Citation
The stability of solutions of phenobarbital sodium

Material Information

Title:
The stability of solutions of phenobarbital sodium
Creator:
Jatul, Bernard Benedict, 1918-
Publication Date:
Language:
English
Physical Description:
1 v. : ; 28 cm.

Subjects

Subjects / Keywords:
Phenobarbital sodium ( mesh )
Pharmacy thesis Ph. D
Dissertations, Academic -- Pharmacy -- UF

Notes

Thesis:
Thesis - University of Florida.
Bibliography:
Bibliography: leaves 157-163.
General Note:
Manuscript copy.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
025061643 ( ALEPH )
20142902 ( OCLC )

Downloads

This item has the following downloads:


Full Text
THE STABILITY OF SOLUTIONS OF PHENOBARBITAL SODIUM
By
BERNARD B. JATUL
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
ThE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILMFNT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
February, 1943




AC KNWLEDMET
The author gratefully acknowledges his indebtedness to Dr.
William J. Husa for his sagacious guidance and gracious assistanee during the course of the present investigation. Thanks are also extended to all those who have aided in the conduct of this work.
ii




TABLE OF COrTENTS
page
ACn0;WLPrJGM1MT .... ........ .......... ...... .. ii
IIe RSETILX. OFT 1E LITERATURE ......o................. 2
A. DFTPRIORATION AND) STA3ILIZATIO)N OF 1ARBITURITTP SOLUTlIONS 2
1. Nature of' the D.eterioration ............. 2
2. Factors in the Deterioration ........... 15
as Hlydrogen Ion Concentration ...o..........o.......o 15
be Temperature ........ ... ....... .. ...... 19 i
3. Stabilization of Barbiturate Solutions ....... 25
a* Adjustment of Hydrogen Ion Concentration ....... 25
be Amines as Stabilizing Agents .......... 25
as Amides as Stabilizing Agents .......... 29
d. Use of Urethane in Conjunction with Other Agents 30
e. Other Nitrogen Compounds as Stabilizing Agents 33
f. Glycols as Stabilizing Agents ................... 54
g. Miscellaneous Stabilizing Agents ........ 38
B. METHODS OF ASSAYING BARBITURIC ACID DERIVATIVES ..... 38
It Procedures Employing Immiscible Solvents ...... 38
a. The U. S. P. XI and U. S. P. XII Methods .o..... 38
b. Method of Rotondaro ............... 9
o. Method of van Itallie and Steenhauer .. ......... 42
d. Method of Glycart ................ 42
2. Procedures Employing Indicators ........... 44
3. Conductimetric Titration ...... ...... 45
4. Procedures in Which Undissociated Barbiturates
Are Formed *.......................... 45
a., Mercury Salts .. ................ 45
be* Silver Salts ... ........... ... 48
59* Colorimetric Methods 5......... .. ... 2
a. Method or Zwikker ...............* 52
b. Method of Oettel ................... 52
c. Method of Koppanyi, et &1* .......... 53
iii




III. EXPERIMENTAL PART .......... ...................... ...... 54
A. SCOPE OF INVESTIGATION ..... ..................... ..... 54
B. AAALTTICAL ME1THODS AND CONTROL .................. 55
1. Materials Used ..................................... 55
2, Determination of Moisture .......................... 55
3. pH Determination ................ ............... 56
4. Determination of Volume of Precipitate ............. 57
5. Use of the Autoclave ............................... 58
C. A COMPARISON OF METHODS OF DRYING PHENOBARBITAL SODIUM
PREPARATORY TO ASSAY 9........... .................*. 59
D. EVALUATION OF REPORTED METHODS OF ASSAY ............ .... 85
1. Comparison of U. S. P. XI and Rotondaro Assays of Phenobarbital Sodium U. S. P. XI .....,,........... 86
2. Comparison of Modified Rotondaro Assays of Phenobarbital and Phenobarbital Sodium .............. 73
3. A Study of Assays employing Silver Nitrate ......... 79 4. Von Babitsoh Titration of Phenobarbital ............ 85
5. Comparison of Assays of a Deteriorated Solution of
Phenobarbital Sodium U. 3. P. XI ............... 87
E. A STUDY OF A NUMBER OF FACTORS IN THE DETERIORATION OF
SOLUTIONS OF PHENOBARBITAL SODIUM U. S. P. XI ..... 91
1. Effect of Heating a 6% w/v Solution of Phenobarbital
Sodium U. S. P. XI at 600 C. ................... 97
2. Effect of Heating 5, 10, and 20% w/v Solutions of
Phenobarbital Sodium U. S. P. XI at 800 C. ..... 100 3*. Effect of Heating 6, 10, and 20% w/v Solutions of
Phenobarbital Sodium U. S. P. XI at 1000 C. .... 107
4. Effect of Heating 5, 10, and 20% w/v Solutions of
Phenobarbital Sodium U. S. P. XI at 11650 C. .... 112
5. Effect of Heating 5, 10, and 20% w/v Solutions of
Phenobarbital Sodium U. 8. P. XI at 1270 C. .... 117 iv




F. RELATION OF pH TO THS DETERIORATION OF SOLUTIONS OF
PHENOBARBITAL SODIUM .........................**4. 123
1. The Use of Monosodium and Disodium Phosphates .eo*.. 123 2. The Use of Sodium Carbonate ............... 127
0. EVALUATION OF SEVERAL SUGGESTED STABILIZING AGENTS ON
SOLUTIONS OF PMRENOBARBIT2AL AIND PHENOBARBITAL SODIUM 130
1. An Experiment on the Stability of Elixir of Phenobarbital U. 8. P. XI and Several Modifications *. 130
2. An Experiment on the Stability of Phenobarbital
Sodium U. S. P. XI in the Presence of Aloohol& Glycerin, and Antipyrine ....................... 133
3. An Experiment on the Stability of Solutions of Phenobarbital Sodium U. S. P. XI Containing Dextrose, 135 IV. DISCUSSION OF RESULTS ...................................... 139
V. SUMMARY AND CONCLUSIONS ................................... 18
VI. BIBLIOGRAPHY ......................
BIOGRAPHICAL ITEMS .......................... 6




I, ITRODUCTION
Since the introduction of diethyl barbituric avoid into medicine
by Fischer and von Mering (1) in 1903, barbituric acid derivatives have assumed a leading place amongst the medicines which assist in alleviating the ills of mankind. Those medicinals are now more widely used than any other class of hypnotics. For oral use, they are generally dispensed in the dry state or in elixir form. however, for parenteral injection, it has usually been necessary to employ aqueous solutions. Since barbiturates as a class rapidly decompose in aqueous media, they are administered soon after preparation of the solution. This instability has prompted a number of workers to conduct research relative
to the nature of the phenomena and to methods of its retardation.
The present work was carried out to extend the knowledge regarding the deterioration and stabilization of barbiturate solutions, limiting the research to one derivative, namely, phonylethyl barbituric acid* Methods of assay applicable to the problem have been studied and compared. Experiments were conducted to investigate a number of
factors in the decomposition and to evaluate and amplify several reported methods of stabilization.




2
REVIEW OF THE LITERATURE
A. DETERIORATION AND STABILIZATION OF BARBITURATE SOLUTIONS
1. Nature of the Deterioration
In 1927, Steenhauer (2) found that aqueous solutions of sodium diethylbarbiturate, when heated at 1000 C., suffered hydrolysis into carbon dioxide and diethylaoetylurea, The latter compound crystallized out on cooling. Confirmation of these findings came from Madsen
(3) in 1934, and Bailey (4) in 1936. Bailey (4) separated the crystals formed in solutions of sodium barbital which had been autoclaved. The crystals were washed with water and then extracted with a chloroformether mixture. The compound obtained from the latter was reorystallized from boiling water. The colorless needles resulting melted at 2070 C. (corr.). No ammonia was formed when the compound was heated with strong alkaline solutions, whether aqueous or alcoholic. Ammonia was formed after long heating with 50 sulfuric acid and subsequent treatment with alkali. The fact that nitrogen was formed in this sulfuric acid solution when treated with excess alkaline hypobromite showed the presence of urea. Analysis yielded the following results
Hydrogen 8.92 Nitrogen 17.69
Carbon 53.44 Oxygen 20.01 (by difference).
The simplest empirical formula derived from the above data is CTH14N202. The molecular weight was found to be 155, when determined by cryoscopie means with camphor as the solvent. Urea and an oily liquid, which was practically insoluble in water, were the products of hydrolysis formed by refluxing this compound with 50% sulfuric acid. The oily liquid




3
was diethyl aoetti acid. The oily liquid had an aoid reaction, a boiling point of slightly over 1900 Co, and a density less than that of water. Fittig in 1880 gave the boiling point of diethylaoetie aid as 194-195o C. and stated that it was lighter than water. This informa.tion, together with other considerations, led Bailey (4) to conclude that crystalline precipitate was diethylasetylurea. Its molecular weight is 168.
No ammonia is produced by heating diethylaoetylurea with strong alkaline solutions because of its great stability. Lassaigne's test cannot be performed because the compound tends to volatilise rather than decompose when dropped on boiling sodium.
According to Fischer and Dilthey (5), ureides of dialkylaoetio
acid are frequently formed as by-products in the preparation of ureides of dialkylmalonio acid in the presence of sulfuric acid. They are also
formed by heating the ureides of dialkylmalonic acid. A third method of preparation is based on the action of phosphorus oxychloride on a mixture of dialkylmalonio acid and urea. These three methods are probably one and the same, namely, the decomposition of dialkylmalonylurea. Fischer and Dilthey (5) remarked that dimethylacetylurea could not be prepared by any of the aforeto mentioned methods. Furthermore, it was unknown at that time (1904).
Diethylacetyluresa was first described by the von Meissen brothers
(6) who stated that it was formed by heating diethylbarbiturie acid to 1020 C. They indicated that the melting point was 207.60 C. (oorr.) and that it was hydrolysed into diethylacetic acid and urea only on prolonged heating with concentrated hydrochloric acid. Fisher and




4
Dilthey (5) prepared diethylaeotylurea by heating a mixture of urea and diethylmalonic acid to which phosphorus oxyohloride had been added. The product was orystallised from hot water, They found that the melting point was 207.5o C. The structure of the compound was learned from its cleavage into urea and diethylaoetic acid upon heating with oonentrated hydroohlorio acid in a sealed tube for seventeen hours. The diethylaetic acid separated as an oil. To identify the latter product, its silver salt was prepared, The silver content was determined and found to correspond to that of silver diethylanetate. The presence of urea was shown by preparing the nitrate salt in order to free it from the hydrochloric acid solution, regenerating the urea molecule, and then analyzing for nitrogen content. The nitrogen content and the melting point were those of urea.
Diethylaoetyluroa dissolves in approximately 120 parts of hot water. it is practically insoluble in sold water, moderately soluble in chloroform and ether, and very soluble in alcohol. It is insoluble in both cold dilute acids and alkalies.
Methods of preparing dipropylacetylurea and methylothylaoetyluroa were also given by Fischer and Dilthey (5). From qualitative tests on purified precipitates formed in solutions of other barbiturates, Bailey
(4) found that the disubstituted acetylurea, corresponding to the barbiturate employed, was produced in eaeh ease.
In addition to the diethylaetyluroa, Bailey (4) found that the following compounds resulted from the hydrolysis of solutions of diethylbarbiturates
Gases carbon dioxide and ammonia




5
Substances in solutions urea, sodium diethylaoetate, sodium diethylmalonate, sodium bloarbonate, and sodium carbonate.
From the products formed, Bailey (4) deduced that the barbituric acid molecule is hydrolysed in two ways s
(1) loss of one molecule of carbon dioxide, opening the ring to form diethylacetylureas
(2) cleavage of the ring to form diethylmalonio acid or its salt and urea. The diethylmalonie aid then loses one molecule of carbon dioxide to give diethylaoetic acid or its salt. The urea molecule decomposes slowly into carbon dioxide and ammonia.
Earlier, Madsen (3) had reported that diethylacetylurea hydrolyszed in considerable quantity into diethylacetic acid and urea, and latter hydrolysing further in a small degree to carbon dioxide and ammonia.
Bailey (4) suggested that neither reactions proceeded to completion, so that both final and intermediate products are obtained. Bailey
(4) presented the equations which follow.
(1)
0 H 0 H
C2 C5 C1 C 2,6 \ Na 0\
-C C=O0 + 2 H20) C C=0 + C =0
Ci 2 \ 2 115HH
q2R / C8 5 / O
O Na
Diethylaoetylurea
0
0 C-ONa
i- H C2 H5 /
CS 5'C + NHSCO NR3+coS
C2 C C -0 + NaHCOsA 3 C / H '
CS
H 6Sodium diethyl
aoetate




(2)
0
C 2 H 0 -N \ C2H5,, / 6- OH
c2H 6 C C-N / C 0 + 2H 2 0 C 2 a 5 C \ C-ON& 4 N"2CO NH2
0 01
0 ft 0
Primary Sodium diethylmlonato
0
c- OH
C 2 H V- H
\C-ON& c2H 5 1-1 c + CO 2
0
Sodium diethylacetate
NH2CO ER2 + JJ20 2 VH3 + CO 2
Aspolund and Skoglund (7) outlined the course of deterioration of substituted b&rbituric &aide an follows
9 0
C -N \ I / 8 NficONH 2 Ile COOH
C 0 --- R c R
2 'COOH 2 C 11-1 COOK
C- N
R 2 CHCONHCONH2
I
R 2 CHCONH 2
R2CHCOOH
Trisubstitutod barbituric saide may break down in either of two ways.




7
CO-N-CO-NH2
l9 pa32 / 6c
/c -a / coon
0a
R2 > g -NR=O 00
CO-NH-CO.NiHCH3
Aspelund and Skoglund (7) separated the decomposition products from one another, as they occurred in deteriorated solutions of barbiturates, by filtration and extraction with ether at successive pH levels. The resuits obtained are sumarised in Table 1 and Table 2 showing the percentage yields of the different products obtained.




TABLE I
PERCENTAGE YIELDS OF DECOMPOSITION PRODUCTS OBTAINED FROM DRT'SRIORATE1) BARBITURATE SOLUTIONS WHICH WERE BOILED FOR THREE HOURS *
Name of the Original Substituted S ubstituted Substituted
Barbiturio Barbiturio Malonurie Acetureide Acetic Acid
Acid Acid Acid
Diethyl b~ao 45 30 7
Ethyl-allyl- 29 15 34
be&*
Diallyl b~a. 14 7 24 12
Isopropyl- 85 1.8 4 1
allyl boa.
Phonylethyl beas 87 11 12
Phenyl-ethyl- 30 0.3 34 7
N-methyl beam,
Cyclohexenyl
methyl-N-methyl 23 4 48
be&*
Ethyl-allyl- 28 25 32
N-Methyl be&.
*Results obtained by Aspeiwid and Skoglund (7)




9
TABLE 2
PERCENTAGE YIELDS OF SUBSTITUTED 2MALOWURIC ACID OBTAINED FROM DETE~RIORATED BARBITURATE SOLMTONiS WiiICdI HAD BEEN STANDING FOR EIGHT DAYS*
Name of the Sarbiturie Percentage Yield of
Acid Substituted Malornuric Acid
Diethyl box* 0.6
Ethyl-allyl boa. 200
Diallyl boa. 4.0
Isopropyl-allyl be&* traces
Phenyl-othyl b.a. 1
Phonyl-ethyl-N-ethyl boa* 24
Cyclohexany1-methyl- 12
N-mthyl bao
Ethyl-al lyl-N -methyl 2.6
b*&.
*Results obtained by Aspelund and Skoglund (7)




Aspelund (8) stated that the instability of the ulonylurea ring in barbiturates is probably due to the negative hydroxyl group* H. studied the decomposition of 6,5-bromobensyl barbituric aid. The resuits obtained are outlined by the equations which follow.9
OH =0 NOHI
II
C 5 i \C h /-1H 5i~ COICOIH2
r Sr9 Boil,
H HS V Br C H
C 5
C UCH 2 C0NHCN CHC2C~iOH
Boiling H OH AlcoholicKO
wih10% 'c CO
H OH
C SH5CH2 COliHCN
C6H5CH CONILCOkIH2




Aspelund and Lindh (9) made a study of the chemistry of the decomposition products of 5,5-bromoalkylN-alkcyl barbituric acids. The results obtained are outlined by the equations which follow*
QR#
R C -_N R- -CONHCONIMO
0 R at room
temperature Heating in alcohol jand water
RO Heating with BrCH
R CONHON alcoholic caustic R"" "'1CONECONHRO
alkal ies
IBoiling with
10% H2S04
[RCH(oH)CONll 2c0




12
TABLE 3
ELTING POINTS OF DECOMPOSITION PRODUCTS OBTAINED FROM
DETERIORATED BARBITURATE SOLUTIONS
Decomposition Product Me. p. 0 C. Worker
Phonylethylaostylurea 150 Bailey (4)
Phonylmethylacetyl uroaib Bailey (4)
N-mothyl-cycooxonyl- 974101 Bailey (4)
niethylaostylures,
Jutylethylacetylures, 154 Bailey (4)
Diethylacetylures, 206-207 Bailey (4)
Diethylmalonuric aid l1.12 Neilson (10)
Phenylethylaotia acid 44-46 Neilson (11)




Although they may keep for 8-15 days as Wetsel (12) claimed, solutions of sodium phenylethyl barbiturate also undergo an analogous hydrolysis into phenylethylacetylurea (m. p. 1460-147O C.) and carbon dioxide. Nielsen (15) found that phenylethylasetylurea was precipitated when he refluxed 20% solutions of sodium phenobarbital for periods of 1/2 to 1 hour. Heating for 3 to 5 hours resulted in hydrolysis of part of the phenylethylaetylurea into phenylethylaoetic acid and urea. Aspelund and Skoglund (14) stated that phenylethylmalonurio acid was very unstable.
The formation of free phenylethyl barbiturice acid is also a possibility. It is to be noted that the solubility of this acid in water is about 1 1 1000, whereas the solubility of free diethyl barbiturie acid in water is about 1 s 130. Therefore, the appearance of a preoipitate of the free barbituric acid in solutions of sodium phenobarbital is a greater possibility than in solutions of sodium barbital. According to Woodward (15), Nielsen (13) has provided evidence that free phenylethyl barbituric may be formed in solutions of sodium phenobarbital on prolonged standing. Both liberation of the latter acid and hydrolysis take place at the same time. When determining phenylethylaoetylurea formed in solutions, Nielsen (13) had difficulty in obtaining deposits and filtrates free from simultaneously formed phenylethyl barbituric acid. The acid was freed by the carbon dioxide absorbed from the atmosphere. To prevent deterioration by carbon dioxide, Cassani (17) suggested that the solution be saturated with nitrogen. Nielsen (13) orevented the formation of the free acid by adjusting the




pH of the solutions with standard sodium hydroxide. Presumably the sodium hydroxide was employed in a manner so as to redissolve any liberated acid.
Mesnard (1) reiterated that solutions of sodium phonylethyl barbiturate hydrolyse into phenylethylbarbituric acid and sodium hydroxide. The decomposition of phenylethyl barbituric acid follows due to the action of sodium hydroxide, forming products without therapeutic value. Measnard (18) gave the following#
C-C-N Ot /NH
C/56\ / \ 02 5 / I
C C=0 aOH C C=0
C-N C-N
0 H 0 H
The rapidity of hydrolysis of some barbiturate solutions is indicated by the length of time during which they are suitable for parenteral administration. The hydrolysis of evipal-sodium is rapid. Cazzssani (17) stated that aqueous solutions should not be used 2 to 3 hours after their preparation. According to Eli Lilly & Company (19,
20) solutions of sodium amytal and of sodium pentobarbital should not be used thirty minutes after their preparation. Abbott Laboratories
(21) stated that solutions of sodium ethyl (1-methyl butyl) thiobarbiturate oan not be safely allowed to stand longer than 4 hours. Aspolund and Skoglund (14) found that N-substituted barbituric acids were more easily decomposed than the 5,6 disubstituted barbituric acids. They stated that the sodium salts of isopropylallyl barbituric acid and cyolohexylethyl barbituri acid and perhaps of diethyl barbituric acid




15
are stable in aqueous solution for a definite length of time.
2. Factors in the Deterioration a. Hydrogen Ion Concentration
In 1933, Nielsen (13)found the rate of hydrolysis of solutions of
sodium phenylethyl barbiturate was decreased when there was an increase in hydrogen ion concentration. The amount of hydrolysis occurring was followed by determining the quantity of carbon dioxide and phenylethylaseetylurea formed. Nielsen (13) stated that the determination of the ohloroform-soluble urea derivative introduces an error due to its high solubility in the original solution and due to the formation of free phenylethyl barbiturioe acid by the carbonic acid released during the hydrolysis. Presumably, Nielsen (13) extracts the urea derivative with chloroform, thus explaining the above mentioned error. The solubility of phenylethyl barbiturice acid in chloroform is 1 Gmn. in 40 co. (22). The results obtained by Nielsen (13) are summarized in Table 4.
TABLE 4
RELATION OF pH TO DECOMPOSITION OCCURRING IN 10% SOLUTIONS
CONTAINING BOTH PHENOBARBITAL SODIUM AND PHENOBARBITAL AT 18.8 (1) C.
pH Number of days required
for 1$ decomposition
8.9 30
9.4 22
9.9 18




16
A 10/ solution of phenobarbital sodium saturated with the free acid has a pH of 8.9. Berasain and Vitali (23) stated that the pH of a 10% w/v solution of the sodium salt was 9..
Berasain and Vitali (23) prepared 10 w/v solutions of phenobarbital sodium in buffer solutions instead of water. The pH values of the solutions prepared are indicated in Table 5. However, they failed to state what proportions of the respective buffer solutions were used.
TABLE 6
EFFECT OF BUFFERS ON THE pH OF 10% SOLUTIONS OF PHENOBARBITAL SODIUM
Buffer Originator pH obtained
N&a0 and glycocoll Sorensen 9.71, 9.36, 8.39, 8.57, 8.24
KH2PO4 and NaOH Clark and Lubs 8.00, 7.80, 7.60, 7.40, 7.20
Veronal and HC1 Michaelis 9.40, 8.60, 8.00, 7.60
Sodiua oitrate and Clark and Lubs 7.00
NaOH
In another experiment, phenobarbital was used to neutralise the free alkali and to lower the pH. The latter and the buffered solutions were subjected to heating by steam for fifteen minutes. A precipitate was formed in each ease but the quantity was less in those solutions, having a pH less than 9.00. The preoipitate appeared as white, small flakes, floooulent at first and then as needles, grouped in a radiating fashion.
Tomski and Waller (24) investigated the effect of varying amounts
of sodium oarbonate on the hydrolysis of solutions of sodium phenylethyl




17
barbiturate. One (1) oc., 2 co., and 3 co. portions of N/10 sodium carbonate were added to 5 co. portions of a solution of sodium phenobarbital. Blank solutions were also made using water instead of the sodium carbonate solution. All the solutions were heated for sixty minutes at 80o C. Bach blank solution showed a loss in percentage of 2.8. The results which Tomski and Waller (24) obtained are summarised in Table 6.
TABLE 8
LOSS IN PERCENTAGE ON EATING A SOLUTION4 PHEIWOBARBITAL SODIUM IN THE PRESENCE OF SODIUM CAR~OUATE
co. Loss in
N/O10 Na2CO3 Percentage
1 5b.5
2 10
3 15
Madsen (3) found that solutions of sodium diethylbarbiturate hydrolysed at a slower rate upon an increase in hydrogen ion concentration. He calculated the degree of hydrolysis by determining the amount of diethylacetylurea and carbon dioxide formed. The results secured by Madsen (3) are given in Tables 7 and 8.




18
TABLE 7
EFFECT OF HEAT ON 10% SOLUTIONS OF SODIUM DIETHYLBARBITURATE
Temperature Duration of Per cent pH Change
Eating Decomposed
1000 C. 1 hour 2.5 10.1 to 9.3
800 2 hours 1.0 10.1 to 9.6
600 3 hours 0.23
TABLE 8
EFFECT OF siEAT ON 10 SOLUTIONS OF SODIUM DIFTHifLARITURAtTE WITH A pH OF 8.9
Temperature Duration of Per cent
Heating Decomposed
1000 C. 1 hour 1.5
800 2 hours 0.32
Bailey (4) discovered from a study of solutions prepared from different samples of sodiu barbital that some showed precipitates on heatIng at 1000 C. while others required a slightly higher temperature to bring about the same result. Upon determine the pH values of all the solutions, he found that those which decomposed at 1000 C. had a slightly higher pH value. This was apparently due to the excess alkali used




19
in preparing the sodium salt of the barbituric acid. Bailey (4) suggested that only 96% of the calculated quantity of sodium hydroxide be used in preparing the salt of the acid. The slight excess of acid then could be filtered off and a product would be obtained whioh would withstand heating at 1000 C. in sealed containers for short periods without decomposition. In this regard it may be pointed out that amines have been used as basic substances to replace the sodium hydroxide. The intention is to replace the strong inorganic alkali by a weaker base. The use of amines is reveiwed in another part of this discussion.
A fall in the pH of solutions of sodium aytal upon aging was reported by Cassani (18). Fresh solutions have pH values of 9.56 to 9.8. The solutions decompose readily on standing. When the pH falls to 9.3, they become opalescent and cloudy.
b. Temperature
Nielsen (13) found that the rate of deoomposition of solutions of sodium phenylethyl barbituric acid varied greatly with temperature. The rate was followed by determining the amount of carbon dioxide and phonylethylaoetylurea formed. The results which he obtained are indicated in Table 9.




20
TABLE 9
EFFECT OF TEM"PERATURE ON 10% AQU4EOUS SOLUTIONS OF SODIUM PHENYLETHYL BARBITURATE ADJUSTED TO A pl OF 9.4 WITH PiENYLETHYLBARBITURIC ACID
Temperature Time Per cent Decomposed
10 C. 2 months less than 1%
200 3 weeks 1
59o 1 month 22
A 1s300 solution of sodium phenobarbital, which is kept at 20 C. is 1% decomposed in 8 weeks, according to Nielsen (13).
Tomaki and Waller (24) measured the rate of decomposition of solutions of sodium phenobarbital by the Regland method (25). This method consists of titrating a solution of the barbiturate, to which sodium carbonate has been added, with N/10 silver nitrate until a faint opalescence is formed which does not disappear on shaking. Investigating the effect of heating solutions of different strengths under the same conditions, Tomaki and Waller (24) have found that the loss of sodium phenobarbitone increases with increase in concentration. The difference was small, however, and for all practical purposes the loss on heating a 20%, 10%i or even a 5% solution is the same, according to these workers. Solutions of 2% to 10% were stored and analysed periodically. A solution kept for four months (Deember-Maroh) lost 6% whereas a solution of equal strength when kept for the same length of time (March-June) lost 9% of the phenobarbital sodium. Additional results which were




21
secured by Tomaki and Waller (24) are given in Table 10.
TABLE 10
20% SOLUTIONS OF FIENOBARBITAL SODIUM IN AMPULS HEATED AT 80o C. FOR VARYING LENGTHS OF TIME
Time in minutes Percentage Lose
50 1.0 to 1.5
60 2.0 to 3.0
120 4.0 to 6.0
180 640 to 7.0
A 20% solution therefore loses 6.0 to 7.0% under conditions similar to those prevailing in the process of tyndallization. A solution of the same strength loses 1.0 to 1.5% when sterilized by the emergency method of the British Pharmacopoeia, 1982 (26). Further results obtained by Tomaki and Waller (24) are indicated in Tables 11, 12, and 13.
TABLE 11
20% (A PROX.) SOLUTIOUS OF PiHENOBARBITAL SODIUM IN AMFULS CREATED IN BOILING WATER FOR FIFTEEN MINUTES
Initial Strength Final Strength Loss
19.5% 19.3% 1.0%
18.7 18.52 1.0
20.1 19.9 1.0




22
TABLE 12
20% SOLUTIONS OF PHEiNOBARBITAL SODIUM EXPOSED TO STLAM AT 1000 C. FOR VARIOUS LENGTHS OF TIME
Time in Minutes Percentage Loss
60 9.5 to 10.5
120 14.5 to 15.5
180 19.5 to 20.5
TABLE 15
20% (APPROX.) SOLUTIONS OF PHENOBARBITAL SODIUM IN AMPULS EXPOSED TO
STEAM AT 1150 C. FOR THIRTY MINUTES
Initial Strength Final Strength Loss
19.8% 16.0% 19.2%
20.1 16.0 20.4
18.7 15.1 19.3
Tomski and Waller (24) obtained a white precipitate of phenylethylacetylurea when a 20% solution of phenobarbital sodium was exposed to any of the methods of sterilisation cited above.
Van Leont (27) stated that sterilization of sodium phenobarbital in the dry state at 1200 C. yielded a product which gradually booame incompletely soluble after this treatment. He suggested that this was due to the action of traces of moisture.




23
Bailey (4) likewise found that with increasing temperature, greater amounts of precipitate were formed in solutions of sodium diethyl barbiturate. He obtained needle-shaped crystals on autoclaving 7%, 10%, and 14% solutions for thirty minutes at 1050 C. Increasingly higher yields were obtained upon autoclaving these solutions at 1100, 1150. and 1200 C., the iszs of the needles varying with the temperature. Upon removing the crystals and subsequent autoelaving an additional quantity of crystals were obtained. At lower temperatures, the yield was influenced by the prolongation of heating* Stioh (28) claimed that 20% solutions withstood sterilisation in steam at 1000 C. He acknowledged that 1 to 5% solutions are hydrolysed by warming to 1000 C. The British Pharmacopoeia, 1932 (26) contains the information that solutions of sodium barbital can be sterilized in an autoclave at 1150 C. for 50 minutes or by tyndallisation at 800 C. Berasain and Vitali (23) stated that solutions which had been subjected to tyndallisation were not suitable for injection. As has been mentioned before, Bailey (4) stated that solutions of the sodium salt which had been prepared by using only 95% of the calculated quantity of sodium hydroxide, would withstand heating at 1000 C. for short periods without decomposition.
Bailey (4) felt that the decomposition reaction was probably reversible. The course of the reaction was influenced by the initial concentration of the solution, temperature, and removal of the precipitated product.
Bailey (4) found that a higher yield was produced when the autoclaving was carried out in an open flask than when conducted in a sealed




24
vessel. This was explained by the escape of carbon dioxide, presumably driving the reaction towards completion by the removal of one of the products from the field of chemical reaction.
From a study of a number of alkyl and aryl substituted malonylurea compounds, Bailey (4) found that their solutions all decomposed at temperatures exceeding 100 0.C and that some decomposed at temperatures below the boiling point of the solution. Ruhkopf (29) studied the behavior of substituted barbituric acid during heating at elevated pressures in the absence of alkalies. When pressures up to five atmospheres were used, the substituted barbiturice acid was almost completely hydrolysed with the formation of approximately equal amounts of the substituted acetyl urea and the substituted acetamide. The formation of the aide was favored with increasing pressures, becoming the only product at a pressure of about ten atmospheres in addition to the carbon dioxide and ammonia formed. At a pressure of five atmospheres, salts of strong acids wore found to enhance the formation of the substituted aeetylurea, whereas salts of weak acids were found to increase the production of the substituted aeetamide. Above the latter mentioned pressure, the amount of the acetylurea derivative decreases under any cirounstances. Hydrolysis of diethyl barbiturie acid in aqueous solution at five atmospheres yielded 47% diethylacetylurea and 40% diethylacetamide. At ten atmospheres, 95% diallylacetamide was formed in the hydrolysis of diallylbarbiturio acid in aqueous solution.




25
3. Stabilization of barbiturate Solutions
a. Adjustment of Hydrogen Ion Concentration
Berasain and Vitali (23) found that decomposition of solutions of sodium phenylethyl barbiturate was decreased when their pH value was lower. The same was found true for solutions of sodium diethylbarbiturate by Madsen (3), Bailey (4) and by Tomski and Wailer (24).
N
Sohlemmer and Torber (30) claimed that solutions of sodium diethyl barbiturate possessing a pH of 5.92 underwent no hydrolysis within the limit of error of 0.4%, when subjected to the usual conditions of sterilisation. Berasain and Vitali (23) attempted the use of buffers to control the pH. The lowest pH value of their solutions was pH 7. A decrease in the amount of precipitate was noted in solutions of pH less than 9.00, when heated fifteen minutes by steam.
Mesnard (18) suggested the use of sodium bicarbonate as a stabilizer in solutions of sodium phenobartitall its effect was supposedly as follows$
Cl Ho1103N (N Na)- H20 --- C12H11101 (NH)+ NaO
NaiCO3 + NaOR - a2Co3 HZO
C1 110N (NH) Na2CO3 c12110N (N Na)+ NaHCO3
Berasain and Vitali (23) stated that the use of sodium bicarbonate was unsatisfactory.
b. Amines as Stabilizing Agents
One of the earliest applications of amines to stabilize barbiturate




solutions made its appearance as a commercial product called Somnifene. Somnifene is manufactured by of man-LaRoche and Company (31) and contains in 1 cc. the diethylamine salts of 0.1 Gm. of diethyl barbituric acid and 0.1 Ga. of isopropyl propenyl barbituric acid. In 1921, Le 'dedicin (32) described it as being soluble and with effects superior to those of the other barbiturate derivatives. New Somnifene it the isopropylallyl barbiturate of diethylamine. According to Cassani (17), it can be sterilized by tyndallisation at 800 C., has good keeping qualities, and is well tolerated.
In a British Patent (33) it is claimed that a 6 to 10,% solution of butylethyl barbituric acid, which is stable at 1000 C. and sterilisable by tyndallisation, can be made by neutralizing the acid with aminoethanol, diethylaminethanol, and diethylamine. Cassani (16) stated that keeping qualities of these solutions are good. Piperazine has also been used in an analogous manner and Cazsani (16) indicated that it too yields solutions with -ood keeping qualities. All of the above solutions are claimed to be well tolerated.
Blok (34) reported that a stable solution of phenobarbital suitable for intramuscular injection can be made by the use of diethylamine. Blok (34) suggested the following formulas
Phenobarbital 10 Gm.
Diethylamine 2.75 Gm.
Petit Mixture q.s. ad 100 cco
Petit Mixture
Alcohol (90%) 28 Gm.
Glycerin 31 Gm.
Water q.s. ad 100 co.




27
The phenobarbital was placed in a sterile flask containing some sterile Petit Mixture. The diethylamine was added and the mixture was shaken and heated at 40a to 50o C. in a closed flask until solution was effeooted. Sufficient Petit Mixture was added to make 100 co. Then the solution was filtered, placed in ampuls and sterilized for a half hour at 600 C. during two successive day*, No appreciable decomposition was found after three months. When heated in flowing steam for one hour, 2%* was hydrolysed. Schulte (35) modified Blok's procedure (34) by neutralizing the barbituric acid with diethylamine dissolved in alcohol and then adding glycerin. In this way, the solution is obtained without the use of heat. Berasain and Vitali (23) subjected Blok's solution (S3) to steam heat for fifteen minutes. The solution became colored and an abundant quantity of precipitate was produced. Glycerin and alcohol would be objectionable for injections. The latter authors pointed out that glycerin produces deleterious effects at the site ef injection. In addition alcohol produces a disagreeable burning sensation when injected.
Van Leent (27) stated that a solution of sodium barbital is best prepared by dissolving barbital in diethylamine and sterilizing for thirty minutes at 600 to 650 C. on two successive days.
E. R. Squibb and Sons have patented (36) a preparation which is
claimed to be stable and which may be sterilized by heating. It is an aqueous solution containing ethyl isopropyl barbituric acid and more than a molecular proportion of diethanolamine.
Meanard (18) stated that sodiuan phenobarbital in solution hydrolyzed




28
into phenobarbital and sodium hydroxide, followed by cleavage of the phenobarbital itself due to the action of the sodium hydroxide formed. To prevent this action, Mesnard (18) suggested the use of ethanolamine acetate in solutions of this barbiturate. The intention was to have the sodium hydroxide liberate ethanolamine which would form a stable salt with barbituric acid. Beraseain and Vitali (23) declared this prooedure was unsatisfactory. They found that 10% solutions of this barbiturate prepared with mono-, di, and tri-ethanolamine (or piperazine) were not stable when subjected to steam sterilization for fifteen minutes.
Nightingale and Morris (37) stated that the amine salts of the
1-phenyl derivative of 5-bromobarbiturie acid are lose stable than those without the phonyl group. In alcohol 5,5' dibromobarbiturio acid reacted rapidly in ethyl alcohol with butylamine to give the relatively stable butylamine salt of 5-bromobartiturice acid. Analogous compounds with a 1-phenyl group were unstable. Cold sodium carbonate caused the latter to break up into butylamine and 1-phenyl 6-bromobarbituric acid.
Brolumal is a product of the Italian Drugs Importing Company, Inc.
(38) which contains hexamethyltetramine. Each ampul of 2 co. contains:
Phenylothyl barbituric acid 0.015 Gm.
Organic Bromine 0.010 Gm.
Rexamethyltetramine 0.010 Gm.
kydroglyceric solution, sterile q.s.
Brolumal is made for intramuscular and intravenous injection. Eli Lilly and Company (39) claimed that stable solutions of amytal could be made by dissolving amytal in hydroalooholic mixtures containing a sufficient




29
quantity of meethenamine. Elizir Amytal, Lilly is stated to be such a stable solution. Each fluid ounce coiitains amytal, 2 gr., in a vehicle containing nethenamine, 2 gr. per fluid ounce, glycerin, alcohol, water, and aromatics. The alcohol content is 30f,.
Elixir Amytal, Lilly (1%, approx.) Amytal gr. iv
Methenamine gr. iv
Glycerin
Alcohol
Water
Aromatics q.s. aa ad 3 i
The alcohol content is 34$.
In a German patent (40) mention is made of the preparation of a soluble compound of phenobarbital by the addition of at least one molecular proportion of diethylenediamine.
Vande Velde (41) studied the influence of neutral salts on the hydrolysis of urea at elevated temperatures. He found that hydrolysis was accelerated by sodium, potassium and phosphate ions and retarded by ammonium ions.
c. Amides as Stabilizing Agents The use of amides to prepare solutions of barbiturates suitable for injection has appeared in a number of patents (42,43). Unsubstituted aides of the lower aliphatic acids and amides of the lower aliphatic acids with one or two alkyl groups attached to the nitrogen atom are used to convert barbituric acid derivatives such as phenylethyl barbituric acid or yoelohexenyl barbituric acid into water soluble compounds




30
whose solutions are claimed to be suitable for subcutaneous injection. Specific amides mentioned are monomethyl and monoethyl acetamide.
In a French Patent (44) a description is given of the use of unsubstituted amides of fatty acids, having at least three carbon atoms in the molecule, as auxiliary agents to obtain 20 % solutions of barbiturie acid derivatives, the 5, 5' substituents of which consist of an aliphatic and a cyclic group. A solvent for cyclohexylethyl barbiturice acid is given ass
Propionamide 82
Betaine 3
Water 15%
A solvent for phenylethylbarbituric acid is given as: Propionamide 50
Acetamide 40
Water 10%
Similar solvents are given in an Austrian Patent (45). A 10% solution of phenylethyl barbiturate in the following solvent is cited, Acetaside 65
Betaine 5
Water 30%
Addition of a little urea and/or ethyl alcohol may be advantageous, according to the patent.
d. Use of Urethane in Conjunction with Other Agents
A British Patent (46) described the use of a mixture of tertiary
chlorobutyl alcohol and urethane as a solvent for substituted barbituric acids. This alcohol and urethane may be mixed in equal parts by weight to give a liquid product which is suitable as a solvent of barbiturates for injection.




51
The Danish Pnarmacopoeir (47) contains a formula entitled "Compound Solution of Phenylethyl barbiturio Acid" in which ethylurethane is employed. The formula is as follows
Phenylethyl barbiturie acid 20 Gmn.
Amylene Hydrate 28 Gm.
Ethyl Urethane 35 Gm.
Distilled Water 7 Gm.
The solution is prepared by keeping the mixture in a well-closed flask at 200 to 300 C. until dissolved. It is then filtered through a bacteriaproof filter and sterilized. Zwikker (48) claimed that this non-alkaline solution is suitable for injection. Van Leont (27) stated that such compounds as amidopyrine and caffeine may be dissolved in the solution. Whether or not they oact as stabilizing agents was not stated. Berasain and Vitali (23) claimed that the Danish formula is unpractical for injection because of the proportion and nature of the ingredients.
Nielsen (49) followed the decomposition of the barbituric acid derivatives contained in Solution Hypnopheni (Ph. Dan., 1933) by measuring the carbon dioxide produced. Storage for a period of one year showed slight decomposition of diethyl barbiturie acid, allylisopropyl barbiturioe acid, and urethane. Nielsen (60) developed the following formula
for injections
Phenylethyl barbiturie acid 5.0
Sodium phenylethyl barbiturate 6.72 Urethane 25*.0
Alcoholic Spirit (Ph. Dan.,1933) 15.0 Glycerin 12.6
Aq. deast. steril. q.s. ad 100.0 co.
The specific gravity of this preparation at 150 C. was 1.056. In three months at 120 C., about 0.55% of the barbiturates were hydrolysed, and at 24o C. in the same time 1.35% were hydrolysed. This decomposition




32
was one-fifth of that occurring in a 10% aqueous solution of sodium phenobarbital, kept under the same conditions. The solubility of the main decomposition product, phenylethylacetylurea, in the solution was 1% at approximately 220 C. This corresponded to a 12.3% decomposition of the sodium phenobarbital. Therefore, Nielsen (50) reasoned that the preparation could be stored for a long time without any precipitation. After 2 hours warming at 80 C. only the primary decomposition was detected and it amounted to about 0.75% of sodium phenobarbital. The decomposition of urethane ensuring during heating at 800 C. was also determined. The decomposition is represented by the following equations,
(A)
N~12
/ H
oc as
25
(B)
NH2
2 O=C +H0 NH2CONH2 2 C2 K -OH + CO0.
0 CH5
Nielsen (50) also gave a formula for a 10% solution of phenobarbital,
suitable for injection, containing diethylamine. Nielsen's formula
(50) is as follows,
Phenobarbital 10.0
Urethane 25.0
Alcoholic Spirit (Ph. Dan., 1933) 15.0 Glyoerin 12.5
Diethylamine 20.0
Aq. dest. steril, ad 100 co.




33
The specific gravity of this solution at 150 C. was 1.035. Nielsen
(50) claimed that this solution was very stable. No hydrolysis was detected after 6 1/2 months storage at 240 C., nor after two hours heating at 800 C. Berasain and Vitali (23) heated this solution for fifteen minutes in steam. It became deeply colored and yielded a precipitate in a few days. They pointed out that Rakieten et al. (51) had found that urethane, in association with some barbituric acid derivatives in narcotic doses, had produced somber effects and alterations in the acid-base equilibrium, tending towards acidosis. The disadvantages of glycerin and alcohol in solutions intended for injection have been mentioned elsewhere.
In a British Patent (52) the claim is made that a stable solution of 6,5 disubstituted barbituric acids can be prepared by dissolving the barbiturate with the aid of at least one molecular proportion of an alkanolamine in the presence of a relatively large quantity of a carbamic ester. For example, 20 Gm. of 5,5 isobutylallyl barbituric acid, 5.5 Gm. of monoethanolamnine, and 50.Gm. of urethane are dissolved in 34 co. of water.
e. Other Nitrogen Compounds as Stabilizing Agents
A British Patent (53) stated that stable solutions of barbiturates for therapeutic use are obtained by adding pyridones or piperidones to their aqueous solutions. The pH of the solutions is 6.0 to 6.8. A solution of allyl isopropyl barbiturie acid containing 1-methyl 2-piperidone is specifically mentioned.
Hasleton, Koppanyi, and Linegar (64) employed antipyrine, glycerin,




34
and alcohol as stabilizing agents for solutions of sodium secondary amyl bromallyl barbiturate. They su-gested the following formulas
Sodium secondary amyl bromallyl barbiturate 10 Gm.
Glycerin 10 GM.
Antipyrine 10 0m.
Alcohol 10 cc,
Water q.s. ad 100 cc.
The solution is intended for rectal administration.
Gruber (55, 56, 57) patented the use of urea, urethane, aoetamide, and such pyrasolon derivatives as dimethylphenyl pyrasolone for the stabilization of aqueous solutions of barbiturate salts.
f. Glycols as Stabilizing Agents
Page and Coryllos (68) claimed that stable solutions of practically all of the barbiturates could be made by dissolving them in ethylene glycol. Merely the concentration of the solvent had to be varied according to the barbiturate to prepare a stable solution. A solution of sodium amytal is prepared in the following manners the amount of amytal needed to make a 10% solution is dissolved with very vigorous shaking in the least quantity of 15% sodium hydroxide, without the use of heat. Ethylene glycol is added in sufficient quantity so that its concentration in the final solution will be 16%. The excess sodium hydroxide is neutralized with dilute hydrochloric acid, the solution being shaken vigorously upon the addition of each drop of acid. The shaking is neoessary to prevent the separation of isoamyl ethyl barbituric acid in clumps which are difficult to redissolve. Vihen a fine cloudiness remains, the addition of acid is discontinued, A minimum quantity of concentrated ammonium hydroxide is then added quickly to clear the solution. If a




large excess of aamonia is present, it may be quickly boiled off over a free flame. The entire procedure must be performed rapidly to prevent hydrolysis. Berasain and Vitali (23) stated that they found this solution to be stable. They remarked that Oettingen and Jerouch (59) had found ethylene glycol to be toxic, hemolytic, and intensely irirtating at the site of injection. Page and Coryllos (58) pointed out that isoamylethyl barbiturice acid is soluble in ethylene glycol and
that it is not essential that the sodium salt be prepared. The ethylene glycol solutions may be heated to boiling to augment the already powerful sterilizing action of ethylene glycol.
In a British Patent (60) and a U. S. Patent (61) it is stated that stable and sterilisable preparations of barbiturate salts can be prepared by the use of alkylene glycols as solvents. The alkali metal salt of the barbituric acid compound is dissolved in the glycol, such as ethylene or propylene glycol, in the absence of any substantial quantity of water. The solution may also be made by dissolving the calculated quantity of an alkali metal in the glycol under anhydrous conditions and then dissolving the free barbituric acid in this solution. Diethyl-, dipropyl-, diallyl-, phenylethyl- and oyclohexenylethylbarbituric acids and their substitution products were specified in the patent. Dumes (62) stated that he had been informed by certain manufacturers that they were using ethylene and propylene glycols to prepare anhydrous solutions of barbiturates of sufficient stability to permit sterilization.
Berasain and Vitali (23) claimed that the following formula provided solutions stable to sterilizations




36
Sodium pheno Larbital 10 Gm.
Propylene glycol 40 cc.
Distilled water q.s, ad 100 cc.
The barbiturate was dissolved in the propylene glycol and sufficient water was added to make 100 cc. The solution was sterilized by heating in steam for fifteen minutes. The solutions thus treated were found to be free of aerobic and anaerobic bacteria by inoculating bouillon and agar and observing after 24, 48 and 72 hours. They claimed that the solutions were stable for the length of time observed, namely, six months. The concentration of the propylene glycol is the minimum required to maintain stability. This glycol was similarly used to make solutions of sodium barbital and sodium amytal. They withstood sterilization with steam for 15 minutes. Braun and Cartland (63) demonstrated that propylene glycol is less toxic than glycerin.
It has been reported that propylene glycol may be substituted for glycerin in any non-official article where glycerin itself may be safely used without increasing the toxicity of the resulting product (64).
In a U. S. Patent (65) the use of aliphatic polyhydric alcohols, such as glycerol, or a glycol, is claimed to stabilize aqueous solutions of barbiturate salts.
A German Patent (66) contained a claim that stable solutions of barbiturate salts can be prepared by using an anhydrous glycol, such as ethylene glycol, as a solvent.
g. Miscellaneous Stabilizing Agents
The use of bases weaker than sodium hydroxide to prepare watersoluble compounds of various barbituric acids led Berasain and Vitali (23)




to try calcium saccharosate. To prepare it, 200 co. of water was added to 10 Gm. of lime. After twventy-four hours had elapsed, the solution was decanted, and a solution of 60 Gm. of sucrose in 200 co. of distilled water was added to the residue remaining. After another twentyfour hours had elapsed, the, calcium saccharosate was obtained by filtration. It was found to be a good solvent for phenobarbital. However, an abu dant precipitate, vhioh did not disappear on cooling, was produced whien solutions were sterilized with steamr for 15 Aminutes.
Page and Coryllos (56) claimed to have prepared a fairly stable solution of sodium amytal, suitable for intravenous injection, by the use of mucilage of acacia. A 20% solution of acacia, in sufficient quantity to make a final concentration of 10, was added to a solution of amytal just before the excess sodium hydroxide used in dissolving the acid was neutralized by hydrochloric acid. The solution was further stabilized by the addition of sufficient ethylene glycol to make 10.
An anonymous writer in the Pharmaceutical Journal (67) made the claim that suspensions of phenylethyl barbituric acid in mucilage of tragacanth retained their therapeutic activity for a long time.
Baldi (68) stated that a stable solution of sodium phenobarbital could be prepared by dissolving it in pure glycerin or a mixture of equal parts of glycerin and water. lie reported that a 4% solution prepared in this way remained perfectly clear end resisted deterioration on heating to !000 C. Baldi (68) gave the following formulas




Sodium Phenobarbital 20 Cm.
Glycerin 20 Gm.
Alcohol 10 Ca.
Distilled water q.s. ad 100 co.
Five per cent magnesium sulfate and 10% mannitol are occasionally used to prevent the dissociation of organic salts. Berasain and Vitali
(23) found that their use in solutions of sodium phenobartital failed to give good stabilisation.
Bush, Dickinson, and Lamson (69) studied the stability of seconal in paraldehyde. They found that a 30% solution of seconal in paraldehyde retained its original anesthetic activity nearly completely after heating at 1200 C. for four hours. Upon intramusceular injection of this solution in some thirty patients, it was found that no discomfort, nor any sign of irritation or injury to tissue was produced. These workers determined the solubility of a number of barbiturates in various substances such as propylene glycol, triacetin, cellosolve, etc. Stability of these solutions was not reported.
B. MFTHODS OF ASSAYIEG BARBITURIC ACID DERIVATIVES
1. Procedures Employing Immiscible Solvents
a. The U. S. P. XI and U. S. P. XII Methods (70, 71)
The salt of the barbiturate (barbital sodium, pentobarbital sodium, and phenobarbital sodium) is dissolved in water and then precipitated as the free acid by the addition of diluted hydrochloric acid. The organic acid is extracted with ether. The ethereal solution is evaporated to dryness. The residue is dried to constant weight. In the U. 8. P. XI




39
assay (70), the sample is dried before extraction, whereas in the U. 8. P. XII assay (71), the percentage of barbiturate recovered is calculated on a moisture-free basis, employing a sample of known moisture content.
Phenobarbital Tablets and Phenobarbital Sodium Tablets of the
U. S. P. XII (71) are freed of such diluents as stearic acid by shaking with a saturated solution of sodium chloride and barium hydroxide. The volatile solvent employed in the extraction of the barbiturate is composed of eight volumes of chloroform and two volumes of ether.
b. Method of Rotondaro
Aspelund and Skoglund (7) separated the decomposition products
from deteriorated solutions of various barbiturates by filtration and extraction with ether at various pH levels, This work showed that all aqueous solutions may undergo appreciable decomposition. Analyses employing immiscible solvents require that the barbituric acid be free from its products of decomposition. Rotondaro (72) pointed out that the N. F. VI Method of assaying Elixir of Phenobarbital (75) eliminated the neutral decomposition product, phenylethylacetylurea, by discarding it along with the aromatics. The acidic decomposition products, phenylethylmalonuric acid and phenylethylacetic acid, are not separated from the barbituric acid. The presence of these acids arising from the decomposition of phenobarbital is indicated by the fact that the melting point of the barbituric acid residue is usually low. Rotondaro
(72) asserted that the contaminant was phenylethylacetic acid and not glycerin as had been expected by others when assaying Elixir of




40
of Phenobarbital N. F. VI (73).
Rotondaro (72) carried out an experiment in which he was able to
isolate several of the decomposition products of a barbiturate. A solution of phenobarbital in 0.2 N potassium hydroxide was refluxed for five hours. Upon dilution with water, a flocculent precipitate was formed. A portion of the mixture was extracted with chloroform. The residue obtained on evaporating the chloroform extracts was syrupy. However, on dissolving in alcohol and diluting with water, a white, flocculent precipitate separated out. Its melting point was the same as that of phenylethylaoetylurea. Another portion of the alkaline solution which had been refluxed was made acid to litmus with dilute hydrochloric acid.
Upon adding excess sodium bicarbonate and extracting with chloroform, Rotondaro (72) was able to secure a residue which had the melting point of phenobarbital. The remaining bicarbonate solution was acidified and extracted with chloroform. A syrupy residue was obtained which Rotondaro
(72) was unable to erystallise. However, by titrating a solution of the residue in alcohol with 0.1 N sodium hydroxide, a titration equivalent of 206 was obtained. The titration equivalent of a 11 mixture of phenylethylacetie acid and phenylethylmalonuric acid is 207.
Rotondaro (72) devised the following method to assay Elixir of
Phenobarbital N. F. VI (73):
"A 50 co. sample was diluted with about 60 cc. of water in a separator, saturated with sodium chloride, and made acid with dilute hydroohloric acid. The mixture was extracted to completion (six or seven times) with 25-35 co. portions of chloroform or about 35-50 co. portions of ether, if emulsions formed. The chloroform extractions were filtered




41
through chloroform-wet cotton into a beaker and the solvent was evaporated to a small volume (4-8 co.) on a steam bath with the aid of a brisk current of air, care being taken that active ebullition of the solvent did not take place. The solution was transferred with the aid of a little chloroform to a separator containing about 20 co. of alkaline salt solution (0.5 N sodium hydroxide saturated with sodium chloride). The mixture was shaken thoroughly and the chloroform was allowed to separate as completely as possible. The chloroform was drawn off into a second separator containing about 3 c. of 0.5 N sodium hydroxide. These were shaken well. The chloroform was allowed to separate completely and was then drawn off through chloroform-wet cotton into a tared dish. The extraction was repeated with three or four more 20 coc. portions of chloroform. The combined chloroform extracts were evaporated to dryness. A few treatments with anhydrous ether greatly assist in completely expelling the last traces of the chloroform solvent. The residue was dried at 90-1000 C. for ten to fifteen minutes. Residue "A" may contain non-volatile aromatic compounds plus any neutral decomposition products.
The alkaline wash water was combined with the alkaline salt solution. A piece of litmus paper was placed in the separator and acidified with dilute hydrochloric acid. An excess of dry sodium bicarbonate was added in small portions. The phenobarbital was completely extracted with 20-25 c. portions of chloroform. Each portion was filtered successively through chloroform-wet cotton into a tared dish. The solvent was evaporated with the usual precautions. The residue is usually syrupy due to traces of chloroform, which are tenaciously held until




42
superheating causes the chloroform to vaporize with explosive violence, thus causing losses by decrepitation. A few small additions of anhydrous ether materially shortens the time required to drive off the chloroform, minimize the loss by decrepitation, and leave dry, granular residues which possessed melting points very close to theory.
The sodium bicarbonate solution in the separator was then cautiously acidified with dilute hydrochloric aid. The mixture was extracted with three or four 20 cc. portions of chloroform or other appropriate solvent, filtered, and the solvent was evaporated as usual. The residue "C" was examined for decomposition or acidic drugs."
co. Method of van Itallie and Steenhauer (74)
The method of van Itallie and Steenhauer (74) is an immisciblesolvent method adapted for the quantitative determination of barbiturates in urine and organ extracts. In addition to the procedure which must be employed to remove extraneous matter, the only difference which is of interest in this discussion is the nature of the volatile solvent. These workers claimed that substituted barbituri acids are incompletely extracted by ether and that ethyl acetate eliminates this error by its greater solvent power. Veronal is more than twice as soluble in ethyl acetate as in ether. Mancani (75) described a procedure which was similar to that employed by van Itallie and Steenhauer (74).
d. Method of Glyeart (76)
The barbituric acid is dissolved in a 2% solution of sodium hydroxide which is saturated with sodium chloride* The solution is shaken




43
with ether. The ether washing is discarded. The contents of the separator are then acidified with hydrochloric acid and extracted with a mixture of two parts of ethyl alcohol, one part of ether and seven parts of chloroform* The extract is evaporated to dryness and weighed. The method of Glycart (76), with slight modifications, was adopted as the official method of assay for barbital and phenobarbital by the Association of Official Agricultural Chemists (77). The official method employs a solvent consisting of two parts of ether and eight parts of chloroform.
The method of Glycart (78) was also employed, with slight modification, in the N. F. VI assays of Elixir of Phenobarbital, Elixir of Barbital, and Tablets of Phenobarbital (73) and in the U. S. P. XII assays of Barbital Tablets and Pentobarbital Tablets (71). In the U. S. P. XII (71) the assay of Elixir of Phenobarbital is changed to an assay consisting essentially of acidification 'vith diluted hydrochloric acid and extraction with chloroform. In the h. F. VII (78) the assay of Elixir of Barbital was modified to the extent of using a solvent composed of one volume of ether and nine volumes of chloroform.
In addition to the above gravimetric methods, there are other types depending upon the formation of complex addition products, such as the complex products formed by the reaction of phenobarbital with mercuric iodide as reported by Montignie (79); the complex salts formed with iodine in an alkaline solution as reported by Bougault and Guillou (80)1 complex mercury salts as reported by Floury (81); and the copper pyridinyl phenobarbital formed by the use of Zwikker's reagent (82). Bell
(53) stated that most of the methods depending upon the formation of




44
complex addition products were unreliable, due probably to the inconsistent or unknown composition of the addition product.
2. Procedures Employing Indicators
Several indicators have been suggested for the direct titration of the replaceable hydrogen atom in the malonylurea group. The acid form of the barbiturate was dissolved in a solvent consisting wholly or in part of an organic substance.
Joensson (84) stated that veronal cannot be titrated directly with sodium hydroxide using phenolphthalein as an indicator. Explanation was made that the pH of a 0.1 molar solution of veronal sodium is pH 10.2 and that the p1 of a 0.01 molar solution of the same compound is pH 9.7, whereas the color change of phenolphthalein occurs at pH 8.2. This difficulty was overcome by the use of an organic solvent such as acetone. The method of Joeneson (84) consisted of dissolving the veronal in 20-25 co. of acetone, adding 0.10 to 0.15 oco. of 1:2000 phenolphthalein solution, and titrating to a faint red color with 0.1 N sodium hydroxide. The method of Sol'ts (85) for the determination of luminal differed only in the use of alcohol as a solvent.
Von Babitsoch (86) employed thymolphthalein as an indicator in
titrating veronal and luminal with sodium hydroxide because its transition limits of pH 9.3 to pH1 10.5 encompass the pH of veronal sodium and phenobarbital sodium. Von Babitsech (86) employed alcohol as the solvent. He obtained results which represented 99.77% of the samples taken. Gervay (87) also reported the use of thymolphthalein as an indicator for the titration of phenobarbital. The solvent in this case consisted of




46
methanol which was neutral to thymolphthalein. Gervay (87) suggested that alizarin yellow R, instead of thymolphthalein, be used as an indicator when titrating veronal. Belentska (88) modified the method of von Babitsch (88) to the extent of using a solvent consisting of two parts of alcohol and one part of water.
Morin (89) titrated barbituric acids, which were dissolved in acetone, with a methanol solution of potassium hydroxide in the presence of thymol blue.
Enell (90) titrated barbital sodium in ether using iodeosin as the indicator.
3. Conductimetrie Titration
Hrynakowski and Modrsejewski (91) stated that the application of conductimetric titration to the determination of weak acids and bases generally gave good results. They claimed to have determined barbital and phenobarbital with an average error of 0.2 to 0.5%. Substances insoluble in water but soluble in ethyl alcohol can be titrated in alcoholic solution. Better results were obtained, however, if the weak acids, after dissolving in excess of alkali, were titrated back with hydrochloric acid.
4. Procedures in which Undissociated Barbiturates are Formed.
(a) Mercury Salts
lonesco-Matiu (92) and von Babitech (86) developed procedures which are dependent upon the insolubility of the mercury salts of barbiturates in slightly acid media. The precipitant is prepared by dissolving




46
mercuric oxide in concentrated sulfuric acid and diluting with water. The precipitate formed by the addition of this reagent to the barbiturate solution is separated, washed, and dissolved in a mixture of sulfuric and nitric acids. The amount of mercury present is determined and from this the amount of barbiturate is calculated.
(b) Silver Salts
Budde (93) stated that the procedure of Stas-Otto (immisciblesolvent method) is troublesome and time consuming because of the repeated shaking out with ether.
Budde (93) claims that an alkalimetric determination of barbituric acid does not lead to satisfactory results. The probable presence of replaceable, enolio hydrogen atoms durinG titration with sodium hydroxide brings about only a slow change in hydrogen ion concentration, so that an exact end point can only be observed with difficulty and urder special conditions.
Budde (93) developed a volumetric determination which is based on an observation of Zernik that mercuric oxide is dissolved by veronal in a presumably alkaline media. Budde (93) assumed that a mercuric chloride of known strength may be titrated into an alkaline veronal solution until an excess of mercury ions is brought into reaction with the alkali with separation of mercuric oxide. Since a standard solution of mercuric chloride is not official in the Usutsohe Arsneibuch 6, Budde (93) proceeded to investigate whether or not silver nitrate solution acted in a similar way, It was found that in the presence of sodium hydroxide, 0.85 to 0.9 mols of silver nitrate was consumed by I mol of veronal upon the appearance of the first turbidity. Silver




47
nitrate is transformed by the sodium veronal to a silver barbiturate which is soluble in the alkaline media. Like mercury cyanide, the silver barbiturate is very difficultly dissociated, so that the excess of sodium hydroxide gradually and incompletely produces a separation of silver oxide.
Rupp and Poggendorf (94) reported that phenobarbital is precipitated by silver nitrate. The precipitate was found to be soluble in anmonia.
W
Viebock and Fucha (95, 96) gave the equations as follows
C121105N2Na Na0H + 2AgNO3 C12i1003N2Ag2 + 2NaNO3 + H20 (phenobarbital
sodium)
C81105)0Na + NaOH + 2AgN053---4 C8O1003NAg2 4 2NaN03 4 H20 (barbital-Na)
Tomski and Waller (97) gave the following as the course of the reactions
2 C12H1100NNa + 2AgNO3-4 C12"1103N2Na C12100N2Ag2 + NaN03+ N03O (a soluble, double compound)
C211 0NNa C1 1003N2Ag AgN034+2 C12 1005N2Ag2+ NaNO3 + N03
Tomaki and Waller (97) state that silver phenobarbital is insoluble, presumably meaning in acid or neutral media.
The first experiments of Budde (93) showed results which were 10 to 15% off. He thought that better results could be obtained by the use of weaker alkalies. However, later workers (83, 95, 96) indicate that sodium hydroxide may be used but the quantity must be regulated.




Budde (93) developed the following methods
About 0.2 to 0.3 Gm. of the sample are accurately weighed and dissolved in 30 cc. of water with 1 Gm. of anhydrous sodium carbonate. To the clear solution, 0.1 N silver nitrate is added until a distinct turbidity is produced, which remains for some time. Von Babitsch (86) found Budde's method (93) completely satisfactory.
The method employed by Kalinowaki (98) involves a titration of an alcoholic or acetone solution of the barbituric acid in the presence of sodium hydroxide. To a solution of 0.2 to 0.3 G*a. of the substance to be analysed in 20 to 25 coo of aeotone, or 30 to 36 co. of ethyl alcohol, 15 to 20 co. of a normal solution of sodium hydroxide and 20 to 30 co. of water are added. This clear solution is titrated with 0.1 N solution of silver nitrate until a turbidity is obtained. Kalinowski (98) claimed the method to be accurate to 0.1%.
Viebook and Fuche (95) explained that phenobarbital and barbital are monobasic toward alkalies and dibasic towards silver nitrate. If, to a solution of phenobarbital two equivalents of sodium hydroxide are added, then two equivalents of silver nitrate, the silver salt is procipitated and the solution is neutral.
C12 1105Na+ NaOH + 2AgNO3 -- C11005N2A 2+ 2NaN0S 20
When luminal or veronal is titrated with sodium hydroxide in the presence of excess silver nitrate, towards the end of the titration, silver hydroxide is formed which does not dissolve and which is easily concealed by the insoluble veronal-silver or luminal-silver. To overcome this difficulty, sodium salicylate is added. The latter replaces the excess




49
silver nitrate with silver salicylate. Silver salicylate has a modorate solubility, sufficient for a quantitative precipitation of veronalsilver or luminal-silver but on the other hand it hinders the formation of silver hydroxide.
The method of Viebock and Fuchs (95) for the determination of salts of barbiturates is as follows
A. Determination of alkali.
Dissolve 0.2 Gm. of the compound in 25 co. of water and titrate with 0.1 N sulfuric acid with methyl orange as the indicator,
This step is probably to regulate the quantity of alkali in the
solution, inasmuch as barbiturate salts occur on the market with varying amounts of alkali (4). In tho next step, a definite amount of alkali is added. This avoids the difficulty experienced by Budde (93) in his use of sodium hydroxide.
B. Determination of Phenobarbital and barbital.
The solution from part A is freed of carbonic acid by heating A measured quantity of 0.1 N carbonate-free alkali is added. For 7.5 co of acid used in the first titration, 14.0 to 14.5 co. of alkali are used in the second titration (generally 0.5 to 1.0 cc. less than double the quantity of acid used in the first titration). About 25 co. of 0.1 N silver nitrate is added (5 to 10 o, more than double the quantity used in the first titration). After a further addition of 0.5 to 1.0 Gm, of sodium salicylate, the solution is titrated with 0.1 N alkali, with phenopnthalein as the indicator. With an absolutely pure preparation, the quantity of alkali used in B must be exactly twice the quantity of acid used in A.




Fuchs (99) also employed the above method to assay diethyl barbituric acid by titrating the free acid, which was dissolved in alcohol, with alkali in the presence of thymolphthalein. The second part of the assay outlined above was unchanged.
U
It would seem from the work of Budde (93), Viebook and Fuchs (95, 96), and Tomski and Waller (97) that the barbituric acids are dibasio towards silver nitrate in neutral media and monbasic towards silver nitrate in alkaline media.
Madsen (3) and Nielsen (13) followed the course of hydrolysis of barbiturates from the amount of dialkylacetylures and carbon dioxide
U
formed. Schlemmer and Torber (30) state that this method is incomplete since hydrolysis of dialkylacetylurea is assumed to proceed further to produce diethylacetic acid and urea, as shown by the equations which follow.
H
CO--N
cCO 20 R o COOH I C02 R a I
O C CO *CO-C"0
CO-N RC "CO- N/ R /C1 CON-CO-NH2
H H H
R H R H
C
R CO-N-CO-NH2 H 0 R N Co00o + NH2CO NH2
H
NH2CONH2 12 2 NH3 + CO2
The values of percentage decomposition obtained by Madsen (3) are lower than those obtained when Sohlemmner and Torber (30 employed Buddets procedure (95). This provides evidence for the criticism of Schlemmer and




51
Torber (30), mentioned above.
Bellts (83) application of Kalinowski's modification (98) of the
Budde method (93) to the analysis of theobromine and phenobarbital mixtures is as follows #
The swiple containing theobromine and phenobarbital is boiled with
2 N sulfuric acid. The mixture is cooled to 400 C. and transferred to a separatory runnele The mixture is shaken out with alcohol-free ether* The extracts are filtered through cotton. The combined ether extracts are evaporated. The residue is dissolved in exactly 100 cc. of a solvent consisting of 50 cc. of ethyl alcohol, 10 co. of water and 40 co. of 1.0 N sodium hydroxide, The solution in filtered and the first quarter is discarded. The aliquot three-quarters remaining is titrated with 0.1 N silver nitrate until a distinct turbidity in evident* The silver nitrate must be added very slowly and toward the end point must be added drop by drop, shaking vigorously after the addition of each drop. The immediate turbidity caused by the addition of each drop must be cleared by shaking before the addition of the next drop. The end point is oharacterized by a persistent turbidity of the solution even
after vigorous shaking. The titrated solution at this point begins to change rapidly to a brownish-black color due to the separation of silver oxide.
Bell (83) states that the alkalinity of the solution of extracted phenobarbital should not be varied appreciably from the concentration
mentioned because the sharpness of the end point is influenced by this factor.
Schulek and Rossa (100, 101) devised an argentometrie assay of




52
barbitur&tes with potassium chromnate as the indicator. The barbiturate (0.1-0.15 Ga.) was dissolved by boiling in 25 co. of a 5% borax solution. Potassium chromate (I co. of a 10 solution) was added. Silver nitrate, 0.1 N, was introduced into the hot mixture until the yellowish-green, turbid liquid attained a reddish color, which persisted even after boiling. Since Schulek and Roszza (100, 101) stated that 1 cc, of 0.1 U silver nitrate is equivalent to 9.250 mg. of barbital, it may be concluded that both hydrogen atoms attached to the nitrogen atoms were replaced.
5. Colorimetric Methods
(a) Mcthod of Zwikker (102).
Barbiturates react with cobalt chloride in anhdrous media to
give a red color. Addition of barium oxide converts this to blue. A red dibarbitalooobaltodiamine has been isolated by Zwikker (103).
(b) Method of Oettel (105).
The sample is extracted with chloroform and filtered. The solution is diluted as necessary. Two (2) cc. of the sample is transferred to each of three test tubes. Then 0.05 coc., 0.1 co. and 0.16 cc. of 0.2% cobalt acetate solution in absolute methanol are added. After shaking, to the same respective tubes 0.05 co., 0.1 cc. and 0.15 co. of a 0.2% solution of lithium hydroxide in absolute methanol are added. These volumes must be accurately measured.
If all three tubes or only the last two give an intense blue color, the chloroform solution contains more than 0.1 mg. per co. of a barbiture acid derivative. If the test is positive in the first two tubes




and negative or disappears within a minute in the last, the chloroform extract contains about 0.05 mE. of a barbituric acid derivative. if only the first is positive, the chloroform extract contains about 0.026 rLC. of a barbituric acid d-rivativo. With practice, fair accuracy can be attained. By comparison with standards greater accuracy should be possible.
If all samples are positive or the two most concentrated are positive, dilute the chloroform extract. A large excess of barbituric acid derivatives cer give a negative reaction or a gray color in all of the tubes. In that case, dilute the extract with chloroform until the blue color is obtained.
(o) Method of Koppanyi et al. (108).
The oolorimetric estimation is based on the blue color produced by cobalt acetate in alkaline solution# When the test is performed
with 1% cobalt acetate and 1% barium hydroxide in absolute methanol, it is sensitive to one part of barbital in 10.000 parts of solution* When the test is performed with 0.2O% cobalt acetate and 0.2O% lithium hydroxide in absolute methanol, it is directly sensitive to one part
of barbital in 100,000 parts of solution and 1 in 2,000,000 after concentration of the chloroform solution twenty times. The accuracy of this method is limited to about 6%.




54
III EXPERIMENTAL PART
A. SCOPE OF INVESTIGATION
The products formed during the decomposition of barbiturates and the manner in which they are produced constituted the research of sevoral workers, namely, Steenhauer (2), Kadsen (3), Bailey (4), Nielsen (10, 11, 13) and Aspelund and his workers (7, 8, 9). From the agreement in the results which were reported by these men, the nature of the decomposition appeared to be amply explained. However, it was felt that an investigation of a number of factors in the deterioration
of phenobarbital and phenobarbital sodium, together with research on several methods of retardation, would add to the existing knowledge concerning tasse phenomena.
Evaluation of several reported methods of assay were made in an
attempt to ascertain their applicability to such a study. Methods employing immiscible solvents and methods employing silver nitrate were studied. Suitable methods of determining the percentage moisture were sought inasmuch as its accurate determination was essential in preparing solutions and in checking the accuracy of the assay used in the experiments on deterioration and stabilization. The experiments on the deterioration of Phenobarbital Sodium U. S. P. XI in various concentrations included the determination of volume of precipitate, change in pH, change in refractive index, and percentage loss of barbiturate, when the solution was subjected to heating at various temperatures for various periods of time. The relation between pH and deterioration was studied by making use of sodium carbonate and phosphate buffers to




55
regulate the pH. A number of suggested stabilizing agents were subjected to tests and assays to determine their influence on the retardation of deterioration.
B. ANALYTICAL METHODS AND CONTROL
1. Materials used
The phenobarbital was purchased from the Mllinokrodt Chemical
works. It was labeled "Phenobarbital U. S. P. XI, Control HRCl." Data given elsewhere in the discussion showed that the drug conformed to the standards of the U. S. P. XI and U. S. P. XII as regards moisture content and melting point. The phenobarbital sodium was likewise purchased from the Mallinekrodt Chemical Works. Two lots were obtained. The labels read "Phenobarbital Sodium U. S. P. XI." The control designations were KKM and K.LM. Data given subsequently revealed that these lots conformed to the standards of the U. S. P. XI and U. S. P. XII as regards moisture content and purity rubric. The particular lot which was used in an experiment is mentioned in the review of the experiment. The source of the other chemicals employed is mentioned in the discussion of the experiments concerned.
2. Determination of Moisture
Unless otherwise stated, the samples of phenobarbital sodium which were directly used in an experiment were of known moisture content. quantitative determination of phenobarbital sodium was calculated on a moisture-free basis, i.e., a lot of drug whose moisture content had previously been determined on separate samples. As subsequently discussed, it was deemed advisable to dry the samples to constant weight




56
at temperatures slightly above 1000 C. The values obtained by drying samples in a vacuum desiccator approached those obtained by drying samples at the latter temperature. The term, dried to constant weight, was interpreted to mean that two consecutive weighings do not differ by more than 0.1 per cent when the second weighing is made after an additional hour of drying* Samples not exceeding 0.8 Ga. were taken for the moisture determinations which were made at approximately 1000 Co This was done in order to assure complete drying by exposing a thin layer of drug to the surrounding air in the oven. A Freas Thermo Electric Oven was the drying chamber employed,
3. PH Determination
The Beckman pH Meter, laboratory model G, was employed in all the determinations of pH, For measurements above pH 9 the Beckman, high-pH, No. 1190-E, glass electrode was used. This electrode is intended for measurements made on alkaline solutions of high pH at room temperature. In highly alkaline solutions this electrode does not attain equilibrium instantly but may indicate a gradually increasing pH for five minutes or more at approximately 260 C, The time required for attainment of equilibrium was shortened by keeping the tip of the glass electrode immersed in pH 10 buffer, while not in use. The asymmetry potential also tends to change slightly during measurement' acting to reduce the pH reading* The usual result is to give a pH curve against time that reaches a maximum in about five minutes followed by a slow decrease in reading. Standardization of the electrodes and meter against a pH 10 buffer was made after every two readings, frequently after each reading When the glass electrode attained equilibrium slowly, pH readings were




57
made at two or three minute intervals until no further upward drift was noted. The buffer standardisation was then made. If the latter reading differed from the p11 value of the buffer, the difference between the two wau either subtracted from or added to the last reading of the sample, as required to make the proper correction. In the event the equilibrium was rapidly attained, the maximum pH1 reading was taken as correct for the sample.
The Beckman 015 type glass electrode was used for measurements below PH 9.
4. Determination of Volume of Precipitate
Inasmuch as the amount of precipitate formed in heated solutions of phenobarbital and phenobarbital sodium became greater with increase in temperature, a method of approximating the volume of precipitate was devised. 'With shaking, the contents of the bottle, which had been subjected to heating, were transferred to a beaker* While the mixture was rapidly stirred by means of an electric stirrer, a convenient volume, usually 10 coo, was withdrawn by means of a pipette. The contents of the pipette were placed in a graduated centrifuge tube. If the mixture could not be drawn into a pipette, the contents of the bottle were Introduced with shaking Into a centrifuge tube. In either case the tube was centrifuged at approximately 1200 ropem. for five minutes as timed by R stopwatch. Time was taken the moment the current was turned on and the moment it was shut off. It took approximately twenty seconds for the centrifuge to cease revolving,
It is not purported that this determination should be taken as
quantitative. However, the method is considered more satisfactory than




58
the use of such words as "large," "moderate," or "small" in describing the quantity of precipitate.
5. Use of the Autoclave
To attain the higher temperatures to which some of the solutions were subjected in this investigation, the bottles containing the solutions were exposed to saturated steam under pressure. The apparatus employed was a "Peerless" pressure cooker. The following procedure was resorted to in order to gauge the time during which the solution was exposed to a certain temperatures the autoclave was heated until steam was issuing strongly from the petcock. The lid was removed and the bottles were introduced into the chamber. The lid was immediately replaced and clamped down. The autoclave was rapidly heated by several Tucker burners in order to quickly saturate the space in the chamber with steam. As soon as all the air was replaced with steam, the petoock was closed, causing the pressure to rise rapidly. When the desired pressure was attained, time was taken by means of a stopwatch. The Meker burners were then shut off and the pressure was maintained at a constant level by means of a low bunsen flame. The variance of temperature was z 10 C. When the period of tire during which the solution was supposed to have been exposed to heating had elapsed, the petcock was opened completely to permit rapid escape of steam. The autoclave was quickly opened and the bottles were immediately inserted in a mixture of ice and water to rapidly cool the solutions.




C. A COMPARISON OF ETO(DS OF DSYIEC PH OBRBITAL SODIUM PREPAEATORY TO ASSAY
The purpose of comparing various methods of drying phenobarbital sodium to constant weight was to detect the presence of decomposition which might occur during drying. Phenobarbital sodium absorbs considerable moisture from its surroundings due to its hygroscopic nature. Inasmuch as solutions of phenobarbital sodium undergo hydrolytic decomposition, there is a possibility that the compound may suffer hydrolysis when heated in the solid state, particularly if it contains an appreciable quantity of moisture. A method of drying to constant weight without producing decomposition was necessary in determining the dry weight of samples which were used in the study of various methods of assay of phenobarbital sodium and in the study of decomposition of solutions of the compound. In the latter case it was essential to propare solutions of definite concentration from samples whose dry weight was known. Not only could the concentration of solutions be determined from the amount of dry sample present in a unit volume but the consistency of the method of assay and the per cent of phenobarbital sodium present in the dry sample could be calculated.
The drug employed was purchased from the 14allinckrodt Chemical %orks. It was labeled Phenobarbital Sodium U. S. P. XI and it bore the control number I1l. Samples were heated to constant weight in an oven at 1000 C. These samples were then heated to constant weight at 1400 0. to determine whether or not an additional loss in weight would result. Additional samples were heated at 1400 C. Other samples were spread out in petri dishes and kept in an evacuated desicator




s0
for one month. Samples which had been heated were, assayed by the !3udde method to detect the presence of decomposition*
The IF. S. P. XI limits the amount of moisture in phenobarbital sodiuma to seven per cant as determined by heating the salt at 1400 C. for six hours# It was stated by van Leont (27) that heating phenobarbital sodiu at 1200 C. produced a product which gradually became incompletely soluble. N~o dif ficulty was experienced during this investigation in dissolving samples which had been dried to constant weight at 142 (271)0 C. during a period of ton hours*




TABLE 14
DETERMINATION OF MOISTURE IN PIFNOBARBITAL SODIUM U. S. P. XI
PART A. LOSS IN WEIGHT ON HEATING PHENOBARBITAL SODIUM U. S. P. XI TO CONSTANT WEIGHT AT 101 (2)0 C.
Weight in Gm. Duration of Heating Loss of Weight Percentage Loss in Hours in (m. of Weight
(1) 2.6459 12 0.0459 1.7
(2) 3.5706 12 0.0580 1.7
(3) 1.o90 8 0.0191 1.8
(4) 1.1427 6 0.0199 1.7
(b) 1.0771 8 0.0188 1.8
(6) 0.8437 6 0.0147 1.7
Average 1.7
PART B. ADDITIONAL LOSS IN WEIGHT ON HEATING THE ABOVE SAMPLES TO CONSTANT WEIGHT AT 141 (-1)0 C.*
Weight in Gm. Lose of Weight Peroentage Loss Total Loss of in Ga. of Weight Weight
(1) 2.6000 0.0443 1.7 3.4
(2) 3.3126 0.0555 1.7 3e.4
(3) 1.0699 0.0187 1.7 3.5
(4) 1.1288 0.0192 1.7 3.4
(5) 1.0583 0.0182 1.7 3 .4
(6) 0.8290 0.0140 1.7 3.4
Average 1.7 3.4
*Six hours of heating were required to attain constant weight




82
TABLE 15
LOSS IN WEIGHT Oi EATING P ENOBARBITAL-SODIUM U. S. P. XI TO CONSTANT WEIGHT AT 141 (l1)o C.*
Weight in Gm. Loss of Weight in Gm. Percentage Loss
of Weight
(1) 1.1225 0.0380 3.4
(z) 1.0635 0.0366 3.4
(3) 1.0041 0.0341 3.4
(4) 1.0206 0.0351 3.4
(5) 1.1007 0.0377 5.4
Average 3.4
*Seven hours of heating were required to attain constant weight.
TABLE 16
LOSS IN WEIGHT ON VACLUM DESICCATION OF PHENOBARDITAL-SODIUM U. S. P.
XI*
height in Gm. Loss of Weight in Gm. Percentage Loss
of Weight
(1) 11.5408 0.1700 1.5
(2) 11.6715 0.1676 1.4
(3) 13.0285 0.1901 1.05
Average 1.5
*The samples were kept for 31 days in a vacuum desiccator evacuated to approximately 0.5 mm Hg.




In the three preceding tables it is seen that the lot of Phenobarbital Sodium U. S. P. XI employed in this study decreased 1.7 per cent in weight on drying to constant weight at 101 (t2)0 C. A further 1.7 per cent loss was undergone upon heating the same samples to constant weight at 141 (1)0 C. The total loss in weight was 3.4 per cent, which figure was likewise obtained on drying samples exclusively at 141 (1)0 C. On drying samples in an evacuated dessicator for one month, the loss in weight was 1.5 per cent or a value approaching that
obtained on drying phenobarbital sodium to constant weight at 101 (-2)0 C. These results indicated either incomplete drying at approximately 1000 C. or the presence of decomposition at approximately 1400 C.
To determine whether or not decomposition occurred during the
drying of samples at the aforeto mentioned temperatures, dried and umdried samples were assayed by the Budde method.




G4
TABLE 17
BUDDE ASSAYS OF DRIED AND UNDRIED PfiFNOBARBITAL SODIUM U. S. P. XI PART A
Sample Sample After Loss in Per Cent
Before Drying at W~eight 0.1N AgNO3 C12HU110YT& of
Drying 103 (1l)0 C. Per Cent CooGm plred
Gm.0S!Ml
0.4588 0.4432 3.4 17.05 0*4332 94.4
0.6120 0.5913 3.4 22.86 0.5758 94.4
0.3506 0.3387 3.4 12.93 0.3286 93.7
AT* 3.4 Ave 94.2
F-ART B
Sample Sample After Lose in Per Cent
Before Drying at Weight 0.1N AeTO3 C12H1103hN21& of
Drying 142 (:ti)0 C. Per Cent Cc G. S1die
Gm. aml
094558 0.4337 4.9 16.78 0.4287 94.1
0.4463 0s4246 4.9 16.50 0.4193 93.9
0*b646 0.5370 4.9 20.83 0.5293 93.8
Av. 4.9 Avo 93.9
PART C
Tndriod Sample 0.1M AgNO3 C12H1103N~a, Per Cent of
Gm, 00. Gm. Undried Sample
0.2821 10.40 0.2643 93.7
0.2634 9.74 0.2474 93.9
0.3211 11.79 0.2996 93.3
Av. 93.8




65
On the basis of undried sample, 94 per cent phenobarbital sodium was present in samples of Phenobarbital Sodium U. 8. P. II before and after drying at 103 (1)o C. and at 142 (1)o C. Therefore, the presonce of decomposition in samples heated at approximately 1000 C. and approximately 140 C. was not shown by the Budde Assay.
Throughout this investigation undried samples were used exeoopt
where otherwise indicated, The equivalent weight of dry sample, upon which the percentages were usually based in this study, was calculated fromr loss of weight sustained on heating other samples to constant weight at temperatures slightly above 1000 C. unless otherwise indicated. This loss in weight at approximately 1000 C. was employed in preference to that sustained at approximately 1400 C. because of possible decomposition which may ooour at the higher temperature.
D. EVALUATION OF REPORTED METHODS OF ASSAY
Several methods of determining the percentage deterioration of
barbiturate solutions have been reported. Nielsen (13) and Madsen (3) followed the process quantitatively by measuring the carbon dioxide
R
and dialkylaeetylurea formed. Sohlemmer and Torber (30) found that the results obtained by such a method were too low because of the existenee of partial decomposition of dialkylasetylurea. Tomaki and Waller (24) determined the amount of undeoomposed barbiturate by titration with a standard solution of silver nitrate in the presence of sodium carbonate. Rotondaro (72), Glycart (70) and Aspelund and Skoglund (7) separated various decomposition products from one another at various pH levels. In addition other methods were reported but their




use was mentioned in connection with assays of undeoomposed barbiturates. The methods employing an immiscible solvent at a single pH level were considered unsatisfactory for the investigation. The methods in which the alkalinity or acidity of the barbiturate derivative was measured by simple titrating with a standard acid or alkali were not applicable because of the presence of alkaline decomposition products in deteriorated solutions. Colorimetric methods were not investigated because of their reported inaccuracies.
The method of Rotondaro was chosen as a method worthy of investigation because a quantitative separation of the undecomposed barbiturate was claimed. The U. S. P. XI method (70) was studied for the sake of comparison. The Budde (93), Viebook and Fuchs (95, 98), and Sohulek and Rossa (100, 101) methods exemplified the several methods employing silver nitrate. The possibility of applying the von Babitech titration
(86) to the determination of the quantity of the free barbituric acid in the residues recovered in some immiscible-solvent methods was investigated.
1. Comparison of U. S. P. X1 and Rotondaro Assays of Phenobarbital
Sodium U. S. P. XI
The purpose of the experiment was to compare the U. S. P. XI assay (70) with the Rotondaro assay (72) and a modified form of the Rotondaro assay.
The U. S. P. XI (70) assay of phenobarbital sodium is given as follows "Dissolve about 0.5 Gm. of Soluble Phenobarbital, dried at




67
1400 C. for six hours, and accurately weighed, in 60 co. of distilled water in a separator. Add 10 co. of diluted hydrochloric acid, and completely extract the liberated phenobarbital with successive portions of 26 *a. each of ether. Evaporate the combined ethereal extracts in a tared dish, and dry the residue to constant weight at 1000 C. The weight of phenobarbital so obtained is not less than 90.4 and not more than 91.4 per cent of the weight of the Soluble Phenobarbital taken for the assay." The assay was modified to the extent of using undried samples of known moisture content. This was done to avoid the possibility of incurring decomposition during the drying of the sample.
The Rotondaro method (72) was discussed in the review of the literature. The a.ethod was modified in some of the assays for the purpose of simplification and elimination of details which did not have to be retained in following the decomposition of barbiturate solutions. Instead of acidifying at the start, the solution of the barbiturate was treated with a proportional amount of normal sodium hydroxide equivalent to the amount of alkali which Rotondaro (72) used to extract the acidic compounds from the chloroform extracts containing both the barbiturate and the decomposition products. The volume of solution prepared or taken for the assay was such that the dilution with the normal solution of sodium hydroxide resulted in a concentration of alkali approximating a 0.5 N sodium hydroxide. At the same time the ratio between the total amount of alkali and barbiturate was maintained as found in the original assay. The alkaline solution was extracted with chloroform to remove the neutral decomposition product, phenylethylaoetylurea. The net result of the modification was to extract the




68
neutral decomposition product from an alkaline solution instead of leaving this product behind in the chloroform on extracting the latter, which contained both the barbiturate and the decomposition products, by means of an alkaline solution. The remainder of the assay was unaltered. Residue I indicates phenylethylaoetylurea, residue II indicates phenobarbital, and residue III indicates the acidic decomposition products, such as phenylethylmalonurie acid and phenylethylaetic acid. The shaking device designed by Mittelstaedt, Horn, and Kaufan (107) was employed to perform the shaking-out with immiscible solvent as required in the experiments subsequently mentioned. Each period of shaking-out was not less than ten minutes in duration.
All the residues obtained by the Rotondaro procedure (72) were
dried to constant weight at temperatures slightly above 1000 C., suecessive weighings being made at fifteen minute intervals.
Samples of Phenobarbital Sodium U. S. P. XI, possessing the Mallinckrodt control designation KLM1, were employed in this study of evaluation of assays. The U. 8. P. XI (70), Rotondaro (72), and modified Rotondaro assays were performed on samples of known moisture content. In addition a number of modified Rotondaro assays of samples which had been dried at 141 (t1)o C. were conducted. The results are summarized in the tables which follow.




TABLE 18
U. S. P. XI ASSAY OF PHENOBARBITAL SODIUM U. S. P. XI*
Calculated Phenobarbital Exitracted Calculated
Day Sample C12il1105N2N
Gm, am,. am.
(1) 0.7157 0s646 90.4 0.707 99.00
(2) 0.5149 0,468 9008 0.512 99.4
(3) 0.6573 0.596 90.7 0.655 99.3
*Undriod samples were employed* The percentage# were calculated on a moisture-free basis* Separate samples, which were dried at 102 (.l)O C., had undergone a loss In weight of 2.2%.




70
TABLE 19
ROTONDARO ASSAY OF Pd;NOBARBITAL SODIUM U. S. P. XI* PART A
Calculated Residue I Residue II** Residue III
Dry Sample
Gm. Gm. % Gm. Gm. %
(1) 0.3411 0.004 1.2 0.263 77.0 0.033 9.5
(2) 0.3179 0.004 1.3 0.253 79.7 0.023 7.3
(3) 0.3160 0.005 1.6 0.2650 79.2 0.023 7.3
PART B
Calculated Sum of the Sum of I, III, with II
C12H103N2Na Three Residues Calculated as C121103N2Na
Gm. % Gm. %Gm. %
(1) 0.287 84.3 0.299 87.7 0.324 95.0
(2) 0.277 87.2 0.280 88.2 0.304 95.7
(3) 0.274 86.7 0.278 88.1 0,302 95.6
*Undried samples were employed. The percentages were calculated on a moisture-free basis. Separate samples, which were dried at 102
(tl)O C., had undergone a loss in weight of 2.2%.
**The quantity of phenobarbital constituting Residue II was determined by the von Babitsch titration (88) which is discussed in the section entitled "Evaluation of Reported Met-ods of Assay."




71
TABLE 20
MODIFIED ROTONDARO ASSAY OF PHENOBARBITAL SODIUM U. S. P. XI* PART A
Calculated Residue I Residue II Residue III
Dry Sample
Gm. Gm. % Go f Gm. %
(1) 0.4951 0.006 1.0 0.590 78.7 0.048 9.2
(2) 0.4907 0.008 1.1 0.5681 73.5 0.017 3.5
(3) 0.5004 0.010 1.9 0.368 75.5 0.028 5.2
PART B
Calculated Sum of the Sum of I, III, with II
C1211031N2Na Three Residues Calculated as C12H1105Nga
(1) 0.427 88.2 0.440 88.9 0.477 94.2
(2) 0.395 80.4 0.383 78.1 0.418 85.1
(3) 0.403 80.5 0.404 80.7 0.439 87.8
*Undried samples were employed. The percentages were calculated on a moisture-free basis. Separate samples, which were dried at 100
(1)0 C., had undergone a lose in weight of 2.2%.




72
TABLE 21
ZADIPIED ROTONDARO ASSAY OF ? PNOBAR-ITAL SODIUM U. 3. P. XI 'N'ICH HAD
BEEN DRIED TO COINSTANT W-EIGliT AT 141 (t1)0 C. PART A
Dry Sample Residue I Residue 11 Residue III
Ga. 0m. GIM. %Ga.
(1) 0.9700- 0.000 0.0 0.824 85.0 0.063 8.5
(2) 0.9856 0.003 0.3 0.852 84.8 0.080 6.1
(3) 1.0843 0o907 83.7
(4) 1.0269 0.877 85.4
PART B
Calculated Sun of the Sum of I, III, with II
C1HI03'291 three residues calculated as C12Hll03N2N&
Ga. % Gm. % am*.
(1) 0.903 93.0 01888 91.6 0*966 99.6
(2) 0.911 92.5 0.895 90.9 0o974 ago8
(3) 0.993 91.6
(4) 0,961 93.5 --* The average percentage lose In weight on heating these saples was 3.4%.




73
The phenobarbital sodium employed in these assays conformed to the requirements of the purity rubric of the U. S. P. XI inasmuch as 91% phenobarbital was extracted during the performance of the U. S. P. XI assay. By means of the Rotondaro method (72), 78.8% phenobarbital was extracted. y means of the modified Rotondaro method, 75.2% phenobarbital was extracted. This indicated the existence of several possibilitiess (a) the original sample was partially decomposed; (5) docomposition took place during the modified and unmodified Rotondaro assays; and (o) incomplete extraction, and/or failure of the assay to separate the decomposition products quantitatively without admixture. The presence of the third possibility was further indicated by the sum of the three residues obtained in the modified Rotondaro assay of samples dried at 141 (ti)0 C. This sum is comparable to the residues obtained in the U. S. P. XI assays inasmuch as the latter does not separate the decomposition products. In the latter set of Rotondaro assays, the sum of the three residues came to 91.2 0 This figure, representing the best results obtained, is within the U. S. P. XI tolerance of 90.491.4%; in the assays given in Tables 19 and 20, the sum of the three residues was lower, showing inefficient extraction.
2. Comparison of Modified Rotondaro Assays of Phenobarbital and
Phenobarbital Sodium
In an attempt to learn why the residues of phenobarbital had been lower in the Rotondaro assays, both modified and unmodified, than in the U. S. P. XI assays performed on phenobarbital sodium, determinations of Phenobarbital U. S. P. XI and phenobarbital, recovered from previous




74
Rotondaro assays, were conducted*
The Phenobarbital U. S. P. XI which was used in this experiment had a melting point of 1760 C# The melting points of the individual residues, which made up the mixture employed in one set of assays, varied from 174.50 C. to 181.00 C. The U. S. P. XII (71) gives the melting point of phenobarbital as 174-1780 C.
In addition the modified Rotondaro assay of Phenobarbital Sodium U. S. P. XI was repeated on the same lot of drug used in the foregoing experiment, namely, allinckrodt KLMI. The number of treatments with chloroform to extract each residue was increased from the number originally employed by Rotondaro (72), namely, six or seven, to twelve, in an attempt to increase the efficiency of extraction. Each period of shaking-out with the mechanical shaker was not less than ten minutes in duration.




75
TABLE 22
MODIFIED ROTONDARO ASSAY OF ?K~O3ARBITAL SODIUM U. S. P. XI*
Sample Number 1 2 5
Calculated Dry
Sample Gm. 1.1545 1.1257 1.3352
Residue I
Ga. 0.012 0.037 0.028
per cent 1.0 5.3 2.1
Residue II
GM. 0.810 0.800 0.949
per cent 70.2 71.1 71.1
Residue III
Gm. 0.163 0.141 0.183
per cent 14.1 12.5 13.7
Residue II
calculated as
C12H1103N2Na
GM. 0.888 0.876 1.040
per cent 76.9 77.9 77.9
Sum of the three
residues with II
calculated as
C12H1105N2Na
Gm. 1.062 1.054 1.250
per cent 92.0 93.6 93.6
*Undried samples were employed. The percentages were calculated on a moisture-free basis. Separate samples, which were dried to constant weight at 106 (-1)o C., had undergone a loss in weight of 4.2%.




76
TABLE 25
MODIFIED ROTONDARO ASSAY OF PH1TOBARBITAL U. S. P. XI*
Sample Number 1 8 3
Calculated Dry
Sample Gm. 1.1600 1.2626 1.3855
Residue I
GMn. 0.007 0.013 0.013
per cent 0.6 1.0 0.9
Residue II
Gma. 0.876 0.978 1.078
per cent 75.6 77.4 77.9
Residue III
Gm. 0.180 0.166 0.182
per cent 15.5 13.1 11.7
Sum of the three
residues
Gma. 1.063 1.156 1.253
per cent 91.7 91.5 90.5
*Undried samples were employed. The percentages were calculated on a moisture-free basis. Separate samples, which were dried to constant weight at 108 (1)o 0., had undergone a loss in weight of 0.2%.




77
TABLE 24
'AJODIIED ROTODARZO ASSAY OF PT,OARI TAL*
Sample Number 1 2 3
Sample W.eight
Gm. 0.7348 0.7026 0.6568
Residue I
Im. 0.205 0.018 0.021
per cent 3.5 2.5 3.2
Residue II
Gm. 0.531 0.538 0.493
per cent 72.2 76.6 75.0
Residue II
Gm. 0.121 0.107 0.076
per cent 16.4 15.2 11.6
Sum of the three
residues
Gm. 0.676 0.662 0.590
per cent 92.1 94.3 89.8
*The phenobarbital employed in these assays represented a mixture of dry phenobarbital residues obtained from previous Rotondaro assays. The melting points of the individual residues varied from 174.50 C. to 181.00 C.




78
By means of the modified Rotondaro assay, 70.80 phenobarbital
was extracted from samples of Phenobarbital Sodium U. S. P. iI. The sum of the three residues averaged 86.4$. These results further substantiated the conclusions reached in the last experiment.
Modified Rotondaro assays of Phenobarbital U. S. P. XI yielded 77.0* recovery of phenobarbital. Modified Rotondaro assays of phenobarbital which had been previously obtained in other Rotondaro assays yielded 75.8/ recovery of phenobarbital.
Although the number of treatments with chloroform to extract each residue was increased, the sum of the three residues obtained in the modified Rotondaro assays of Phenobarbital Sodium U. S. P. XI was be* low the U. S. P. XI tolerance of phenobarbital extracted. Moreover, the average of the sun of the three residues obtained in the modified Rotondaro assays of Phenobarbital U. S. P. XI and phenobarbital which had been previously obtained in the other Rotondaro assays, was only 91.7%. When 75.8% phenobarbital remains undecomposed, the loss of carbon dioxide from phenobarbital amounts to 2 .9% of the phenobarbital taken. The difference between the sum of the three residues given in Table 24 and 10/o recovery is only partially accounted for by the loss of carbon dioxide.
From the results obtained in this experiment, the existence of several possibilities in both the modified and unmodified Rotondaro assays is indicated (a) occurrence of decomposition during the assay, (b) incomplete extraction, and (c) failure of the assay to separate the decomposition products quantitatively without admixture. Rotondaro (72) had claimed that 98% of the phenobarbital actually present




If9
in a seazple could be recovered by the use of his assay.
From the results obtained, it was concluded that the Rotondaro method, modified or unmodified, was not suitable for the quantitative determination of deterioration occurring in solutions of phenobarbital and phenobarbital sodium.
3. A Study of Assays Employing Silver Nitrate
The methods of Sudde (93), Viehock and Fuchs (95, 96), and Schulek and Rotsa (100, 101) exemplify the reported methods of assay employing silver nitrate for the determination of barbiturates. A comparison of results obtained by applying these assays on phenobarbital sodium constituted the purpose of this experiment. In addition an insight into the chemistry of the Budde assay (93) was sought.
The Budde assay (93) was conducted in the following ways About
0.2 to 0.3 Gm. of phenobarbital sodium was accurately weighed and dissolved in 30 cc. of water, together with 1 Gm. of anhydrous sodium carbonate. Andydrous sodium carbonate of analytical reagent quality was employed throughout the investigation whenever this compound was called for. The solution was prepared in a 50 cc. beaker which was supported by a universal clamp, attached to a ring stand at a convenient height. An electric stirrer was used to stir the solution vigorously throughout the titration. By means of a microburet, 0.1 N silver nitrate was introduced drop by drop. The dropping was regulated so that the turbidity produced by a drop of silver nitrate had disappeared before the introduction of the next drop. It was found that a time interval of one to




30
two seconds was sufficient to permit clearance of the solution. A stopwatch was employed to assist in maintaining a constant rate of introduction of the silver nitr-Ite. The rate was maintained at one drop every one or two seconds. The apparatus was set up in a darkened room. A strong beam of light emitted by a microprojector was directed through the solution. The solution was viewed at a right angle to the ray of light. In this manner a faint trbidity was easily detected which would not be observable in diffused light. The appearance of a distinct turbidity which persisted for at least one minute constituted the end point, Budde (93) explained that a silver compound of the barbiturate was formed which was soluble in this media. Budde (93) stated that the silver barbiturate was very difficultly dissociated. This elucidated why the silver nitrate reacted with the barbiturate instead of reacting with the sodiun carbonate. Silver carbonate and silver oxide are well dissociated. The silver ions are removed from the field of reaction by the formation of a poorly dissociated organic compound, namely, the silver barbiturate derivative.
In order to learn more concerning the assay, several tests were
carried out. To show that silver phenobarbital is only slightly ionized in an alkaline media, a solution of Phenobarbital Sodium U. S. P. XI, was treated under conditions existing in the Dudde assay. To 20 co. of a 1.3% solution of Phenobarbital Sodium U. S. P. XI, 5 oc. of water,
0.83 Gm. of anhydrous sodium carbonate, and 5 cc. of 0.1 N silver nitrate were added. The resulting solution was clear as observed with the aid of a strong beam of light. Upon the addition of 1 co. of 1% sodium chloride, no turbidity was produced as observed under the same conditions.




81
,hen 1 c. of 1% sodium chloride was added to a mixture of 25 cc.' of water and 5 c. of 0.1 N silver nitrate, a very marked turbidity was evidenced. These results showed that silver phenobarbital is weakly dissociated.
The solubility of silver phenobarbital was investigated. The addition of 0.02 co. of 0.1 N silver nitrate to a saturated solution of Phenobarbital U. S. P. XI resulted in the formation of a distinct turbidity which did not disappear after five minutes of stirring. The addition of 0.83 Gm. of anhydrous sodium carbonate caused a disappearance of the turbidity. The addition of sufficient silver nitrate solution again restored the turbidity. It was concluded that silver phenobarbital is insoluble in water but soluble in the presence of sodimn carbonate. The appearance of a turbidity on adding silver nitrate the second time was due to reaction between the excess silver nitrate and the sodium carbonate. The removal of silver ions from the field of reaction was shown by the disappearance of turbidity on adding a saturated solution of Phenobarbital U. S. P. XI to a suspension of silver carbonate which had been formed by adding 0.02 cc. of 0.1 N silver nitrate to a solution of 0.83 Gm. of anhydrous sodium carbonate in 25 co of water.
To determine whether or not the pH was lowered when phenobarbital was added to a suspension of silver carbonate, the following experiment was conducted The conditions existing in the Budde assay (93) were duplicated. Determination of pH was made at various stages. The pH of a 3.3% solution of anhydrous sodium carbonate wam 11.56 at 250 C. Upon adding 0.02 cc. of 0.1 N silver nitrate, the pH was found to be




82
11.54 at 250 C. The solution was distinctly turbid. The pH after addition of 1 co. of a saturated solution of Phenobarbital U. S. P. XI was found to be 11.56 at 250 C. Therefore, suspensions of silver carbonate were cleared by phenobarbital without a lowering of the pH.
As mentioned in the review of the literature, the precipitation
of phenylethylaetylurea occurs in deteriorated solutions of phenobarbital sodium. This decomposition product is slightly soluble in water as was shown by evaporating a filtrate obtained from a mixture of phenylethylaoetylurea and water which had been shaken for two hours. The reaction of silver nitrate with this compound as it occurs in deteriorated solutions of phenobarbital sodium is a possibility in the Sudden assay (v3). To determine whether or not the presence of dissolved phnylethylacetylurea in such solutions would introduce an error in the Budde assay, a suspension of this compound was treated with sodim carbonate in the concentration employed in the assay. To 35 cc. of a saturated solution of phenylethylaetylurea containing an excess of the compound, 1.2 Gm. of anhydrous sodium carbonate was added. The mixture was shaken for two hours and then filtered. To 25 cc. of the filtrate, 0.1 N silver nitrate was added. A distinct turbidity was produced by 0.04 co. This amount of silver nitrate would otherwise indicate the presence of 1 mg. of phenobarbital sodium. The error introduced thereby would be insignificant in the experiments on deterioration of solutions of phenobarbital sodium subsequently discussed.
The possibility of an introduction of error due to the presence of other decomposition products remained to be settled. It was deolded to heat a suspension of phenylethylacetylurea under the most




83
drastic conditions to which solutions in the subsequent experiment on deterioration were subjected, namely, heating at 1270 C, for two hours at pH 9.9. In this way the maximum breakdown of the compound into phenylethylacetio acid and urea would occur. To 35 co. of water,
0.25 Gm, of phonylethylaoetylurea was added. It was found to have a pH of 7,7. Sufficient 0.1 1 sodium hydroxide was added to raise the pH to 9.9. The mixture was heated in a glass-stoppered pyrex bottle under the conditions just mentioned. The pH fell to 7.5. Ina the contents of the bottle, 1.2 Gm. of anhydrous sodium carbonate was dis* solved. The mixture was filtered and 25 co. of the filtrate was titrated with 0.1 N silver nitrate. The entire experiment was repeated to provide a cheek. The average amount of silver nitrate required to produce a distinct turbidity was 0#50 coc. This amount would otherwise indicate the presence of 0.05 Gm. of phenobarbital sodium in 100 co. It was concluded that the Budde assay was sufficiently accurate to follow the decomposition of solutions of phenobarbital sodium.
Phenobarbital Sodium U. S. P. XI, which had been purchased from the Mallinckrodt Chemical Works with the control designation KLM1, was assayed by the methods of Budde (93), Viebock and Fuchs (95, 96), and Schulek and Roase (100, 101). The details of each method were given in the review of the literature. The results which were obtained are outlined in the table which follows.




84
TABLE 26
COM.PAPISON OF SEVERAL ','PTHC)DS OF ASSAY OF PlinVOBARBITAL
SODIUM U. S. P. XI EMPLOYING SILVER NITRATE
Caloulated Phenobarbital Phenobarbital Sodium
Dry Sample
Gs. GM. Gm.
BUDDE ASSAY
(1) 0.3073 0.275 89.6 0.,301 98.1
(2) 0.2596 0.233 89.8 0.255 98.3
(3) 0.2583 0.230 89.2 0.252 97.7
VIEBOCK A1LD PriCHS A6SAY
(4) 0.2661 0.235 86.4 0.258 96.8
(5) 0.2637 0.238 90.1 00250 98.8
(6) 0.2738 0.243 58.8 0.266 97.2
SCIIULEK AND ROZSA ASSAY
(7) 0.1683 0.146 86.6 0.160 94.8
(8) 0.1798 0.151 84.1 0.166 92.1
(9) 0.2074 09177 85.4 0.194 93.5




The Budde assays showed the presence of 89.56 phenobarbital in
the samples of Phenobarbital Sodium U. S. P. XI. The Viebook and Fuchs assay showed the presence of 89.1% phenobarbital. The Schulek and Rossa Assay showed the presence of 85.4% phenobarbital. The U. S. P. XII tolerance (71) for phenobarbital is 89.0 to 91.5%. Therefore, the Budde method was considered satisfactory. Although the method of Viebook and Fuchs (95, 96) gave equally accurate results, it was difficult to determine the end point because of the pink color produced upon the addition of silver nitrate in the second part of the assay. The method of Schulek and Rossa (100, 101) was considered unsatisfactory both from the standpoint of the low results and the difficulty in detecting the end point.
4. Von Babitseh Titration of Phenobarbital
A method was sought for the quantitative determination of the
phenobarbital in residue II of the Rotondaro assay (72). The purpose was to eliminate the work involved in drying the residue to costant weight by substituting a volumetric determination of the phenobarbital after the chloroform had been evaporated off. In this experiment the samples employed were residues of phenobarbital obtained in previous Rotondaro assays. The melting points of the individual residues were within the melting point range given in the U. S. P. XII (71). The assay consisted of dissolving 0.1 to 0.2 Gm. of sample in 20 co. of alcohol, adding 2-3 drops of 0.1% alcoholic solution of thymolphthalein, and titrating with 0.1 N sodium hydroxide to a blue color. The results obtained are given in the table which follows.




Be
TABLE 26
VOL' BA&.8ITSCII TITRATION OF PUENOBiARBITAL
Sample Weight of Amount Found For cent of
Number M~pe0 C, sample Sample taken
Ofl. Gm*
1175 0.2328 0.234 100.4
2175 0.2405 04241 100.0
3 175 01.2095 0.209 9906
4 176.6 0.2460 0.242 98.9
5 176.6 0.1821 0.162 99.8
6 176.6 0.1334 0.132 99.00
7 176.1 0*1759 0.177 100.3
8 176.1 0.1845 0.184 9909
9 178.1 0.2124 0.212 10000
Average 99.8




The volumetric determination of pure phenobarbital by the method of von Babitsch (86) gave an average result of 99.6. It was coneluded that the method was sufficiently accurate for the titration of residue II as obtained in the Rotondaro assay (72).
5. Comparison of Assays of a Deteriorated Solution of Phenobarbital Sodium
By assaying a deteriorated solution of phenobarbital sodium
with the U. S. P. XI (70), Rotondaro (72), and Budde (93) methods, it was desired to obtain a comparison of the percentage decomposition as indicated by these methods.
Undried Phenobarbital Sodium U. S. P. XI, which had been purchased with the Mallinokrodt control designation KI2M1, was used to prepare the solution. The per cent of calculated dry sample in the solution was
4.00 % w/v. Into four-ounce, glass-stoppered, pyrex bottles were intro4uced 25 coc. portions of the solution. Therefore, 1 Cm. of calculated dry Phenobarbital Sodium U. S. P. XI was in each bottle. The bottles were placed in an autoclave and heated at 1160 C. for thirty minutes. Time was taken when the pressure had reached 20 pounds. Eighteen minutes had elapsed since the introduction of the bottles into the chamber.
The U. S. P. XI assay (70) was carried out on each member of a set of three samples by merely doublint the quantities called for in the official procedure, inasmuch as the latter employs 0.5 Gm. of the sample. The phenobarbital was extracted with twelve 25 cc. portions of ether, each period of shaking being not less than ten minutes in




duration.
The modified Rotondaro assay was performed on another set of sam* ples. The method employed was previously described in the section on evaluation of reported methods of assay. The samples were quantitatively transferred to separators with the aid of small portions of water which totaled approximately 33 co.
In carrying out the Budde assay (93) 3.33 Gm. of anhydrous sodina carbonate was dissolved in the contents of each bottle. The mixture was quantitatively diluted to 100 cc. and filtered through a dry filter. The first 20 cc. of filtrate were discarded. Of the subsequent filtrate, 25 cc. ere titrated with a standard solution of silver nitrate. The assay was repeated in each case on another portion of the filtrate.
The results obtained in this experiment are sumnarised in the following three tables.
TABLE 27
U. S. P. XI ASSAY OF A 4% W/V SOLUTION OF PHENOBARBITAL SODIUM
U. S. P. XI AFTER -EATIG AT 1150 C. FOR THIRTY LilUTES
Phenobarbital Calculated Strength of Per Cent
Extracted C1 1 103N Na Solution DeeompoAfter iHeating sition
Gm. Gm. %
(1) 0.870 87.0 0.952 95.2 3.81 4.8
(2) 0.881 88.1 0.964 96.5 3.86 3.6
(3) 0.880 88.0 0.963 96.4 3.85 3.7




89
TABLE 28
MODIFIED ROTONDARO ASSAY OF A 4% W/V SOLUTION OF PHENOBARBITAL SODIUM
U. S. P. XI AFTER DATING AT 1150 C. FOR THIRTY MINUTES
Sample Number 1 2 3
Residue I
Gm. 0.052 0.024 0.042
per cent 5.2 2.4 4.2
Residue II
Om. 0.605 0.595 0.588
per cent 60.5 59.5 58.8
Residue III
Gm. 0.216 0.173 0.195
per cent 21.6 17.3 19.5
Calculated
C12l1103N2Na
Gm. 0.662 0.652 0.643
per cent 66.2 65.2 64.3
Strength of
Solution after
Heating 2.65% 2.61% 2.57 %
Per cent
Decomposition 33.8 34.8 35.6




90
TABLE 29
BUDDE ASSAY OF A 4% W/V SOLUTION OF PEN O3ARBI'AL SODIUM U. 8. P. XI AFTER HiATING AT 1150 C. FOR THIRTY MINUTES
Sample so. of C12H11 3N Nsa Strength of Per Cent
Number 0.1 N AgNO3 Solution DocompoAfter Heating sition
Gm. %
1 8.20 0.209 85.5 3.34 1665
8.19 0.208 83.4 3.33 16.6
2 8.20 0.209 83.5 3.34 18.5
8.20 0.209 85.5 3.34 16.5
3 8.18 0.208 83.3 3.53 16.7
8.18 0.208 85.3 3.33 16.7
The U. S. P. XI assays showed 3.7% decomposition in the 4% w/v solution of Phenobarbital Sodium U. S. P. XI which had been heated at 1150 C. for thirty minutes. The modified Rotondaro assay showed 34.7% decomposition and the Budde assay showed 16.6% decomposition. Therefore, checks were not obtained for the determination of deterioration by the several methods mentioned.
The U. S. P. XI method gave results for percentage deterioration which are obviously lower than the true decomposition, since in this method most of the products of deterioration are evaluated as undooomposed phenobarbital. The results for percentage deterioration by the modified Rotondaro method are considered to be too high, since previous results indicated that the modified Rotondaro method, as carried out in the present investigation, failed to recover all the unchanged phenobarbital. The Budde method, which gave results for percentage




91
deterioration between those obtained by the U. S. P. XI method and the modified Rotondaro method, was selected for further use in determining percentage deterioration in the present investigation. From the studies made of the Budde method, it appears that it gives results that are approximately correct or that are sufficiently accurate for determining the relative degree of deterioration in comparative tests.
E. A STUDY OF A NUIHER OF FACTORS IN THE DETERIORATION OF SOLUTIONS
OF PHENOBARBITAL SODIUM U. S. P. XI
It was desired to investigate the effects produced by variations in time, temperature, and concentration on such factors in the deterioration of solutions of phenobarbital sodium as percentage deterioration, pH, volume of precipitate, and refractive index. This work was prompted by the results reported by Nielsen (13), Madsen (3)* Aspelund and Skoglund (7), and Tomski and Waller (24). Madsen (3) reported results regarding pH change and percentage decomposition obtained on heating a 10% solution of diethylbarbiturio acid at 600 800, and 1000 C. Nielsen (13) showed the relation of pH to decomposition of a 10% solution of phenobarbital sodium at temperatures not exceeding 590 C. Aspelund and Skoglund (7) reported the quantitative isolation of various decomposition products of phenobarbital sodium in solutions which had been boiled for three hours, and in solutions which had been stored for eight days. Tomski and Waller (24) investigated the percentage deterioration of solutions of phenobarbital sodium at various temperatures for various periods of time. In an attempt to verify and extend the data presented by those workers, experiments were designed




92
to include, in whole or in part, an investigation of factors previously studied and at the same time to amplify, and to fill in the hiatuses, and to asutdy additional factors in the deterioration of solutions of one barbiturate, namely, phenobarbital sodium. The temperature and periods of exposure were selected to encompass most of the conditions previously employed by the aforeto mentioned workers and to include for the most part the conditions of sterilization recommended for solutions intended for parenteral use as given in the N. F. VI (73), the British Pharmacopoeia, 1932 (26), and the United States Dispensatory (108). In this manner the changes occurring under conditions of sterilization and conditions reported by other workers were investigated, providing at the same time an insight into deterioration taking place during prolonged storage through the agency of these accelerated tests.
In this series of experiments, 5, 10, and 20% solutions of Phono* barbital Sodium U. S. P. XI were employed. In every case, 25-co, portions of the solution were heated in 125-c, glass-stoppered, pyrex bottles. Unless otherwise indicated each of the titrations given in the tables represents the results obtained from the titration of the contents of a separate bottle. The group of determinations which ineludes volume of precipitate, refractive index, and pH were made on the contents of a separate bottle. In the event a precipitate had been produced, the clear supernatant liquid which remained after centrifuging was employed in making the determination of refractive index. The instrument used in making this determination was the Abbe Refractometer the temperature of whose prisms were thermostatically controlled by the flow of water, maintained at 25 (0.5) 0. through the surrounding




93
Jackets. The use of the autoclave, the determination of pH, and volume of precipitate were previously described.
From the investigation conducted on the evaluation of methods of assay, it was considered advisable to employ the Budde method in following the deterioration. The original Budde method (93) was reported as follows: About 0,2 to 0.3 am. of the sodium salt of the barbiturate is weighed and dissolved in 30 oo. of water, together with 1 Gm of anhydrous sodium carbonate. To the clear solution 0.1 N silver nitrate is added until a persistent, distinct turbidity is produced. One molecular weight of silver nitrate is equivalent to one molecular weight of
the barbiturate.
Inasmuoh as solutions of various strengths were employed, appropriate volumes had to be taken and appropriate dilutions had to be made in order to duplicate the conditions of the Budde assay. The procedures followed in all of the "says of this series of experiments are outlined in the table ihich follows.




94
TABLE 30
PROCEDURE FOLL3WED I! THE BJDDE ASSAY OF 5, 10, and 20% SOLUTIONS OF PHENOBARBITAL SODIUM U. S. P. XI
Percentage Strength of the solution
studied 5 10 20
Volume of each solution placed in each bottle (cc.) 25 25 26
Dilution made for each 25 co. of solution (cc.) 125 250 500
Amount of anhydrous Na2CO3 added before dilution (Gm.) 4.17 8.33 16.67
Alternative amount of anhy*
drous Na200CO to add to each 25 .oo aliquot (Gm.) 0.83 0.83 0.83
25 cc. of the dilution was taken for the titration. This volume represented (x) cc. of the original solution 5 2.5 1.25
In all of the experiments of this series, undried samples were used to prepare the solutions. The per cent of moisture in the drug was determined on separate samples before preparation of the solutions. The strength of each solution was based on the calculated amount of dry sample dissolved. The solutions were assayed before and after treatment. If a precipitate was present, the contents of the bottle,in which the solution was kept during the treatment, were quantitatively transferred to a volumetric flask. Before dilution the prescribed amount of anhydrous




95
sodium carbonate was dissolved in the mixture. The sodium carbonate was dissolved before dilution in order to redissolve any possible phenobarbital which may have been precipitated. This amount of anhydrous sodium carbonate will dissolve the maximum amount of phenobarbital titat can be formed from the amount of phenobarbital sodium which was present in the aliquot part of any solution studied. In a previous experiment using I Gm. of Phenobarbital U. S. P. XI, 2.8 Gm. anhydrous sodium carbonate, and 25 cc. of water, it was found that a clear solution was produced after stirring for four minutes. Eight minutes of agitation of the sodium carbonate with the contents of the bottle was considered ample time in which to dissolve any free phenobarbital which may have been present. After dilution a portion of the mixture was filtered through a dry filter. The first 20 cc. of the filtrate was discarded. Of the subsequent filtrate 25 cc. was taken for the titration. In the event no precipitate had formed in the solution during the heating, sodium carbonate was not added until the filtration had been performed, at which time a proportional part of the sodium carbonate as given in Table 30 was added to the aliquot part of the solution which was to be titrated.
It was deemed advisable to determine whether or not there would
be any significant difference between the results obtained by the Budde method if the deteriorated solution in the bottle was quantitatively passed through a filter before dilution instead of diluting the doteriorated solution containing the reocipitate to the same volume. Towards this end, four 25 cc. portions of a 10% solution of Phenobarbital Sodium U. S. P. XI (lallinokrodt Control KKM) were heated at 1150 C.




Full Text

PAGE 1

THE STABILITY OF SOLUTIONS OF PHENOBARBITAL SODIUM By BERNARD B. J ATUL A DISSERTATION PRE S ENTED TO THE GRADUATE COUNOL OF THE UNIVERSITY OF FLORIDA I N PARTIAL FULFILMENT OF TI-IE REQUIREMENTS FOR TI-IE DEGREE OF DOCTOR OF PHILOSOPHY UN I VE RSITY OF FLORIDA Fe bru ary, 1943

PAGE 2

ACKNOWLEDG MEN T The author gr atefully acknowledge his indebtednea1 to Dr Wi lliam J. liu aa for his sagacious gu idance a n d gracious assistance during the couree of the p resent inveati g ation. Thanks are also exte nde d to all those who have aided in the co nduc t of this wo rk. ii

PAGE 3

I II. TABLE OF CONTENTS ACI
PAGE 4

III EXP F,R IM EI~ T.AL PART s, A SCOPE OF INVESTIGATIO N ............................... 64 B A N ALYTIC.AL tJS T HO DS Ai'I D CO N T R OL 55 1. 3 4 Mat erials U d ~ Determin a tion of M oisture pH Determ i nat Determinati o n of Volume of P r cipitate 56 55 56 57 5 U ot the Autoclave 68 c A COMPARISO N O F ME THODS OF DRYI NG P HENOBAR B IT AL SO D IUM PR E PARATORY TO ASSAY 69 D EVALU A TION OF RE PO R T E D M ET H ODS OF AS S AY 66 l Comparison of u S P X I and Rotondaro Assays of Phe n obar bit al S odi um U S P X 66 2 Co m parison o r M odified R otondaro Assay o f Phen o ba.rbital and Phenobarbital Sodium 73 3 4 A Study of Asaays &n. ployin g Silver N itr Von .Babitsch Titration ot Phenobarbital te 79 85 6 Comparison of As1ay1 of a Deterio r ated Solution or Phenobarbital Sodium U .S. P 87 E A STUDY O F A N U MBE R O F F A CTORS IN THE D E T ER IORATION OF SOLUTIONS OF PHENO.BARB I r AL SODIUM U. S P X I 91 l E ffect of H eatin g a 5 % w/v Solution of Phen obarbital Sodium U S P X I at so C. 97 2 E ffect of H ea t in g 5 10 and 2 0% w/v Solutions or Phenoba rbital Sodium u s. P XI at ao 0 c 100 3 E ffe ct ot Heating 5, 10, and 20% w/v Solutions ot Ph enobarbital S odium U $ P X I at 100 O 107 4 Ef fect of He ating 5, 10, and 20% w/v Solution, ot Phenob arbital Sodium u S P X I at 116 C 112 6 E ffect or H eating 6 10 and 20% w/v Solutions of .Phenobarbital Sodium u s P XI at 121 C 117 iv

PAGE 5

F RELA T ION OF pH TO TR S DETERIORAT IO N OF SOLUTIO N S OF P HENOBARB I TAL SODIUM 123 l. The Use of M onosodium and Dieodium Phoaphat ea .,.,.,. 1 23 2. The U ae ot Sodium 121 a EV ALU AT I ON OF S EVE RAL SUG GES T E D STABILIZING AG EN TS ON SOLUTIONS OF PID.;NO.BARB I 'rAL AN D PHEl'lOBARBI'fAL SO D I UM 130 1 An Experiment on the Stability of E l ixir of Ph eno ba.r-bital u s. P X I and Several M od if ic ations 130 2. An E xperiment on the Stability of Ph nobarbital S odi um U s P XI in the Pr ese n ce of A lcohol, G lycerin, and .Antipyrine 133 3. An Expe riment on the S tability of Solutions of Pheno barbital Sodium u s. P X I Containin g Dextrose 136 IV DISCUSSION OF R B S UL TS 139 V SU MMAR Y .AND CONC LUSIONS ~ 148 V I. B IBL IOGRAPHY 1 6 1 B IOGRA PH ICAL IT EM 166

PAGE 6

I. TRODUCTION Si n oe the i n troduction of diethyl b a rbiturio acid into m edicine by F ischer and von M e ri ng (1) in 1903, barbiturio aoid derivatives have assumed a leading place amongst the m edicine which assist in allevi ating the ills of mankind These medi cinals are now mo re widely used than any other class ot hypnotics For oral u ee they are generally dispensed in the dry state or in elixir form How ever, tor parente ral injection, it hae usually been ne e ssary to employ aqueous solutions Since barbiturates as a ola s rapidly decompose in aq ueous media they are administered soon after prepa r ati on ot the solution This insta bility has p ro mpted a number of workers to eonduot re search relative to the nature of the phenomena and to method ot ita r tardation The p resent work wa. carried out to extend the knowledge re g ard ing the deterioration and stabilization of barbiturate solutions lim iting the research to one derivative namely phenylethy l barbiturio acid M ethods of assay applicable to the problem have been studied and compared Experiments were conducted to inveeti g ate a numbe r ot factors in the decomposition and to evaluate and amplify several re ported m ethod of st bilization

PAGE 7

2 REVIE'N OF THE LIT ERATURE A. D E T ER IORATIO N AND STABILIZATIO N OF BARBITURATE SOLUT IO N S 1 Na ture of the Deterioration In 1927 Steenhauer (2) round that aqueous solutions ot sodium diethylbarbiturate when heated at 100 c suffered hydrolysis into carbon dioxide and diethylacetylurea The latter com p ound crystal lized out on cooling Co nf irmation of these find in g came from Ma dsen (3) i n 1 934 and Bailey (4) in 1936 Bailey (4) separated the crystals formed in solutions of sodium barbital which had been autoclaved Th cr ysta ls were washed with water and then extracted with a chloroform ether mixture The com p ound obtained from the latter was reor yst al lized from boiling wate r The colorless needles resultin g me lted at 207 c (corr ). N o amm onia was formed when the compound was heated with strong alkaline solutions whether aqueous or alco h olic Ammonia was for~ed after lon g heating with 60% sulfuric acid and subsequent treatment with alkali The fact that nit ro g en wa s f ormed in this sul furic acid soluti o~ when t reated with excess alkaline hypobromite showed the presen ce of urea Ana lye is yielded the fo llowin g resul tu Hyd ro gen 8 92 N itrogen 17 69 Carbon 53 44 Oxygen 20 01 (by d if fe rence) The simplest empirical formula derived from the above data is c 7 a 14 N 2 o 2 The m olecular weight was found to be 166 when determined by oryosoopio means with ca mp hor as the solvent Urea and an oily liquid wh ich was practically insoluble in water were the p roducts of hydrolysis formed by refluxin g this co mp ound with 60% sulfuric aoid Th e oily liquid

PAGE 8

3 was diethyl aoeti~ aoid T h oily liquid had an acid reaction. a boiling point ot slightly over 190 o and a density less than that of water Fittig in 1 880 gave the boiling point otdiethylacetie acid as 194 c and stated that it was l ighter than water Thia informa tion. to g ether with other considerations led Ba.il$y (4) to conolude that crystalline p r oipitate was diethylaoetylurea Its mole oular weight ia 168 N o ammonia is produced by heating diethylaoetylurea with strong alkaline solutions because of ita gr eat stability Lasa ignes test cannot be pe rformed becau e the compo\Uld tends to volatilize rather than decompose when dropped on boiling sodium According to Fische r and Dilthey (6) ureides of dialkylaoetic aoid are frequently formed as byproducts in the preparation of ureid of dialkylma.lonio acid in the p rese nce of sulfuric acid They are also formed by heating the ur ides of dialkyl.malonic aoid A third m ethod of p reparation is baaed on the action ot phosphorus oxychloride on a mixture of dialkylmalonio aoid and urea These thr ee methods re p rob ably one and the same namely, the decomposition ot dialkylmalonylurea Fischer and Dilthey (6) re marked that dim thylaoetylurea could not be pr pared by a:ny ot the ato r eto ment ioned m thoda Furthermore it was un known at that time (1904) Diethylacetylurea was fi r st deaoribed by the von N eis1en brother (6) who stated that it was form ed bf heating diethylba. r biturio a c id to 102 c They indicated that th melting point was 207 6 c (corr ) and that it waa hydroly&ed into diethylaoetio acid and urea only on prolonged hea ting with concentrated hydrochloric acid Fischer and

PAGE 9

4 Dilthey (6) p rep red dieth yl acetylurea by h eating a mixture ot urea and diethylmalonic acid to whi c h ph osphorus oxyohloride had been add ed The p roduct was crystallised tram. hot ter. They fo und that the m elt 1ng point was 207 6 C The structur e of t he com p ound was l earn ed t'rOlll. its cl eavage into ure and diethylaoetic acid upon hea tin g with concen .. trated hydroohlorio acid 1n a see.led tube for seventeen h ours The diethy laoe tic acid separ at d as an oil To ident ify the latter product its silver salt was p repared. The silver content wa determin d and found to correspo n d to that of silver diethylaoetate The p reae no e of urea w as s hown by p re par i ng the nit rate sal t in o rd er to tree it from the hydrochloric ao14 solution re g n er at ing the urea m olecule and then a na lyzin g for nitrogen content Th e nitrogen content and th melting point were those of urea. Diethylacety lur dissolves in appro xi mately 12 0 p arts of h ot wat er it i p raotically i n soluble i n cold w ater, mo de r at ely solubl~ in chloro torm and ether and very sol ub le in alcoho l It ia insoluble in both oold di l ute aoids an d alkalies M ethods of p re paring dip ropylaoety l urea and me thylothyl a oetylurea were also g iven by Fisch e r and D ilth ey ( 5 ). F ro m qualitative tests on purified pre cipita te s formed in solutions of othe r barbiturates Bailey (4) round that the disubstituted aoetylurea, correspo n din g to the ba.rbi turate employ ed, was produced in ea.oh case I n addition to the diethylacetylurea Bailey (4) round that the following compound resulted from the hydrolysis of solutions of di ethylbarbiturate a Gaeeaa carbo n d ioxide and amm o n ia

PAGE 10

6 Subst an ces in solutions urea sodi um diethylaoetate sodium di e t h ylma l onate so d i um b icarbonate and sodium carbonate. From th e p roducts f ormed Ba iley (4) de du ced t h at t h e barbituric acid molecule is hydrolysed 1n two ways, (1) los of one m olecule o t carbon dioxide o p enin g the rin g to f orm d iet h ylacetylurea J (2) cleava g e of t h e rin g to form diethylme.lonio acid or its ealt an d u rea Th e d iethylmalonio aoid then loses o n e m olec u le ot carbon dioxide to g ive diethylacetic acid or its s a l t T h e urea m ol e cule de co m.p oses slowly into carbon dioxide and ammonia Ear lier Ma dsen (3) ha d reported th at diethylacetylurea h y drolysed in co n sider a ble quantity int o diethylaoetic acid and urea, an d latter hydrolysin g fur t her in a small de g ree to carbo n d i o xide an d amm on i a Ba iley (4) su g g ested t ha t n either reactio n s p roceeded to oom p le tion so t h at b oth f inal an d i n te rm ediate p roduct are o b t a i n ed Ba. ile7 (4) p rese n ted the eq u ations w h ich f ollow (1) S odium d iethyl aoetate

PAGE 11

(2) Primary Sodium d1ethylmalonate Sodium die thy lacetate Aspelund and Sko g lund (7) outlined the co urs e of deterioration ot substituted ba rbi turio ao1da as follow s s 0 H 0 II II .,, C N \ / C NH CO NH 2 ,,, ,, C OOR R 2 > C C = 0 --> R > C R C "2 2 ,. "CO OH "/ COOR C N 8 a l R 2 CRC0 NH C0 NH 2 l T risubstituted barbiturio aoids may b r eak down in either ot two ways

PAGE 12

7 CO N -CO N H 2 I c~ R 2 > C ______, \ COOH Aspe l und and SkoglWld (7) separated the de c omposition products rom one another as they occurred 1n dete r iorated solutions ot barbituratea by filt r ation and extraction wi th eth rat suoceaa1v e pH levels The re aults obtained are summarised in Table 1 and fable 2 showing the p e r oe n ta g e yields ot the ditterent p roducts obtained

PAGE 13

8 TABLE 1 PE RC EN TAGE YI E LDS OF DE C OMPOSIT IO N PRODUC TS O BTAI N ED FROM DE T E RIORAT E D BARBITURATE SOLUT IONS WH ICH WE RE B O I LED FOR THREE HOUR S N ame ot the O r igina l Substituted Subetituted Substituted Barbituri o .&rbituric M alonuric Acetureide Aceti c A ci d A cid Acid A cid Diethyl b a 46 3 0 7 E thyl-allyl 29 1 6 34 b a Dia llyl b a 14 7 24 12 Is o propyl 83 1 5 4 a.llyl b a Pheny l ethyl b a 61 ll 12 Pheny lethyl 30 0 3 34 1 N -methyl b a Cyc lo hexenyl m ethylN methyl 2 3 48 b .a.. E thyl all y l 28 25 32 N M eth yl b a Results obtained by Aspe l und and Sko g l und {7)

PAGE 14

9 TABLE 2 P E RC EN TAGE YI E L D S OF S UBSTIT U T E D .MAL O NUR IC ACI D O JJ.l' A I N E D F R O DE T E R I ORATED BARB IT URATE S OLUT I ON S W H I CH H,AD BEEN STAND ING FOR E IGHT DAYS N ame of the Ba.rbiturie Acid Diethy l b a E thy lallyl b a Diallyl b Isop ro py l-a.llyl b .a. Ph enyl-ethyl -m e thyl b a Cyclohexenyl ethy l N ethyl b a. Ethyl all yl N m ethyl b .a. Peroentage Yield ot Substituted Malonuric Aoi d 0 6 2 .0 4 0 trace 1 2 4 12 Reaulte obtai n ed by A pelund and Sko g lund (7)

PAGE 15

1 0 Aspelund ( 8 ) stated that the instability of the ma.lonylurea ring 1n barbiturates is p roba bl y due to then ga tive hydroxyl grou,. He studied the decomposition ot 5 5 bromoben1yl barbiturio acid The re .. sul ts obt a ined. a.re outlined by the equaUona which f' ollow .. 0 B 0 N Br '/ \ C C :: 0 c 6 H 5 CH~ -\ / ;, 0 N Boiling with 1 0% H 2 so 4 ,, 0 H N a.OH ) l fu il H eating w ith Alcoholic KOH

PAGE 16

11 Aspelund and Lindh (9) ma de a study of the ch emis try of the de compositio n p roducts of 5 a 6 bromoalky N --alkyl bar bi turio acids The results obtain ed are outlined by the equations which f ollow HO ......._ H c / R / '-... CONIICN 1 Bo ili ng w ith 1 0% H z S0 4 N aOli at room temperature H ea ting w ith alco h olic oa u stic al ka liea l li eatin g in al co hol and w ater Br _.,,.. H c R / 'CO N H C ONliR

PAGE 17

1 2 TABLE 3 il! ELT G .POH! TS OF DECOMPOSIT I ON PRODUCTS O BTAINED FROM DE1'ER I ORATED BAR B I TURATE SOLUTION S Decomposition Product 0 W orker m P c Phenylethylacetylurea 1 50 Bailey Ph enylm e thyla o etyl urea :;. 59 Bailey N methyl oyc l ohexeny l 97-10 1 Ba iley methyla c etylurea Butylethylacetylurea 154 Ba iley Diethylacetylurea 2 06 207 Bailey D iethyl.malonuric aoid 1 61 1 5 2 N eilsen Phenylethyla o etio acid 44 46 N eilsen (4) (4) (4) (4) (4) (10) (ll)

PAGE 18

Although they may keep tor 8-15 days as W etzel (12) claimed, ao lutiona of sodium phenylethyl barbiturate also under g o an analogous hy drolysis into phenylethylaoetylurea (m P 146-147 c ) and carbon di oxide N ielsen (13) found that phenylethylaoetylur a was precipitated when he refluxed 20% solutio n s ot sodium phenobar b ital for p eriods ot 1/2 to l hour Heating for 3 to S hours resulted 1n hydrolysis of p art o f the phenylethylactylurea into phenylethylaoetic aoid and urea. Aspelund and Skoglund (14) stated that phenylethylmalonurio acid waa very unstable. T he form a tion ot free p henylethyl barbituric aoid is lso a poa sibil1ty lt is to be noted that the solubility of this acid in water ia about l 1 1000, whereas the solubility of tree diethyl barbiturio acid in water is about l 130 Therefore the appearance ot a p re cipitate of the tree barbitur1c aoid in solutions ot sodium phenobar bital ii a g reater poasibility than i n solutions ot sodium barbital According to W oodward (15) N ielsen (13) h&s provided evidence that free phenylethyl b arbituric may be formed in sol utions of sodium. pheno barbital o n prolon g ed standin g Both liberation of the latter acid ard hydrolysis take place at the same time Wh en determinin g phenylethyl aoetylurea formed in solutions N ielsen (13) had difficulty in obtain. ing deposits and filtrates tree from simultaneously formed phenyl e thyl barbiturio acid The a c id was freed b y the carbon dioxide absor;ed from the atmosphere To p revent deterioration b y carbon dioxide Cazzani (17) su g g ested that the solution be saturated with nitro g en N ielsen (13) p revented the f ormation of the f r ee acid by adjusting the

PAGE 19

l4 pH of the solutions with standard sodium hydro x ide Presumably the sodium hydroxid wa1 employed in a manner ao as to redissolve any lib erated acid M esnard (l ) reiterated that olutions of sodium phenylethyl bar biturate hydrolyze into phenylethy-lbarbiturio acid and odium hydroxide The decomposition ot phenylethyl barbiturio acid follows due to the action of sodium hydroxide forming products with o ut therapeutic value M esnard (18) gave the following, N aOH The rapidity of hydrolysis of some barbiturate solutions is in dicated by the length of time during which they are suitable for p ar enteral administra t ion The hydrolysis of evipal sodium is rapid Cazza.ni (17) atated that aqueous solutions should not be used 2 to 3 hours after their preparation According to E li Lilly & Company (19 20) solutions or sodium amytal and of sodium pe n tobarbital should not be used thirty minutes after their preparation Abbott Laboratories (21) stated that so l utions or sodium ethyl (1-methyl butyl) thiobarbi turate oan not be aafely allowed to stand longer than 4 hours Aspe l und and Skoglund (14) found that N aubatituted ba.rbiturio acids were mo r e easily decomposed than the 6 6 disubatituted barbiturio aoida They stated that the sodium salts ot isopropylallyl barbiturio acid and cy clohexylethyl barbituric aoid and p erhaps ot diethyl barbiturio acid

PAGE 20

15 are stable in aqueous solution tor a definite le ngt h or tim 2 Fa ctors in the Deterioration a Hy drogen Ion Concentrati o n In 1933, Nie lsen (13) found the rate of hydrolysis of solutions of sodium phenyle thyl barbiturate was decreased whe n there was an increase in hydrogen ion co n centration The amount ot hydrolysis occurring wae followed by det e rmin i ng the quant i ty of carbon dioxide and phenylethyl aoetylurea formed N ielsen (13) stated th a t the de termina tion of the ohlorof o rm so l uble urea derivativ i n troduoes an error due to its high so l ubility in the ori g inal solution and due to the f ormation of free phenylethyl barbituric a o id by the c a r bonic acid released durin g the hydrolysis Presumably N ielsen (13 ) extracts the urea derivative with chloroform thus explainin g the above mentioned error T h e solubility of phenylethyl barbituric oid in chloroform isl Gm in 40 oo.{22) The results obtained by N ielsen ( 13) are summ a rized in Table 4 TABL E 4 R E L A TION OF pH TO DECO M POSITIO N OC C URR GIN 1 0% ~OLUTIONS C ON TAINING BO T H PHENOB.ARBI rAL SODIUM AN D PHENOBARB ITAL AT 1 8 e ( ~ l C pR N umbe r of days r equi r ed tor 1% de c ompositi o n 30 2 2 18

PAGE 21

16 A 1 0" /4 solution of p henobar b ita l so d ium saturated with the free a oi d has a pH of 8 9 Be r aaain an d V i ta li ( 23 ) s t at e d that the pH of a 1 0% w / v sol u tion o f th e sod i um salt w as 9 6 Be r asain and Vitali ( 2 3) p r pa r ed 10% w / T soluti o ns of phen o barb i ta l s od i um in buffe r solutions instead of water The pH values of th e s o luti o n s prep ar ed a r i nd i c a t ed in Table 6 How eve r they fai l ed to st at what proport i ons ot the respe c t i v e buffer s ol utions we r e uaed T ABL E 6 EFFEC T OF BUFFE R S ON THE pH OF 1 0% SOLUT I ONS OF PHENOBARB I T AL SOD IUM Buf fer O ri g inato r p H obtained N a Oli an d g l y coooll S ore n sen 9 71 9 3 6 8 3 9 8 57 a 24 KH 2 P0 4 an d N a o C lark and L uba e o o 7 8 0 7 60 7 4 0 1 20 V ero n l an d H Cl M iohaelia 9 4 0 8 60 a oo 1 6 0 S odium citrate and C l a rk and Luba 1 00 .N a OH In anothe r ex p erime n t phen o ba r b ital wa s used to n e utr a lize the free l ka li an d to l ower th e pH The latter an d the b uf fer e d sol ut ions we re s ubj ected to h e a t in g b y steam for fifteen m inut e s A p reoi p itate waa fo rm ed i n e a o h c a se b u t t h e quantity w ae les i n t h ose sol ut ions havin g a pH leas t h an 9 00 T h e p re ci pitate a pp ear e d aa w hite small f l a k es floooulent at first and t h en a s n eedles g ro upe d in a radiatiJl g fas h ion Tom ski and W a ll e r (24) i n vesti g ated the ef f e c t of va r yin g amo unt o f sodium. carbo n ate on the hydro l ys i s of solutions of s o d ium p henyl t h y l

PAGE 22

17 be.rbiturat On e 1) oc ., 2 co and 3 co portions of N /1 0 sodium carbonate were added to 5 cc portion of a solution of sodiU111 pheno barbital Blank s ol utions w e re also made using water instead of th e odi um car bonate solution. All the solutions were h eated tor sixty minu tes at so 0 c Ea.ch b lan k solution showed a loss in p ercentage of 2 a The resu lts which Tom ski and W aller (24) ob tained are summarized in Tab l e 6 TABL E 6 L O SS I N P ERCENTAG E ON HEATING A SOLUTION PHENOBARB ITAL SODI PR E S EN C E OF SODI CAROO ~!A TE l 2 3 Lou in Pe rce n ta g e 6 6 10 15 IN THE Ma dsen (3) fo\llld that solutions of sodium diethylbarbiturate hy drolyzed at a slowe r rate upon an in crea e i n hyd ro gen ion co n centra tion H e calculated t h e degree ot hyd ro l ysis by determinin g the amount of diethylao etylurea and c a rbo n d ioxide form ed The res u lts secur d by M adsen (5) are g iven 1n Tables 7 and a

PAGE 23

1 8 T A BL E 7 EFFEC T OF HEA T ON 10% SO L U TIO N S O F S O D I UM D I E THYLBAR B IT 'URA 'l' E Te m perature D uration of P er cent p H Chan g e H eating Decomposed 100 c ~ l hour 2 6 10 1 to 9 3 so 0 2 hour 1 0 10 1 to 9 6 60 3 hours 0 23 TA B L E S EFFE CT OF i1EA T ON 1 0% SO L UTION S OF SOD I UM D I E THYLBAR .B I TUR A T E W I TH A p H OF 8 9 Te m perature Duration of H eating l hour 2 ho u rs P er ce n t Decomposed 1 s 0 32 Bailey (4) discovered from a study of solutions p repared from dif f e rent sa m ples of sodium barbital that som s h owed p recipitates on h eat i ng at 100 c while others req u ired a sli g htly hi gh er temperature to bring about the same result Upon determin g the pH values of all the solution he :found that thoH whioh decomposed at 100 c had a slight ly hi g her pR value Thie was apparently due to the exoeaa alkali used

PAGE 24

19 in preparing the sodium salt of the barbituric acid Bailey (4) sug g ested that only 95% of the calculated quantity ot sodium hydroxide be used in preparing the salt ot the acid The sli g ht excess of acid then could be filtered off and a product would be obtained whioh would ith stand heating at 100 C in sealed containers for short p eriod without deco m position In this re g ard it may be pointed out that ami n ea have been used as basic substanoes to replace the sodium h.ydroxid The in te n tion is to replace the strong inorganic alkali by a weaker base The use ot amine is reveiwed in another part ot this discussion A fall i n the pH ot solutions ot sodium amytal upon a g ing was re ported by Ca11ani (16) Fresh solution have pH values ot 9 6 to 9 8 The solutions decompose readily on standin g W hen the pH f'alla to 9 3 they become opaleacent and cloudy b Temperature N ielsen (13) found that the rate ot decomposition of solutions of sodium phenylethyl barbituric aoid varied greatly with te m per a ture The rate was followed by determinin g the amount ot carbon d ic-xide and phenylethylaoetylurea formed The results which he obtained are in dicated in Table 9

PAGE 25

20 TABLE 9 Eft'FE CT O F T E~PE RA T URE ON 1 0% AQUEOUS S OLUTIONS OF S OD IU M PHEN YLETHYL BARBITURATE ADJUSTED TO A pH OF 9.4 W ITH PUE.N YL E THYLBAR B ITU R IC ACID Tempe rature Time Z m onths 3 weeka l m onth Pe r cent De composed less than 1 % 1 22 A 11300 solution of aodium phenobarbital wh ich 1s k ept at 20 c is 1 % deco mp osed in 5 weeks according to N ielaen (13) Tomski and W aller (24) m easured the rate of decomposit i on of s o lu tions of sodium phenobarbital by the H e g land metho d (25) This me thod co n sists ot titratin g a solution ot the barbiturate to which eodiWll cubonate haa been a dd ed with N /10 silver ni trate until a taint opa eeoence is formed which cloea n ot diaappear on s haking Investigating the effect of h e ting solutions of different str ength s under the same co nd itions, Tomski and W aller (24) have found th at the lose ot sodium phenobarbitone increases with i n crease in co n ce nt ration The difference was sma ll. howe ver and f or all practical purpose the loss on heating a 20% 1 00 ; or e en a 6% solution is the same according to t h ese wo rkers Sol H t:l.ons of 2 % to 1 0% were stored and analysed pe riodically A solu tion kept for tour m o n ths (Decembe r -March) loat 6% whereas a solution of equal stre ng th when kept for the same length of time ( Ma reh June) loat 9% of the phen obarbital sodium Additional result1 which we re

PAGE 26

21 secured by Tomski and W aller (24) are g iven 1n Table 10 TABLF. 10 20% SOLUTIO N S OF P HENOBARBITAL SOD I UM I N A M PULS HE AT E D AT a o 0 c FOR V AR YING LENGTBS OF TIME T im e in m inutes 30 60 120 180 P erce n ta g Loss 1 0 to 1 5 2 0 to 3 0 4 0 to 5 0 6 ,0 to 7 0 A 2 0% solutio n therefore loses s o to 7 ~ under co ndit ions similar to t h ose p revaili ng in the p rocess of t ynda llization A solution of th e aame stre ngt h lo1e1 l O to 1 5% when sterilized by the em er gen cy meth od ot the B ritish Pharmacopoeia 19S2 (26) Further results obtained by Tomski and Wall r (24) are i nd ic at ed in Tables 11 12. and U TABLE ll 2 0% ( A .?P ROX .) SOLUTIO N S OF P HEN OBARBITAL SOD I UM I N A M P ULS RB ATED IN BO I LING W ATER FQR F IFTEEN M INUTES Initial Strength 1 8 7 20.1 Final S trength 19 9 L osa 1 0 1 0

PAGE 27

22 TABL E 12 2 0% SOLUTIONS OF PHEN OBARBITAL SO DIUM E XPOSED TO ST EAM AT 100 C FOR VARIOUS LENG T HS OF TIME Time in M inutes 6 0 120 1 8 0 TABLE 13 Pe rce n tage L oss 9 5 to 1 0 5 14 .6 to 1 6 5 1 9 6 to 2 0 6 20% (A PPROX .) SOLUTIONS OF P ti EN OBARBITAL SOD I UM IN AM PULS EXPOSED TO ST~ A T 116 C FOR THIRTY M I N UT E S Initial Stren&t h 20 .1 1 8 7 F inal Strength 1 6 0 16 l L o a 1 9 2 % 20 4 Tom&ki and W al ler (2 4) obtain ed a white p recipitate ot phenyl et hyla cet y lurea when a 2 0"/4 solution of phenoba r b ital sodium w as exposed to any o f the metho ds ot sterilisation cited above Van Leent (27) stated that s te rilizati on of sodium phenobarbi tal in the dry s tat e at 120 c yielded a p roduct wh ioh g ra dua lly became incompletely soluble after this treat men t He s uggest ed th at this was due to the action ot traces of m oisture

PAGE 28

2 3 Ba.11 y (4) likewise found that with 1 n oreas1ng te mp rature greate r ounta of p recipitate wer tormed in solutions of sodium diethyl bar biturate H e obtained needle shaped cr y stals on autoclaving 7 % l" and 14 % solutions for thirty m inutes at 105 C Increasingly higher yields were obt ained upon autoclaTing these olutiona at 11 0 116 and 120 c the size of the needles varying with the temperature Up o n removing the crystals and subsequent autoclaving an additional quan tity of cr y stals were obtained At lower tem p eratures, the yield waa influenced by th p rolo ngation of h eatin g Stich (2 8 ) cl aimed that 20% s9lutions withstood sterilisation 1n steam at 100 c H e ac knowledg e d that l to 6% sol ut ion are hy drol y zed by w arming to 100 c Th e British Ph armacopoeia 1932 (2 6) co n tains the i n formation that solutions of so dium barbital can be sterilised in an autoclave at 115 c for SO min utea or by tyndalliz tion a t ao 0 c Be rasai n and Vitali (23) stated that solutions whioh had b een su bje cted to tyndallisation w re n ot suit able fo r injection As has b een m entioned before Bailey (4) stated that solutions of the sodium salt which had been p repared by us ing on ly 95 % ot the calculated quantity of sodium hydroxide w ould w ithstand heating at 100 c for short per iods without d composition Bai ley (4) felt that the decomposition reaction w as probably re versible The course of the reaotion was influenced by the initial c on c en tration of the solution, te mp erature, and re m oval ot th e p re cipitated p ro du ct Be.ilty (4) found that a hig her yie ld w s p roduced wh en the auto clavin g waa carri d out in an open flask t han wh en conducted in a sealed

PAGE 29

24 vessel Th is was explai n ed by the e s cape of c a rbo n d ioxide, p resumably drivin g th e reaction towards c om pl e tion by th e re m oval of one or the pro d ucts fro m the f ield of chemical reaction From a stu d y of a n umber of alkyl an d ryl su b stituted m alo n ylurea compounds, Ba.iley (4) found t hat th e ir solutions all dec om posed at tem pe ratures exceedin g 100 c and that some deco mp osed at temperature below the boilin g p oint of the sol u tion R uh k opt (29) st u died the be havior of substituted barbiturio acid durin g he a ti ng at elev a ted p res sures in the a b se n ce of alkalies 'Wh en p ressures up to five atmos p heres were u sed, the substituted b a rbituric acid was a lm ost oo m pl e t ly hy drol y zed wit h th e form a tion of a pp roximately equal am ounts of t h e sub stituted aoetyl urea and th e substituted aoet emi de The formation of the amide was favored w ith i n c r easin g p ressures, bec om i ng t he only product at a p ressure of about ten at m osp h ere in addition to the car bon d ioxide an d amm onia f ormed At a p ressure of f ive a tm osp h eres sa lts of s t ro ng acids were found to en h ance the fo rma tion of the su b sti t uted acetylurea w h ereas salts of weak acids were found to increase th e p rod u ction of t he substituted oetaraide Above the latte!" m entioned p ressure the am ount of t h e aoetylurea deriva t ive deoreaaea under any circumstances H ydrolysis o f diethyl be.rbituric acid in aqueous solu tion at five atmospheres yielded 47 % 4iethylaoetylurea and 40 % diethy acetam.ide At t~ a tm os ph eres 96 % diallylac tam.ide was for m ed in the hydrolysis of diallylbarbiturio acid in aqueous solution.

PAGE 30

25 3 S tabilization of Ba rbitura t e Solutions a Adjustment of Hyd ro g en Ion C o n ce n tration Berasain and Vita li (2 3) found that decomposition of s olution ot sodium ph e nylethy l barbiturate waa deoreaaed when the ir pH value was low e r The same wa s r oun d true tor so l ut io ns of sod i UDl diethy lb&r b i turate by V i.ad se n ( 3 ) Bailey (4) and by Tomski and Wa ller (2 4} S chl emme r and Torbe r (30) claimed that sol ut io ns of sodium diethyl barbiturate po ss e ssi ng a pH ot 5 9 2 underwent no hydrolysis within the l imit of error of 0 4 % when subje cted to t h e usual conditions ot star 1liaat1on Be rasain and Vital i (2 3 ) attempted the use of buffe rs to control the pH The lowest p H value of t h ir so lution was pH 7 A decrease i n the amount of ~ reoipitate was noted 1n s ol ut ions of pH les s than 9 oo when heat ed fiftee n m in utes by steam M esnard (18} su gge sted the uee ot sodium b icarbo na te aa a stabilise r in solutions of sodium phenoba rtitalJ its effect waa supposedly as f o l lows, N aHC03 1N a OH. N a2C 0 3 + H zO C 12 H 11 O g N ( N H) tN a2C03 __, C12 H 11 03N ( N N a) + NaHCO z Be raeain and Vitali (23) stated that the u se of sodium bi oarbo na te was u n satisfactory b Am i n es as S ta bi lizin g Ag e n ts One of the earliest a pp l i cati o ns of am i n es to stabilize barbiturate

PAGE 31

26 solutio ns mad e its appea r an c as a co mme rcial p roduct called Somni fe n So mn itene is manuf actured by H offmanLaRoche and Company (31) and con tains i n l cc th e diethylam.ine salts of 0 1 Gm of diethy l barbiturie aoid and 0 1 G m o f isopropyl propenyl barbituric cid I n 1921, Le M ed i cin (32) described it as being soluble and w ith e ff ects superior to those of the other barbiturate derivatives N ew Somnif ne ii t h iso propylallyl barbiturate of dietbylamine A ccording to Ca zzani (17). it oan be terilized by tyndallization at 8 0 c ., h as g ood keeping quali ties. and ie well tolerated In a British Patent (33) it i olaimed that a 6 to 10"' ~ solution ot butylethyl be.rbituric acid which is stabl at 100 c and steril i&able by tyndallization, can be ma de by neu tralizin g the acid with aminoethano l d iethylami ne thanol and diethy l amine Cazzani (16) stated that keeping qualities ot these solutions ar g ood P i p erazine has also been used in an analo g ous manne r and Cazzani (1 6) i nd icated that it too yields solutions with ood k eepin g qualities All of the a b ov solu tiona are claimed to be well tolerated Bl ok (34) reported that a stable sol ut ion of p henobar bi tal suit able for i ntramuscular i njection can be m ade by the uae ot diethylamine B lok (34) su gg ested the fo llowin g form ulas Phenoba rbital Diethylamine P etit M ixtur 4 ad Pe tit M ixture A lcohol (90%) G lycerin W ater q s ad 10 Gm 2 76 Gm. 100 cc 26 31 100 Gm Gm co

PAGE 32

27 The phenobarb tal was placed 1n a sterile flask co n tainin g some steril~ Petit M ixture. 'l'he d iethyl amine was a dd ed and th mixture was shaken and heated at 40 to 50 c in a closed fl sk until solution was ef tected. Sufficient Petit M ixture was added to !ll8.k e 100 co. Then t he solution was filtered, placed i n ampuls and sterilized for a ha lf how at 60 c durin g two successive days N o ap p reciable decomposition was found after three months Wh en hea ted in flowin g steam for one ho u r 2% was hydro lyzed. Schulte (35) modified Blok s procedure (34) by neutralizing the barbituric acid w ith diethyla.mine dissolved in l cohol and then adding glycerin In this way the solution is obtained without the use of hea t Berasain and Vitali (2 3) subjected Blok s so lution (5 1 ) to ste am heat for fifteen m inutes T l w solution became colored and a.n abundant quantity of p r ecipitate was produced G lycerin and alco h ol would be objectionable for injections. The latter authors pointed out that glycerin produces deleterious effects at the site of injection In addition alcohol produces a disa gree able burnin g senaa tion when iiljected Van Leent (27) stated that a solution of sodium barbital 1s best prepared by di ssolvin g barbital in diethylamine and sterilizing tor thirty minutes at 00 to 66 c o n two successive days E R Squibb and Sons have patented (36) a p reparation which is claimed to be stable and which may be steriliz d by he ating It is an aqueous solutio n containin g ethyl isopropyl barbituric aoid and mor than a molecular proportion of diethanclamine M esnard ( 1 8 ) st ted that sodium p henobarbital in solution hyd r o l yzed

PAGE 33

28 into phen obarbital and sodium hydroxide ; followed by cleava ge of the phenobarbital it elf du to the action ot the sodium hydroxide formed To prevent this action, M esnard (18) su g g ested the use of ethanolamine aoetate 1n soluti ons of this barbiturate The intention was to have the sodium hyd roxide liberate ethanolamine which would form a stable aalt with barbiturio acid Be rasain and Vitali (23) declared this pro oedure was unsatisfactory They found that 1 0 % solutions of this bar biturate prepar d with mono di and tri ethanolamine (or pipera zine ) were n ot stable when subjected to steam sterilization for fifteen min utes N i g htingale and M orris (37) stated that the amine salts of the 1-phenyl derivative of 5 bromobe.rbiturio acid are less stable than those without the phenyl g roup In alcohol 6 6 dibromoba.rbiturio acid re acted rapidly in ethyl aloohol with butylamine to g ive the relatively stable butylamine salt ot 6 bromobartituric acid Analogous compounds with a 1-phenyl g r oup were unstable Cold sodium carbonate ca used the latter to break up into butylam.ine and lp henyl 6 bromobarbituric acid Br olumal is a p roduct of the Italian Drugs Importing Company Inc (38) which co ntains hexamethyltetramine E ach ampul of 2 co container Ph enyl e thyl ba.rbituri o acid 0 015 Gm Organic Bromine 0 010 Gm Hexam.ethy ltetramine 0 010 Gm liydroglyeeric solution sterile q s Bro l umal is made for intramuscular and intravenou injection Eli Lilly and Company (39) claimed that stable solutions of amytal could be made by dissolving amytal 1n hydroalooholic mixtures containing a sufficient

PAGE 34

29 quantity of m ethenamine E l i xir .Am yta l, Lil l y is st a ted t o be s uch a stable s ol ut io n Ea c h f l u i d oun c e c on ta i n s am yt al 2 g r i n a vehi cle co n t a i n in g m e t henamin e 2 g r pe r f l u i d o un c e g l yc eri n a lco h ol wa ter and ar o mati c s The al oo h o l c o n tent i s 50% E lixi r Am ytal L illy (1 % a pp rox ) Am ytal M et h e nam ine G lycerin A lcoho l W ater Aromatic g r iv g r. iv q s aa ad 3 i T h e a l co h ol co n te n t is 34 % In a German p ate n t (40) men tio n is m a d e of th e p r ep aration ot & solu b l e co mp o un d of ph e n obarbital b y the a d dition of at le a st on e m o lecular p r op ortion of diet h yle n edi am i n e Van de V elde (41 ) s tu died t h e i nf lue n ce of n e u tral salte o n the hy drol y sis of urea at elevated t em per 1 tures H e f otm d tha t hydroly1i1 was a c celer at ed by 1odium p otaaai wn and p h o sp ha te io n s an d r e tarded by amm o n i um ions c Am i des a s Sta b ilizi ng Ag e n ts Th e u se of ami dea to p re p are s ol u tio n s of ba r bi tur at es s ui ta b le for inje ction ha appe ar ed in a num ber ot p at en ts ( 42 43 ) Unsu bsti tuted am ides o f th e l o wer ali ph atic acids an d a m id es of the l ow er ali ph a t ic acids w it h o n e or tw o a l ky l g ro up s at tac h e d t o th e ni tr o g e n at om are u sed to co nv ert ba r b ituric aci d deriv a tives su ch as ph e n yleth y l bar b ituric acid or c y olo h exe n yl bar b ituric a ci d int o wa ter solu b le c o mp o un ds

PAGE 35

30 whose solutions are claimed to be suitable for subcutaneous injection Specific amides mentioned are mono methyl and m onoethyl acetamide In a French Patent (44) a description is g iven or the uee ot un substituted amides ot fatty acids having at least three carbon atom, in the m olecule as auxiliary agents to o btain 20;4 solutions of barbi turic acid derivative, the 6 5 subetituenta of which consist or an aliphatic and a cyclic group A solvent for cyclohexylethyl be.rbiturio acid ia g iven as, P ropionamide 82 Betaine 3 W ater 1 5% A solvent for phenylethylbarbiturio acid is g iven as, Propionamide 50 Acetamide 40 W ater lo% Similar solve n ts are g iven in an Au strian Pa.tent (45). A 10% solution of phenylethyl barbiturate in the followin g solvent is cited, Ace ta.aide 65 Betaine 5 W ate r 30% Addition of a little urea a n d/or ethyl alcohol may be advantageous, according to the p atent d Use of Urethane in Conjunction with Other A g ents A British Patent (46) described the uae or a mixture of tertiary ohlorobutyl alcohol and urethane as a solvent tor substituted barbiturio acid This alcohol and urethane may be m ixed 1n equal part by wei g ht to g ive a liquid product whi o h is suitable as a solvent of barbiturates tor i n jection

PAGE 36

31 The Dani h P n armacopoeiF (47) co n tains a formula entitled "Com pound Solution or Ph enylethyl barbiturio Acid" in which ethylurethane is employed The formula is as follows Ph enylethyl barbituric aoid Amylene Hy drate E thyl U rethane Distilled W ater 20 Gm 28 Gm 35 Gm 7 Gm The solution is prepared by keepin g the mi xture in a well closed flask at 20 to 30 c until diaaolved It ia then i' iltered throu gh a bacteria proof filter and sterilized Zwikker (4 8 ) claimed t ha t this non alkaline sol ut ion is suitable for injection Van Leent (27) stated that suoh com pounds a amidopyrine a n d ca f feine may be dissolved in the solution Wh ether or n ot they act as stabili,in g agents was n ot stated Beraaain and Vitali (23) claimed that the Danieh f ormula is unpractical f or in jection because of the p roportion and nature ot the ingredients N ielsen (49) followed the deoompo ition of the barbituric acid de rivatives co n tained in Solution Hypnopheni (Ph Dan 1933) by measuring the carbon dioxide produced Storage tor a period of one year showed slight deco mp osition of diethyl barbituric aoid allylisopropyl barbi turic ai,id and urethane ~ Ni elaen (50) developed the followin g formula for injection Phenylethy l barbituric aoid S odium pheny l ethyl barbiturate U rethane Alcoholic Spirit {Ph Dan ,1933) Glycerin Aq dest. steril q a ad 3 0 6 72 2 5 0 16 0 12 5 100 0 cc The specific g ravity of this preparation at 15 c was 1 066 I n three m onths at 12 c ., about 0 65 % : ) the barbiturates were hydrolyzed and at 24 C in the same time 1 35% were hydrol y zed This decomposition

PAGE 37

32 was one fifth of that occurring in a 10% aqueous solution or sodium phenoba rbital, kep t under the same co ndi tio ns The solubility ot th e ma i n decompo sition pr oduct, phenylethylacetylurea i n the sol u tion was 1 % at approximate l y 22 c T h is correspon d e d to a 12 3 % deco mp osition of the sodium phenobarbital Therefore Nie lsen (60) reasoned th t the preparation could be stored for a lon g ti me without any p recipitation Arter 2 hour war m in g at eo 0 c only the p rimary decomposition was de tected and it amount d to about 0 75 % ot sodium p h enobarbital The deco mp osition of urethane en surin g d urin g h eatin g a t s o 0 c was also dete.rmi n ed The decomposition is represe n ted by the following equations, (A) / NR z O == C \ (B) N i elsen \60) also g ave a formula fo r a 1 0% solution of ph enobar bi ta l, suitable tor injection, containing diethylamine Ni elsen's f or m ula (50) is as follows, Phe nobarbita l U rethane Alcoh ol i c Spirit (Ph Dan 1 933) G l y cerin Diethyl amin e Aq dest steril ad 10 0 25 0 15 0 12 5 20 0 100 co

PAGE 38

33 The specific g ravity of this solution at 15 c was 1 035 Nie lsen (50) claimed that t h is solution was very stable No hydrolysis wae detected after 6 1/2 m o nth s stora g e at 24 c nor after two hours heating at so 0 c Beraaain and Vitali (23) hea ted tnis solution for fifteen minutes in steam It became deeply colored and yielded a pr e cipitate in a few days They pointed out that R akieten et al (51) had found that urethane in association with some ba.rbituric acid deriva tive in n arcotic doses had produced so mbe r effects and alterations in the acid base equilibrium tending towards a.oidosil The disad vantage of g lycerin and alcohol in solutions intended for injection have been mentioned elsewhere In a British P atent (62) the claim is m ade that a stable sol u tion of 5 5 disubatituted barbituric acids can be p r ep ared by dissolvi ng the barbiturate with the aid ot at least one m olecular proportion of an a l kano lamine in the p reae n oe of a relatively lar g e quantity or a carbamio eater F or ex amp le 20 Gm of 5 5 isobutylallyl ba.rbituric acid 5,6 Gm of m o n oetha.nolamine and 50 Gm of ur ethan e are dissolved in 34 o o of water e Other N itro gen Compol.Dlds as Stabilizing Agen ts A Bri tish Paten t (53) stated that stable solution of barbiturates for therapeutic use are obtained by adding p yridones or p i p ridonea to their aqueous solutions The pH of the solutions ia 6 0 to 6 8 A so l ut ic:m of allyl iaopropyl ba.rbiturio aoid co n taining 1-meth.yl 2 piperidone is s p ecifically m entioned Ha&leton K oppanyi and Linegar (64) employed antipyrine glycerin

PAGE 39

34 and alcohol as st abilizin e a gen ts for solutions of sodium se con dary amyl bromallyl barbitur at e. They su g ge sted the fo llowin g form ulas S o dium secondary amyl b romallyl barbiturate G lycerin Antipyrine Alcohol W ater q .s ad The solution is intend ed fo r rectal administration 10 Gm 10 Gm 10 Gm 10 cc 100 cc Gruber (65, 56 57) pa te nt ed the us e of urea, urethane, aoetamide and such pyra zolon derivatives as dimethylphenyl pyra zolon e tor the stabilizatio n ot aqueou s ol utions of barbiturate salts r Glycols as Sta b ilizin g A g ent P a g e and Coryllos ( 68) claimed that stable solutions of p ractical ly all of the barbiturates oould be made by dissolvin g them in ethylene g lycol M erely the co n ce n tratio n of the solv en t had to be varied ac cordin g to the barbiturate to p re p are a stable solution A solution ot sodium amytal is p repared i n the .following manne rs the amount of amytt.l need d to make a 10% solutio n ia dissolved with v ry vi g orous ahaldng in the least quantity of 15 % aodium hydroxide wi tho ut the use of h eat E thyl ne g lycol is added in sufficient quantity ao that its ooncent r a tion in the final solution wi ll be 16 % The excess sodium hydroxide ia neu tralized with dilute hydrochloric aoid t he solution be in g shaken vi g orou ly upon the addition of each drop of a cid The shaking is ne oes sary to p revent the separation of iaoamyl ethyl be.rbituri o acid i n clumps which are difficult to redissolve Wh en a f ine clou din ess re main s, the addition o f acid is disco n tinued A minimum quantity of ooncentrat>d ammonium hydroxide ia the n added quickly to clear the solution If a

PAGE 40

35 large ex cess of a m monia is presen t, it m ay be quickly boi led off ov er a f ree flame The entire p rocedure nust be pe rfor m ed r apid ly to pre vent hydrolysis Derasain a.nd Vitali (23) stated that they found this solution to be stable The y remarked t hat O etti ng e n ~nd Jerouch ( 59) had found ethylene g lycol to be toxi c hemo lytic, and intensel y ir irtating at the site ot i nje ction Page and C oryllos (66) p ointed out that iso am yl e thyl ba r bi t urio acid is solu b le i n eth ylen e g lycol and that it is no t es se nt ial t hat the sodium salt be p r epare d The et hy lene g lycol sol ution s may be heated to boilin g t o augment the already powe r ful st e rilizin g action of ethylene g lycol In a Bri tish Paten t ( 60) and a U s Paten t (61) it is stated that stable and sterilizable prep ar a tions of barbitur a te salts c an be p re p ared by the u se of alkyl ene g lycols as solve n ts The alkali m e tal salt of the barbiturie a ci d com p o und is dis s olved in the glyc ol, sueh as ethylene or p ro py l ene g l yc ol i n the abse n ce of any subs tantial quan tity ot water Th solution may also be ma de by diss ol ving the calcu lated quantit y of an alkali m etal i n the g lyco l under anhyd rous con di tions and then dissolvin g the free barbituric aoid in this solution D iethyl dipropyl -, d inllyl phen ylethyl and cyclo h exenylethyl barbituric acids and their substitutio n products were specified in t he patent Dum es (62) stated tha t h e had b en i nf ormed by certain manu fa ctur e rs that t hey were usin g e t hy l ene and propylene g l y cols to p re pare anhydrous s olutio n s of ba rbitur a tes of sufficient stability to permit sterilisation De r sai n and Vi tali (23) claimed that the followin g formu l a pro vided solutions stable to ste rilization s

PAGE 41

36 S odium pheno l arbital Propylene g lycol Distilled w ater q.s ad 10 Gm. 40 oc 100 co. The barbiturate was dissolved in the p ropylene g lyool and su.f'fioient water was ad d ed to make 100 cc. The solution was sterili& d by heat ing in steam for fifteen m inutes The solutions thus treated were found to be free of aerobic and anaerobic bacteria by inoculating bouillon and a g ar and observing a f ter 24 48 and 72 hour, Th y claimed that the solutions were stable for the len gth or time observed, name ly six months The concentration of the p ropylene glyco l is th m inimum. req u ired to maintain stability This g lycol was similarly used to make solutions ot sodium barbital and sodium am.ytal They withstood st ril ization with steam for 16 m inu t es Braun and Car tland (63 ) demonstrated that propylene g lycol is less toxic than g lycerin It has been reported that p ropylene glyco l may be substituted tor g l ycerin in a:ny non offioial rticle whe re g lycerin itself m ay be safe ly us ed without increasing the toxicity of the resultin g p roduct (64) In a u s Pa tent (65) the use of aliphatic polyhyd ric aloohola such as glyce rol or a g lycol is claimed to stabilize aqueous solu tions of barbiturate salt A G erman Patent (66) co nt ained a claim that stable solution ot barbiturate salts oan be prepared by using an anhydrous glyco l, such a ethy lene glycol as a solvent g M iscellaneous Sta b ilizin g Agents The use of bases weaker than sodium hydroxide to p repare water soluble compo un ds of various barbituric acids led Berasain and Vitali (23)

PAGE 42

37 to try calcium saccharosate. To prepare it 200 co of water was added to 10 Om of lime Afte r twenty fo ur hour s had elapsed the solutio n was decanted and a sol uti on of 60 Gm of s ucr ose it 200 co of di tilled water was ad d ed to t he residue remainin g Ai'ter anothe r twenty four hours had elapsed the c~lcium sacc harosate was obtained by f1 1 tration It was f ound to b e a go od solvent for phenoba r b ital aow ever an abun. dan t p recipitate w hich did not d isappear o n cooling was pro duced when solutions were sterilized with stef.\I:', for 1 5 m i ni; tas P a g e a nd Coryllos (5 6 ) claimed to h ave p repared a fairly stable solution of sodium 9.lnytal suitab le for i n travenous injection by the use of muoilag of acacia A 20% sol uti on of acacia in sufficien t quantity to make a final concentration of lO ;i was added to a solu tion ot e.mytal just before the exc ss sodium hydroxide used in dis solvin g the acid was neutralized by h ydr oc hloric acid The solution was further stabilized by the additi on of sufficient ethylene glyc ol to mak e 1cr ; An anonymous writer in the Pharma ceu ti cal Journal (67) mad e the claim that suspensions of phenylethyl barbiturio acid in m ucilage of tragacanth retained their therapeutic activity for a lon g time &ldi (68) stated that a stab l e solution of sodium phenoba r bit al could be p repared by dis solvin g it in pure glyo a rin or a m ixture of equal parts ot glycerin and wate r He reported that a 4 % solution prepare d in this way remained p erfectly clear and resisted dete riora .. tion on heating to i oo 0 c Baldi (68) g ave the fo llowi ng formu lai

PAGE 43

3 6 S o d i um P h e n oba r bi tal G lycerin A l coh ol D istilled water q s ad 20 Gm 20 G m 10 Gm 100 cc J.i' ive p er ce n t magnesium sulfate and 1 0% mannitol are oocasionallY: used to p revent the d issociation or organic salts Berasai n and Vitali (23) found that t heir use in solutions of sodium phenobartital failed to g ive g ood stabilization. B ush Dickinson and La m son (69) studied th sta b ility of Seconal i n p araldeh y d They found that a 30% solution of seoonal in paralde hyde retained its or i g inal anesthetic activity nearly compl e tely after h eatin g a t 120 c for four hours U pon i n tramuscular i n jection of t h is solution in some thirty patients it was fowd that no disco m fort n or any sign of irritatio n or i n jury to tissue was pro du ced These workers determi n ed the solubility of a n um ber of barbitur a tes in va rious subs t a n ces such as p ropylene g lycol triacetin o e llosolve eto Sta b ility of t h ese soluti o ns was n ot reported B M E T HO D S OF A S SA Y ING BARB IT UR IC ACID D E R I VAT I VE S 1 Pro c ed ur e s Em ployi ng Imm isci b le Solve n ts a Th e U s P ,. XI a n d U S P X II M e thods (70 71) Th e s al t o f th ba r b it u r a te ( ba r b ital s odium p e n to be. r b ital sodium and p h en o ba r bi tal s o d i um ) i s di sso l ved i n water and th e n p reci p itated as the f ree a c id by t he addi tion o f diluted hydroohlorio acid T he o rganic acid is extract d w ith e th e r '.t'h e eth e real sol u tion i s e va p o ra ted to dryne s s The r e sid u e is dried to co n stant wei g h t I n the u S. P X I

PAGE 44

39 assay (70), the ample is dried before extraction, w her ea e 1n the u s P X II assay (71), the percenta g e of barbiturate recovered is calculated on a m oisturefre be.sis, e m pl o yi ng co n t e nt sample ot known m oisture P h enobarbital T ablets and Ph enobarbital Sodium Tablets of the u s P X II (71) are freed of such dilu e nts as stearic acid by aha.king with a saturated solution of sodium chloride and barium hydroxide The vol a tile solvent e m ployed in the extraction of the barbiturate is com posed of ei g ht volumes ot chloroform and two volumes of ether b M ethod of Rotondaro Aspelund and Skoglund (7) separated the decomposition products from deteriorated sol u tions of various barbiturates by filtration and extraction with ether at various p H levels This work showed that all aqueous solutions m ay undergo appreciable decomposition Analyses em ployin g immiscible solvents require that the barbituric acid be free fro m its pro d ucts of deco m position Roto n daro (72) poi n ted out that the N F VI M ethod of assayin g E l ixir of Phenobarbital (73) eliminated the n eutral decomposition product ; phenylethylacetylurea by discard in g it along with the aromatics The acidic deco m position products p h enylethylmalonuric acid and p h enylethylacetic acid ar not eparated from the barbiturio acid The presence of these acids arising from the deco m position or p henobarbital ia indicated by the fact that the melt in g point of the barbituric acid residue is usually lo~ R otondaro (72) asserted that the co n taminant waa phenylethylaoetic acid and not g lycerin as had been expected by others when assaying E lixir ot

PAGE 45

40 of Ph enobarbital N F VI (73) Ro tondaro (72) carried out an experime n t in wh ich he was able to isolate several of the decomposition products of a barbiturate A solu tion ot phenobarbital in 0 2 N potassium hydroxide was ref lu xed f or fi ve hou rs Up on dilution with w ater a flocculent p reci pitat e was f ormed A portion o f the mi xture was extracted with chlorofo rm Th e residue ob tai n ed on e vapor a tin g the chloro f orm extracts w as s y ru py H owever on dissolving i n alcohol and dilutin g with wa ter, a wh ite, tlocc u lent p re cipitate se pa rated out Its m eltin g p oint was the s am e as t ha t of ph e ny lethylaoetylurea Anothe r p o r tion of the alkaline solution wh ich h a d been refluxed was made acid to lit mu s w ith dilute hydrochloric acid U po n add in g excess sodium bicarbonate and ext r ac tin g with c h loroform R oto n daro (72) was a b le to secure a residue which had the m elting p oint of p henobarbital The remainin g bicarbonate sol u tion wae acidified and extr a cted with chloroform A s y rupy residue was obtained wh ich R oto n d aro (72) was unable to crystalli&e Ho wever b y titratin g a soluti o n ot the residue in alco h ol w ith 0 .1 B sodium hydroxide a titration equivalent or 206 wa obtained The titration equiv a le n t of a 1,1 m i xt ure of phenylethylacetio aoid and ph enylethylmalonuric acid is 207 Rot o n daro (72) devised the following meth od to assay E lixir ot Phen obarbital N F V I (73)a "A 5 0 co sam p l e was diluted with about 50 cc of water in a sepa rator satur a ted w ith so d i um chloride and m ade aoid with dilute hydro c h loric acid The mi xture waa extracted to completion (six or seven ti m es) with 25 35 cc portions of chloro f orm or about 36 50 oo portions of ether if emulsions formed The chloroform extractions we re filtered

PAGE 46

41 through chlorofor m-wet cotton into a beaker and the solvent was evap orated to a small volume (4S co ) on a steam bath with the aid or a brisk current of air oare being taken that active ebullition ot the solvent did not take place The solution was transferred with the aid of a little chloroform to a separator containing about 20 cc of alka line salt solution (0 6 N sodium hydroxide saturated with sodium chlo ride) The mixture waa shaken thoroughly and the chloroform was al lowed to separate aa completely as possible The chloroform wa. drawn off into a second separator co n taining about 3 oc of 0 6 N sodium hy droxide These were shaken we ll The chloroform was allowed to sa p rate completely and was then drawn off throu gh chloroform-wet ootton into a tared dish The extraction was repeated with three or four mo re 20 cc portions of chloroform The combined chloroform extracts were evaporated to dryness A few treatments wi th anhydrous ethe r g reatly assist in completely expelling the last traces of the chloroform solvent The residue was dried at 90 C for ten to fifteen m inutes Residu e "A" may contain non-volatile aromatic oompoWJ.ds plus any neut ral de composition product s The alkalin wash water was combined with the alkaline salt solu tion A pieoe of litmus paper wa p laced in the separator and acidified with dilute hydroch loric acid An excess of dry sodi'Ulll bicarbonate was added in small portions. The phenobarbi tal was completely extracted with 20 co. porti ons of chloroform E ach po rtion was filtered sue cessively through chloroform-wet cotton into a tared dish The solvent was evaporated with the u ual p recautions The residue is usually syrupy due to tr a oea or chloroform. which are tenaciously held until

PAGE 47

42 super h ating causes the chloro f orm to vaporize with explosive violence thus causing losses by decrepitati o n A few small additions or anhydrous ether material ly shortens the time required to drive off the chloroform minimize th loss by decrepitation and leave dry granular residues which possessed m elting points very close to theory The soditnn bicarbonat solution in the separator was then cautious ly acidified with dilute hydrochloric acid The mixture was extracted with three or four 20 co porti ons of chloroform or other appropriate solvent, filtered and the solvent waa evaporated aa usual The residue "C" was examined for deco mp osition or acidic drugs." o M ethod or van Itallie and Steenhauer (74) The method of van Itallie and Steenhauer (74) is an immi cibl solvent m ethod adapted for the quantitative determination of barbitu rates in urin and organ extracts In addition to the procedure which must be employed to remove extraneous matter the only differen ce whi e h is ot interest in this discussion is the nature of the volatile solvent These workers claimed that substituted barbituric acids are inoomplete ly extracted by ether and that ethyl acetate eliminates this e rror by its greate r solvent power Ve ro na l is mo re than twice as soluble in ethyl acetate as in ethe r. Man cani (75) described a pr ocedure which was simil a r to that employed by van Itallie and Steenhauer (74) d M ethod of Glycart (76) The be.rbiturio aoid is dissolved in a 2% solution of sodium hy droxide whioh is saturated with sodium chloride The solution is shaken

PAGE 48

43 wi th ether The ether washin g 1 diecarded The contents of the sep rator are then acidified with hydrochloric aoid and extr a cted with mixture ot two pa rts ot ethyl alcoho l, one p a rt or ether and seven parts of chloroform The extract is e v apo r ated to dryness and wei g hed The method of Glycart (7 6 ), with sligh t modifications was adopted as th official method of a say for barbital and phenobarbital by the Assoc ia tion of Of f icial Ag ricultur al C h mi sts (77) The official method em ploys a solvent cons isting of two parts of ether and eight p arts of chloroform The method ot G lyc art (7 6) was also employed with slight m odifi cation i n the N F VI assays of E lixir of P h enobarbital Elix ir ot Ba rbital, and Tablets of Phenobarbital (73) and in the U s P XII as sa ys of Barbi tal Tablets and Pent obarbital Tablets (71) In th u s P X II (71) the assa y ot E l ixir of P henobarbital ia changed to an assay co nsi sti ng essentially or ac idifica tio n ~ 1th diluted hydr o o hlo r ic cid and extraction with chloroform I n the N F VII (78) the assay or E lix i r of Barbital was modifi ed to th e extent or using solvent composed ot one volume of ether and n ine v o l ume s or chloroform In addition to th above g rav imet ric methods the re are other types depe n ding upo n the formation of complex addition p ro du cts such a the complex p r odu cts formed by the reaction of phenoba rbital with m ercuric iodide as reported by M ontignie (7 9 )J the complex salt formed with iodine i n an alkaline solution as repor t ed by Boug ault and Guill o u ( 8 0)J c omp lex me rcury alts aa reported by F leury ( 8 1)1 and the copper pyri di ny l phenob rbital f ormed by the u se of Zwikk er's rea gen t ( 82 ) Bell { 8 3) stated that mo at or the m ethods depending upon the f ormation ot

PAGE 49

44 complex addition products wer unreliable due probab ly to the incon sist nt or unknown composition of th addition product 2 Pro oedur s Emp loying Indicators Several indicators have been uggested for the direct titration of the replaceable hydrogen atom in the mal onylurea g roup The a cid form of the barbiturate was dissolved in a solvent consisting wh olly or in part of an organic substance Joensson (84) stated that veronal cannot be titrated directly with sodium hydroxide using phenolphtha lein as an indicator Explanati on was made th.at the pH of a 0 l mo l a r solution of veronal sodi'Ulll is pH 10 2 and that the pR of a 0 01 molar solution of the same compound ia pH 9 7 whereas th e color change of phenolphthalein occurs at pH 8 2 This difficulty was overcome by the use of an organic solvent such a acetone The method of Joensson (84) consisted of dissolving the vero nal in 20-26 co of acetone adding 0 10 to 0 15 co ot ls2000 phenol phthalein solution and titrating to a faint red color with O l N so dimi hydroxide The method of Sol ts (86) tor th determination of luminal differed only in the use of alcohol as a solvent Von Babit sch (86) empl oyed thymolphthalein as an indicator in titrating veronal and lumin l with sodium hydroxide because its transi tion limits of pH 9 3 to pll 10 5 encompass the pli of veronal sodium and phenobarbital sodium. Von Ba.bitsoh (86) employed alcohol as the sol vent He obtained results which re p rese nte d 99 77% of the samples taken Oerv ay (87) also report d the use of thymolphthalein as an indicator for the titration of ph enobarbital The solvent in this case consist dot

PAGE 50

46 methanol which was neutral to thym.olphthalein Gervay (87) suggested that ali,arin yellow R instead ot thymolphthalein be used as an in dicator when tit r ating veronal Delen i te ka (88) modified them thod of von Babitsch (86) to the extent of using a solvent consistin g ot tw o parts of aloohol and one part of wate r M orin (89) titrated barbituric acids which were dissolved 1n aoe tone with a m ethanol solution of potass}um hydroxide i n the p resence of thymol blue En ell (90) t it rated barbital sodium i n ether using iodeosin as the i n dicator 3 Conduotimotrio Tit r ation H ryn kowald and Mo drzejew s ki (91) stated that tho application of co n ducti me tric titration to the determination or weak acids and bases gen erally gave g ood results They claimed to have determined barbit&l and ph enobarbital with an average e r ror or 0 2 to 0 6 % Substances in soluble in water but soluble in ethyl aloohol can be titrated in alco holic solution Better results were obtained however if the weak aoids after dissolving in exoes of alkali were titrat d baok with hydrochloric acid 4 Procedures i n which Undissociated Barbiturates are Form ed (a) M ercury Salts Ionesco Matiu (92) and von Babitsoh (86) developed procedures which are dependent upon the i n solubility of the me r cury salts of barbiturates i n sli g htly acid media The p recipitant is prepar d by dissolving

PAGE 51

46 mercuric oxide in co n centrated sulfuric acid and diluting with water The p recipitate formed by the addition of this rea gent to the ba.rb1tu rate solution is separated washed, and dissolved in a mi xture ot sul furic and nitri c acids The amount of me rcury pre se nt is determined and fro ra this the amount of barbiturate is calcul ated (b) Silver Salts Budde (93) tated that the procedure of Stas-otto (immiscible solvent m ethod) is tro u blesome and time consuming because of the re peated shaking out with ether Budde (93) claims th a t an alkalimetrio determination of barbituri o acid does not lead to satisractory results The probab le p resence of replaoeable enolio hydrogen atoms during titration with sodium hy droxide brings about only a slow chan g e in hydrogen io n concentration so that an exact en d point oar only be observed with difficulty and u n der speoial conditions Budd e (93) developed a volumetric determination whi ch is be.eed on an obs e rvation of Zernik that m ercuric oxide is dissolved by veronal in a p resumably alkaline media Budde ( 93) assumed that a mer curic chlo ride of known str ngth may be titrated into an alkaline v ronal solu tion until an excess of m ercury ions is brought into reaction w ith the alkali with separation of me rcuric oxide Since a sta n dard solution of m ercuric chloride is not official in the Deutsche Ar1neibuch 6 Budde ( 93) p roc eed ed to i n vesti ga t e wheth r or not silver nitrate eolu tion acted in a similar way It was found that in the pr esence of so d1um hydroxide o as to o s mo ls of silver nitrate was consumed by 1 mol of veronal upon the appearance of the first turbidity Silve r

PAGE 52

47 nit ra te is transformed by the sodi um veronal to a silver ba rbiturate whi c h is a olu b l e in the alkaline media Like me rcury cyanide, the sil ver ba r biturate is very diffi oultly dissociated ., so that th e excess ot sodium hyd ro xide gradual ly and inc o mpletely produces a separation of silver oxide Rup p and Poggendo rt (94) r epo rted t hat pheno barbital is p re cipitated by silver n itrate. The p recipitate wa s f o lmd t o be soluble i n auun onia. Vieb ~c k and Fuch s ( 96 96) g ave the equations as follows, c 12 H 11 o 3 N 2 N a 1N aO H + 2A gN0 3 c 12 H 10 o 3 N 2 Ag 2 + 2 N a N 0 3 + H 2 0 (p h enobarbital sodium) c 8 H 11 0 3 NiN a + N a OH + 2A gN0 3 ---,. C 8 H 10 o 3 N 2 A g 2 -+ 2 NaN 0 3 + H zO (barbital a) Tomski and W aller (97) g av e the following as t he course of the reaction a 2 CizHuOs N aNa + 2AgN03 C12H1103 N 2 N a CizH1o03 N 2Ag2 + NaN Os + BN 03 (a soluble ., double co mp ound) Tomski and W aller (97) state that silver ph enobarbital is insoluble ., presumably mean i ng in acid or n eutral me dia The first xperiments ot Budde ( 93 ) showed result which were 10 to 16 % oft H e thought that better results could be obtained by the use of weaker alkaliea However ., later workers ( 83 95 ., 96) indicate that sodium hydroxide may be used but the quantity m ust be re g ulated

PAGE 53

48 Bu dde (93) developed the following m ethods About 0.2 to 0 3 Gm. of the sample are accurately weigh ed and d is solved in 30 cc of water with l Gm of anhydrous sodium carbonate. To the clear solution, O l N silver nitrate ie added until a d istinct tur bidity i produced which remains for some time Von &.bitsch (86) fotmd Budde 's method (93) completely satisfactory The method employed by Kalinowski ( 98) involves a titration of an alco h olic or acetone soluticpi of the barbituric acid in the p resence ot sodium hydroxide To a solution ot 0 2 to 0 .3 Gm. of the substance to be analyzed in 20 to 26 co of acetone or 30 to 36 cc of thyl alcohol 15 to 20 oc of a normal solution of sodiwn. hydroxide and 20 to 30 cc of water are added This clear solution is titrated with O l N s olution or silver nitrate until a turbidity is obtained Ka linowsk1 (98) claimed the meth od to be accurate to 0 1 % II Viebook and Fuohs (95) explained that phenobarbital and barbital are monoba.aic toward alkalies and dibe.sic towards silver nit rate If to a solution of ph enobarbital two equivalents of sodium hydroxide are added. then two equivalents of eilver nitrate the silv r salt is pre cipitated and the solution is n eutral W hen lumi na l or veronal is titrated with sodium hydroxide in the presence of exci,ss silver nitrate, towards the end of the titration silver hy droxide is fo r m ed which does not dissolve and which is easily concealed by the i n soluble veronal silver or luminal silver To overcome this difficulty sodium salicylate is added The latter replaces the excess

PAGE 54

49 silver nitrat with silver salieylate Silver salicrlat ha a mod erate solubility sufficient for a quantitative precipitation of veronal silver or luminalsilver but on the other hand it hi n ders the formati on of silver hydroxide The m ethod of Viebook and Fu chs (96) for the determ.in tion of alt of barbiturat s i as follows: A D etermination of alkali Diss olve 0 2 Gm of t he co mp ound in 25 cc of w at rand titrat with O l N sulfuric acid with methyl oran g e as the i nd icator This step is p robably to re g ulate the quantity or alkali in the sol u tion inas m uch as barbiturate salts occur o n the marke t with vary in g amounts of lkal1 ( 4) In tho n ext step a definite amount of al kali i added T h is avoids the difficulty experien ced by Budde (93) in his use of sodium hydroxide B Determ i nation of Phen obarbital and Ba rbital The solution from part A is freed of oarbonic acid by h e ating~ A m easured quantity of O l N carbonate free alkali is added Fo r 7 6 co of e.oid used in the first titration 14 0 to 1 4 5 oc of alkali are used in the second titration (generally 0 5 to l O oo less than double the quantity of acid used in the fi r st titration) About 25 co of 0 .1 N silver nitrate is added (5 to 10 oc more than double the quantity used in the first titration) Afte r a further addition of 0 5 to 1 0 Gm of sodium salicylate the solution is tit r ated with O l N a l kali with ph enop 11 thalein as the indicator W ith an absolute l y pure prepa. r tion the quantity ot alkali used in B must be exactly twice the quantity of aoid used in A

PAGE 55

50 Fuchs {99) also employed the a b oTe me thod to assay diethyl bar bituric acid by titr a tin g the fr ee acid which was dissolved in aloohol, with alkali i n the presenoe or thymolphthale1n. The second part of the assay outlined above was unchanged It would aeem from the work ot Budde (93) Viebo ok and Fuchs (95 96 ), an d To m ski and W aller (97) th.at the barbituric acids are dibaaio towards silver nitrate in neutral media and monba ic to wards silver ni trate 1n alkaline medi a Madsen (3) and N ielsen (13) fol lowed the co u rse of hydroly sis ot barbiturates from the amotmt ot dialkylaoetylurea and carbon dioxide formed. Schlemmer and Torber (30) state that this m ethod is incomplete since hydrolysis of dialkylaoetylurea ie assumed to p roce d further to produ oe diethylacetio aoid and urea as shown by the equations whioh follow Ii CO N R / '\ / c co R \ / CO N H 1i O ) 2 The values of pe rce n tage decomposition obtained by Mad sen (3) are lower ff than tho e obtained when Schlemmer and Torber (30 employed Budd e"s pro cedure (93) This p rovides evidence tor the criticis m of Schlemmer and

PAGE 56

51 Torber (30). mentioned above Bel l's ( 83) applicatio n of Ka li no~ ski' s modification (98) of the B udde method (93) to th analysts o:f' theobromine and ph enobarbital mix tures is as follows, The sa m ple co n taining theo'bromine and phenoba r bit l is bo iled with 2 N s ulfuric acid The mixture is coo l ed to 40 c and transferred to a s eparatory funne l T he mixt ure is shaken out with alcoho l .. f'ree ether The xtracte are filtered throu gh cotton T h e c ombined ether e xt raota are evaporated The residue is dissolved in exact l y 10 0 oc of a sol vent c onsisting ot 50 cc of ethyl al co h ol 10 co. o f wate r and 40 cc ot 1 0 N sodium hy droxide The s ol ut i on is f ilter d and the f irst quarter is disca rded The aliquot t h ree quarters r emaining is titrated with 0 1 N silver nitrate unti l a d istinct tur bid it y is e vid en t Th e silver n itrate must be added ve ry s l owly and toward the e nd p oint must be added drop by drop shaking vigorouely after the addition ot eaoh drop The imme diat t ur bid ity caused by the addition of eaoh drop must be cleared by shaking befo re the ddi tio n of then xt drop The e n d point i cha racterized by a pe rsistent turbidity of the sol ut ion even a:f'te r vi g orous shaking The titrated solutio n a t t h is p oint be g ins to c hang e ra pid ly to a brownish black color due to the se p aration 0 sil ver oxide Be ll ( 83 ) states that the alkalinity of the solutio n o f extracted phen obarb i tal should not be varied appreciably f ro m the co n ce nt ration mention ed be c ause the s h arp n ess of the end p oint is i nf lue n ced by thia fa ctor Schulek and Rozza (100 101) devised an ar g entom e trio assay of

PAGE 57

62 ba rbitur ~ tes w it h p otas s i um ch ro .ma t e as t he in d icator Th e barbitu r a te (O l 0 1 5 Gm ) was dis s o l ved by b oili ng in 26 cc of a 5 % borax so lu tion P otassium c h ro ma te (1 cc of a 1 0; 1' s olution ) wa s added S ilver nit rate, 0 1 N was i n tro d uced into the hot m ixture until the y ello w ishg reen turbid liquid attained a reddish color, which per sieted eve n after boili ng Since S oh u lek and Rozea (100 101) stated t h at l cc or 0 1 N sil v er n itrate is equivalent to 9 250 mg ot bar b i tal it ma y be concluded that both hydrogen atoms attach d to the ni tro g n atoms were replaced 5 C olor ime tric M ethods ( a ) Me t h od of Z wikker (102) Ba rbitur a tes react with cobal t c h loride i n anh y drou s m e dia to g ive a red color Addition or barium oxide converts this to blue A red dibarbitalooobaltodiamine has been isolated by Z wikker (103) (b) M ethod of Oettel (106) The sample is extracted with chloroform an d filtered T h e solution is di l uted as necessary Two (2) co of the sample ia transferred to e oh of three teet t u bes Th en 0 06 cc ., O l cc and 0 16 co of 0 2 % cobalt acetate solution in absolute methanol are added After shaking to the sam res pecti ve t ubes 0 06 oo ,, 0 1 cc and 0 16 cc of a 0 2% solution of lithium hydroxide in absolute m ethanol are added These volumes must be aoourately m easured If all three tubes or o n ly the last two g ive an i n te n se blue colo r, the c h loroform solution co n tains m ore than 0 .1 m g per co of a be.rbi turio acid derivative It the test is p ositive i n th first two tubes

PAGE 58

63 and negative or disappears wi thi n a minute in the last, the chloroform extract co n tains about 0 05 mg of a barbituric acid derivativ. It only the first is positive, the chloroform extract cor.ta ins about 0 025 mg of a barbitu ric acid d s rivative With practice, fair aeeur cy can be attained. By comparison wi th sta n dards greate r accuracy should be possible If all sa.rcples are positive or the tv,o most concentrated are posi tive dilute the chloroform extraot. A l arge excess of barbituric o1d de rivatives c P.n give a negative reaction or a gray color in all of the tubes In that case dilute the extract with chloroform until the blue co lor is obtained (c) Me thod of Koppanyi et al (106) the oolorimetric estimation is based o n tho blue color produo 4 by cobalt acetate in alkaline solution Wh en the test ie performed with 1 % cobalt aoetate and 1 % be.rium hydroxide 1n absolute methanol it is sensitive to one part of barbital in 10 000 parts of soluti o n Wh e n the test is pe rformed with 0 2 % cobalt acetat and 0 2% lithium hydroxid in absolute methan ol it is directly sensitive to one pa rt of barbital in 100 000 par ts of solution and l in 2 000 000 a.f't r con ce nt ration of the ohlorororm solution twenty times The accuracy ot thie method is limited to about 6%

PAGE 59

54 III EXP g R I MEN TAL PAR T A SCOPE OF I NVEST I GATION The p roducts fo r me d crurin g the deco mp ositio n of b rbiturates and the manne r in which they are p roduced co n stituted the research ot ev eral w orkers name ly Steenhauer (2), Mad en (3) Bailey (4) N ielsen (10, 11 13) and Aspelmid and his cow orkers (7 8 9) From the agree m e nt in the results which were reported by the s e m en the nature of t he decomposition a pp eared to be amp ly exp lained H owever it was felt t h at an i n vesti g ation ot a number of factors in the deterioration ot ph e n o barbital and ph enobarbital sodium to g ether with research on several me thods ot r et ardation would add to the existing knowledge concernin g t n ese ph e n ome n a Ev al uat ion of sev e ral re po rted m ethods of assay we re m ade in an atte mpt to ascertain their applicability to s uch a study M et h ods e m ploying immiscible solvents and m et h ods employin g silver ni trate were st ud ied Sui ta b le m ethods of det erminin g the p erce n ta g e m oisture w re sou g ht i nasm uch as its accurate deter m i na tion was esse n tial in p re pa rin g solutions and in checkin g the accuracy of the assa y used in the exp e riments on deterio r at io n and stabilization The experiments on the deterioration of Phen obarbital Sodium u s. P X I in v a riou1 concen trations included the determinati on ot volume o~ p recipitate chan g e in p H chan g e in refractive index and p erce n ta g e loss of b arbi t ur a te when the solution wa1 su bj ected to h eating at various temperatures for various pe riods of time The relation between pH and deterioration was studied by making use ot sodium carbo nat e and ph osphate buff ers to

PAGE 60

56 re gula te the pH A number of sug g ested stabilizing agents were sub j oted to teats and assays to determine their i n flue n ce on the retarda tion of deterioration B A. ALYTICAL ME T HO DS AND C ON TROL l Mat erials used The phenobarbital w as p urchased from the M allinckrodt Che m ic al works It was labeled "Phenobarbital u S. P X I, Control HR Cl." D ata g ive n elsewhere i n the discussio n showed that the drug conformed to the sta n dards of the u s P X I and U s P XII as re ga rds moisture oon tent and m elting p oint The phen obarbital odium was likewise pur chased from the Ma llinckrodt C hem ical W orks Two lots we re obtained The labels read Phenob ~rbital Sodium u s P XI The control des ignations were KKM and KLM l. Data g iTen subseque n tly revealed that these lots co nf ormed to the sta n dards of the u s P X I and U S P X II as re ga rds moisture content and purity rubric The particular lot which was us ed i n an experiment i s men tioned in the review of the ex p er 1men t Th e source of the other che mi cals emp loyed is menti o n ed in the di c u ssion of the e xpe rime n ts concerned 2 Det er m ination of M oisture Unless ot herw is stated the sam p les of phenoba rbital sodium which were d irectly used i n an exp riment were ot known m oi st ure conte n t ~ uantitative d etermi na tion ot ph enobarbital sodium was calculated on a m oisture tree basia i e ., a lot of drug wh ose m oisture content had pr evioualy been determined on separate s amp les As subsequently die cussed it was deemed advisable to dry the samples to co n stant weight

PAGE 61

56 t temper ture slig htly above 100 c ~ The values obtained by drying samples in a vacuum desiccator approached thoae obtained by drying sampl sat the latter temperature The term, d ried to constant weight was i nt erpreted to m ean that two consecutive weighings do not diffe r by m ore than O l per cent when the second weighing is made atter an addi tional hour of drying + Sam p les not exceeding o a Gm were taken for the moi sture determinations which were ma de at approximately 100 c Th is was done in order to as s ure oo mp lete dryin g by exposing a thin layer of drug to the surrounding air i n the oven A Freas Thermo Ele ctric Oven was the drying chamber employed 3 pH De termination The Be ckman pH M eter, laboratory mode l G wa s emp lo yed in all the determinations of pH For m easurements above pH 9 the B eckman high-pH N o ll90 E g lass electrode was used Thia electrode is i n te n ded for .meas ure m e n ts made on alkaline sol ut ions of h i gh pH at room temperature In highly alkaline sol u tions this electrode does not attain equilibrium i n stantly but may i n dioate a g radually increasing pH tor five minutea or rn ore at a pp roxi ma tely zs C The time r quired to r attainment ot equilibrium was shortened by keeping the tip of the g lass electrode i mm ersed in p H 10 buff er. while not in use, Th e asymmetry potentia l also tends to chan g e sli gh tly during me asurement, actin g to re du oe the pH readin g The usual result is to g ive a p H curve against time that reaches a maximum in about five m inutes followed by a slow decrease in readi ng Standardization of the electrodes and m eter against a pH 10 buffer was ma de after every two readin g s, frequently atter each reading Wh en the g lass electrode attained equilibrium. slowly pH readin g s were

PAGE 62

67 made at two or three minut intervals until no further upward drift was noted Th buffer standardization was then made If the latte r reading differed from the p H value of the butter the difference be twee n the two was either su b tracted f ro m or added to the last reading or the sampl as required to m ake the proper correction In th event the equilibriwn was rapidly attai n ed. the maximum p H readin g was taken as correct for the sample The Beckman 015 type g lass electrode w as used for m easurements be low pH 9 4 Determination or Volume of P recipitate Inasmuch as the amount of p recipitate form din he ated solutions of p henobarbital and p h enobarbital sodium be c ame greater with inore se in te mp erature a m et h od of approximating the volume of precipitate was devised W ith shakin g the contents of the bottle which had be e n subjected to heating wer tra n sfer r ed to a beaker Wh il the mixture was rapidly stirred by m eans ot an electric stirre r, a convenient volume u ually 10 oc ., wa withdrawn by m eans of pipette The contents ot the pi p ette we r e placed in a g raduated centrifu g e tube If the m ixtur e could n ot be drawn into a pi p ette the co n tents of the bottle were in t r o d ucad with shakin g into a ce n t ri fuge tube In either case the tube was ce n trifu g ed at approximately 12 00 r. p m for five minutes as timed by, stopwatch Time wa taken the m om e nt the current was turned on and the mo m ent it was shut ott It t o ok approximately twenty seconds tor the centrifuge to cease revolvin g It i n ot p urported that this deter m ination should be taken ae quantitative H owever the m ethod is considered more satisfactory than

PAGE 63

58 the use of such words as "large," "moder a te or "small" in describing the quantity of precipitate 5 Use of the Autoclave To attain the hi g her tomperatures to which some of the solutions were subjected in this investi g ation the bottles co n tainin g the solu tions were exposed to saturated steam under p ressure The apparatus employed was a "Peerless" pressure cooker The followin g procedure was resorted to in order to g au g e the time during which the solution was exposed to a certain temperatures the autoclave was heated until steam was issuin g strongly fro m the p etcock The lid was re m oved and the bottles were i n troduced into the chamber T h e lid was immediately replaced and clam p ed down The autoclave was rapidly neated by several !.ie ker bur n ers in order to quickly saturate the spac i n the chamber with steam As soon as all the air was replaced with steam, the p etcock a closed causin g th e p ressure to rise rapidly Wh en the desired p res sure was attained time was taken by m eans of' stopwatch The M eker burners were the n s hut off and the pressure was maintained at a con sta.nt level by m ean s of a low bunsen !'lame T h e varianc ot tem p erature was :! 1 c When the period ot tim e d u ring w h ich the sol u tion was up posed to have b een exposed to h eating had elapsed the pe tcock was opened co m pletely to p ermit rapid esca p e of steam The autoclave was quickly o p e n ed and the bottle were i mm ediately i n serted in a mixture of ice a n d water to rapidly cool th e solutions

PAGE 64

59 C. A CO?. PARISON OF ME T IIODS OF D l: Y IKG Pli:.:TJ : OBARBITAL SOD it m PRPARATO RY TO A S SAY The purpose of comparin g various me thods of drying phenobarbital sodium to constant weight was to detect the pre se nce of decomposition 4 which might oecur during drying Phenobarbital sodium absorbs eon siderable moisture from its sur roundin gs due to its hygroscopic natur e Inasmuch as solutions of phenobarbital sodium undergo hydrolytic deoom position there is a possibility that the compound may suffer hydro l ysis when heated in the solid state ., particularly if it contains an e.p preciab l e quantity of moisture A method of drying t o constant weigh t without producing decomposition was necessary in determinin g the dry eight of sam p les wh ich we re u sed i n the study of various m ethods of assay of p henobarbital sodium and in the study of de co mposition of so lutions of t he compound In the latter case it was essential to pre pare sol ut ions of definite concentra t ion from samples whose dry weigh t was known N ot only could the concentration of solutions be de termined from thG amount of dry sample p resent in a unit volume but the con sistenoy of the method ot assay and the per ce nt of p henobarbital so dium present i n the dry sample could be calculated The drug e mp loyed was pu rchased fro m the M allinckrodt C h emical Work s It was labeled Phenobarbi tal Sodium u. s P XI and it b ore the oontrol number lrui l Samples were a eated to constant ,veig ht in en oven at 100 C These samples ere then heat ed to constant weigh t at 140 c to determine hether or not an additional loss in weight would result Additional samples were hea ted at 140 c Other sam ples were spread out in pe tri dishes and kept in rm evacuated de eicato r

PAGE 65

60 for one m onth Samples w h ich had bee n he a ted w ar assay e d by the Budde m ethod to detect the p resence of deco mp osition The u s P X I limits the am ount of m oisture in p h e n obarbital so dium to seve n p er cent as determined by heatin g the salt at 140 c for six hours It ~s stated by van Leant {27) that heatin g p henobarbital sodium at 120 c produced a i_) roduot which gra d ually became i n oom p l e t ly soluble N o dif f iculty w as e xp eri e nced durin g this i n vesti ga tion i n d issolvin g samples whioh had bee n dried to constant wei g ht at 142 { :!: 1) 0 c durin g a period ot ten hours

PAGE 66

6 1 TABLE 14 DETF..RMINAT I ON OF tO I STURE IN PRF.NOBARBITAL SOD I UM. U S P X I PART A L OSS IN WEIG HT on HEATING P HENOBARB I TAL SOD I UM u s. P X I T O C ON ST AN T WE I G HT AT 101 ( t 2) C W eight 1n Gm Duration or H eating Loss of W eight Percenta g e Lou 1n B our1 in Gm or W eiht (l) 2 6459 12 0 045 9 1 7 (2) 3 3706 1 2 0 0580 1 7 (3) l O b 90 e 0 0191 1 a (4) l 1427 6 0 0199 1 7 (6) 1 077 1 6 0 0 1 88 1 8 (6) o 8437 6 0 0147 1 7 Av e ra g e 1 7 PA.~T B. AD DI T I ONAL LOSS IN WE I GH T ON HEAT IN G THE Aa'.)VE S AMPLES TO C ON ST AN T WE I G HT AT 141 ( ~1 ) 0 c We i g ht in Gm Loss or W ei g ht P ercenta g e LoSI Total Lo88 of 1n Gm or W eight W eight% (1) 2 6000 0 044~ l 7 3 4.: (2) 3 3126 0 .0 656 1 7 3 4 (3) 1 0699 0 0 1 87 1 7 3 .5 (4) 1 1288 0 0 1 92 1 7 3 4 (6) 1 0683 0 0182 1 7 3 ~ 4 (6) 0 8290 0 0140 1 7 3 4 Average 1 7 3 4 Six ho u rs of heatin g were required to attain constant wei g ht

PAGE 67

62 TABLE 16 LOSS IN WE I GH T Oli HE ATING PHfillO BARBITAL SO D IUM U S P XI TO C ONSTAN T WE IGHT AT 141 ( ~ 1) C W eight in Gm L oss or W eight in Gm Pe rce n tage Lose of' W eight (1) 1 1223 0 0380 3 4 (2) 1 0635 o osss 3 4 {3) 1. 0041 o 034 1 3 4 (4) 1 0206 0 0351 3 4 (6) 1 1001 0 0377 3 4 Averag e 3 4 Seven hours of h e ati ng we re r equired to attain constant weigh t TABLE 1 6 LOSS IN WE I GH T ON VA C lJ !JM DESICCATION OF PilENOBARB ITAL OD I OM U S. P X I il ei g ht in Gm Loas of W eight in Om P ercenta g e Lou of W eight (1) 11 5408 0 1100 1 6 (2) ll 6 71 5 0 1 676 1 4 (3) 13 0286 0 1901 1 5 A verage 1 6 ,ttThe samples were kept t or 31 days in a vacuum desi coator evacuated to app ro ximate ly 0 6 mm Hg

PAGE 68

In the three p recedi ng tabl it is see n t ha t th e lot of Ph eno barbital Sodium u s P X I employed i n this study decreas ed 1 7 pe r ce nt i n wei gh t on dryin g to constant weight at 101 ( 2 C A :t'u rth r 1 7 pe r cent loss waa undergone up on heating the sam. s amp les to co n stant weight at 141 ( ~ 1) 0 c The total loss i n wei g ht was 3 4 pe r c nt whi ch figure was likewise obta i n ed on d ryi n g samples exclusively a t 141 ( 1) 0 c On dryin g sam p les i n an evacuated dessio at or f or one mon th the loss in wei gh t was 1 6 pe r oe nt or a va lue approachin g that obtai n ed o n drying ph enobarbital so d ium to constant wei g ht at 101 ( T 2) 0 c These results i n dicated either incomplete dryin g at approximately 100 c or the p rese n ce of decomposition at app r oxima tely 1 4 0 c To determine whether or no t decomposition occurred durin g the dr y i ng of s amp les at the aforeto m entioned te mp eratures, dried and un dried samples we re assayed by the Budde me thod

PAGE 69

G4 TABLE 17 B U D D E A S SAYS OF DR I ED A N D UND R IE D P ftF.NOBAR BITAL S O DI 'CT !lt U S P X I PAR T A Sample S am ple After Bef ore D ryin g at D rying 103 ( Z l)o C Gm 0 4588 0 4 432 0 6120 0 6913 o. s sos 0 3387 toss in W e ight F er Cent Av 3 4 Sample Sample Arte r L o88 in Before D ryin g at W ei g ht D r ying 142 ( ~ 1 ) 0 c Per Cent Gm 0 4558 0 4337 4 9 0 4463 0 4246 4 9 0 6646 o 63 70 4 9 Av 4 9 Und ried Sample O lN A gN 0 3 Gm cc 0 2821 1 0 40 0 2 63 4 9 74 0 3211 11 79 O l N A gN0 3 cc 17 06 22 66 12 93 P .AR T B O l N A g N 0:5 cc 16 78 16 50 20 83 PAR T C C 12 H 11 0 3 11 2 Ne. G m 0 2643 0 2474 0 2996 c 12 H 11 0 3 N 2 M a Gm 0 4332 0 5768 0 32 8 6 Per Ce n t of Un dried Swn. ple 94 4 93 7 Av 94 2 P e r Cent c 12 H 11 o 3 N 2 N ot Gm Und ried Samp le 0 4287 9 4 l 0 4193 9 3 9 0 5293 93 8 Av 9 3 9 Per Cent of Und ried Sample 9 3 7 93 9 93 3 A v 93 6

PAGE 70

66 On the basis of undried sample 94 p er c nt phen obarbital sodium was present in aamplea of Phenobarbi tal Sodium U s P X I before and after dryin g at 103 (~1) 0 c and at 142 ( t1 ) 0 c Therefore. the pre ence of deco m position in samples h eated at approximatel7 100 c. and approximately 140 C was not sho,m by the Budde Aaaay Throughout this investi g ation undried samples were used except where otherwise indicated T h e equivalent wei g ht ot dry sample upon which the pe rcenta ge were usually based in this study was calculated fro m loss of weight sustained on heating other sam.ples to co nstant weight at te mpe rature slightly above 100 c unless otherwise indi oated This loes in weight at approximately 100 c was employed in prefere n ce to that austa1ne4 at a~proximately 1 40 c. beoa~se ot pos Bible deco mp osition which may occur at the higher temperature D EVALUAT IO N OF R E PORTED -l ET H:O DS OF ASSA Y Several methods ot determining the percentage deterioration ot barbiturate solutions have been r eported Nielsen (13) and Mad1 en (3) followed the p rocess quantitatively by mea suring the carbon dioxide II and dialkylaoetylurea formed Schlemmer and To rber (30) round that the results obtained by such a m ethod were too lo w be cau1e ot th e ex istenoe ot p artial decomposi~ion ot dialkylaoetylurea Tomeki and W aller (24) determined. the amount of undecompoeed barbiturate by ti tration with a standard solution of silver n itrate in the p resence ot sodium carbo nate R otondaro (72), G lycart (70) and Aspelund and Skog lund (7) separated various decom p ositio n products f ro m one another at various pH levels In addition other m ethod were re ported but their

PAGE 71

86 use was men tioned in co nne ction with assays of undecom p osed be.rbitu rate The m ethods emp loy ing an immiscible solvent at a si ng le pH level were co n sidered un a atisfactory f or the i nv esti g ation Them t h od a 1n whioh the alkalinity or acidity of the ba.rbiturate derivative was measur d by simple titr at in g with a standa rd acid or alkali were not applicable because ot the p resence o f alkaline decom p ositio n p roducts i n de teriorated solutions. Colorim e tric m et hod were n ot investigate d because ot their reported i na ccuracies The meth od of Rot o n daro was choaen as a me thod worthy o f i n veati ga tion b e cause a quantitative separation of t he unde oo mpo aed ba.rbitu rate was claimed The o s P X I metho d (70) was studied tor the sake of co mp arison The Budde (93) Viebook and Fuchs ( 96 96). and Sohulek and R ozsa (10 0 101) met h ods exe mp li f ied the sevral m ethods employing silver nitrate The po1sibility of a pp lyin g the von Babitsch titration ( 86 ) to the det ination of the quantity of the tree ba r biturio acid in the r sidues reoov e red in some imm. isciblesolvent m ethods was in vesti g at d 1 Co mp arison of U s P XI and Ro tondaro Assay s ot Phenobarbit al Sodium U S P X I Th e p urpose of the experiment was to co mp are the U s P XI assay (7 0) with t he Rotondaro &aaay (72) and a mo dified torm ot the R oto nd aro aseay The U s P XI (70) assay ot phenoba r bital sodium 11 g iTen a1 fo llows, nDiseolve about 0 6 Gm ot Soluble Ph enobarbital dried at

PAGE 72

67 140 c for six h ours, and a ccur at ly weigh ed 1n 50 oc ot distil led water in a se p arator Add 1 0 00 of di luted hydr ochloric ao id and co mple tely ext ract the liberated phenoba rbital wi t h s u ooessiv p ortions of 26 oc each of e ther Evap orate the combined ether al ext raots in a tared dish, and dry the reaidue to constant weight at 100 C Th weight of ph enobarbital so obtained is n ot less than 90 4 an d n ot mo re than 9 1 4 p er cent o f the wei g ht ot the Soluble Ph nobarbital taken f o r the assay The asaay was mod ified to the extent of u ain g undried aam plea ot known m oisture co n tent This was done to avoid the possibility of incurring d eco mp ositio n durin g the dryin g of th e aam p l Th e Rot o nd aro me thod (72) was d isous1ed in the review ot the lit erature Th e m t ho d was m odified i n some of th e assay tor the p urpo of simplificatio n and l imina tio n of details whi ch did not h ave to be ret in ed in f ollowi ng the decom p osition of barbitur a te solutions In stead of ac i difying at th e start the solution of the barbitu rate was treated with a proportional amount o f n ormal sodium hyd roxide equiva lent to th amount ot alkali wh ioh R oto nda ro (72) used to e xtract the aoidio compounds fro m th chloroform extracts containin g both the bar biturate and the decomposition p roducts The volume ot solutio n pr e pared or taken f or the assay w as such that the dilution with the norma l sol ut io n of sodium hydroxide resulted i n a concentration of alkali ap proximating a o 5 N sodium hydroxide At the same time the ratio be tween the total amount of alkali and ba.rbi turat was m aintained. aa found in the ori g inal asaay The alkaline solution wae extract e d with chloroform to remove the neutral deco position product phenylethyl aoetylurea The n et r esult of the m odificati o n wa1 to extract the

PAGE 73

68 neu tral dec o mpositi o n product from an alkaline solution ins t e dot leaving this pr od uc t behind in th chl oro form on ext r acting tho lat te r, which co nt a ined bo th the barbitura te and the decomposition products by means ot an alkaline solution The remainder or the ass ay was unalte r ed R esidue I in dica tes phenylethylao etylurea residue I I indicate p hen o barbital and residue III indicate s the acidic decompos it ion products such as phenyle t hylma lo nur io eid and pheny l ethylaoetic acid. The shakin g de vic e designed by M ittel tae d t Ho rn and Kautman (107) wa.a emp loy ed to pe rfo rm the shaking-out w ith immiscib l e solvent as requi re4 in the experiments s ubsequent ly mention ed E ac h pe riod of shaki ng -out wa s not less than ten m i nu tes in duration ill the resid u es obtained by the Rot o nda ro p rocedure (72) were dried to constant weight at te mp er a tures slightly above 100 c ., sue oesaive weighings being m ade at fifteen m inute interval, Sample s or Phenoba r bi tal Sodium O S P Xl possessing the l linokrod t control designa tion KLM l were e mp loyed in this study ot evaluation of assa y s The u s P X I (70) R otondaro (72) and m odi fied R otondaro assays were p erformed on sam p les of known moi sture oon tent In a dditi on a number ot m odified Rotondaro assays of samples which had been dried at 141 ( t 1) 0 e were conducted The re sults are summarized in the tables which follow

PAGE 74

69 TA BLE 1 8 U s P X I A S SAY OF P HEN OBARBI TA L SODIUM U S P X I C al cu l a ted Phen o barbi t a l E rt r a oted C a lcu l a ted D ay S amp le c 1 2 B 11 o 3 N ~a G m Gm % Gm % {l) o 11s1 0 64 6 90 4 0 101 9 9 0 (2) 0 6 1 4 9 0 4 68 90 8 0 6 12 9 9 4 (3) 0 6673 0 6 96 9 0 7 0 66 3 99 3 Un d r i e d e am p l e s we r e em p lo yed T h e p e ro en ta g es we r e oalou l at ed o n a m oi st ure f r e e basis S e p ara t e s amp les wh ioh we re dried at 102 (:t..1) 0 c had und er g o ne a loss 1n w e igh t of 2 2 %

PAGE 75

7 0 TA B L E 1 9 R OTO N D ARO ASSAY OF Pd EN OBARB ITAL S O D I UM U S P X I PA R T A Calculated R esidue I R esidue nu R esidue III D r y Samp l e G m G m % Gm % Gm % (1) o 3411 0 004 1. 2 0 263 77 0 0 033 9 S (2) 0 3179 0 004 1 3 0 253 7 9 7 0 023 7 3 (3} 0 31 6 0 0 005 1. 6 0 250 79 2 0 023 7 3 PART B C aloulate d Sum of the Sum of I I II w ith I I C 1 2 H 1103 N z N a Three R esidue s Calculated aa c 12 H 11 o 5 N z N a Gm % Gm % G ::n % (l) 0 2 8 7 84 3 0 299 8 7 .7 0 324 95 0 (2) 0 277 8 7 2 0 2 8 0 86 .2 0 304 9 6 7 (3) 0 214 8 6 7 o 21 e 8 8 .l 0 302 95 6 Un dried s amp les w e r e e mp loyed T h e p erce n ta g es were calc u lated on a m oisture t ree basis Separate s amp les wh ich w ere d ri e d at 102 ( ) 0 c ., h ad un d e r g o n e a loss in wei g ht of 2 2 % .-:I"h e quantity of p he n obarbital constit u tin g R es i due I I was de t e r m i n e d b y the von .Bab itsch tit r ation ( 86 ) wh ich is discussed in the section en titled "Evaluation of R epo r ted M et h ods of Assay

PAGE 76

71 TABLE 20 MO DIFIED RO TO N DARO ASSA Y OF PHENOBARBI TAL SOD IUM U.S. P X I PAR T A Calculated R eaidue l R eaidue II Res idue III Dry S am ple Gm Gm % Gm % Gm % (l) 0 4961 0 006 1 0 o 390 78 7 o 046 9 2 (2) 0 4907 0 0 06 1 1 0 361 73 5 0 011 3 6 (3) 0 6004 0 010 1 9 0 368 73 5 0 026 5 2 PART B Calculated Sum ot the Sum of I I II with II C12H 11 o 3 N 2 Na Thr ee R eeidue1 Calculated as c 12 H110:sH2Na Gm % Gm % Gm % (1) o t21 86 2 0 4t0 88 9 0 ,11 94 2 (2) 0 396 so 0 383 78 l 0 418 85 l (3) 0 403 80 6 0 404 8 0 7 0 439 8 7 6 Und ried sam p les were e m ployed The pe roe n tages were calculated on a m oisture rree basis Separate samplee whi ch w ere dr ied at 100 (:! 1 ) 0 c ., had under g one a loss 1n wei g ht of 2.2 %

PAGE 77

72 TABLE 21 .:V U DIPIED R\)TOJllDARO ASSAY OF PiiSNOBA.-q : .iJTAL SOD IU J\:1 U S P .X I WH ICH HAD BF.EN DRI E D TO CO N S TANT WE IGHT AT 141 { t. l C PAR T A Dry Sample Residue I Reeidue II R eeidue III Gm Gm. % Gm % Om % (1) 0 9700 0 000 o o 0 ~ 824. 85 0 0 063 6 6 (2) 0 9856 0 003 0 3 0 832 84 6 o.oeo 6 1 (3) 1 0843 o eo1 83 7 (4) 1 0269 o a11 85 4 PART B Calculated Sum of the Sum of I I~ I with II C1 2 H 11 0 3 N 2 Na three r esidues calculated as c 12 H 11 o 3 N2 N a Gm % Gm % Gm % (1) 0 903 93.0 0 888 91 5 o 966 99 6 (2) 0 911 92 6 0 895 90 9 0 97, 99 8 (3} 0 993 91 6 (4) 0 961 93 5 The average percentage loss 1n weight on heating these aamplea was 3 4 %

PAGE 78

73 The phenobarbital sodium e m ployed in these assays conformed to the require m ents of the purity rubric of t he U S P XI i n as m uch aa 91 % phenobarbita l was extracted durin g the p erformance of the U S P X I assay By m eans of the Rotondaro m ethod (72) 7 8 6% phenobarbital waa extracted By m eans of the modified Rotondaro method 75 2% pheno barbital was extracted This i n dicated the existence of several posai bilitiesa (a) the original sample waa p artially deoomposedJ (6) de compos i tion took place durin g the modified and unmodified Roto n daro assaysJ and (o) incomplete extraction and/or failure of the assay to separate the deco m positio n p roducts quantitatively without admixture The p reae n oe of the t hird possibility was furth r i n dicated by the sum of t h e three residues obtained i n the m o d ified Roto n daro assay of aam ples dried at 141 ( -.tl. ) 0 c Thia sum is co m parable to the residues ob tained in the u s. P X I assays i n as m uch as t h e latter does no ts pa rate the decomposition products In the latter set of R otondaro asaaya the sum of the three residues came to 91 2% This fi g ure, re p resenting the beat results obtained is w ithin the u S. P X I tole r ance ot 90 4 91 4 % J in the assays g iven in Tables 19 and 20 the sum of the three residues was lower showing inefficient extraction 2 Comparison of M odifie d Rotond.aro Assay of P henobarbital and Phenobarbital Sodium In an attempt to learn why the r esidues of p henobarbital had been lower in the Rotondar o assays both m odified and unmodified than in the U s P. X I assays p erform do n phenobarbital sodium d etermi n ations of Phenobarbital U S P X I and p henobarbital recovered fro m previous

PAGE 79

74 Rotondaro assays, were cond u cted. T h e Phe nobarbital U. s P X I wh ic h w a s u sed in this experiment had a m elting point ot 176 c The m elting p oints ot the individual res id ues w hich m ade up the m ix t ure e mp loyed in one set of a says va ried fro m 174 5 c to 1e1 0 c The u S. P X II (71) g ives the m elt in g p oint of p he n obarbital as 174 178 c In additi o n the m odi f ied Rotondaro a say ot P he n obarbital Sodium u s. P XI was re p eated o n the same lot of drug uae4 in the lo regoing experime n t namely, Ma llinckrodt ICLM l T h e number of trea.tm nta w ith chloroform to extract each residue was increased from th e n\Ullber ori g i n ally employed by Rotonda.ro (72) namely eix or seven, to twelve 1n an attempt to i n crease the ef f iciency of extraction E aoh p eriod ot shakin g -out with the m echanical shaker was n ot l ess than te n m i nu te in d uration

PAGE 80

75 TABLE 22 MOD I FIED ROTONDARO A SS AY OF PHENOBARBITAL SOD I UM U S P X I Sample N umber Calculated Dry Sample G m R esidue I Gm per cent R e idu II Gm per oent Residue III Gm per cent Residue II o aloulated ae c 12 H 11 o 3 N 2 N a Gm per cent Sum of the t h ree residuu with II oaloulated ae C 12 H 11 0 3 N 2 Na Gm per oent 1 1 1545 0 012 1 0 o a10 70 2 0 163 14 1 0 888 76 9 l 062 92 0 1 125'1 0 037 3 3 o aoo 71 1 0 141 12 5 0 876 77 9 1 054 93 6 1 3362 0 028 2 1 0 949 71 .1 0 183 13 ? 1 040 77 9 1 250 93 6 Undried samples were e m ployed The p ercentages were calculated on a moisture tree baaie Sep rate samples. which were dried to con atant weight at 106 (~1) 0 c had undergone a l o aa in weight ot 4 2 %

PAGE 81

76 TABLE 23 MOD IFIED ROTONDARO ASSAY OF PHFlWBARBITAL U S P X I Sam ple N umbe r l 2 Calculat e d D ry Sample Gm 1 1600 1.2 626 1 3835 R e idue I Gm. 0 001 0 013 0.013 pe r cent o.6 1 0 o 9 Re sidue II Gm 0 876 0 978 1 078 p er ce n t 7 5 6 77 4 77 9 Residu e III Gm 0 180 0 166 0 162 per c e n t 1 5 6 13.l ll.7 Sum of the th ree resid u e& Gm 1 063 l 1 56 1.2 53 per cent 91 7 9 1 5 90 6 Undried s amp les were e m ployed The p eroentagea were calculated o n a mo i s tur e t ree basis Separate e amp les. wh ich were dried to oon stant weight at 106 0 c had under g one a loas i n weight o~ 0 2 %

PAGE 82

77 TABL E 24 t O DIFI F D ROTO !:DAR ASSAY OF P il E!J OBA R BITAL Samp le N umber 1 2 3 S a m ple W ei g ht Gm o 7348 0 1026 0 6 568 R esidue I Gm 0 205 0 018 0 0 21 per oe n t 3 5 2 6 3 2 R esidue II G m o 531 0 538 0 493 p er oent 72 2 76 6 7 5 0 R esidue I II Gm 0 121 0 101 0 076 p er cent 1 6 4 1 5 2 11 6 Sum of t h e t h ree residues Gm 0 61 6 0 6 62 0 6 90 per ce n t 9 2 l 9 4 3 89 8 *Th e ph en o barbital emp loyed in t h ese assays re p rese n ted a m ixt u re of d ry ph eno barbital residues obtai n ed fro m p revious R otond a ro assays The m eltin g p oi n ts of the i n dividual residue varied fro m 174 6 C to 1 a 1 0 c

PAGE 83

78 .By mean s of the modified R oto n da.ro as say 70 8 % phenobarbital was extracted from sam ple s of Ph e noba rbital S odium u S P XI Th e sum of the thre r sidues avera g ed 86 4% The se results further sub stantiated the conclusions reached in the last experiment t do dified R otondaro assays of P henobarbital U S P XI yielded 11 wi recovery of phenobarbita l. M odified Rotonda ro a say of ph eno barbital whi ch had been p reviously obtained in othe r Rotonda ro assays yielded 76 8 % recovery of phenoba rbital Alth ou gh then ber of treatments with chloroform to xtraot each residue was increased the sum of the three residues obtained in the modified Rot o nda ro ssays of P henobar b ital Sodium V s P X I was be low the U S P X I tolerance of phenobarbital extracted M oreove r, the avera g e of the sum of the thre re idues obtained i n the modif ied Rotonda ro assays of Phenobarb ital u S F X I and phenobarbital which had been previous ly obtained in the other Rotonda ro 1t ssays was only 91 7% When 76 8 % phenobarbital remains undecomposed the loss of oar bon dioxide from phenobarbital amounts to 2 9% of the phenobarbi tal taken Th e differen ce between the sum of the three residues gi ven in Tab le 24 and 100" ~ recovery is only partially accounted for by the loss of car b on dioxide F rom the results obtained i n this experiment the e xistence of several poseibilities in both the modified and unmodified Rot. ondar o assays is indica ted a (a) occurrence of decomposition during the as say, (b) i n co mp l e te extraction and (c) failure of the assay to se p a r ate the decomposition products quantitatively without a dmix ture R o tondaro (72) had claimed that 98% of the phenobarbital actually p rese nt

PAGE 84

79 in a s am ple could be recov e red by the use of h is assay From the results ob ta in e d it was concluded that t h e Ro to n daro method mo dified or unmodified was not suitable for the quantitative d et e r~i na tion of dete rioratio n occurrin g in solutions of phenobarbita l and phenobarbital sodium 3 A S tudy of A ssays Emp lo!rin r; Silver N itrate The m ethods of Bud de (93) Vie hock and Fuch s (95 96) and Schulek and Rozsa (100 101) exempli f y the reported me thods of assay employing silver nitrat e for the determination of barb~turates A comparison of results obtained by applyi ng these assays on phen obarbital sodium oon stituted the p urpose of this experiment In addition an insight into the c hemis try of the Budde assay (93) was sou g ht The Budde assay (93) was conducted in the followin g way: Abou t 0 2 to 0 3 Om of phen obarbital sodium w as accurately wei g hed and dis solved in 30 cc of water to g ether with 1 Gm of anhydrous sodium car bonat Andydrous sodium carbonate of analytical rea gent quality was employed throughout the investi gat ion whenever this compound was called for The solution was p repared in a 50 cc beake r which was supporte d by a universal clamp attached to a ring stand at a convenient hei g ht An electric stirrer was used to stir the solution vigorous l y throughout the titration By means of a microburet 0 1 N silver nit r a te was in troduced drop by drop The droppin g was re gula ted so that the turbidity produced by a drop of silver ni tr ate had disappeared before the intro duction of the next drop It was found that a time interval of one t o

PAGE 85

80 two s e conds was sufficient to permit olearanee of the solution A stop watch was em1ployod to assist in 1Ml.i n taining a constant rate of intro duction of the silver nitr ~ te The r ate was maintained a t one drop every one or two seconds The apparatus was set up 1n a darkene d room A strong beam of light e~itted by a mioroproJeotor was directed through the solution The solution was view ed at a right angle to the ray of light In this manner a faint t1irbidity was easily detected which would not be observable in diffused light The appea ra n ce of a distinct tur bidity which persisted f or at least one minute constituted the end point Budde ( 93) e xp lained that a silver oompot.md of the barbitur a te was formed which was soluble in this media Budde (93) stated that the silver bar biturate was very diffioultly,dissociated This elucidated why the si l ver nitrate reacted with the barbiturate instead of reactin g with the so dium carbonate Silver carbonate and silver oxide are we ll dissociated The silver ions are removed from the field of reaction by the formation of a poorly d1Sf3ocia.ted organic compound namely t he silver barb:i .turate derivative In order to learn more concerning the ssay s1Sveral tests wer oarried out To show that silve r phenobarbital is only slightly ionized in an alkaline media a solution of Phenoba rbital Sodium u s. P XI was treated under conditions existing in the Budde assay To 20 oo of a 1 3% solution of Phenobarbital Sodium u S F XI 5 cc of water 0 83 Gm of anhydrous sodium carbonate and 6 ec or 0 1 N silver ni trate were dded The resultin g solution was clear as observ d with the aid or a strong beam of li g ht Up on the addition of l co or 1 % sodium c hlo ride no turb idity was p roduced as observed \mder the same conditions

PAGE 86

81 W hen l oc ot 1 % sodium chloride was added to a mixture of 25 cc. ot water and 6 co of O l N silver nitrate a very m arked turbidity was evidenced These results showed that silver phenobarbital is weakly dissociated The solubility of silver phenobarbital wa s investi ga ted The ad dition of 0 02 cc of O l N silver nitrate to a saturated solution of Phenobarbital u s P X I resulted in the formation of a distinct tur bidity wh ich did not disappear after five minutes of stirrin g The addition of 0 83 Gm of anhydrous sodium carbonate caused a disap pe ar ance of the turbidity The addition of sufficient silver nitrate so lution again restored the turbidity It was concluded that silve r phenobarbital is insoluble 1n wate r but sol ub le in the p resence of so dium carbonate The a pp eara n ce of a turbidity on addin g silver nitrate the second time was due to reaction between the x~ess silver nitrate and the sodium carbonate The re m oval of silver ions from the field of reaction was shown by the disappearance ot turbidity on adding a satu rated solution of Ph enobarbital U s P XI to a suspension of silve r carbo n ate wh ich had been formed by adding 0 02 co of O l N silver ni trate to a solution of 0 83 Gm of anhydrous sodium carbonate in 26 cc of water T o deter m ine whether or not the pH was lowered when phenobarbital was added to a suspension of silver carbonate the following ex pe riment was oo n ducteds The co ndi tions existin g in the B udde assay (93) we r e duplicated Determination of pH was m ade at va rious sta g es The pH of a 3 3 % solution of anhydrous sodium carbonate w a~ 11 56 at 25 c Upon adding 0 02 cc of O l N silver nitrate the pH was f ound to be

PAGE 87

82 11 64 at 25 c The solution was distinctly turbid The pH after addition ot l cc of a saturated solution ot Phenobarbital U S P XI was found to be 11 56 at 26 c Therefore suspensions of silver carbonate were cleared by phenobarbital without a lowering ot the pH As mentioned in the review ot the literature the precipitation of phenylethylaoetylurea occura in deteriorated solutions of phenobarbi tal sodium This decomposition product is slightly soluble in wat r a1 was shown by evapor a ting a filtrate obtained from a mixtur of phenyl ethylaoetylurea and wa.ter which had been shaken for two hours The reaction of silver nitrate with this com p ound as it occurs in de teriorated solutions of phenobarbital sodium 11 a poasibility 1n the Budde assay ( ~ 3) To determine whether or not the p resence ot dis solved phenylethylacetylurea in suoh solutions would introduce an error 1n the Budde assay a suspension ot this compound was treated with so dium carbonate in the oonoentration employed in the assay To 36 cc of a saturated solution of phenylethylaoetylurea containing an exoe a of the compound 1 2 Gm of anhydrous sodium carbonate was added The mixture was shaken for two hours and then filtered To 25 cc ot the filtr a te 0 l N silver nitrate was added A distinct turbidity was pro duced by 0 04 cc This amount of silver nitrate would otherwise indi cate the presence ot 1 mg of phenobarbital sodium The error intro duced thereby would be insignificant in the experiments on d teriora tion of solutions of phenobarbital sodium subsequently discussed The possibility or an introduction of e r ror due to the presence or other decomposition products remained to be settled It wa.s de oided to heat a suspension of phenylethylaoetylurea under the most

PAGE 88

83 drastio conditions to which solutions in the subsequent experiment on deterioration were subjected namely heating at 127 C tor two hours at pH 9 9 In this way the maximum breakdown of the compound into phenylethylaeetic acid and urea would occur To 36 co. ot water 0 25 Gm or ph enylethylaoetylurea was ad d ed It was f ound to have a pa of 7 7 Sufficient 0 1 N sodium hydroxide was added to raise the pH to 9 9 The mi xture was hea ted in a g lass-stopper d pyrex bottle under the conditions just ment ioned The pH fell to 7 6 In the con tents or the bottle 1 2 Gm of anhydrous sod ium carbonate was dis solved The mixture was f iltered snd 25 co ot the f iltrate was ti trated with O l N silver nitrate The entire experiment was r13 pe ated to p rovide a check The average amount ot silver nitrate required to produce a distinct turbidity was 0 60 cc This am ount would otherwise i n dicate the p resen c e of 0 05 am ot phe nobarbit l sodium in 100 co It was co n cluded that the Budde assay was sufficiently accurate to fo low the decom p osition or solutions ot p henobarbital s odium Phen obarbital Sodium U s P X I which h d been purcha ed. from the Mal linckrodt Chemical ~ ork with the co n trol desi gn ation KI.M l was assayed by the m ethods of Budde (93) Viebock and Fu chs ( 95 96) and Sohulek and R oz ea (100 1 101) The details ot each meth od were g iven 1n the review ot the literature The results which we re obtained are out lined i n the table whi ch f ollows

PAGE 89

84 TABLE 26 CO MP A R ISON OF S EVERAL :iAE TH O DS OF A S SAY OF PllENOBARB ITAL SODIUM U S P X I EMP LOYING S IL VER N ITRATE Calculated Phenoba rbital Phenobarbi tal Sodium D r y Samp le Gm Gm % Gm % BUDDE A S SAY (1) o 3073 0 275 89 6 0 301 98 l (2) o.2596 0 233 89 8 0 26 6 98 3 (3) 0 2583 0 230 89 .2 0 252 9 7 7 VIEBOCK AND F U CHS A S SA Y (4) 0 266 1 0 255 8 8 4 o 2ss 96 8 (6) 0 2637 0 238 90 l 0 260 98 6 (6) 0 2738 0 243 68 8 0 266 97 2 SCRUL EK AND R OZSA ASSA Y (7) 0 1 683 0 146 86 6 0 160 94 8 ( 8 ) 0 1798 0 1 5 1 84 l 0 166 92 l (9) 0 2014: 0 177 86 4 0 194 93 6

PAGE 90

85 The B udde assays showed t h e p rese n ce of 8 9 5 % p h e n obarbital in the samples of Phenoba r bital Sodium U s P X I. assay showed the p r esence of 8 9 1 % phenoba r bital Assay showed the p rese nce of 85 4 % p henoba r bital erance (7 1) for phenobarbital is 89 0 to 91 6% n The Viebook and Fu chs The Schulek and R ozaa T h e u S P XII to l T h erefore the Budd e m ethod was considered satisfactory Alt h ou g h the m ethod of Viebock and Fuchs (96 96) gave equally accurate results it was difficult to de termine the end point because of the pink color produced upon the addi tion of silver nitrate in the second part of the assay The m ethod ot Schulek and Rozsa (100, 101) was o~nsidered unsatisfactory both from the standpoint or the low results and the difficulty in detecting the e n d point 4 V on .Babitsch Titration of Phenoba rbita l A m ethod was sought for the quantitative determination of the phenobarbital in re idue II of the R oto n daro assay (72). Th purpose was to eliminate the work involved in drying the residue to constant weight by substituting a volumetric determination of the phenobarbita l fter the c hl oro form had been e va porated oft I n ~hia experiment the sample emp loyed were res idues of phenobarbital obtained in previous Rotondaro assays T h e m elting points of the individual r esidu es we re withi n the m elting point ran g e given in the u S P XII (71). The assay consisted of dissolving 0 1 to 0 2 Gm of sample in 20 co ot alcohol, adding 23 drops of 0 1 % alcoholic s olutio n of thymolphthalein and titrating with 0 1 N sodium hydroxide to a b l ue color The results obtained are giv en in the table which fo llo ws

PAGE 91

86 TABL E 26 VO N BA.BITSC ll TITRATIO N OF P HENOBAR BITAL Samp le Weight of Amount Found Pe r cent of Numbe r m p 0 c Sample Sample taken Gm Gm l 175 0 2328 0 234 1 00 4 2175 0 2406 0 241 1 00 0 3 175 0 2095 0 20 9 99 8 4 176.6 0 2460 0 242 98 9 5 176 6 0 1621 0 162 99 8 6 1 76 6 0 1334 0 1 32 99 0 7 176.l 0 1759 0 177 100 3 8 176 l o 1e45 o 1e4 99 9 9 1 76 l 0 2124 0 212 100 0 Average 99 8

PAGE 92

87 The volumetric deter minat io n of p ur e phenobarbi t al by the me thod or von Babitsch ( 86) gave an average result of 99 8 % It was con cl ude d that the me thod was sufficiently accurate for the titration of residue II as obt ained in the R oto n daro assay (72) 5 Co mparison of Assays of a De teriorated Solution ot Phen barbital Sodium By assaying a deteriorated solutio n of phenoba rbital sod1\llll wi th the u S P X l (70). Ro tondaro (72). and Budd e ( 9 3) me thods, it was desired to obtain a comparison of the pe ro en t ag e decomposition a s indicated by thes e m ethods Un dried Phenoba r bital Sodium u s P XI whi ch had be en purchased with the W .a l l i n ckr odt co nt rol designation KL ~ l w as us ed to p re pa re the sol ut ion The pe r c ent of calculated dry sample in the s ol u tion was 4 00 % w/v Into four ~o unoe g lass-sto ppe red pyr x bottles were in ~ tr o iu oed 25 cc. porti o ns of the solution Therefore 1 Gm of calcu lated dry P h e n obarbital Sodium U. s. P XI w as in each bottle The bottles were p laced in an a u toclave and h eated at 115 C f or thi rty .minut es Time was ta k n when the pressure h ad reached 2 0 p ounds E ighteen minu tes had elapsed s i nce the i n tro d uctio n of the bottles in ~ to the cham be r The U S P X I assay (70) was carried out on eaoh me mber of a set of t h ree sample by me rely d oublin g the quantities c a lled fo r in the of f icial p rocedure i n as mu ch as the latter employs 0.5 Gm of th e sample The ph enobarbital was extracted w ith twelve 25 cc portions of ether each pe riod of shakin g being not lesa than ten minut es in

PAGE 93

88 duration The m odified Roton d aro assay w as performed o n another set of sam ples The m ethod e m ployed was p reviously described i n the sectio n on evaluation of reported m ethods of assay The samples were quantita ti v e l y t r ansfer r ed to separators with the aid of small portions of wa t er which totaled a p proximately 33 co In carrying out the Budde assay (93) 3 33 Gm of anhydrous sod i um carbonate was dissolved i n the contents of eaoh b ottle Th mixtur e w as quantitatively diluted to 100 cc and filtered throu g h a dry filte r The first 20 ce of filtr a te were discarded O f the subse q uent filtr ate 25 cc ~ere titr~ted with a standard sol u tion of silver nitrate The assay repeat~d in each case on another portio n of the filtrate The results obtained i n this experiment are summarited in the followini; three tables TABLE 21 U S P. XI ASSAY O F A 4 % W /v SOLUTION OF PHENOBARBITAL SODIUM U s P. X I AFT ER lll.:A TI NG AT 1 1 6 C. FO R T HIR TY M INUT E S Phenobarbit al Calculat e d Stren g th ot Pe r Cen t Ext ra ct ed c12H11ll a S ol uti on D eo omp o Mte r Hea tin g aiti on Gm % Gm % (l) 0 ,, 8 70 8 7 0 0 95 2 95 2 3 8 1 4 8 (2) 0 88 1 88 1 0 9 64 9 6 5 3 86 3 6 ( 3 ) 0 88 0 ss o 0 96 3 9 6 4 3 86 3 ,, 7

PAGE 94

89 TABLE 28 M ODIFIED R O TO N DARO ASSAY OF A 4% W /v SOLUTIO N OF PHEN OBARBil'AL SODIUM U. S P X I AFT E R tl ~TING AT 115 C FOR T H IRTY M il W TES Sample Numb er l 2 3 R esidue I Gm 0 062 0 024 0 042 per ce n t 5 2 2 4 4 2 R esidue II Gm 0.606 0 595 0 688 per oent 60 6 69 6 68 8 R esidue III Gm 0 216 0 .1 73 0 195 per oe n t 21 6 17.3 1 9 6 Calculated c 12 Hll 0 3 N 2 Na. Gm 0 662 0 662 0.643 per ce n t 66 2 65 2 64 3 Strength ot Solution after H eating 2 66% 2 61% 2 57 % P er cent De oompoeition 3 3 8 34 8 36 6

PAGE 95

Sample Num ber l 2 90 TABLE 29 BUDDE ASSAY OF A 4 % W/v SOLUTION OF Plil,'NOBARB I I AL SOD IUM U s. P X I AFTER H ~ ATING AT 116 C FOR THIRTY MIN UTES co ot c 12 H 11 0 3 N 2 N a Strength of Per Ce n t 0 1 N AgN03 S olution Deoompo After H eating sition Gm. % % a 20 0 209 8 3 5 3 34 16 5 8 .1 9 0 20a 8 3 4 3 33 16 6 8 20 0 209 83 6 3 34 16 5 8 20 0 209 83 5 3 34 16.6 8 18 0 208 83 3 3 33 16.7 8 .1 8 0 208 83 3 3.33 16 7 The u S P XI assays showed 3 7% decomposition in the 4% w/v so lution or Phenoba rbital Sodium U s P XI which had been heated at 116 C. for thirty minutes The m odified Rotonda ro assay showed 7 % decomposition and the Budde assay showed 16 6% deco mp osition The re fore checks were not obtained for the determi n ation of deterioration by the several methods mentioned The u s. P XI method gave results for pe rcentage deterioration which are obviously lower than the true deco mpo sition since in this met h od most of the p roducts of deterio r ation are evaluated aa unde composed phenobarbital The results tor p ercentage deterioration by the mo difi d Rotondaro me thod are considered to be too high since pre vious results indicated that the modified Rotondaro method as carried out in the present investi g ation. failed to recover all the u n c hang ed phenobarbital The Budde method which gave results for pe rce ntag a

PAGE 96

91 deterioration between those obtained by the u s P XI method and the modifi ed Rotond a.ro m ethod wa.$ s elect ed for furt h er us in deter m inin g pe rce n tage deterioration in the p resent i n vesti g ation From the st ud ies ma de of th Budd e m ethod it appears that it g ives results that re a pp rox im ately c orrect or that are s ufficien tly accurate tor determini ng the relative de g ree of deteri oratio n in comparative tests E A STUD Y OF A NUMBER OF FA CT OR S I N THE DETER IORA T IO N OF SOLUT IO N S OF PHENOBARBI T AL SOD I UM U.S P X I It was desired to i n veati g ate the etfeots produced by variation i n time tem p erature, and co n centration o n such factors in the de terioration of sol u ti011a of ph enobarbital sodium as perce n tage de terioration, pH, volume ot p recipitate, and refractive i n dex This work was p rom pt ed by the results reported by N ielsen (13), Ma dsen (3), Aspelund and Skoglund (7), and To m ski and W aller (24). Madsen (3) re ported results re g ardin g pH chan g e and p erce n ta g e deco mp osition ob tained o n h eating a 1 0% solution of diethylbarbituric acid at 60 a o 0 and 100 C N ielsen (13) showed the relation ot pH to decomposition of a 1 0% solutio n of phenobarbital sodium. at temper a tures not exceeding 39 C. Aapelund and Skog lund (7) re ported the quantitative isolation of Yarious decomposition produ cts of phenobarbital sodium in solutions which had been boiled for three hours and in solutions which had been stored for ei g ht days Tomski and W aller (24) inveati g ated the pe r ce n ta g e deterioration of solutions of phenoba rbital sodium at varioua temperatures for various pe riods of time In an atte m pt to verify and e x tend the d ata p resented by these workers experiments were designed

PAGE 97

92 to include, i n hole or in part, an investigation of taotor pr evious l y t udied and at the same time to amplify, and to fill i n the hiatuses and to sutdy a dd itional tactora i n the deterioratio n of s olutions or o n e barbiturate, nam ely, phenobarbi t a l sodium Th e tempe r ature and periods o r exposure were el e cted to e n co mp ass m oat of the conditions previou sl y emp loyed by the afor to m e n tioned w orkers an d to i n clude for the m ost p art the co n ditions of sterilization r~ omm e n ded tor solu tions i ntend ed for paren teral use as g ive n in the N F VI (73) the British Pha rmacopoeia 1932 (2 6 ) and the Uni ted States Dispensatory (108) In this mann er th e changes occurrin g und er co nd itions or ateril ~ iaatio n and conditions re p orted by o th er workers were i n vesti g ated, pro vi d in g at the same time a n i n si g ht i n to d terioration taking plaoe dur i ng p rolo ng ed sto ra g e throu gh the a g e n cy or these accelerated teat, In this aeries or experiments 6 lO and 20% solutions ot Ph eno barbital Sodium u s P XI were mployed In every case; 25 oo, por tione of the solution were h eatecl in 125 00,, g laesstop pe red, pyrex bottles Unless ot herw iae i n ioated ea.oh o f the titration& g iven in t h e tables represe n ts the res u lts ob tained fro m the tit ration of the co n tents of a se p arate bottle The g roup or dstel'mi na tions wh ich in eludes volume of p recipitate r efractive i n dex ; and p H w e re m ade on the co nt ents of a se p arate bottle In the event a p r e cipitate had been p ro duced the clear supernatant liquid which remai n ed after c en trifu g in g was employed in makin g the determination of ref r active index The in atr umon t used in mak in g thie determination waa the Abbe Ref ractom e ter t h e tempe rature of whoee prisms were thermostatica l ly controlled by the -+0 flow of water ma i ntain ed at 25 { 0 5 ) c ., through the surrou n ding

PAGE 98

93 jackets The use of the auto-claw, the determination of pH, and wlume of precipitate were pr viously d scribed F r om the investigation condu c ted on the evaluation of methods of sa.y, it was considered advisable to employ th Budde method in tol lowing the deterio r ation The ori g inal Budde method (93) s r ported as followsa About 0 2 to o a Gm, of the sodium salt of the barbiturat is weighed and dissolved in 30 cc ot ter, together with l Gm, of an hydrous sodium carbonate. To the clear solution 0 1 N sil~ r nitrate is added until a persistent, distinct turbidity is produced, One mole cular weight of silver nitrate ia equivalent to one mol oular weight of -' the ba.rbitura.te, Inasmuch as solutions of various tren g ths re employed, appro priat volumes had to be taken and ap p ropriate dilutions had to be made in order to duplicate the oonditio n s of the Budde assay T h e procedure followed in 11 of the s a.ya of this series of experiment are outlined in the table which follows

PAGE 99

9 4 TABL E 3 0 PR O C -'-' DURE FO LL OWED m T HE B U DD E ASSAY O F 6 1 0 and 2 0% S O L U T I ON S OF P HEN OBARB I I A L SOD I UM U S P X I P erc en ta g e Strength of t h e sol ut ion studied 5 10 20 V olume of each solution p laced i n eac h b ottle {oc.) 25 25 26 D il u tion ma de f or e a oh 2 5 co of s ol ut ion {cc ) 125 260 5 00 Am ount of anhy drous H &2C 0 3 a d de d b e f o r e dilu t ion (G m ) 4 .1 7 8 33 1 6 67 A lter n at ive ani o tm t of anhy drous N a 2 c o 3 to add to each 26 cc aliquot ( Gm .) 0 83 0 83 o aa 25 oo. of th e dilution was ta k e n f or th e t itra t ion Thi a vol um e re p rese n ted {x) co o f th e ori g i n al s olution 6 2 6 1 2 5 In all of the experiments ot thi1 aeries undrie d sample were used to p re p are the sol u tio n s T h e p er cent of m o i sture i n the d r ug waa de termi n ed o n aeparate aam p lea betore p reparation of t he sol u tions The str eng th of eac h sol u tion was baaed on the calcul at ed am ount of d r y sam ple d i ssolTed T h e sol u tions were assa y ed b efore and after t rea tm e n t If a p recipitate waa p rese n t t he co n te n ts of th e b ottle,i n wh ich the sol u tion waa k e p t d urin g t h tr e at m e n t we re quantit ati vely t r a n sf e rred to a vol wne trio tlask Before dilutio n t he p rescribed am oun t ot anh ydrous

PAGE 100

95 sodium carbonate was dissolved i n the m ixture. The sodium carbonate w as dissolved before dil u tio n in order to redissolve any po sible p h no barbital which may have been p recipitated This amount of anhydrous sodium carbonat wi l l dissolve the m aximum amount of phenobarbital th.at can be formed fro m the amount of phenobarbital sodium which was pres ent i n the aliquot p art of any solution studied In a previous e xp eri m ent using 1 Gn: of Ph enoba rbit al U .S. P XI 2 8 Gm anhydrous sodium car bonate and 26 cc of w ate r. it was found that a clear solution was p ro duced after stirl'ing for four minute, E ight m inutes of a g itation of the sodium carbonate with the co n tents of the bottle was considered am ple timo in which to dissolve any free phenobarbital which may have been present After dilution a porti on of the mixtu re was filtere d through a dry filte r. The firet 20 cc ot the filtrate wae discarded Of the subsequent filtrate 25 cc was taken for the titration I n the event no precipitate had form din the so l ution durin g the heating so dium carbonate was not ad d ed until the filtration had been performe d at which time a p roportional pa rt of the sodium carbon ate t.s giv n in Table 30 was added to the aliquot pa rt of the solution which was to be titrated It wa d emed advisable to determine whether or not there would be any i gnitieant diffe re n ce between the results obtained by the Budde metho d if the deteriorated sol ut ion in the bottle wae quantitatively passed through a filte r before dilution instead ot diluting the de teriorated solution containing the recipita.te to the same volume To wards this end f our 25 cc. portions ot a 1 0% solution of Phenobar bital 0 Sodium u s P XI ( M allinckrodt Control KKM ) were heated at 115 c

PAGE 101

96 for one h our dur in g w hi c h time approxim a tely 16 cc of p recipitate per 100 cc o f solution waa formed The actual volume of the p recipi tate was p robably co n siderably less be cause the latter m e nt io ned de termination ot vo l ume of pr eci p itat e did not take into consideration the occluded solution n or the free ph nobarbital wh ich may have been present and which wou ld have been dissolved by the sodi um c a rbo na te Two ot the sam p les were diluted to 250 ec befor e filtrat io n and two were filtered bef ore dilution In e ac h case the sod ium c a rbo nat e waa first ot all di ssolved in the co nten t of the bo ttle Portions of the solution were a sayed before the treatment The results are g iven in the table which fol l ows TA B L E :U E F F E CT O N THE B Ui ; D E ASSA Y O N R EMO VING THE PREC I PI T ATE FROM A D E TERIORATED SOL u "TI CN OF PH EN OBARBITAL SODIU' .:A U S P. XI IN TWO DI FFE R EN T WA YS P recipitate f ilt e red P recipitate f ilte r ed out before dilution out afte r dilution P eriod ot H e ting M i n ute 0 80 0 60 co ot O l N AgNO i for 2 6 cc aliquo 9 59 7 57 9 69 7 52 7 56 7 53 Per ce n t c 1 2 H 11 o 3 N 2 N a in solution 9 75 7 69 9 76 7 65 Ap pa re n t deterioration per cent 21 .1 21 6

PAGE 102

97 No appreciable difference in results oonoerning pe roe ntage de terioration of a 10 % solution ot Phenobarbital Sodium U s P XI which had been heated at 115 c fo r thirty minutes was found on em ploying the Budde me thod in two diffe r ent ways namely (a) re mo val of the precipitate formed i n deteriorated solution before dilution and (b) re m oval of the precipi t ate after dilution The diffe r ence be tween the results on pe rc entage deoompos1tion was o.t% Therefore it we,g co nclud ed that the determination of pereent ge deterioration could be made by transferring the deteriorated solution to g ether wi th the p recipi tate to~ volum etr ic flask wi th subsequent filtration after dilution without incurring appreciab l error l. Effe ct of He atin g a 6 % w/v Solu tio n of Phen obarb ital Sodium U S P XI at 60 C, The deterioration and chan g e 1n pH of a 5% w/v solution of Ph e n o barbital Sodium u s P X I upon exposure to a temperature ot 60 (~0,6) 0 c for perio~s of time r anging fro m 1 5 to 90 minute s were determined4 The lot of drug employed was pu rc has ed with the Mal linckrodt con trol designation ICLMl The solution was hea ted 1n a thermo tat The thermostat mention ed in this and the subsequent experiment which waa Ci) carried out at 80 C ., was re g ulated by a De Khotinsky Therm.oregulatar The temperature was observed by m ean of a m eroury thermometer cali brated in tenths of a de g ree The wate r in the thermostat was kept 1n co n stant m otio n by m eans of an e l ectric stirrer On separate o c casions f our solutions of the same stren g th we r e prepared from the same lot of dru g Before h eatin g each solution was

PAGE 103

98 assayed three times and the pH of each was determined a. t least twi oe In 12 5 -c o ., g la.sssto ppe red p:rrex b ottles we re p laced 25 -o c porti on or the so l ution Each result given i n Tab le 32 concern i ng the p er oent phenoba r bit al sod i um remaini ng in a treated solution represents the averag e of six assays con d uct ed twicon the contents of e a.ch of th ree bottles Time was taken when the bottles wer e i mm ersed in the th e rmo stat They we r e withdrawn at the end ot the time interval an d imme diately inserted in t o a. mixtur e of ice and wate r to cool the solution rapidly The determination of pH wa s mad e o n sepa r ate 25-cc po rtione 0 the s ol ution whi ch ha d been t r eated in the manne r just de scr ibed From the an alysis of the unt re ated solutions it we.a found that the p er oent of phenoba rbital sodium in the lot e mp loyed waa 97 4 % In the p revious experiment on comparison of asaaya e mp loying silver nitrate, t he Budde assay ga ve 98. 0% phenobarbital so d ium The results obtained i n this experiment are sumrnar i, ed i n the table which fol lo ws

PAGE 104

99 TABLE 32 EFFE CT OF HEA'U N G A 6 % W /v SOL T.J T I Ol OF PHEN OBARBITAL S O D I UM U S P X I AT 60 (:to.6)0 C Time So lution Per cent Per oent M inutea N umbe r C l2 li ll o 3 N ~a Deteriora 0 1n solution tion pH at c 0 l 4 9 9 53 29 0 2 4 9 9 5 1 28 0 3 4 9 9 49 28 0 4 4 9 9 48 30 Average 4 9 o o 9 51 29 15 l 4 8 9 5 1 28 15 2 4 9 9 51 28 Avera g es 4 9 o o 9 51 28 30 2 4 9 o .o 9 49 21 60 3 4 8 2.0 9 43 29 90 4 4 8 2 0 9 50 30

PAGE 105

1 00 Fr om T able 32 it ie seen that 5% solutions of Phenobe. r bit l Sod i um u s P X I underwent no deterioration when heate d at 60 C fo r 15 o r 30 minutes Percen ta&e deterioration a.mounted to 2. 0% after exposure fo r 60 e.nd 90 minutes The pH of the s olution remained th e same at the end of 15 and 30 m i nutes The determinati o n of pH at th e end of 60 and 90 minutes was in doubt because of the l a o k of re gular ity in r e sults obtained R o precipita te was evident in any or th e expos ed solutions The ?! F V I dire ct s tha t fr acti o n al sterilization a.t low t8l'i 1 pe ra ture b e ar ri d out by h ea ting the solution at 60 to 10 c ~or 30 m in u-te s on th r ee successi ve days T he refore. it was concluded that 2 0% deterioration m i gh t occur i n this particular sol u tio n when treated at so 0 c under the co n ditions r eferred to in the N F VI (73) 2 Ef fect of H eating 6 10. and 20% w/v Solutions of Phenoba rbital Sodium U S P X I at 80 c The purpose of the experiment was to determine the ef f ects pr d uc ed by heating s 1 0 and 20% solutions of P henobarbital Sodium u s P 0 XI at 80 c for 15 3 0, and 60 minu tes. As an e m erge n cy m ethod of sterili&ation the B ritish Pha rmacopoeia 1932 (26) dir e cts h eating the solution at 80 c for not less than 30 minu tes. In the p revioue experiment on the deterioration of a 5% solution at 60 C ., th sample employed was that purchased from the Mallinckrod t Chemical W orks bea ri ng the co nt rol designation KIJU In this experiment the sample use d was pur ohaeed from the aame com pany but it had the con trol designation K KM In the event that there m ight be some ditferen oe

PAGE 106

101 between the samples a 6 % solution of the former sample was p repared The results of 1ta assay and determination of deterioration after heat 0 in g at 8 0 C for 15 minutes were com p ared with correspo n din g results obtained from a 5 % solution of the latter sample Table 33 indicates the results obtained with a 6 % solution pre pared from the sample which was purchased with the Ma llinckrodt control designation Kt.M l. T A BL E 33 EFFE C T OF HEA T IN G A 6 % W /v SOL U TIO N OF PHEN OBAR B I ':AL S OD I UM U S P Time cc M inutes O .l N Ag N 03 for 5 cc aliquot 0 0 9 63 9 6 1 Averages 9 62 1 5 1 6 Av e ra. g ee 9 66 9 54 9 55 X I AT 80 ( t 0 3) C P er cent C1z H 11 o 3 N 2 Na in solution 4 89 4 86 Per cent D eteriora tion 0 6 pH at 0 c 9 38 31 9 36 31 --9 37 31 9 31 32 9 31 32 9 31 32 The P h enobarbital Sodium u S P X I was purchased with the Ma llinckrodt control designation KLMl The amount of phenobarbita l sodium i n the calculated dry 1ample used to prepare the solution waa 97 8 % N o precipitate was evident in any of the solutions

PAGE 107

102 TABLE 34 EFFECT OF HE AT G A 6 % w jv SOLUTION OF PII r.:M O.B.A.RBITAL SODIU M u s P x r AT so ( o 6) 0 c Tim e co O .l N Pe r ce n t P er cent M inute A gN0 3 for c 12 H 11 0 3 N aJJa De teriora 6 cc i n solution tio:n aliquot EH at oc 0 9 58 9 52 30 o 9 58 9 54 30 A v e ra g e s 9 68 4 8 7 9 53 30 15 9 53 9 59 29 15 9 54 9 60 29 1 5 9 54 A vera g es 9 54 4 86 0 4 9 6 0 29 30 9 48 9 48 29 30 9 51 9 48 29 30 9 49 Ave ra g es 9 49 4 8 2 1 0 9 48 29 60 9 43 9 36 29 60 9 41 9 38 29 6 0 9 42 Averages 9 42 4 79 l 6 9 37 29 The Phenoba rbital Sodium U S P XI was p urc ha sed with the M allinckrodt co n trol designation KKM The am ount of ph nobe.rbital sodium in the calculated dry sample u sed to p re pa re the solution wa.s 97 4% N o precipi tate w as evident in s:ny of the solution ex ce pt those wh ich had b ee n he ated for 60 minut es The latter po ssessed a very sli g ht amount of pr eci p itate

PAGE 108

103 TABLE 35 EFFE CT OF HE ATING A 1 0% W/v SOL U TI ON OF ?.:-H..:.N OBAR B ITAL SODIUl 1 U s P x i AT so ( :t o s) 0 c Time P er oe n t P er cent P recipitate M inut s c l2 H ll o,l 2 N a De teriora ~ 00 /10 0 cc in s olution tion pH at oc 0 9 70 0 9 63 30 0 9 68 0 9 62 30 Averages 9 69 0 9 63 30 15 9 68 0 9 60 30 16 9 69 0 9 60 30 16 9 69 0 Av e rages 9 69 o o 0 9 60 30 30 9 66 0 9 56 29 30 9 66 0 9 56 29 30 9 64 0 Ave r ag es 9 65 0 4 0 9 56 29 60 9 51 l 9 40 29 60 9 52 l 9 4 2 29 60 9 51 Avera g es 9 61 1 9 l 9 41 29 The Ph e n obarbital Sodiwn U s P X I was purcha sed wi th the Ma lli n ckrodt co n trol d e si gn a t ion K.KM Thee.mount of ph enobarbital sodium i n the oalculated d ry sample u sed to p re pa re the sol ution was 96 9%

PAGE 109

104 TABLE 36 EFFECT OF HEAT G A 2 0% W /v SOL U TIO N OF P H E M OBARBITAL SODIUM U S. P X I* AT 80 (~ 1) C T ime Per cent M inutes c 1 2 H 11 o 3 NaN a in solution 0 0 Averages 16 1 6 16 Averages 30 30 30 Averages GO 60 Avera g es 19 62 19 62 19 62 19 46 19 50 19 46 19 47 19 36 19 :n 19 3 1 19 33 1 9 03 19 02 19 03 Pe r oent Deteriora tion o s 1. 6 3 0 P recipitate cc /100 oo 0 0 0 0 0 0 0 0 0 2 3 3 pl:l at 0 c 9 70 9 72 9 71 9 69 9 69 9 69 9 51 3 1 3 1 31 30 30 30 30 30 30 30 30 30 h e Phenobarbita l Sodium u s P X I wae purchased with the M allinckrodt co nt rol designation KKM The amount of ph enobarbita l sodi um i n th calcul a ted dry sample used to prepa re the solution was 98 1 %

PAGE 110

106 TAB L E 37 PE R C EN T D E T ER I ORAT I O N AND p H CHA NGE O F 5 10 AND 2 0% SOLUT IO N S OF P HEN OBAR B ITAL SOD I U ii.-t u s P X I HEA T E D AT 80 ( l) 0 c Stre ng th o f S oluti o n 6 % 1 0% 20% Unheated S ol uti on pH 9 53 9 63 9 71 15 m inutes P er eent dete rior ati on 0 4 o o o e pH 9 60 9 60 9 69 30 minut es P er cent deter ioration 1 0 0 4 1 6 p H 9 48 9 68 9 66 60 m inutes P er c ent det e rioration 1 6 1 9 3 0 pH 9 37 9 41 9 5 1

PAGE 111

10 6 The results obtained by means of t he Budde assay of two 5% so lutions of Pheno bar bi tal Sodium U s P XI prepared from. two dif ferent lots of drug both of wh ich solutions were h eated at 80 C for 1 5 minut es showed 0 6% deterioration in the solution p re pa red from the lot desi gna te KI.M l and 0 4 % deterioration in the solution p repa r ed fro m the lot desigr..at ed KKM It wa co nc luded that thee results in dic a ted no si gnif icant difference between the two lots of drug which would p r L vent comparison of r esults obtained in this study of deterio r a tion In Tab l e 37 the r esults obtained on heating 6 10, and 20 % solu tions at eo 0 c for 16 3 0 an d 60 minutes are summari&ed An in creasing drop in pH was found to have o ccurred o n h eating each solu tio n at successively l o n g er p eri o ds ot time An exception to this is to be not ed in the case of the 6 % sol ut ion he ated for 16 minutes It was felt that this re p rese n ted an error Likewise an increase in p er ce n tage deterioration was n oted for ea c h separate solut i on Table 37 i ndi cates that 20% sol ut ions showed so m ewhat g reater de t erioration than 6 and 10 % solutions The results of percenta g e deterioration of the 2 0% solution were c ompa rable to those obtained by Tom ski and W aller (24) These workers f o,md l O to 1 6 % deterioration at the end ot 30 m inutes and 2 0 to 3 0% deterioration at the end ot 60 m inutes In the p rese n t study 1. 6 % and 3 0% deterio r ation respectively were found As an eme r gen c y method o f steri l izat i on the B ritish Pharma copoeia 1 932 (26) directs heating solutions at so 0 c for 30 minutes Under these c o n ditions 6 % and 20% solutions underwent 1 0% and 1 6% deterioration respective l y

PAGE 112

107 3 E ffeot or He ati ng 5 1 0 and 20"fo /v S olutions of Phenobarbi tal Sodium U S P X I at 1 00 C Th e purpos e of the ex pe ri m ent was to determine the effects of h eating 6 10, and 20% w/v solutions of Phen obarbital Sodium u s P XI at 100 c f or 16 30 and 60 minu tes The bottles containin g the solutions were exposed to steam f or the period of time m e n tio n ed The bottles were set in a tray abov e boiling water oontained i n an autoclave Th e lid was i mme diately re plac ed after the introduction of the tray As soon as the tem pe rature of th e surroun ding s above the water h ad reached 1 00 c ., as indicate d by a mr cury thermomete r. the t ime was taken by m eans of a stopwatch The temperature was rapidly b rou gh t to 1 00 c by hea tin e the a utocl ave wit h the aid of several M eker burners Th petcock was allowed to re main op en d urin g the e nt ire he atin g period to al low escape ot steam At the end o t the period the bottles w e re immediately removed and cooled by immersion into a m ixtur of ice and water

PAGE 113

10 8 TA B LE 38 EF:F'.,!, CT OF llF.ATDJ G A W /v SOLtlTIOU OF PHENOBARBITAL SODIU' ~ U S P XI* AT 100 (:t o s) 0 c Time Pe r ce n t Ppt Refractive :.: inutes C12H1103 N aN& Dete rioracc pe r Index in solution tion 100 cc pH at oc at 26 c 0 4 87 0 9 53 31 1 3432 0 4 87 0 9 63 31 1 343 1 Ave ra ge s 4 67 0 53 31 1 3432 15 4 77 0 9 42 30 1 3432 1 5 4 77 0 9 43 29 1 3432 1 5 4 78 Averages 4 77 2 1 0 9 43 30 1 3432 30 4 67 2 9 23 28 1.3 429 30 4 66 2 9 22 28 l.3 429 30 4 67 Averages 4 67 4 1 2 9 23 28 1.3429 60 4 39 9 01 28 l.3 428 60 4 39 9 01 28 1.3428 60 4 37 Averages 4 38 10 .1 4 9 01 29 1.3428 *The eubstanoe used to p repare the so l ution bore the M allinckrodt co n trol desi gna tion KKM The amount of phenobarbital sodiwn in the calculated dry sample used to p repare the solution was 97 4 % Ten (1 0) co or so lution were heated in a corked centrifuge tube

PAGE 114

1 0 9 TABLE 39 EFFE CT 1 )F JB: ATING A 10 % W/v SOL u TION OF PE ENOBARBITAL S O DIUM U. S P XI AT 100 (: o 5) C Ti me in P9 r cent Per cent ~ inutes c 12 H 11 o 3 N 2 N a Deteriora P:reoipitate in solution tion 00 /100 cc p.a at oc 0 9 77 0 9 63 31 0 9 77 0 9 61 31 Averages 9 77 0 9 62 31 15 9 58 0 9 40 31 15 9 58 0 9 41 31 15 9 56 Averages 9 57 2 1 0 9 41 31 30 9 29 4 9 23 31 30 9 28 4 9 23 31 30 9 28 Averages 9 28 5 l 4 9 23 31 6 0 8 81 6 9 04 29 60 8 80 6 9 06 29 60 8 78 Avera re s a so 9 9 6 9 05 29 -the substance used to p repare the so l uti on bore the Mallinc krodt co n trol n\mlber KKM The amount of phenobarbital sod ium in the oalou l a ted dry sample used to p repare the so lution was 97 7%

PAGE 115

110 TABL E 40 EFFE CT O F H E AT IN G A 20% W/v S O L U T IO N OF P H EN OBARB I T AL S OD I UM U. S. P. X I* AT 100 ( !. 0,5) C Time Per cent % P pt R efraotive M inutes cl2 H ll03 N 2 N a D eteriora cc per Index 1 n solution tion 100 co p H at oc at 26 c 0 19 66 0 9 8 0 28 1 3750 0 1 9 .64 0 9 80 28 1 3760 Avera g es 1 9 56 0 9 80 28 1.3750 16 19 13 l 9 60 28 1 3760 16 19 15 1 9 62 28 1 3750 16 19 17 Avera g es 19 15 2 1 l 9 61 28 l 3750 30 18 52 4 9 37 28 1 3740 30 18.62 6 9 38 28 1 3740 30 18 52 Avera g e 18 52 6 3 5 9 38 28 1 3740 60 17 67 14 9 14 28 l 3722 60 17 67 13 9 15 28 l 3721 60 17 69 Averages 17 67 9 6 14 9 15 28 1 3722 h e substa n ce used to p repare the solution bore the M allinckrodt co n trol d esi g nation KKM T h e amount of phenobarbital s odium 1n the calculate d dry sample used to p repare the solution was 97 8 %

PAGE 116

111 TABLE 41 EFFECT OF HE ATING 6 10 and 20% wjv SOLUTIONS OF PHENO BAR B ITAL SODIOK U .S. P XI AT 100 (t0 5) 0 c S treng th ot Solution 6% 10% 20% Unheated Solution pH 9 53 9 62 9 80 Precipitate cc /100 co 0 0 0 15 minutes Per c ent deterioration 2 1 2 1 2 1 pH 9 43 9 .41 9 61 Pr ecipitate 00 /100 co 0 0 l 30 m inutes Per oent deterioration 4 l 5 1 6 3 pH 9 23 9 23 9 38 Pr ecipitate cc /100 co 2 4 5 60 m inutes Per cent deterioration 10 1 9 9 9 6 pH 9 01 9 05 9 16 P recipitate oo /100 oc 4 6 14 In Table 41 the re1ulta obtained on heating 5 10, and 20% 10 lutiona at 100 C for 16 30 and 60 minutes are summarized A oon tinuoue drop in pH waa found to h ave occurred on hea tin g each solution for successively lon g er pe riods ot time The p erce ntage deterioration at a single period of hea tin g remained a pp roximately the same re g ard less of the strength of the original solution An increase in the amount of precipitate was no ted with i nc rease i n co n centration of the ori g inal sol u tion A co m paratively large decrease in refractive index of the sol u tion appeared to take place upon the formation of a large amount of p recipitate To msk i and W aller (24) stated that 1 0 %

PAGE 117

112 deterioration took place in a 20% solution heated at 100 c for 15 minutes and 9 0 to 10 5% deterioration at the end or 60 minutes o t heating In the p rese n t study 2 1 % and 9 6% deterioration, respeotiv ly were found under t h ese co n ditions 4 E ff ect of H eatin g 5 10 and 20% w/v Solutions of Phenobarbital Sodium U S P XI at 1150 c The p urpose of the experi men t was to determine the effects ot heat ing 6 10, and 2 0% w/v solutions ot Phen obarbital Sodium U s P X I at 115 ( 1 C for 1 5 30 60 and 120 minu tes In asmuch &a t h e N F V II (18) p r e scribes steri l ization of a ~p ule at 115 5 c it was deemed a pp ro p r la te to select this tem p erature at which to heat the solutions T he res ults obtained are g iven i n T bles 42, 43 and 4 4.

PAGE 118

113 TABLE 42 EFFECT OF HEATIN G A 5 %W/v SOLUT ION OF PHENOBARB ITAL SODIUM U S P X I* AT 115 ( z l)O c. Time Per oent Ppt Re traotive M inutee c 12 a 11 o 3 N 2 N a Deteriora oc pr Index in so l ution tion 100 oo pH at oc at 25 C 0 4 90 0 9 48 27 1.3432 0 4 88 _JL 9 47 27 1 3 432 Average& 4 89 0 9 48 27 1 3432 15 4 70 2 9 21 27 1 343 1 15 4 70 2 9 23 27 1.3430 Averages 4 70 3 9 2 9 22 27 l 3431 30 4 26 4 a 92 27 1. 3427 30 4 27 4 8 9 1 27 1,3428 Averages 4 27 12 7 4 8 92 27 1 3428 60 3 81 5 8 68 28 1 3422 60 3 82 6 s ss 28 1 3422 Averages 3 82 21 9 6 s ss 28 1.3422 120 3 os 9 8 61 28 1 34. 1 2 120 3 06 10 8,52 28 1 .. 3411 Avera g es 3 08 37 0 1 0 8 62 28 1. 3412 ~he substance u sed to prepare the solution bore the M allinckrodt contr o l numbe r K.KM The amount of phen obarbital sodium in the calcu lated dry sample used to p r ep are this solution was 97 8%

PAGE 119

114 TABLE 43 E'1l' FE CT OF tlE A T ING A 1 0% W /v SOLUT I O N OF PUElWBARB ITAL SOD I U'ai U S P X I* AT 116 { 1) C Time P er cent % Ppt Refractive M inutea cl2lill 03W 2 N a Deterio ra co p er Index 1n solution tion 100 cc pH at oc at 25 c 0 9 .79 0 9 62 28 1 3540 0 9 80 0 9 62 2 8 1.3 540 Averages 9 80 0 9 62 2 8 1.3640 16 9 31 4 9 .2 8 28 1 3537 15 9 .26 3 9 28 28 1. 3637 Avera g es 9 29 5 2 4 9 28 28 1.3 637 30 8 71 9 8 .91 28 1 3523 30 8 11 10 8.92 28 1.3 523 Avera g es 8 71 11 1 10 8 92 38 1.3 623 60 7 62 16 8 99 .. 29 l.3505 60 7 59 14 8 29 1 3504 Averages 7.61 22 3 15 a 99 .. 29 1.3 506 120 6 38 18 8e93H 28 1 3488 120 6.35 14 s 930 28 1 348'1 Avera g es 6 37 36 0 16 8 93 28 1 3488 The substance used to p repare the solution bore the .M allinckrodt control number KKM The amount of phenobarbital sodium 1n the ealeu l ated dry sample used to prepare the solution was 98 0% The p resence of free ammonia was detected by odor and by moisten ed litmus paper suspended over the solution The presence of carbon di oxide was detected by Elte1te rea gent pape r (22).

PAGE 120

116 T ABLE 44 EFF E CT O F HEAT ING A 2 0% W j'I SOL UT I ON OF PHENOBARB I TAL SOD I UM U s P T i m e M i n utes 0 0 A ve ra g es 1 6 1 5 Av e ra g es 30 3 0 Avera g es 6 0 60 Aver ag ea 120 12 0 Av e ra g es P er ce n t c 12 n 11 o 3 ti z,N a i n solution 1 9 74 1 9 75 1 8 6 2 1 8 6 2 17 61 17 5 2 17 5 7 15 5 3 16. 6 5 1 5 5 4 1 2 69 12. '71 12 70 X I* A T 11 6 ( :!: l C % De t er1ora tion 6 2 11 0 3 6 8 P pt cc per 100 cc 0 0 4 4 4 9 13 11 33 31 32 48 48 p H a.t 9 7 6 9 ,7 6 9 3 5 9. 34 9 35 9 13 9 10 oc 30 3 0 3 0 30 30 30 29 2 ~ 29 8 8 5.. 2 9 8 86 .. 29 8 85 .. 29 a sou 2 9 s s 20 2 9 8 6 1** 29 Re f r a ctive I n d ex t 2 5 c 1 3 7 5 6. 1 3 7 6 6 1 3 766 l.3748 1 3737 l 3732 1 3735 1 3687 1 3 68 9 1 3 68 8 1 3 6 31 l 3 6 31 $fh e su b st an ce used to p re p are t h e sol ut io n bore t he M allinckrodt co n trol number KKM Th e amount of p henobarbital sodiwn i n the oalcu late d dry s am ple used to p re p are the solution was 98 6% ** Th e p H was dete rm ined immediately atter ope n in g the bottle The p H rose fro m the fi g ure g iven in the table to a f i g ure f ro m p H 0 3 to pH 0 6 hi g her T h e maximum p H was not aso E:11 tained The presenoe ot ammonia was d e tected by suspe n ding m oistened litmus paper over the so lution

PAGE 121

116 '!'ABLE 45 EF F E CT OF HEATING 5 10, and 20 % W/v SOLUTIOMS OF PHENO BARBITAL SODIUM U. S P. XI AT 115 ( l C Strength of Solution Unheated Solution pi Precipitate ec /100 co 15 minut es Per oent det rioration pH Pr ecipitate ec ./100 cc 30 m inute Per eent deterioration pH P recipitate cc /100 co 60 m inute P@r o ent deterioration pH P recipitate cc ./ 100 00 120 m inutes Per cent deterioration pH P recipita te cc / 100 cc 9 48 0 3 9 9 22 2 12 .7 8 92 4 21 9 8 68 5 3 7. o 8 62 1 0 1 0% 20% 9 62 9 .7 6 0 0 s 2 6 2 9 28 9 35 .f: 4 11 1 11 0 8 92 9 .12 1 0 11 22.3 21.2 8 99 8 85 15 32 35 0 35 6 8 93 8 61 1 6 48 The failure of the pH to continue to fa ll below that found at the end of thirty minut es of heatirg waa p robably due to the pr eae n oe of ammonia which was detected up on openi ng the bottles containing the solution

PAGE 122

11 1 In Table 45 the results obt ai n ed on heat i ng 5 1 10 and 2 0% solu tions at 115 ( 1) C for 1 5 J O, 60 and 120 minu tes are s umma r i zed A oontinuous drop in pH was found o n heat i ng 6 % and 20 % aolutio n s at suooessively lon g er p eriods of time In the case of the 1 0% sol u tion the pH failed to fall below that found at the end of 30 minutes of heating This was p robably du e to formation of amm o n ia N o marked differe n ce in pe rce n ta g e deterioration at a single period of heat ing was found to ooour i n solutions of different stre ng t h s An increase in am ount of p reoipit ate was gene rally i nd icated w ith increase in con o en tration of the ori gin al solutions Tomski and W aller (24) stated t ha t 19 0 to 20 6 % deterio r a tion t ook place in approximately 2 0% solu tions heated at 1 1 6 C for 30 minutes In the p resent i n veati g ation 11 0 % deterioration at the end of 3 0 minutes an d 21 2 % deterioration at the end ot 60 minutes was detected on heating 20% sol ut ions at 116 c The N F VII (78) p rescribes the sterilisation of ampuls at 115 6 c f or 30 m inutes 5 E ffect of H e atin g 5 10 and 20% w/v Solutio n s of Phen obarbital Sodium U .S. P X I at 121 C Th e purpose of the experiment was to dete r m i n e the effects of h eatin g 6 10 and 20% w/v solutions of Phenoba rbital Sodium u s P X l at 127 ( tl ) 0 c These solutions were h eated at t h is temperature because it is one of the te mpe r a tures p rescribed by the N F V I (73) and N F V II (78) for sterilization of ampu ls The results obt ained ar e g iven in Tables 46 4 7 and 48

PAGE 123

118 TABL E 46 E FFE CT O F HEA TING A 5% W /v SOL UT IO N OF PHEN OBAR B ITAL SODIUM U S P x x AT 127 ( :t l)O c. Time Per cent % Ppt Re fractive M inute Cl2Hll03 N .;1a D eteriora oc. pe r Index in solution tion 1 00 co pH at o c at 25 c 0 4 88 0 9 59 29 1.3433 0 4 88 0 9 60 29 1.343a Averages 4.88 0 9 6 0 29 1.3433 16 4 .20 5 a so 29 1.3424 1 5 4 .20 5 8 82 29 1.342 5 Ave ra g es 4 .2 0 13. 9 5 8 81 29 l 3425 50 3 73 9 6 28 1 3418 30 3 .72 8 8 28 1.3419 Averages 3 .7 3 23 6 9 8 65 28 1.3419 60 3 09 10 8 43** 28 1. 3 411 60 3 .0 6 10 8 44** 28 l,3411 Averages 3 07 37 l 10 8 44 28 1.3411 120 2 12 17 8 16** 29 1.3400 120 2 .11 Averages 2.12 66 6 17 8.16 29 l.3400 4
PAGE 124

119 T ABLE 47 EFFEC T OF HEATiliG A 1 ()%W/v SOLUT I ON OF PHENOBARB I TAL SOD I UM U S P X I* AT 127 ( 1) C Time Pe r cent % M inutes c 12 H 11 o 3 N ~a Deteriora in solution tion 0 9 76 0 9 76 Averages 9 76 15 8 63 1 5 8.68 Averages 30 30 Avera g es 60 60 Averages 120 120 Avera. g ee 8 66 7 63 7 66 7 55 6 21 6 18 6 20 4 53 4.52 4 53 22 6 53 6 Pp t cc per 1 00 co 0 _Q_ 0 ll 8 10 17 18 18 21 26 24 28 26 27 p H at 0 c 9 73 2 5 9 72 25 9 73 25 9 26 24 9 23 24 9 24 24 8 73** 23 24 8 73 24 8 52 .. 24 a 24 8 51 24 8 24 8 64 0 24 B 63 24 Ref ractive Ind ex at 26 c l 354 1 l 3540 1 36 41 1 3624 1 353 3 l 3529 l 3600 1 3601 l 3501 1 3480 The substance used to p repare the solution bore the M allinckrodt co n trol number KKM The am ount of phenoba rbital sodium i n the ealou lated dry sample used to p re p are the s o lution was 97 6% h e pH was determined immedia tely after openin g the bottle. The pH rose from the fi g ure g iven in the table to a fig ure from pH 0 .1 to pH 0 7 hi g h r The maximum pH was no t ascertai ne d The p rese n ce of ammonia was detected by sus pen ding m oiste n ed lit mu s p aper over the sol ut ion

PAGE 125

120 TAB LE 48 EFFEC T OF H E A T ING A 20% W /v SO L U TIO N O F PHENO BAR B ITAL S ODI UM U. S. P. XI* AT 127 ( 1) 0 c Time Per cent Fpt Refractive M inutes cl2 H ll 03 N 2 N a Deteriora oc per Inde:x in solution tion 100 cc pH at oc at 25 c 0 19 64 0 9 90 25 1 3757 0 19 56 0 9 92 26 1 3766 A rages 19 56 0 9 91 25 1 3757 16 17 26 13 9 48 25 1 3720 15 17 24 1 2 9 49 26 1 3720 ATera g es 17.25 11 8 13 9 49 25 1 3720 30 16 21 18 9.28 2 6 1 367? 30 16 1 9 16 9 30 25 1 3680 Averages 15 2 0 22 3 17 9 29 25 1 5679 60 12 28 30 9 0 8 25 1 3617 60 12 30 40 9 11 2 5 1 3 6 18 Aver g es 12 29 37 l 36 9 10 25 1 3618 120 9 0 1 8 37 25 1 358 2 120 8 97 ** 8 34 26 1 3668 Av e rages 8 99 64 0 8 36 26 1. 3576 The substance used to prepare the solution bore the M. allinokrodt co n trol number K KM The amount of phenobarbital sodium in the calou lated dry 1amp l e used to m ake the so l ution wa s 97 8% The precipitate had a gg l erated into large hard m asses

PAGE 126

121 TABLE 49 EFFE CT OF H EA T ffiG 6 10. and 20% W/v SOLUTIONS OF PHEN 0 3 ARBI T AL SODIUM U.S. P X I AT 127 ( 1) C Str en gth of Sol uti on Unheated Solution pH Precipitate cc /100 cc 15 minut es Per oent deterioration pH Precipitate cc./100 cc. 30 minutes Per oe n t deterio r ation pH P recipitate co /1 00 cc. 60 minut es Pe r ce nt deterioration pH Pr ecipitate cc./100 co. 120 minut es Per cent d terioration pH Pre cipitate oo /100 cc 9 60 0 13.9 s. a 1 6 23.6 8 66 9 37 .l 8 44 10 56 6 8 16 17 10% 9 .73 9 91 0 0 11.3 11 8 9 24 9 49 10 13 22 6 22 3 8 73 .. 9 29 18 17 36 5 37 .l 8 9 10 24 35 53 e 64 .0 8 63 .. 8 36 24 .-rhe p rese n ce of ammonia was detected. The pH was determined immediately after openin g the bott le. h e p re en oe of a mmon ia was d etected The pH wa& determined im. 'll ediately after openin g the bott le. The pH rose from th e fi g ure in dioated to a fi g ure from pH 0 1 to pH 0.7 h i g her The m aximum p H was not aao rtained *Th e p recipitate had a gg lomerated into large. hard m asses

PAGE 127

122 In Table 45 the resu l ts obtained on heating 5 1 0 and 2 0% so lu tions at 127 (:t 1) C for 15 30, ao and 120 minutes are summariged. A continuous drop in pH was found on heating 5 and 2 0% solutions at uooessively long r period of time In the case of the 10% solution the pH failed to fall be l ow that found at th e end of 60 inutes of heating This was p robably due to the form~tion of ammonia Am.. m onia was detect ed in all so l ut i ons a fte r heating for 30 60 and 1 20 minu te periods No marked differenc e in per o entag deterioration at a sing le period of heating was found to ooour in solutions of different st r ength s. An inorea e in amount of p r ecipitate was indicated with increase in c o n centration of the original solutions As a method of sterili z ati o n o t tilled ampuls, the N F VII ( 78) p r esc r ibes heating at 126 5 c f or 15 minutee Under these conditions 14 % 11 % and 12 % deterioration resulted in 6 10 and 20% solutions resp ctively

PAGE 128

12~ F. RELATION OF pH TO TH E D E TERIORATIOll OF SOLUTIO l {S 01~ PHENOBARBITAL SODiti'M Inasmuch o.s several workers (2, 3 lO, 16) indicated tha.t the de ... terioration of solutions of barbiturates decreased upon lowering the pH, an attempt was made to ascertain Whether or not there was particu lar pH at which solutions of phenobarbital sodium could be stabilised Moreover it was desirable to learn the pH at which a solution of pheno barbital sodium would precipitate phenobarbi t;al. l. The Use of' M onosodium and Disodiurn Phosphates. A buffer mixt1lre was sought to lower the pH of e. 5% solution of Phenobarbital Sodium u s P n Monosodium and disodium phosphates were seleoted as components of such a. buffer mixture beoe.use these com pounds are official in the u. s. P. XI (70) and because phosphates are present in the blood stream,. Anhydrous d.ibasio sodium phosphate A. R and monoba.sio sodium phosphate A. R. (Nall 2 P0 4 H20) were purchaaed from the Mallinckrodt Chemical W orks In conducting the experiment a 15 ;' w/v solution of Phenobarbital Sodium U s .P Xl was prepared To 20 cc,. portions of this s.olution various amounts of each of the components of the bu.ffer mixture re added so that the final concentration of the phenobarbital sodium ns 5% w/v The pH of the resulti ng solution wae d termined By substitut .. 1ng water 1n each instance for the solution ot phenobarbital sodium, comparison between the pll of the buffer mixture ~d th pH of the buffer d solutions of the barbiturate s afforded. When the formation of a pre cipitate was noted a determination of the volume of precipitate -was

PAGE 129

124 undertaken For this purpose the buttered solution of the barbitur te were prepared in 60 cc., grad ted centrifuge tubes. The remainder of the proeedure VJas followed as previously described for the determination of volume of precipitate. Ma.int&.ining the same ratio of buffer and pheno barbital sodium, 26 cc samples were prepared in four-ounce, g laas stoppered pyrex bottles. The samples were heated at 115 ( ,! 1) 0 c tor thirty minutes It was found that a 5% solution of Phenobarbital Sodium u s P XI produced approximately 10 cc. of precipit&.te per 100 co Therefore, if a marked decrease in amount of' precipitate ooourNd, a stabilizing effect might be indicated. In addition, these conditions ot treatment were chosen because of their known applicability to steriliza tion. Accordi.~g to Husa (109) moist heat t 1150 c. kill both bacteria and spores L"l less than 30 minutes. The pH of' various proportions ot the components of the buffer mix ture is g iven in Table 60. The result obtainod in the u e or these ame buffer mixtures with a 6% w/v solution of' Phenob rbital Sodium u. s. P. XI are given in Table 51 In Table 61 it is seen that the phosphate butf rs lowered the pH of a 5 1~ /v solution ot Phenobarbital Sodium u. s. P. XI from pH 9,7 through various pH levels to pH 8 7 Precipitation of phenobarbital from 6% solution w a shown to occur t pH a e The faet that the a.mount ot pre cipitate was g reater in some instances without any distinct difference in pH wa probably aocounted for by the increas in buf'f'er cape.city in those solutions 1ich had larger amount ot sodiUlll biphospha.te The pH of e ch of the buffered solutions was lower a.fter heating. Sample which had approximately 12 to 14 cc of precipitate per 100

PAGE 130

125 TABLE 60 pH OF MIXTURES OF MONOSODIW AND DISODIUM PHOSPHATE SOL U TIONS cc co. cc. 0 .1 0 11 1 M Water pH at 0 c NaH2P04 NazHP04 20 0 10 4 47 26 18 2 10 6 81 26 16 4 10 6 13 26 14 6 10 6 40 26 l2 8 10 6 60 26 lO 10 lO 6 .7 4 26 8 12 10 6 91 26 6 14 10 7.11 25 4 16 10 1 ss 26 2 18 10 7.69 25 l 19 10 a 01 25 0 20 10 9 .-2 0 26

PAGE 131

TABL E 61 RE L A TI ON OF P H O S P HA T E BUFFER TO TH E P RE CIPITA T I ON IN A 5 % W /v SOLUTIO N OF P H l!llO BARB I T AL S OD I UM U S. P X I BEFORE AND AFTER .tiEA T IN G AT 115 ( 1 C FOR 30 MINUTE S cc cc cc Before R eatin~ After H eating O .l M O .l M 15 pe r cent Ppt cc Ppt cc N a. li 2P0 4 N a2BP04 C 1z l:I 11 03 N 2 N a pli at oc per 100 cc pH at oc per 100 cc 0 40 20 9 57 24 0 8 70 26 8 4 36 20 8 18 25 0 a so 24 6 8 32 20 8 82 24 3 8 62 26 4 12 28 20 a ~ao 26 6 8 68 26 6 16 24 20 8 79 25 9 8 63 26 6 20 20 20 8.80 25 9 8 49 26 6 24 16 20 8 80 2 5 12 8.64 26 28 12 20 8 79 25 13 8 51 26 32 8 20 8 78 25 13 8 57 26 6 36 4 20 8 76 25 14 8 62 25 8 40 0 20 8 74 25 14 8 52 2 6 6 ~he pH of a solution of Phenobarbital Sodium u s P X I prepared with the use of water instead or the buffer mixtures was 9 70 at 24 C The volume of p recipitate produced by such a solution heated under these conditions was 10 co / 100 cc Indefinite results due to separation of a liquid and to formation of crystals during centrifuging 126

PAGE 132

127 oo showed no precipitate immediately e.!'ter removal .fro m the utoolave Althou.gh some p:recipita.te was formed aft r cooling the amowt waa lea than originally pr sent This indicated that a lowering ot the pH may have decreased the QJllOunt of' deterioration Ine.smuoh as none of the solutions re rre of pr-eoipitate such solutions were not oonsidered satisfactory for any practical use 2 The Use ot Sodium Carbonate An attempt was a.de to regulate the pH of a solution of pheno barbital sodium by emp l oying various amounts of sodium carbonate abo-ve and below the amount neee sary to neutralize the amount of phenobarb1 .. tal equivalent to that theoretically present in 5% solution of pheno barbital sodium Inasmuch as 25 cc. portions were to be pla.oed in tour-ounce, g la.ss stoppered, pyrex bottle the amount ot Phenobarbital u. s P. XI, which was taken for each portion was equivalent to the theoretical. 8lll0unt ot phenobarbital in 26 0 0 of a 6% w / v solution of phenobarbital sodium Therefore l 142 Gm of Phenobubital u. s. P XI s aoourately weighed and be c ause of stickiness of the me.terial transferred to the bottl ith the aid of 95% ethyl loohol The alcohol was evaporated off in water bath To determine whether or not a:ny decomposition had taken place mider these conditions, other s amp l es were utjeeted to the same treatment and the m lting point of the residues was determined The melting point -was found to be 177 60 c The melting point of the origin al drug was the same Inasmuoh as this approximates the highest melting point given by the u s P XI (70) it was felt that no significant de composition had taken pl oe.

PAGE 133

128 The sodium carbonate emplo~d -was purehased in the form of the anhydrous e.nalytiea.l r gent f'ro m the Mallinokrodt Chemioa.l Works In conducting the experiment, w.rious amounts of sodium carbonate wer e.dded in tho form ot a. 2.s % w/v solution to 1,142-Gm. port10li8 ot Phettoba rbital u. s. F XI contained in pyrex bottle The a.mount ot sodium carbonate required to neutralize this amount of phenobarbital 1s o.521 Gm. ter added we.s equiva.l nt to the ditt' rence between 25 cc. a.nd the -volume of sodium carbonate solutien used. Several sam p le thu prepared were aesayed by the Budde method. Sut'ticient sodium carbonate was added to bring its conoen.tn.tion up to the level employed in thie UaAY The remainder of the &uay e conducted as previously de cribed. Before heating, the mixtUNta \fl.Ire shaken meohanica.lly for one hour. A snall amount of phenobarbital remained undissolved in eaoh bottle. Howe ver solution was brought bout 1n each o se because evel"y solution was clear upon removing .from the autocla. The bottle immediately ir.mll.ersed in mixture of ice and water. The amplea w re assayed in the same manner that the controls re aa o.yed. The results obtained are urmnarbed in Tabl e 52. F.e.oh ti.gure giwn. in the table is an ve.rage of.' two detenninat1ons. Although the a.mourits of.' sodium o rbonate add d were bove and below the amount neoeasaey to neuti-a.lize the amount of phenobarbital present, the pH ot the mixtures remained at pH 9 o. The pH ot eaoh &alllple -.a lowered during treatment. The pH tell from pH 9.0 to pH s .s 1n. those samples which h&d an exc. as of sodium oarb11nate, The pH fell trom 9 0 to pH 8 7 111 those samples which had an inauttioien.t amount ot sodium oa.r bonate to neutral1&e the amount of phenobarbital present, The deter1ora

PAGE 134

129 TABLE 52 EFFECT OF SODIUM CARBONATE O l~ THE DETERIORATIO N OF AN APPROXI MATE LY 5% YI/V SOLUTION OF FliffiOBARBITAL SODIUM HEATED AT 116 (!, 1) 0 c. FOR THIRTY M INUTES Gm. pH e.t 24o c. pH at 24o c. % % NazC03 before ater 0 1ti110zN2Na. DeterHeating Hee.ting ioration o.ss 9.02 s.a1 ihl8 14.9 0.54 9,02 a.az 4.20 14.5 o .53 o.oo e.11 4.20 14.6 o.52 9.00 a.11 4.21 14.3 o.s1 9.02 s.14 4.23 1s.s o.so s.oa a.12 4.14 11.1 0,.49 9.00 8,69 4.27 1s.o o.48 s.oo s.n 4+29 u.e Assays ot unheated solutions indicated that the amount or phenobarbital present was equivalent to the theoretical amount of phenobarbital in 25 oo. 0 a 4.91% solution of phenobarbital sod1un. amount ot a~ou sodium carbonate A. R. mentioned above was dd.ed in the form of' a 2.5% r/v olution to l.142 am. of Pheno barbital u s P. XI. 1'he volum ot water added was equivalont to the difference betwe n 26 co. and the -volume ot odium ce.rbonat solution used.

PAGE 135

tta e omewhat les in the latter solutions~ A satis fa ctory de ... crease in deterioratio n was not obtained by the use of soditm1 carbonate in connection with phenobarbital. G. EVALUATION OF SEVERAL SUGGESTED STABILIZI NG AGENTS ON SOLUTIONS OF Pil,'NOBARBITAL AlID mmroB.ARBI'lAL SODIUM 1 An Experiment on the Stability of Elb:ir of Ph enobarbital u s. P. XII ands veral M odifications E lixir of Phenobarbital u s. P. XII (71) represents the only liquid preparation of phenob rbital which 1 official It wa.a decided to in vestigate it sts.b1Uty and the stability of sewral modified t'orme ot this solutio n through the a ge ncy of the Budde Assay The s vere.l formulas studied are as follows, F oFmula A. Formula B F ormula C E lixir of Phenobarbital U s :p XII Phenobarbital Tincture 0 Sweet Orang e Peel Solution of' Amara.nth Alcohol 95 % v/v Glycerin Syrup Wa ter, distilled q ad Phenobarbital Alcohol. 95 % v/v Glycerin Syrup Vil.ter, distilled 4 ad Phenobarbital Alcohol 95% v/v Glycerin V iator, distilled 4 s d 0 4 Gin s. oo. l. oe 12.6 oc. 45. co 15 oe 100. oo 0 4 Gm 14 9 CC 111 ,s. co 15, co 100 oc 0 4 Gm 14~9 co 4S cc. 100, oe

PAGE 136

Form.ul D Phenobarbital thenamine Alcohol 95 % v / 'f/ Glycerin Syrup 181 Water, distilled. 4 ad 0.-4 am.. 0.4 Gm 14~9 co. 45 OCh 15. co 100. oc. The amount of 96 % v / v -. l oohol in the tonnulas devoid of t h e tincture adjusted to mai n tain the sam oonoentratio n o f alcohol as occur in th offio.ia.l elixir. For the sake of better oontrol th quivalent amo'Ullta of alcohol. g lyo rill and syrup wre weighed. inst ad of mea ured J4 ethenamine u s. P XI wa.s employed in Formula. D bee&use it wa claimed by E li L illy and Com~ (3-9) that stable solutions of amyt&l could be mad by dissolvi ng amytal in hydroalcoholic mixtures co n taining a uf'ti cient quantity 0 methenamine. The aniOU!lt o f methene.mine employed n equivalent to t h e concentration of the barbiturate. In a p revious experiment it s found that the B udde method as de cribed before, could be employed in a.asaying phenob rbita.l dissolved i n suoh solvents as 95 % ethyl alooh.ol an d G lycerin u s P XI. Aoourate results were n ot obtained When the phenobarbital wa dissol-qd 1n an queous sol u tio n ot Methenemine u s P XI. I n the lat t er case th e.ppearanc o f' th turbidity in the titratio n g radual, making the end point unoertain Four, 35 -c e portions of ea.oh of the a.bove solutio n s were plao din tour-ounce, g lass-stoppered pyre :x bott l es Two bottle of each solu tion wer heated for each of two periods The oondi tiQna were 16 min utes at 100 ( o s)o c. and 30 minutes a.t 116 ( -t 1) 0 c The Budde assay was conducted in the maumer previou s ly described wit h the exception that

PAGE 137

182 25-oo. portions were withdrawn from ea.o h bottle and pipetted into a solution of the required amount of sodiUll carbona.t in 25 co. of water. The avera.ges of' the l".esult obtained on a. saying solutio:U prepared according to ~ormulas B nd Care indicated in Table 53 TABLE 53 THE STABILITY OF MOD IFICATIONS OF ELIXIR OF HiENOBARDITAL, U. s. P. XII Formula B C B .fore Hea ting % c1a11u 0 ~2 in solution 01399 0.402 Ater Heating &t 100 (!, o.s) o. tor lS minutes % c 12 B 11 o 3 N 2 Ba in solution. o .399 Oe399 % Deterioration o .o o.a Aft r Heating at 115 (!, 1) 0 c for 30 minutes C1zH110~2Na in solution 0 400 0 400 % Deteriora.tion o o o.s An ttempt was 100.do to say the otfioi 1 lixir before d a.fter the periods of heating. Satiafi otory results could not be obtained be cause a daFkxtning of the elution was produced due to the otion of the sodi\m carbonate on the amaranth. In addition to the resultant d er ue 1n light intensity of the re.y ot li ght paesi.ng through the olut1on under 'Which conditioruJ th end point is observed, the solution was turbid befor the addition of dlver nitrate solution. .A$eay of the oluti.on

PAGE 138

133 containing methenamine wa. also un.sucee stul. beoa.use of the gn.dWll appearance of a turbidity during the titration as previously encoun tered when assaying in the presence ot methe%18mine. However no preci pitate was produced in any of the solutions after heating for the two periods mentioned. This did not necessarily indicate stabilizin g action because of possible solubility of the decomposition products in &.lcohol. The Budde ssa.y of the solutions cU..ftering trom the official elixir in the deletion of the t1noture e.nd the solution of amaranth indicated no deterioration after both periods of h ting. Approximate ly 1 % deterioration .t'ter both period. ot heating wa.s indio :ted 1n the solutions which differed from the o!"fioi l elixir in that the tinoture, solution of ama.ranth and yrup were dele-ted. a. An Experiment on the Stability ot Phenobarbital Sodium u. s. P. XI in the Presence of Alcohol, Glycerin, &nd Ant1pyr1ne. To continue the investigation of the value of alcohol and glyoerin as stabili~in & agents an experiment on the stability of 10% w/v olu tions of Phenobarbital Sodium u. s. P. XI 1n the preaence o-f alcohol and glycerin was conducted Baldi (68) claimed that a table solution of phenobarbital sodium could b made by di olving it in. e. mixture of qual pe.rta of gly()erin attd water Formula. C given below was stated by Baldi (68) to be a stable solution. Formula E was modeled after a formula designed by Hazelton Koppany1, and Linegar (54). The tollewing formula.a were selected in studying the et'fect of alcohol and glycel"in, alone and in oombinationc

PAGE 139

Formula A Formula B Formula C F ormula D Formula E 1S4 fhenobarbital Sodium Alcohol 95% v / v ter distill d 4 a. ad Phenobarbital Sodium Glycerin Water distilled q. s ad Phenobarbital Sodium Glycerin Alcohol, 95 % v / v Water, distill d 4 d Phenobarbital Sodium Glycerin A lcohol 96 % v / v W ater distilled 4 a ad Phenobarbital Sodium O lyeerin Alcohol 95 % v/
PAGE 140

135 sufficient water to make 250 oo_. Ot the dilution 25 cc. was ta.ken tor the titration. The results ~e summarized in Table 54 E ach figure g ive n in this table represents the average of two determinations F 'rom the results given in Ta.ble 54., it appeared that 100./4 w /v sol ution of Phenobarbital Sodium u s. P XI in approximately 46 % w /v alcohol is more stable s.t 115 c. tfUUl solutions eonta.ining glyo ori,n alo ne or gly ceri n and alcohol in the combinations and percentages studied. The solutio n containing 465~ w/v alcohol showed 1 6% deteriora tion after heating a.t this temperature for 30 minutes, whereas the other solutions sirailarly treated s h owed approximately 8% deterioration whioh was only slightly less than that shown by the co n trol solution of 10 % w/v Phenobarbital Sodil.ml U s P. XI in water alone. Somewhat less dewriora tion than in the control \'V9.S shown by all of the formulas mentioned in the t ble when heated at 100 c for 15 minutes. The solution containing the e.ntipyrine could not be assayed by the Budde ~ethod inasmuch as the res ~ lts obtained with the untreated solution were inconsistent and were in marked difference with the untreated control solution. Although this solution .formed no precipitate after heating at the shorter period of time and only approximately l co. of precipitate per 100 cc. after heating at the longer period of time ., no conelusion drawn concerning its stability bee use of the possibl e solution of the decomposition products in the alcohol present 3. An Experiment on the Stability of Solutions of Phenobarbital Sodium u s P XI containing Dextrose An experiment was conducted to learn whether or not a 10 ~, w/v solution

PAGE 141

136 TABLE 64 THE STABILITY OF A 10% W/v SOLUTION OF PHENOBARBITAL SODIUM U. s. P. XI IN THE PRESENCE OF GLYCERI?J, ALCOHOL .. A.'ID ANTIPmum Formula Control A B G D B fore Heating % C12H1103NzNa in sol u tion 9 76 9 76 9 75 9 76 9 77 Ppt co./100 oc 0 0 0 0 0 Arter Heating at 100 (!, o s) 0 c for 15 minutes % o 1 iHMo 3 N 2 Na. in sol ion 9.51 9 62 9 61 9 61 9 66 % det r-ioration 2 6 1 s l 4 1 5 1.1 Ppt cc./100 oo l 0 0 l l After Heating a.t 116 ( + 1) 0 c. for 30 minutes % Ci_2H1103NzNa in solution 8 88 9 60 8 96 9 02 8 96 % deterioration 9 0 1.5 8 1 7 6 a a Ppt. 00 ./ 100 co 9 0 4 6 l The amount of phenobarbital sodium in the calculated dry sample used to make this solution was accordingly 97 6%

PAGE 142

137 of phenobarbital sodium could be stabiliz d by replacing a portion of the solvent with dextro e. Several formulas containin g 50 % w/v D extrose u. s. P. XI in the presence oi' water, alcohol, and g lycerin were studied The Phenobe. rb i tal Sodium U S. P. XI was purchased with the Mallinckrodt control designation K.KM }'ormula B Formula C Phenobarbital Sodium Dextrose Water, distilled 4 s d Phenobarbital Sodium Dext rose Alcohol. 95 1; v / v W ater, distilled q s. ad Phenobarbital Sodium Dextrose Glyc rin ter, distilled q s ad 10 Ch. 50 Gm 100 co. lO Gm. 50 Gm 10 Gm 100 oc 10 Gm 50 Gm 10
PAGE 143

138 TABLE 56 THE STABILITY OF 10 % W/v SO l. UTIONS OF PHENOBARBITAL SODIUM U.S. P XI CONTAINi l iG 50% W/V DEXTROSE U S. P XI F ormula Co n tro l A B C Before Heating % C12H1103 N 2Na in solution 9 74 9 76 9 71 9 '13 After H eating at 100 ( ,! o 5) 0 c. tor 15 minutes % C12H1103 N 2 N a in solution 9 61 8 87 9 54 9 04 % deterioration 2. 4 9 0 1 a 1.1 The amount of phenobarbital sodium i n the calculated dry sample used to make the sol u tio n was acoordi ng ly 97 4%

PAGE 144

sta.bili ty IV. DISCUSSION OF RESULTS The present investig tion was oonoerned with study of a number of factors in the deterioration of solutions of phenobarbital sodium, to c;athe r with research on several suggested methods of retardation of this process. Method s ot assay applicable to the problem were compared. !11e thods of Assay The hygroscopic nature 01 "" phenobarbital sodium necessitated a method of determining the percente.ge moisture present. Comparison of results obtai n ed on dryillG samples at temperatures sli gh tly above 100 c. and t temperatures slightly abo'Ve 14()0 c. showe d 1 7% and 3 4 % loss 1n weight respectively Vacuum desic<,ation a.t room tem.peratuN rewa.l d a loss in weight of 1 5 % which is value approximatin g the loss &:lstained t the former temperature. Budde ass aye did not reveal e:a.y decrease in the amount o f phenobaTbital sodium ori g inally present 1 n the lot of drug em .. ployed. regardle s of' which of the two temper tures were used in drying. Numerous methods of e. sa.y wn reported in the literature Certain methods were selected e.s worthy 0 study Ree. ons for not investig ting the other methods have e.lr~dy been g iven. The lot of Ph enobarbital Sodium u. s P. XI employed oonfol"med to the official purity rubric in asmuch as 91% "phenobarbital" s extraoted by means of the official assay. By means of the modified and unmodified Rotondaro a.saaya, ?5 Z% and 78 6% phenobarbital respectively was extracted from samples ot the same lot or drug. The ave r ge sum of the three residues obtained in

PAGE 145

140 thee Rotondaro u ys was below the u. s. P XI tolerance. The number of treatments wit h chloroform to extract each residue in the moditi d Rotoncla.ro assay was increased nearly twofold. Th sum of the three reside was again found to be below the u. s. P XI tolerance for nphenobarbite.1 11 dified Rotondaro assays of' Phenobarbital u s. P, XI and of' phenobarbital previousl y reoowr d in other Rotondaro ae.y ge.w e.n average reoo-aery of phenobarbital more than 20% below the original amount taken, Monover the sum of the three residues only a-vera g ed 81.7%. Several methods of esay employing ail ver nitrat were reported but all a.re exemplified by the Budde (93), Viebook and Fuohs {95, 96),. and Shulek and Roz.sa {100, lOl) methods, These methods o f a.sse.y shovmd respectively 89.5%, 89,1%, and 86.4;~ phenobarbital pre ent in t..1-ie lot of Phenobarbital SodiUll'l U, s. P. XI employ,td, The U, S. P. XII tolerance 1 89,0 to 91.&fo phenobarbital, The end points in the latter two method s were diftic ult to determine 'When assaying this drug. The chemistry o .f' the B uddo aa.y was inve tigated. Silver in th form of silver barbitur te, dissolved in a IIJ olution containing sodium carbons. te, snot precipitated upon the ddition of' aodian chloride. The compound for.med on adding silver nitrate to phenobarbital was insolubl in te:r but soluble in the pre enoe of sodium ce.rbonate. Addition 0 phenobarbital to a suspension of silver carbonate containing exc ss odium carbonate resulted in a clear solution. The possible interference or d composition products in th Budde assay was investigated, No significant error was revealed on titrating a solution of pbenylethylacetylurea nor on titrating a solution 0 phenyl et hy la.cetylur a which had been heated under the most drastic conditions

PAGE 146

141 to 'Which solutions in subsequent experiments wre subjected. The maxi mum b reakdown possible in the investigation was thue duplioated to assure formation of phenylethylaoetio aoid and urea. The von Babitsch titra.tion (86) of pure phenobarbital uaiag stan dard alkali in the presence of thymolphthalein ga ;ve an average result of 99.8%. In a oompe.rison ot various assays on a deteriorated solution of phenobarbital sodium, the u s P. XI assay showed s.1 1, deterlor tion in a 4 % \v /v solution of Phenobarbital Sodium u. s. P. XI heated at 115 c. for 30 minutes. The modi.fiecl Rotondaro a.asa.y showed 34 7 % deterioration and the Budde say showed 16.6% deterior tion in the same solution. JEffeet of ilea.ting Solutions ot Phenobarbital Sodium u s p XI Solutions of Phenobarbital Sod1wn u s. P. XI in conoentrationa of 5, 10, and 2C)QJ were ubjected to 15, 30, 60, e.nd 120-minute period& ot heating at 600, 800 100, 116, and 127 c These oonditiona were selected to encompass most 0 the conditions previously mployed by other workers a n d to inolud for the most part, the off1ci l conditions ot sterilization reowmn.onded for olutions intended for par nteral u The results are sunmarized in Table 56. A comparison of these re ult with the results obtained by Tomski and Willer {24) under some of theae conditions is afforcied by Table 57 By the result given. in Table 56, it indicated that sterilization of solutions through the genoy of heat is accompanied by deteriora tion, the extent of which is dependent upon temperature nd duration ot exposure. Hus (109) reported that moist he t t ao 0 c. for 60 minut a kil l 11 living bact ria but not their spor s. The latter are killed in

PAGE 147

142 TABLE 56 EFFE CT OF HEATING SO LUTIONS O F PHENOBARBITAL SODIUM U. s. P. XI* 60 c ao 0 c 100 c 115 c 127 c. 5 % w/v Solution O min. 9 51 9 63 9 53 9 48 9 60 15 min. o.o 0 4 2 1 3 9 13 9 9 51 9 60 9 43 9.22 a al 0 0 0 2 5 30 min. o o 1 0 4 1 12 7 23 6 9 49 9 48 9 23 8 92 a 0 0 2 4 9 60 min 2 0 l 6 10 1 21 9 37 l 9 43 9 37 9 01 8 68 8 0 0 4 5 1 0 120 min 37 o 56 6 8 52 8 .1 10 17 10 w / v Solution O min 9 63 9 62 9 62 9 78 15 min 0 2 1 6 2 11 3 9 60 9 41 9 28 9 24 0 0 4 10 30 min 0 4 5 l 11 1 22.6 9 56 9 23 8 92 8 0 4 10 18 60 min 1 9 9 0 22 3 36.5 9 41 9 05 8 99** 8 5 1 l 6 l5 24 120 min 36 0 53 6 8 8a63U 16 27 The sets of three fitures after each period and below a partioular temperature represent 1n order : a) per oent deterioration, b) pH and c) volume of precipitate ., cc ./ 100 o o. Blank: spaces indicate that de terminations e r e not mad After 0 min only the pH of the original solutio n is given uAmmonia wa detected

PAGE 148

143 TABLE 56 {continued) 60 c so 0 c 100 c. 115 c. 127 c. 20% w/v Solution 0 min 9.71 9 80 9.7G 9.91 15 min. a.a 2.1 s.2 11.a 9 69 9 61 9,35 9.4 $ 0 l 4 13 30 min 1 s 5 3 11 0 22 3 9 65 9 38 a.u 9 0 5 11 17 60 min 3 0 9 6 a1 .2 37.l 9 51 9,16 s.a6 9.10* 3 14 32 35 120 min. 35 6 54.0 a.61 s ie-. 48 The sets of three figures a:f'ter eacm period and below a particular temperature represent in orders ) per cent deterioration. b) pH. and o) wlume of precipitate co./ 100 oo. Blank: spe.ees indicate that de terminations were not mad After "O min only the pil of' the original solution i given Ammonia w s detected

PAGE 149

144 TABLE 57 PERCENTAGE DE'lERIORATION FOUND BY TOMSKI A.ND WALLER O N HEATING 20% SOL U TIO NS O F PHEN OBARBITAL SODI UM CONTAI NED IN AMPULS 15 minutes SO minutes 60 minutes 800 C 1 0 to 1 6 2 0 to 3 0 100 c 1 0 19 a to 20 4

PAGE 150

145 30 minutes at 115 c., according to Hua (109). By the results given in Table 56, it is indicated that uoh sterilization is accompanied b y small amount of deterioration under the milder treatment mentioned and consider ble deterio tion under the stronger treatment mention ed. The res ults obtained on heating 20% solutions oorresponded tor the most part to the results obtained by Tomski and W aller (24) under ome of the e conditions. Ditferenees were not d in peroentage deterioration on heating 20 ,i solutions at 100 c tor 15 minutes and a.t 115 9 c. for 30 minutes Tomski and W aller (24) reported l.0 % alld 19.2 to 20. 4% deteri oration under the respective conditions. The results of the present study were 2% and 11% deterioration respeotiwly The Budde assay was employed excl usively in det nnining th percent age deterioration in this study of the effect ot heating solutions of Phenob rbital Sodium u s P n The amount of phenobarbital odium 1n the calculated dry sample us d to prepare each solution was determined fl'om the asse.ys on the unheated solution The wrage of all the result on per cent of phenobarbital odium in the calculated drys ple wa 97 ,. 7% with an average difference between assays of o s%. Thee figure apply to both lot of Phenobarbital Sodium u. s P XI used 1n the present study Relation of@ to Deterioration U pon lowering the pH of a 5% w / v solution o:f Ph nobarbi tal Sodium u s. P XI by means of various a.mounts ot monosodium and disodium phos phates, it was found that phenobarbital was pr oipitated at pH a a th amount of pr cipitate increasing with increa. e in the amount of mono sodium phosphate added. Five per cent solutions buffered t pH 8 8 to

PAGE 151

146 a. 7 With 12 to 14 co of precipitate per lOO oo,, shovl9d 6 to 8 oc. ot preo1p1tate per 100 oo. after hea.ting at 116 C, tor 30 minutes, wh ea a solution of the same stNngth in tel' alone showed lO cc. pre~pitate per 100 cc at'tE,r slmilar heating, Five per cent solution buffered at the a.me pH but with le-se mom odilB phosphate showed no rlced cleo e in amount of precipitate present be:tore and &i'ter heating Various amounts of odium carbonate re added to phenobarbital, abow and below the amount n eoea ary for neutralization. The amounts ot odim carbonate add d vari between maxi.mlJDl 0 6 abo'99 and m1n1m\lll ot 1.1% below the amount of sodium rbonate nece aa.ry to neutr lie th a.mount ot phenoba.rbit l present in an qui"v&.lent wlum.e of 5% solution of' ph noba.rbital sodium. The pH of the mixtures remained at pH 9 0 with the quant1tiea of odi"Um carbonate used. After heating t 116 o. for 30 minutes, the pH fell .from pH 9 to pH a.a in those sempl a which had exce e of aodi\111 carbonate Th pH tell from pH 9.o to pH a.7 1Zl tho sample which had an insuffioient amount tor neutralization.. The per cent deterioration under these oonditione ranged from 14.9 to 12 &%, being lower 1n the solutions having the smaller amounts 0 sodium carbonate. Evaluation of S veral Sugested Stabilizing Agent The value o f s veral suggested stabilizing agents wa 1nvestigated Elixi r of Phenobarbital U. s. P XII with and without 0 4% methenamine oould not be a sa.yed by the Budd method, but no preoipi te.te was ev1dent a.fter he ting t lOOo c for 1 6 minute and af'ter heating e.t 1150 c. for 30 minutes. The solutions differing from the official elixir 1n the dele tion ot the tinoture and the solution of a.mare.nth shcmed nod t r1orat1on

PAGE 152

14'1 after both periods of heating. Approximately 1% deterioration after both periods of heating was indioate
PAGE 153

148 percentage dete r1ora.tion than a co n trol solution of' 10 % w/v solution of Phenobarbital Sodium in tr alone V SUUMARY AND CONCLUSIONS Nature of the Det rioration The products formed during t h e deoomposition of barbiturate solu tions and the manner in which they are produ ced were revealed by several workers, namely Ste nhauer (2), dse n (5), Bailey (4), N ielsen (10, 11. 13) and Aepelund and his coworkers (7, a 9). Their findings may be summariz.ed by t he outline which appears on t h e following pe.g From the agreeme nt in t h e results reported by the atoreto mentioned worker, the nature of the deco m position ppea.red to be amply explained An investigation wa carried out on a number of factors, such a time, temperature, and concentratio n which re involved in the deteriora tio n o f sol u tions ot phenobarbital odim The effects produced in pH, volume of preo1p1tate, and r &active ind.ex had not been reported tor thia compound und r the conditions of the present tudy. In addition an ex tensio n o f the wo r k on peroentage deterioratio n performed by T9mak1 and W aller (24) s made. Me thods of Assar Meth ods of determining percentage moisture 1n Phenobarbital Sodium u s. P XI we.re studied and compared Compe.riaon ot re ults obtained on "Vacuum desiccation with those obtained on heatin g at 100 c and 140 c indicated either 1neomplete drying at 100 c or t~e presence ot d composi tion at 140 c. Budde says of a.mp les heated at either temperature

PAGE 154

149 5, 5' d.isubstituted malonylurea 1 ~0 2, 2 1 di ubatituted malonurio acid / 00 2 ~20 H R 1 c N \ \ C C : O / \ H N / H2 o<. 1 disubsti tuted acetylurea o( "'' disubstituted e.oeta:mide ~o R COOH ""' / C ,/"' R COO R I 2, 2 disub tituted malonio a.oid 2, 2 disubstituted acetic aoid

PAGE 155

150 f: iled to show any deoompo ition. However, it was deemed a.d.vise.ble to avoid the introduction ot an error due to posaibl deterioration 'Which might have been undetected in samples heated at 140 c. Undried samples of lmom moisture eontent re gene rall y uaed in the 1nve tiga.tion. Numerous methods of a.a ay of barbiturates were found in the litera ture. Nielsen (13) a.nd Madaen (3) calculated th percentage deteriora tion by determining the a.mount of dialkylacetylurea formed. Thia appeared to be at fault because of possible further deoomposition ot dialkyla.oetylurea which 10uld consequently lead to lower res ults as re vealed by Sohlemner and T8rber (30). The method employing an inmiaoi ble solwnt at a single pH level re considered unsati factory because ot possible simultaneous extraction ot decomposition products. The meth od in which alkalinity or a.eidity of the barbiturate derivati'VG a mea ured by aimply titrating with a standard acid or alkali were not applicable because of the existence ot alkaline and acidic decomposition products in deteriorated solutions. Colorimetric methods were not tn .. vestigate4 beoaus ot their reported i:na.eounciea. Experiments were oonduoted on the modified and unmodified Roton.duo aa ays and the u s. P. XI assays of Ph nobarbital Sodium u s. P, n. Both type of Botondaro aasa.ys riled markedly in duplicating the amount ot phenobe.rbita.l recovered by the u. s. P n a.asay Thi bdioated tbe existence ot several possibilit1 sa a) the original sample s p rtially decomposed, b) deoomposi t1on took place during the modified and unmodi fied Ro tondaro ssays, e) incomplete extraction, and d) failure of the as ay to sept.rate the deoomposltion products quantitatively without dmixture, Iner se in the number of shaking -out with chloroform tailed

PAGE 156

151 to produce e. recovery of phenobarbital oomparable to that obtained in the u s. P XI as a.y Therei'ore modified R otondaro e.ssa.ys ere carried out o n samples 0 fhenobarbital u. s P XI and samples ot phenobarbital previou ly obtained from other Rotondaro aasa.ys. The amount of pheno barbital reoovered was greatl y below the uount ta.ken. The pr senoe of poaa1b1liti: s b, o, and d were indicated. Rotonda.ro ( 72) claimed that 98 % phenobarbital could be r oowred by means of his aaaay s ffr&l methods of aasa.y employing ilwr nitrate were compared Th methods designated. by Viebiek and Fuchs (95 96) and by Shu l ek and Roz (100, 101) wre not consiclel"ed applicable beoaus of difficulty experieneed in determining the end point when assayin g phenobarbital ao ... d1wn. The Budde assay yielded re ults within t h e u s. P XII tolerance. The chem! tt"y of' the Budde us y s investigated In support of th claims me.de by the ori g inator, re ult wre obtained which indicated tl'at a silver deriw.tive of phenobarbital is formed during t h e. aay whioh is soluble in the presence ot sodium carbonate and which is weakly disso oie.ted in this medi Additional tests indicated that up to the end point the addition of silwr nitrate to a solutio n of phenobarbital odi\111 in t h e preaenc of sodium carbonat. 1s not accompanied by the formation of silver carbonate or silver oxide but rather by t h e formation ot a silver barbiturate derivatiw Thia a ppeared to be explained by removal of sil'Yer ions .from the field ot reaction due to their combination with the barbi .. turate to torm a weakly di &ooiatecl coapound Experiments were pert'ormed to leam whether or not the presence of decomposition produ c t would interfere with the Budde assay The tite r

PAGE 157

152 0 a saturated solution of phenylethylacetylur was found to be so low that no significant c,.rror would be introduced. Similar resultG were obtained on titrating fl solution of phe,nylethylaeetylurea which had been heated to produce the maxi m um amount of phenylethyla.cetic a.cid and urea possible under the conditions of the present in'flU!ltiga.tion. The Budde assay yielded oonaiatent results in the numerous aasays wh1ohwere per formed during the experiJ!:l.ente on the etteot of heat on solutions ot pheno barbital sodium. liere the average of all the results on per o nt ot phenobarbital sodium in the o loulated dry samp l e used to prepa.re ea.oh solution was 97 77; ,, with an awn.ge. difference between &says of o 3 % These figures apply to both lots of P henobarbital Sodium u. s. P. XI used in the present atudy A deterion.ted solution ot phenobarbital sodium wae assayed by the modified Rotondaro method, the u. s. P. XI method and the Budde method T h e res.ul ts dittered Wid ly The u. S P XI yielded resul ta which wore too low beeause the deOQmpoaition products a.nd the phenobeu-bita.l re ex .. tn.oted as one. The results yielded b y the modified. Itotondaro aasa.ys were concluded to be too hif;ll ~cause it had been previoualy indieated. that the recovery of phenobarbital was incomplete in tM.e uaay It was concluded that the Budde method gives results that are approximately correct or thQ.t are suff'ioiently accurate for d.etermiaing the relative degree of deteriontion in oompru-a.tive tests Effect of Heating So l ut1omof Phenobarbital Sodium u. s. P XI The effeot of such factors as time. temperature, and oonoentra.tion were studied with regard to th per cent deterioration, pH change, change

PAGE 158

153 in r i'ractive index, a.nd volume of precipitate which occurred i n solu tion of phenobarbital sodium The temperature and perioda of exposure were elected to encompa.as most of the oondi tions previously employed. b y other workers and to include, for the moat part the otfici&l condi tions of sterilization r aommended tor solutions intended. for parenteral use An insight into the deterioration which might take place during prolonge d storage was afforded by means of theee ao.oelerated test. It wa.s .found the.t an increase 1n tempen.ture was a.coompani d by a CU"OP 1n pH ., an increase in amount or preo1pitate, and an illorease in pereenta.ge deterioration o n comparing solutions of equal ooncentration, heated for one period ot time. Keepin g the temperature ooute.nt and 1noreaeing the period o f xposure produoed 1mile.r change Madsen ( 3) found an increue in pH -drop an d an increaae in peroenise.ge deterioration on beating lo% solutio n s of barbital sodium C&aza.ni (16) reported that olutions ot amytal sod1\ID displayed a decrease in pH on aging Bailey (4) found that greater amounts o f precipitate were formed on incre i ng the temperature o r solutions of be.rbi tal sodium and on prolonging the period of hea.t1ng Keeping the temp ra.ture and the time oonsta.nt ., the percentage deteriora tion generally bowed little change on increase in conoentra.tion or the ori g inal solutio n under the oo n di tiona studied This -.a 1n agi'e nt with the f indings or Tomski &nd l/ aller (24) A might be expeoted the refractive indes underwent a de c rease with ea.eh inol'e&ae in amount ot pre cipitate

PAGE 159

154 Rel ti'ln ot @ to Det~ni9l"ation ot ~olutio~ of Phe~~bi:!!f.A Sodium From study on the d terion.tion ot solution of Phenobe.rbU:al Sodium u. s. P. XI, it W&S tound that th pH level et 6, 10, and 2()% w / solutions ot thi s drug were pH 9 S, 9.7, and 9 8 respeotively hra..se.in and Vitali (21) stated that the pl of l.0% solution 9-.5 No mention wa found in th~ l1te~tU1"e ot the p1t at Whieh phenobal-b.ttal u pl"eolpita-ted from solwtions or phenobubita.l i,odium. Upon &ddin.g .,...-S.Oue amounts ot phosphate buffer to 6% olut ion, a precipitate e f"irat noted e.t pH s.a In ccnti:nnatio-n of the finding Qf Bera.aa.1D and Vitali. (2.3), it -.. tetmd that a. deoreue in am,umt ot pn0ip1te.te a e'ddeat on heating e.olutiona which had a lowr pH than an unbuttered elution. L1keW'1 e Nielsen ( 13) atatc that the rate of hydrolyid,a ot olutions ot ph no'barl;,i tal odium 1'19.: d&OX-fUed when the~e an inore 1n hydroge ion eono ntration Tho d oc,rea.ae in the amouut or _pneipita.te to d on heating solutions buttered With monoaodlum and d1aod1\2111 phosphate us mt autttcient to rr nt conolulion that any :marked stabilization had b een brought about, '?he use ot sod:i,n oa.rbonate in amounts 1uuft1eient to comp l etely neutralie-e th QmO-\Ult of phenobarbital re tllted in mixtuffa ot lowr pH, The per c nt deterlGX'a'tion was only slightly decreased: ther by .Value ot Se-vera.l Sug,etted $te.bill z 5 A1ea,a ~erime.ts -.re c eaducted. on the stabili:ty of Elixir of Phobub11ia. l tr s. P XI and acn,,n.l of its modifica:~iona Little or no deterion. tion found on hes.ting sol.uhions which differed from th ot'tieial elixil' 1n the deletion of 'the tin ture and th solution oir amaranth or 1n solut1ona

PAGE 160

ditt'ering in t}t-, deletion Gt the two a.torieto mentioned illgredienta to gether with th syrup. The condition ot heating we..100 c. tor15 minutes and 115 c. tor 30 minute Although the percentage deteriora tion of the ott1o1al lixir,_ with and without 0.4% methenam\ne., oould not be determined by the Budde a...sa.y:., no _pr,cipite.te evident il,1't:er subjecting the pr&pe.n.tions to th oonditions mentioned.HI,) r ao de1'1n1:te oonol usion was drawn oonoerning their sta.btll~ b oe.use of the possible olution ot the d oo.m.poaitio.n products in the aleohol pr a t. Eli Lilly and COlnpB.lV (39) claimed th t stable olutiona of ~l could be made by d18aolving the d:rug in hyd.roalcoholie mixtun oonte.bung thenemine:. The per nte.ge deterioration or a 10% w/v olutiDn 0 phe.m barbital sodi eollte.hd,ng 46% w/v thyl aloohol showed markedly 1 pe,r eent aeterion.t1on the.n t t sh01m. by solutions contaiaing 60% wfr glycerin or solutions contaia1ng alcohol end glycerin.. Th s-e soluti.o~ were heated a.t U6 9 c. tor 30 minute BalcU {68) t.ted that stable solutions of phenobe.rbital sod1ua could be ma.de by d:1asolving th dr\\g in a. mixture of equal pe.rt ot glyce rl.n e.nd wa.ter. Bacel'fK>n 1 loppanyi and Linegar (M) employed 10% w/vea.ch or g lycerin, alcohol, nd anti pyrine as at b1li Jing g nte for 10% solutions of th odi\lm. ult of eooJ.ielar:y uvl bromallyl ba.rbi1nUl'ate A solution of 10% ph81'10bani't41 sodium was prepared according to th~r tormul.&.. The Budde a.a ay could not b carried out but it. wa .found that no pr oip1te.t was tormed atter heating at 100 c. for 16 minutes a.nd only e.ppro:.dma.tely l co. preoipi tate per 100 co. after hee,.1d,ng at 116 o. tor 30 minutes-. No oonoluaion was drawn coneerning 1 ta ata.bi.U:ty

PAGE 161

lit ~ use ot 60 % /T de:atrou in l.Of. w / v ,ol.utions ot pheaob&r'h1tal ,od1a Wi8at1af'a.otory 4 .' ue \o the .formatio n ot a. 'bl'Ollll oolol' on he ti ng at 100<> o. tor l.i mlnutes and on boa.lug at 116 8 c tor IO mi.nut

PAGE 162

167 VI. BIBLIOGRAPBY (1) F1eoher and von MeriJlg; Therap Gegemr., l9a&.i through reteino (19). (2) Steenhauer, A,. J., Phann. Weekblad 6.4, 1154-6 (1927)1 through C A. 22, 30 1 (1928). (3) Jladaen e. J. Tott, Dan k Tide Fann.., 8, '2 71 (1934h through Pharme Ztg., Tl., 289 (1934) (4) Bail Arthur E., Phann. J., 13 6, 620-1 (1936). (5) F1eoh r, Emil and Dilthey, illNd; Ann, fil, 334.68 (1904). (6) German Pa.t.nt 11.4,431, 1903, TOD S.eaeen 'broth re, through reter noe (4) (7) Aspeluna, Helie and Skoglund, Lennart, Noti blad, 46 81, 98 \1937)a th.rough Quart J. Ph&.nl Pharmaool,., 11,72'91 2 (1938) (8) Aapelund. Helge Aota cad,. Aboen i Math et Phys., .!2, 20 PP (1936>, through c A ., ll., 6630 (1937). (9) Aapelund, H lge and Lindh Lennart Acta Ao ci ,. A'boeneis lfath. t ~ .!!, 19 PP (1937h through C. A., 680l (1939). I (10) lfieleo, Le Dan.ak Tide. r l 137 !,I (1933h through ret re (72). (ll) U.el en L o Dulak Ticla Fara., 11, 1;, (1937), through reterenoe (72). (12) W etzel, through ret rence (17). (13) 11el n, Leo Danek Tida ,. Farm ., 1 137 52 (1933), through c. A., 'Zl, 5146 {1933h Jour. A Ph A ., 22, 2~ (1933h ref rence tr5). (14) (15) (16) Aap lund Helge and S k.ogluncl, Lennart Aota Aoad A.been 11, Math et Pqs., .!2_, 20 PP {1937h through c. ., ,1, 6634 5 (1937). Wood.ward W ,. A ., Pha J,., 135, 199 (1935). Ca11ani Ugo "Ipodermat rapia", 2nd d., Ind u atrie Gr tich Stucohi Milan 1939, PP 293 94

PAGE 163

158 (17) Ibid., PP 302 10 (18) nard; Pierre Bull oo. phann. Bord aux J.!, 11.,4 (1938) ( 1 9) ttsodi 1" (circular) Bli Lilly t\nd Co Indianapolis (n d ) t.6 PP (20) "Pentobarbi l :odi Lillytt (olroular) 1i Lilly d co ., Indianap l1 (n d ) 30 PP (2 1 ) "Pe :tothal s dium (ciroular) Abbott Laboratorie lonh Chicago (n 1t) 11 PP(22) (23) "The lie~ k In dex 5th e el Merck and C Im, Rahway W J 1940 P 421 D ra ain Bayd. e u and Vl t 1 1 R o tor H (1 939) (24) Toma H W e.nd e.ller L J ., Ph J ., 1 39 421 ( 1 937) (25) Heg l d J. K A ., Ph We kblad '[!, 128 9 (1935h through C A ., !2_ 2659 (1935) ( 26) "The Bri ti h Ph 7 1 3 PP opoeia 1932 Constable & Co Ltd London (27) van Leant F J ., Pha lf kbl d 'Ji, 906 (1936), through Quart J Phann Pna ri: 11a oo l., 2, 741 ~6). ( 28) Sti c h, Conr a d Phan!. Z tg J2, 313 (1934) (29) iuhkopt, Ban B r ., 73 938 (194oh throu Squibb Abatr Bull 11. 1550 (1940) .(30) Soh l er, r and Torbo r c ., Deut Apoth Zt ., 22. 646 53 (1938). (.~l) An Sehw 1 a. Ap th Ztg ., 2!, 93 5 (1923h through C. 17 2171 (1923) (32) An o n ., Le Med.1 e 1n D o 192 1 1 through An o n ., P tion r Feb ., 1922 1481 throu h c ., _!!, 2756 (1922) ; Phum ., 1 08. 112 (1922). (33) Brtti h Pe.tent 271 637 1 through I' ter ne (16) (34) Blok, c J ., Pha iY ekbl d 72 1121 5 (1935) through Quart J Phana Phe col 2, l47\ 1 936h c A ., l2, 236 (1936) (35) chu 1 t Pharm ekb la cl 1274 J through reterenc (17) (36) u s P t e nt 1 956 609 :y 15 1934 Walt r G Christiana n and Alfred B Jurist (to E R Squibb d Sona), through c. A 28 4540 (1934)

PAGE 164

159 (37) Xightingale Dorothy d Morn U01d c., J. Am Ch Soo. .2!,. 1469 70 (1936). (31) Italian Drug Importing co. Inc.. (c talog) 1937..S, New York, p 22.(39) Amyta1 and Ellxil" .Amytal", ( iroulv) 1111 Lilly and co. India.na polia (n 4 ) 23 PP (40) Germ.an Pat nt 688,604 F b 15, 194<), Reinnolcl GnterJ thl"ough O A ., ~ 3651 (1~) (41) Vancl V lde Ab J J ., Nat.uunr Tijd1ohr 15, 192 8 (1933), through C. A 26 1584 5 (1934) (.42) Britt h Patent J25 847 ov 30 1928, I o. Farbenin4J threu h C. A 4121 (1930) (43) Pr noh Pa.tent 677_ 597 June 28, 1929, I. G. Farbenind.J through 0. A ,24., 3~7 (1930) (44) French Pat at 182 424 June 4, 1935, E ooh a J'e.brlk oh u phanl praparate Pa s auJ through c. A., B.2, 6904 (1935). (45) Autrian Pat nt 140 431, Jan. 25 1935 Eg e hemia Fabrik ch und ph pr parate PatzauJ throu h c. 29 3761 2 (1935) (46) British Patent 211 77 1, June 1., 1923 Soc. Anon pour L Ond Chim. a laleJ through o. A ., 18 1880 (1~). (47) .Dani h Pharmacopoe1aJ through Anon., Jour A. Ph. A S 79 (1933) (48) Z wilcker, J J t ., Phann Weekbla4, 72, 1273-4 (1935), through Ce A i! 8 11 ( 1 936) (49) Bi l ien, Leo Danek Tida F ., 1 1, 133 9 (1937) 1 'through O. A 31 6411 (1937), tr s P. Ab~ Sci Lit 1937. 29. (50) 11e l aen Le o, Dul e k Tide Fam ., 11. 65 (19'7h through Phana. Z'bg ~ 5~ ( 1 937 ) Phann 'D"str ., l 330-7 (1937) (51) (52) (53) Raldeten li et a l., J Pharmaoo l., ~ 326-35 (1934). Briti h Patent 538 927 Aug 21 1941. J through c. A., 3631 ( 1 942) British Patent ,381 0 12, De c. 4 1931 F Hottman Le.Roche ancl Co through C A ., 27 3291 ( 1933)

PAGE 165

160 (54) Ha s leton Lloyd w ., op~ !h odore, a.n4 Lin ar., Cluu'lea Jour A.. Ph A .. 4 ;6 ( 1940 >~ (55) u Paten 2 ., 067 3 1 7 Ju 12, 1937, Heinrioh G:rub J"J thr o ugh c A ~ 1555 (1937) ,. (56) Gel'JUD P a tent 630 705 June 4 19'6a through a,. A. 11 21 8 (1937). (57) G l"lll&ll Patent 670 089 Ju. 11 1 939 Hainri.oh Gruber, thN>ugh a A. ll, 3078 ( 1 939) (56) Page IMine Li and Corylloe, Pol J Pharmaoel ., !I. 189-200 (1 92 6) (59) "rOn O ttingen w F and J1rouoh E A ., Bull. oo phann ., 12., ( 1 932) through r teren o e (2J) (6o) B ritish Pated 384. 1 70 Apr 29 1931 Flo r ene Forbing (to I G P Mnind), th r ough C' A ., !l., 3292 (1933) (61) u. s. Pat nt 1 11 984. 753 Dee 18, 1935 John w Forbi (to lnthNp (62) (63) Ch C .), through C, A ., 29 889 (19'5) z A. G ., Jour A. Ph A ., 28 41 21 (1939). Braun H erbert and Cartland, G eorge F ., Jour. A Ph, A 25 74'> 9 (1936) (64) Anon ., Drug. r Ae oo ., Bull 42 104 l (Jun l 1942). (65) tJ s Pat e nt 2 067 ,18 ,. Jan 1 2 1937, H inrloh Gruber, throu g h c A ., il, 1555-6 (1937) (66) German Patent 650 7~ 1, cot. 5, 1937, I q FU'bem.nd., throu g h c. A ., ~ T;O ( 1 938) ( 6 7) Anon., ha J 135 2~8 (1935). (68) B aldi, Giovanni Boll. Chim. Fe.rm !5., 545 (1925h through refer e n ces (16 23). (69) Bu h Mi l ton T ., Diokineon n L ., and Laa on, Paul D J Pha 001 ., (70) (71) ]2., ~9 (l~) T h Pha o o p o eia of the United St tea or eria: a Eleventh Decen nial R.-vtaion Mack Printing Co Be.awn Pa ., 1 936 6 76 PP Tu Phe.rml.copo ia of the United States of America", elt-th R T1 ion ck Pri n ting Co ., Ea ton~ Pa ., 1~, 880 PP (72) lleton daro, Felice A ., J Asao Official Agr. C h ._. m 82 (1940)

PAGE 166

161 (73) 9 '1'h tion&l. o ul ", S1:nh Edition, ok Printir:,.g Co., Ea ton P 1935,. 556 PP (74) van ltalU II L. St nhauer A. J., P kblad 1062 (1921), through o. A., 12., 3363-4 (1921). (75) (76) (77) oani, M azzi Rend. da adunnnae dell' aead. d. tis. 10r nt1 Sperlm nt l, Tl, 178 9 (1923h R1T Ont. Olin 24, ; (1923), through o. A. 18, 6 9 (1924) Gl vt, C, I., Je A e, ofti f.e.1 Agr, Oh 8, 47 9 (1924) "Oftioi l and fentli 1 M th d of Analysis or the A e 1 tion of Oftici l Agrloultllre.l Oh t ", Washington, n. c ., 1936, P 582 (78) "Th W :tion l Formuhry: S venth dition aok Printing Co., Easton, Pa., 194,2, 690 PP (79) llont1gnie, E Bull soc ehim., 2,. 373 (193,h through ret renc (83). (80) ouge.ult and Gulllo com.pt r cl ., ~ 463 ref r c (83) (1931), throu h (81) Fleury, P ul Dull oo obi ., '21,. 165' ence (6;). (1925}, 'through J"efer(82) J l& r, Pha We kbl&d, 68 975 (1931h through refer no (83). (83) 11, C W ., Jour A. :Ph, A ., 240 (1941,). (84) Joen on, August, Fa e'Vy, 1930, 32 .51 through Ph ~g. 12., 1101 (1930)1 Pharm J., 125, 431, (l930)a u. s. Pe Ab tr. Soi Lit., 1930, 15,. (85) Soltta L Ukrain. o darat, Ia t. Ekaptl (Kh.arko ) terialy, 1939, o 7, 216J through c. A., K.onsul tatsiomcy-e i!, 2999 (1942) ( 6) -von &bit s ., Pharm. :Monatsh, Kl, 8 7 (19.36) (67) Gervay Vilma gya. Gyogyezeralt 4. '?arsaaaf' Ef't 1 1.toye, 17, ,67 .. 74 (194].)J through C. A ., ~ 7651 (1941). (86) (89) Bel nit tka D. s ., Ukratn. Go1ud r t In t. 1:mtapt1, arm. (B'.hal"kov) on ul tata ioanye aterialy, 1939, No. 4 108 -111 through C Ae, 3629 (191..2). rin, Oh lf ., J, pharm chim., 22 59 (1935).

PAGE 167

162 (90) Enell He nrie z. Anal. Chi ,22., 452 (1916) (91) l':li-ynakowski Kenstanty and. Modr1ejnsld Feliks, 1adm.o oi ., 62 Wif 9, 441 3, 457 60 (1935h through c. A., 30, 1183 1T93 6 ). (92) Ion H otiu, z. J pharm ebim !2, 551 (1932). (9~) Budde, Hans Apcth, Ztg., 49, 295 (1934). (94) Rupp,& and. Poggendcrt', A ., Areh Pha.rm 269 607 (1931). (95) .. Vl bo o k F and Fuohs K ., Pham Br., 10 Ho. 1, 5-6 (19'5)J thl'Gugh Chi.Tide und hldu trle, .2.!, 6'f11 C A., !2, 8233 1935) (96) ft Viebo o k P and Fuchs, x Pha.rm Uonatsh ., .!2, 39-40 (19J4). (97) 'l'omski, H. w. and Wall r, t. J., Ph J., .!22,, 421 (1937) (98) Kali1.1owsld imierz. Wie.dmoaoi F ~. 63; .5, 647 9 (1935), through c. A., 30, 3946 (19J6). (99) Fuchs, z.., Soientie. Pham., 1934, No. 91 through Phann. Zentral h&.lle 76, 193 (193;) (100) chul k, El rand Rolla Pal MAc,ar Gyegyazer &tud. Taraasag rt i toj 108 { 19'& >, through e. ~ 3901 (1938). (101) Schul k l ar and Ro 1 1a, Pal, Z, Anal Ch ;f 404 (193B)s through Squibl> Abstr. Bull 1 1 1157 (l~ (102) Zwikk .. J.J. L Phe.na Weekblad, 68 975-83 (1931h threu h refer ne (104). (10,) Zwikker J J. L Phe.rm eekblad .22., 1178 refer ne. (104). (104) Snell. Fo 'ter .e and Snell, Con,.elia 111., "Colorimetrle theds or Analyaia", :o Van otrand C o ., lne. l'iew Yol"k,, 19'7, 815 PP (105) Oett l, u A roh. Ph&rm., !'li, 1-10 (19'6). (106) Kopp&n1'1, Theodore t al Jour. A h. A., ., 1074--9 (1934). (107) Mitt l staed t, s o Hom GertN d e M and Kaufman, K. L ... J. Ae Ph. A 29. 183 (1940).

PAGE 168

163 (108) Wood, Hor tio c,. et 1 The Diepen a ry t the t1nited. State, of A'ln ncn l "; Oentemdal (22nd} lldltion, J. B. Llppincott C pey, Phila elpbla, 19J7, 1894 PP (109) lflla., illi J., "Ph& l"ll&Otttt1oa1 1Jisp sing 2nd 7t:J4 PP tion. 1941,

PAGE 169

1 64 BI O GRAP H ICAL IT EM S Bernard B Jatul was born on January 16, 1918 in Boston M assachu setts Unde r graduate Studies Re graduated from the M assachusetts College of Phtu in June 193 9 and attended the 193 8 summer session at Ha r vard Unive rsity. Graduat e Studies The ster ot Science degree was oonterred by Purdue University in Au g ust 1 940 M r Jatul attended the Unive rsity of F lorida from Septembe r, 1940 to Feb ruary, 1943 E xperience .B: e served as a G ra d uate Assistant in Ph armacy while pursuing his studies at the Unive rsity of Florida H onors. At the M assachusetts College of Ph armacy he received the M c Kesson and R obbins Sch olars hip the Scholarship of M u Chapte r or Kappa Psi the Leon c E llie P rize, and the Kappa Psi G rand Couneil Scholarship Key He served as P reside nt of the Iota C h a p ter of the Rho Chi Society at the Unive rsity of Florida durin g 1 940 941

PAGE 170

This disee r tati o n was p repared under the direction of the C h airman of the candidate's S u p ervisory Co mm itt e and h as been ap p roved by all me m be r s of t h e Co mm ittee It was u bm itted to the G ra du ate Council an d was a p proved as p artial f ul f il m ent of the re qu i re m e n ts for the de g ree of D octor of Ph ilosophy D ate ;J;Jr s; / 9 1/ J I S UP E R V I S OR Y C O MM ITT E E r ~ };~ c ~C c/' ~ 1J H Dean


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EG2MX38XM_Q3K6KQ INGEST_TIME 2018-07-06T20:23:07Z PACKAGE AA00062994_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES