, sodium methoxide in methanol and benzene as
titrant, and thymol blue, phenolphthalein or thymolphthalein as indicator. We could
estimate only Allobarbital and Methylphenobarbital with reliable results, from the
four barbiturates examined by the visual titration; even in the case of Methylpheno¬
barbital the end point was not very sharp. The results of the potentiometric titrations
were not satisfactory, however Hexobarbital and Methylphenobarbital showed a good
potential jump at the end points.
y)Chatten's method?
49)
In this method ' chloroform as solvent, potassium hydroxide in methanol as titrant
and thymol blue as indicator were used. We obtained concordant results in all the
examinded barbiturates; the end points in all the assays were sharp.
6) Ryan et al. method:
Ryan et al. ' used dimethylformamide as solvent, lithium methoxide as titrant, and
thymol blue as indicator. We obtained concordant results in the assays of Allobarbital,
Cyclobarbital, Hexobarbital and Methylphenobarbital carried out visually. In regard
to the potentiometric titrations a slight potential jump at the end point was obtained
in the assays of Allobarbital, Hexobarbital and Methylphenobarbital. In case of Cyclo¬
barbital assays no potential jump at the end points was observed.
2. Argentometric methods
When the barbituric acid derivatives, dissolved in alkaline solutions, are titrated
against silver nitrate solution, the corresponding silver barbiturates are formed which
are soluble in the alkaline solution and variably soluble in water. An excess of the sil¬
ver will precipitate the oxide, borate or carbonate depending on the kind of alkali used
e. g. sodium hydroxide, borax or sodium carbonate respectively.
a)Budde's method:
The barbituric acid derivatives are dissolved in sodium carbonate solution and
titrated against silver nitrate solution till a clear turbidity occurs for a certain time.
Budde described the non applicability of his method to the estimation of the nitrogen
methylated derivatives. We carried out the visual turbidimetric method and the
potentiometric method (using a silver and calomel electrode combination and the
Metrohm apparatus. The non applicability of this method to the nitrogen methylated
- 129 -
derivatives was proved for Methylphenobarbital and Hexobarbital. The method of
Budde gave concordant satisfactory results only in the case of Cyclobarbital. For
Pentobarbital sodium we recorded a big error amounting to + 11.64%.
b) Bodin's method 57^:24)
It is based on the same reaction principle of Budde ' and is carried out
potentiometrically. Titration was continued till the potential compared with that of
a standard blank is reached. We applied this method to AUobarbital and the effect
of using different concentrations of sodium carbonate was also studied. It was found
that the 1% sodium carbonate solution was the best in detecting the end point by the
standard potential method. In all the solutions studied the estimation of the end point
by the differential titration curves were low and not to be recommended; since the
potential jump in all cases were low. The turbidimetric determination of the end
point gave good results in the 2% sodium carbonate solution. The method as recommen¬
ded by Bodin was found to be inapplicable and the percentage of error was very high.
c)Mangouri and Mil ad method ':
These authors dissolved the barbituric acid derivatives in sodium acetate
solution an ammonia, the excess of ammonia being expelled by boiling. A known
excess of 0.1 N silver nitrate was added and the solution was boiled again cooled
and filtered. The excess of silver nitrate was determined in the filtrate. We obtained
no concordant results in any of the assayed acids. Low percentages were obtained in
all the assayed acids.
d) Chavanne and Marie method ':
In this method, the barbituric acid derivative is dissolved in a mixture of
ethanolic potash, boric acid and water. This solution is titrated against silver nitrate
using potassium chromate as indicator together with a comparison solution containing
potassium chromate we carried out visually and potentiometrically titrations. Our
results of the assays of allobarbital visually gave more or less concordant results
and an error of -2.4%. In the other 4 barbiturates studied bigger error was recorded.
Only for Allobarbital a moderate potential jump was obtained; in the other cases very
faint potential jumps were noticed near the end points.
- 130 -
3. Mercurimetric method (Pedley's 'method)
Pedley recommended the use of mercuric perchlorate as a precipitant of the
barbiturate. The excess of mercuric perchlorate in the filtrate being titrated against
ammonium thiocyanate. For the insoluble barbiturates, he dissolved them in sodium
hydroxide, neutralized with acetic acid and then mercuric perchlorate was added.
The concentration of the barbiturate should be maintained between 2. 5 - 6. 25 milli-
molecules per liter. We obtained only reliable concordant results in case of Methyl-
phenobarbital. In all the other barbiturates used the results showed a great deviation.
4. Bromometric method
The barbituric acid derivatives possessing unsaturated radicals could be deter¬
mined by treating with a known excess of potassium bromate bromide solution in
presence of dilute acid. By adding potassium iodidie and titrating the liberated iodine
against sodium thiosulphate the used bromine could be calculated. We could theoreti¬
cally determine Allobarbital and Hexobarbital using this method. Cyclobarbital gave
us low results with an error of -5.27%.
31)5. Colorimetric method (Nuppenau ')
31}The colorimetric estimation of the barbiturates (by Nuppenau ') depends on the
development of a red violet colour with cobalt acetate and isopropylamine as a base,
in non aqueous media. Other cobalt salts e.g. cobalt nitrate, and other bases e.g.
ammonia, diethylamine, piperidine etc. were also used by different authors. The
stability of the colour is very poor in methanol solution but much better in solution
containing large amounts of chloroform. A solvent consisting of 12.5% (v/v) methanol
in chloroform is preferred. Moisture destroys the colour. Our Curve 47, showed
straight lines and the quantitative estimation of amounts ranging from 1-5 mg. per
25 ml. reaction mixture of the corresponding barbiturate could be estimated from
these curves. However, fresh standard comparison solutions of the examined barbi¬
turate should be prepared and curves plotted each time before the estimation is carried
out. It impossible to give definite standard curves, but a comparison curve should be
worked out the same day it is to be used under the same conditions. The composition
i. e. the chemical formula, of the measured barbiturate should be known otherwise
serious inaccuracies in the quantitative determination occur.
- 131 -
6. Kjeldahl's method
We carried out this method under different experimental conditions. We found
that when the reaction mixture was left to digest on cold overnight and then heated
before the addition of the alkali sodium hydroxide; concordant excellent results were
obtained for all the studied barbiturate.
7. Spectrophotometry method
The spectrophotometric method is best carried out in buffer solutions of pH 9.5.
We studied the absorption spectra of Allobarbital, Cyclobarbital, Hexobarbital, Pen¬
tobarbital sodium and Methylphenobarbital in buffer solution of pH 9. 5 and in 0.1 N
sodium hydroxide. It is difficult by absorption methods to distingiush between indivi¬
dual barbituric acid derivatives. The peak absorption of the studied barbiturates being
from 238 - 244 m ju., and the maximum molecular extinction coefficients being generally
within the limits 8500 - 10 000. The peak absorption of the studied barbiturates in 0.1 N
sodium hydroxide being from 242 - 252 m u. and the maximum molecular extinction
coefficients being within the limits 7000 - 9490. From curves 9, 10, 11, 12, 13, the
quantitative estimation of the corresponding barbituric acid derivative could be carried
out.
Proposal of the recommended methods
After our experience we recommend the following methods for the quantitative
determination of barbiturates for pharmocopoeial requirements to be included.
Although, the nitrogen determination methods of the barbiturates gave us reliable
results these methods are rather complicated, time consuming, require special mani¬
pulation and an expert to carry out the analysis, So they could not be recommended as
analytical methods to be carried out earily in pharmacies, simultaneously to be in¬
cluded as reliable methods for pharmacopoeial use.
The non aqueous method proposed by Chatten gave good reliable results in all
the barbiturates examined, when we used double the prescribed weights of the barbi¬
turates. As one of the best non aqueous titration methods which gave good results also
51)we propose the adoption of the method recommended by Ryan et al. visually as it
has several advantages over the other methods namely, its minimum amounts of solvent,
suitable indicator and the best titrant i.e. lithium methoxide.
In mixtures of barbiturates, the unsaturated radicals in barbiturates could be
estimated successfully bromometrically e.g. Allobarbital and Hexobarbital.
- 132 -
Zusammenfassung
In unserem experimentellen Teil dieser Arbeit hatten wir die quantitativen Be-
stimmungen der Barbitursaurederivate untersucht. Vergleiche zwischen verschiede-
nen Methoden von Analysen ergaben die folgenden Resultate:
1. Azidimetrische Methode
a) Wasseriges Milieu
Wegen der Unloslichkeit der Barbitursaurederivate in Wasser wurde eine Mi¬
schung von Weingeist und Wasser als Losungsmittel verwendet. Im allgemeinen sind
nicht alle Barbitursaurederivate in der empfohlenen Wasser-Weingeistmischung 18s-
lich. Die Anwesenheit von Weingeist im Titrationsmilieu verandert die Farbe des In-
dikators, besonders wenn Thymolphthalein gebraucht wurde. Nach unseren Erfahrun-
gen und in Uebereinstimmung mit anderen Autoren sind gute Resultate mit der wasse-
rigen, azidimetrischen Methode erreichbar, wenn man kohlensaurefreie Lauge, koh-
lensSurefreie Weingeist- und Wassermischung, Thymolphthalein und Kupfersulfat-
und Kaliumdichromat-Losung als Vergleichslosung beniitzt, was auch Poethke und
Horn '
empfehlen.
Um den Korrekturfaktor zu finden, sollten Versuche zu Vergleichszwecken aus-
gefuhrt werden. Allobarbital, Cyclobarbital und Hexobarbital (das letztere in kleinen
Mengen) konnen mit der von National Formulary 1955 empfohlenen Methode erfolg-
reich quantitativ bestimmt werden. Methylphenobarbital ist unter diesen Bedingungen
wegen der UnlcSslichkeit in der gebrauchten Mischung von Wasser und Weingeist nicht
zu bestimmen. Im allgemeinen ist diese Methode schwieriger durchzufiihren als die
Methoden in wasserfreiem Milieu.
b) Wasserfreies Milieu
Bei der wasserfreien Titration verwenden wir als L6sungsmittel Dimethylfor-
mamid, Pyridin und Chloroform. Als Titranten beniitzten wir Natriummethylat, Li-
thiummethylat und Kaliumhydroxyd. Thymolblau wurde bei alien angewendeten Metho¬
den als Indikator empfohlen. Der Farbumschlag des Indikators andert sich von gelb
fiber griin nach blau; das griine Umschlaggebiet gibt zusatzliche Vorteile fur seine
Verwendbarkeit. Der wirkliche Endpunkt ist erreicht, wenn eine klare blaue Farbe
auftritt. Sowohl visuelle als auch potentiometrische Titrationen wurden ausgefiihrt.
- 133 -
Die folgenden Methoden wurden auf unsere Barbitursaurederivate angewendet:
\ ir j T7 <48).
ex) V e s p e und Fritz :
In dieser Methode wurden Dimethylformamid als Losungsmittel, Natriummethylat
in Methanol und Benzol als Titrant und Thymolblau als Indikator verwendet. Sowohl
visuelle wie potentiometrische Bestimmungen ergaben gute Resultate bei alien unter-
suchten Barbitursaurederivaten, namlich AUobarbital, Cyclobarbital, Hexobarbital
und Methylphenobarbital.
P) Heiz46):
Dieser Autor hat Pyridin als Losungsmittel, Natriummethylat in Methanol und
Benzol, ferner Thymolblau, Phenolphthalein und Thymolphthalein als Indikatoren ver¬
wendet. Nach unseren Erfahrungen bei den visuellen Titrationen gaben nur AUobarbi¬
tal und Methylphenobarbital gute Resultate; die letztgenannte Substanz zeigte aller-
dings keinen scharfen Endpunkt. Die Resultate der potentiometrischen Titrationen
waren nicht befriedigend, obschon Hexobarbital und Methylphenobarbital einen guten
Potentialsprung aufwiesen.
y) Chatten4^:
In dieser Methode wurde Chlorform als Losungsmittel, Kaliumhydroxyd in Me¬
thanol als Titrant und Thymolblau als Indikator verwendet. Wir erhielten iiberein-
stimmende Resultate mit all unseren gepriiften Barbitursaurederivaten. Die Endpunk-
te aller Versuche waren scharf genug.
51)6 ) Ryan und Mitarbeiter
Diese Autoren verwendeten Dimethylformamid als Losungsmittel, Lithiumme-
thylat als Titrant und Thymolblau als Indikator. Uebereinstimmende Resultate erhiel¬
ten wir in den viesuellen Versuchen mit AUobarbital, Cyclobarbital, Hexobarbital und
Methylphenobarbital. In Bezug auf die potentiometrischen Titrationen wurde ein mittle-
rer Potentialsprung am Endpunkt erreicht im Falle von AUobarbital, Hexobarbital und
Methylphenobarbital. Bei Cyclobarbital wurde kein Potentialsprung am Endpunkt be-
obachtet.
- 134 -
2) Argentometrische Methode
Wenn die Barbitursaurederivate, aufgelOst inAlkalien, mitSilbernitrat titriert
werden, entstehen die entsprechenden Silberbarbiturate, welche in Alkalien ISslich
sind, aber im allgemeinen unlSslich in Wasser. Ein Ueberschuss von Silberionen er-
gibt einen Niederschlag von Silberoxyd, Silberborat Oder Silberkarbonat, abhangig
von verwendeten Alkalien.
a ) Budde24^:
Die Barbitursaurederivate werden in Natriumkarbonat-LQsung gelSst und mit
Silbernitrat fur eine gewisse Zeit titriert, bis ein Niederschlag entsteht. Im Falle
von N-Methyl-Derivaten wird die Anwendung dieser Methode von Budde nicht empfoh-
len. In unserem Verfahren hatten wir die visuelle turbidimetrische Methode und die
potentiometrische Methode (unter Verwendung einer Silber- und Kalomel-Elektroden-
Kombination mit dem Methrohm-Apparat) angewendet. In Uebereinstimmung mit
Buddes Angaben stellten wir ebenfalls fest, dass diese Methode nicht verwendbar ist
fiir die N-Methyl-Derivate, z. B. Methylphenobarbital und Hexobarbital. Im Falle von
Cyclobarbital wurden befriedigende Resultate erzielt. Fiir Pentobarbital-Natrium fan-
den wir Fehler bis zu + 11, 64%.
b ) Bodin57):
Diese Methode basiert auf dem gleichen Prinzip wie diejenige Buddes und wurde
nur potentiometrisch ausgefiihrt. Die Titration wurde fortgesetzt, bis das Potential
demjenigen einer VergleichslSsung entsprach. Wir gebrauchten diese Methode fiir
Allobarbital. Weiter wurde die Beeinflussung durch verschiedener Konzentrationen
von Natriumkarbonat untersucht. Es wurde festgestellt, dass eine 1%-ige Natrium-
karbonat-Losung am geeignetsten ist um den Endpunkt bei der Standard-Potential-
Methode zu erreichen. In alien von uns untersuchten LSsungen war der Wert des End-
punktes unter Verwendungvon Differential-Titrationskurven zu niedrig. Die turbidime¬
trische Bestimmung des Endpunktes ergab gute Resultate unter Verwendung von 2fc-
iger Natriumkarbonat-Losung. Die Methode, welche von Bodin empfohlen wurde, ist
unbrauchbar, da die Fehler zu hoch waren.
53)
c) Mangouri und Mi lad ';
Diese Autoren lfisten die Barbitursaure-Derivate in Natriumacetat- und Ammo-
niak-LOsung auf. Der iiberschussige Ammoniak wurde durch Kochen beseitigt. Ein
bekannter Ueberschuss von 0,1 n Silbernitrat wurde der Losung beigefiigt und noch-
- 135 -
mals gekocht, abgekuhlt und filtriert. Der Ueberschuss von Silbernitrat im Filtrat
wurde bestimmt.
Wir erhielten bei alien Barbitursauren keine iibereinstimmenden Resultate, die
Werte waren durchwegs zu niedrig.
55)
d) Chavanne und Marie ':
Bei dieser Methode wurden die Barbitursaurederivate in weingeistiger Kalilauge
gelBst, dann fiigten wir eine Mischung von Borsaure und Wasser zu. Diese Losung
wurde zusammen mit Silbernitrat unter Verwendung von Kaliumkarbonat als Indika-
tor mit einer Vergleichslosung titriert. Beide, das visuelle und potentiometrische
Verfahren, wurden angewendet. Die Resultate der visuellen AUobarbital-Bestimmung
waren mehr oder weniger iibereinstimmend mit Fehlern von -2,4%. Bei den anderen
4 Barbitursaurederivaten wurden grBssere Fehler festgestellt. Nur fiir Allobarbital
wurde ein brauchbarer Potentialsprung erzielt, in den anderen Fallen wurden sehr
geringe Potentialspriinge in der Nahe des Endpunktes beobachtet.
3) Mercurimetrische Methode P e d 1 e y'
Pedley empfahl die Anwendung von Quecksilberperchlorat als Fallungsmittel fiir
die Barbiturate. Der Ueberschuss von Quecksilberperchlorat im Filtrat wurde gegen
Ammoniumthiocyanat titriert. Fiir die unlBslichen Barbitursaurederivate empfahl er
Natronlauge und neutralisierte mit Essigsaure unter Zusatz von Quecksilberperchlo¬
rat, wie beim gewohnlichen Verfahren. Die Konzentration der Barbitursaurederivate
sollte zwischen 2, 5 - 6, 25 Millimol per Liter gehalten werden. Wir fanden gute Re¬
sultate nur im Falle von Methylphenobarbital. Bei alien anderen Barbitursaurederi¬
vaten zeigten die Analysen grosse Abweichungen.
4) Bromometrische Methode
Bei ungesattigten Barbitursaurederivaten wurde, in Anwesenheit von verdiinnter
Saure, eine Ueberschussmenge von Kalium-Bromat-Bromid zugefiigt. Unter Beimi-
schung von Kaliumjodid wurde das entstehende Jod gegen Natriumthiosulfat titriert.
Allobarbital und Hexobarbital ergaben sehr gute Ergebnisse; Cyclobarbital da-
gegen lieferte zu tiefe Resultate mit einer Abweichung von - 5, 27%.
- 136 -
5) Kolorimetrische Methode Nuppenau '
Diese kolorimetrische Bestimmungsmethode der Barbitursaurederivate beruht
auf der Entwicklung einer rotvioletten Farbe mit Kobaltacetat und Isopropylamin im
wasserireien Milieu. Andere Kobaltsalze, z.B. Kobaltnitrat und andere Basen, z.B.
Ammoniak, Diathylamin, Piperidin u. s. w. wurden ebenfalls von verschiedenen Au-
toren empfohlen. Die Stabilitat der Farben ist sehr gering in Methanol allein, viel
besser dagegen in Chloroform-Methanol-Mischung (12, 5 V/V) Methanol in Chloro¬
form ist vorzuziehen). Feuchtigkeit zerstort die Farben. Unsere Kurven zeigten ge-
rade Linien, aus welchen die Moglichkeit der quantitativen Bestimmungen von Men-
gen von 0-5 mg in 25 ml Reaktionsmischung hervorgeht. Es mtissen immer neue
Vergleichslosungen fiir die zu untersuchenden Barbitursaurederivate hergestellt und
neue Eichkurven aufgenommen werden.
6) Stickstoffbestimmung nach K j eldahl
Kjeldahls Verfahren wurde unter verschiedenen Bedingungen durchgefiihrt. Wir
fanden, dass, wenn die Reaktionsmischung iiber Nacht kalt stehengelassen und vor
dem Zufugen der Lauge wieder erhitzt wurde, gut iibereinstimmende Resultate fiir
alle Barbitursaurederivate zu erreichen sind.
7) Spectrophotometrische Methode
Diese Methode wird am besten in PufferlSsungen von pH 9, 5 ausgefiihrt. Wir
studierten die Absorptionsspektren bei Allobarbital, Cyclobarbital, Hexobarbital,
Pentobarbital und Methylphenobarbital in Pufferlosung von pH 9, 5 und in 0,1 n Na-
trium-Hydroxyd. Es ist schwer, mit Hilfe dieser Methode die einzelnen Barbitur¬
saurederivate zu unterscheiden; die spitzen Absorptionskurven der untersuchten Bar¬
bitursaurederivate liegen alle zwischen 238 - 244 mu und der maximale Molekular-
Extinctionskoeffizient liegt im allgemeinen zwischen 8500 - 10 000. In Natronlauge
z.B. liegt das Maximum der Absorptionskurve zwischen 242 - 252 mu und die maxi-
malen Molekular-Extinctionskoeffizienten zwischen 7000 - 9490.
- 137 -
Vorschlage von Methoden als Pharmakopoe-Verfahren
Obschon die Stickstoffbestimmungsmethoden der BarbitursSurederivate zuver-
lassige Resultate brachten, sind diese Methoden eher kompliziert, zeitraubend und
eriordern spezielle Manipulationen sowie eine langere Einarbeit. Aus diesem Grunde
konnen sie nicht als Pharmakopoen-Methoden eingefuhrt werden.
Die von Chatten vorgeschlagene Methode ergab sehr gute Resultate bei alien
untersuchten Barbitursaurederivaten, wenn die von ihm vorgeschlagenen Einwaagen
der Barbitursaurederivate verdoppelt wurden. Als eine der besten wasserfreien Ti-
trationen, welcheuns sehr gute visuelle Resultate ergab, schlagen wir die Methode
von Ryan undMitarbeiternvor. Diese Methode benotigt ein Minimum an LSsungsmittel
und den besten Titranten, das Lithiummethylat.
In den Mischungen von Barbituraten konnten die ungesattigten Barbitursaurederi¬
vate z.B. AUobarbital und Hexobarbital bromometrisch erfolgreich neben den iibrigen
Barbituraten bestimmt werden.
- 138 -
V BIBLIOGRAPHY
1) Pharmacopoea Helvetica 1933, Supplementum primum, secundum and tertium
(Ph.Helv. VSuppl. I, n& m).
2) The British Pharmacopoeia 1953 and 1958 (Brit. Ph. 1953 & 1958 or B.P. 1953
& B.P. 1958).
3) The Egyptian Pharmacopoeia (E.P. 1953).
4) The Pharmacopoeia of the United States of America (USP XV).
5) Pharmacopoea Danica 1948 and Addendum 1954 (Ph.Dan. IX & Add. 1954).
6) Codex Medicamentarius Gallicus 1949 (Pharmacopge Francaise 1949) and
Supplement 1954 (Codex Gall. 7 Suppl. 1954).
7) Deutsches Arzneibuch 1926 (DAB 6).
8) Nederlandsche Pharmacopee 1958 (Nederl. Ph. VI).
9) The British Pharmaceutical Codex 1954 (Brit. Pharm. Codex 1954) or B.P.C. 1954.
10) X. Perlia, BeitrSge zum Nachweis therapeutisch wichtiger Barbiturate, Diss.
ETH Zurich 1953.
11) A.I. Biggs, J.Chem.Soc. 2485 (1956).
12)H.Erlenmeyer et al., Helv.Chim. Acta, 19, 354(1936).
13) J.K.Wood, J.Chem.Soc. Transactions, 89, n, 18 31(1906).
14) W. Poethke and D. Horn, Archiv der Pharmazie, 287, 487 (1954).
15) G.H.A. Clowes, A.K. Keltch and M.E. Krahl, J.Pharmacology and Exper.Therapy, 68, 312 (1940).
16) W.F. Elvidge, Quart. J. Pharm.Pharmacol., 14, 134(1941).
17) W.F. Elvidge, Quart.J.Pharm.Pharmacol., 13, 219(1940).
18) L.R. Goldbaum, Anal.Chem., 24, 1604 (1952).
19) R.E. Stuckey, Quart. J. Pharm. Pharmacol., 14, 217(1941).
20) R.E. Stuckey, Quart.J. Pharm. Pharmacol., 15, 370(1942).
21) L.N. Mattson, J.Amer.Pharm.Ass. (Sci.Ed.) 43, 22 (1954).
22) E. Schulek and P. Rozsa, Z. Anal.Chem., 40, 415(1938).
23) B. Danielson, Svensk farm.T., 55, 125 (1951).
24) H. Budde, Dtsch. Apoth. Ztg., 49, 295 (1934).
25) C. Stainier etal., Ann.pharm.franc, 14, 384 and 476 (1956).
26) E. Pedley, J. Pharm. Pharmacol., ^, 39 (1950).
27) J.J.L. Zwikker, Pharm.Wbl., 68, 975 (1931).
28) W. Parri, Boll. chim.farm., 36, 401 (1924).
29) G.Robles, Bol.Soc.Quim. Peru, 15, 70 (1949).
30)H.Bagesgaard-Rasmussen and B. Jerslev, Dansk T.Farm., 25, 29(1951).
31)H.Nuppenau, Dansk T.Farm., 28, 194(1954).
32) E. Flotow, Pharm. Zhalle, 88, 198 (1949).
- 139 -
H.Gomahr and H. Kr esbach, Scientia pharm., 19, 154(1951).
G.E. Turfitt, Quart.J. Pharm. Pharmacol., 21, 1(1948).
H. Palme, Pharmaz. Ztg., 75, 1347 (1930).
S. Babitsch, Chem.Zbl. 3146 (1936) H.
V.Gervay, Pharmaz. Zentralhalle, 83, 494(1942).
Ch. Morin, Chem.Zbl., 3948 (1935) II.
Quart.J.Pharm. & Pharmacol., 8_, 691 (1935).
L.Hamilton and S.Simpson "Calculations of Analytical Chemistry" Inter¬
national Chemical Series 5th. Edition Mc Graw-Hill Book Company (1954)p. 176.
W. Horsch, Pharmazie, 12, 122, 212(1957).
I.M. Kolthoff and H.A. Laitinen, "pH and Electro Titrations", 2nd Edition,John Wiley & sons, p. 30, (1949).
C.W. Pifer et al., J. Amer.Pharm. Ass., Sci.Edit., 42, 509(1953).
L.P. Hammet, Physical Organic Chemistry, McGraw Hill Book Co., New York
1940, page 261.
P.C. Markunas and J.A. Riddick, Anal.Chem., 23, 337(1951).
J.S. Fritz, Anal.Chem. 23, 589 (1951).
R. Heiz, Dansk Tidskr. Farm., 26, 69(1952).
J. Berger, Dansk Tidskr. Farm., J27, 53(1953).
V. Vespe and J.S. Fritz, J. Amer. Pharm.Assoc. 4L, 197(1952).
I.G. Chatten, J. Pharm. and Pharmacol., 8_, 504(1956).
C.J.Swartz andN.E.Foss, J. Amer. Pharm. Ass., Sci.Ed., 44, 217(1955).
J.C.Ryan et al., J. Amer. Pharm. Ass., Sci.Edition, 43, 656 (1954).
G. Schill, Chem.Abstr., 6022-8 (1946).
H.A. Mangouri andL.Milad, Quart.J. Pharm. & Pharmacol., 2£, 109(1947).
K.Kalinowski, Chem. Zentr., _I, 2391(1936).
P. Chavanne and H.Marie, J.Pharm. & Pharmacol., _5_, 1045(1953); Ann.
pharm.franc, 11, 91 (1953).
G. Vastagh and E.Scaboles, Arzneim.Forsch., 8_, 355(1958).
J.I. Bodin, J. Amer. Pharm. Ass. Sci.Ed., 45, 185(1956).
A.M. Mattocks and E.C. Voshall, J. Am. Pharm. Ass. Sci. Ed., 49, 28 (1950).
J.I. Bodin and A. Taub, J.Am. Pharm. Ass.Sci.Ed., 44, 296(1955).
B.Danielson, J.Pharm. & Pharmacol., _5_, 435 (1951).
H.Brauniger and G. Gor gwar dt, Pharmaz. Zhalle, 93, 266 (1954) and 93,299 (1954).
~~
W. Parri, Boll.Chim.Farm., 36_, 401 (1924).
A.S. Ljunberg, Collect.Pharm.Suecica (1952).
H. Griffon and R. Le Breton, Ann. pharm. Franc., 5_, 393(1947).
- 140 -
65) T. Koppanyi, W. Murphy and S. Krop, Arch. int. Pharm.Thgr., 46,76(1933).
66) J.Kjeldahl, Z.anal.Chem., 22, 366, (1883).
67) J.W. Gunning, Z.anal.Chem., 28, 188(1899).
68) O. Forlin and L.E. Wright, J.Biol.Che., 38, 461 (1919).
69) C.T. Dowell, W.G. Friedemann andD.C. Cochrane, Ind.Eng.Chem.,13_, 358 (1921).
70) F.C. Koch andT.L. McMeekin, J.Am.Chem.Soc., 4£, 2066 (1924).
71) B. Mear s and R.E. Hussey ,J. Ind.Eng.Chem., 13, 1054 (1921).
72) H. Wilfarth, Chem.Zentr., 56, 17, 113(1885).
73) Organic Analysis III, Interscience Publishers, N. Y. page 137 (1956).
74) Z.Winkler, Angew.Chem., 26, 231 (1913).
75) Bentley and Driver's Text-Book of Pharmaceutical Chemistry, Fourth Edition
page 249 (1946), Geoffrey CumberlegeOxford University Press. London.
Curriculum vitae
On March 15th 1926 I was born in Shoubra Cairo, Egypt, as the third son of
the late Saleh Adel Bey and the late Naima (Radwan Fahmy). In Heliopolis, Cairo, I
attended the Primary School and I obtained the Primary School Certificate 1938. From
1938 to 1943 I was a student in Farouk El-Awal Special School where I obtained the
first grade Secondary School Certificate (El-Thakaffa) 1942 and finally in 1943 the
second grade of Secondary School Certificate (El-Tawgihia-Matura). I began the
University studies at the end of 1943 in the faculty of Science, Cairo University, as
the preliminary Natural Science year. Then I passed 3 years study in the School of
Pharmacy, Faculty of Medicine, Cairo, and practised work in a private pharmacy for
800 hours in vacations. In November 1947, I obtained the Bachelor's degree of
Pharmacy and Pharmaceutical Chemistry with grade "good". In February 1948, I
joined the staff of the faculty of Pharmacy as a demonstrator in the practical Pharmacy
and Dispensing Dept., under the direction of Mr. Brunskill. At the end of 1949, I
visited the 4 semester courses of Bacteriology under the supervision of Professor Dr.
M. Abdel Hamid Gohar . After this I worked under the practical supervision of
Assistant Professor Dr. M.M. Mohammed El-Mekkawy and Professor Gohar for the
Master's Degree of Pharmacy, covering the subject of "A Search for Antibiotic
producers from Streptomyces in Egyptian soil". In November 1954, I obtained the
Master's degree with grade "good" from a counsel of Professor Dr. Gohar, Fahmy
and Abdel-Shaffy from the faculty of Medicine, Cairo University. At the end of 1955,
I had the permission from Professor Dr. A. Abdel Rahman to make further studies
in the field of Pharmaceutics under the direction of Professor Dr. J. Biichi in the
Swiss Federal Institute of Technology, Zurich, where I carried out this thesis of "The
quantitative estimation of the barbituric acid derivatives".