Determination of Diphenhydramine and Certain Related Compounds by Ultraviolet Spectrophotometry JACK E. WALLACE, JOHN D. BIGGS, and ELMER V. DAHL Forensic Toxicology Branch, USAF Epidemiological laboratory, lackland Air Force Base, Texas A rapid method for determining diphenhydramine and some related compounds permits analysis for those substances in biological materials in the presence of other alkaline drugs without preiiminary separation. In the procedure, acid hydrolysis of the ether linkages in the compounds i s followed by oxidation of the resulting secondary alcohol to yield steam-distillable products which, in contrast to the original compounds, have a high molar absorptivity and a well-defined absorption maximum in the ultraviolet region. The method is specific for a group of compounds which have a diphenylmethyl ether functional group and are extractable as bases. It is sufficiently sensitive to determine concentrations of the drugs in urine after therapeutic doses.
T
ANTIHISTAMISES have a wide medical use. Most of them have a low toxicity but they do, nevertheless, frequently demand attention in a forensic toxicology laboratory and there is a need for analytical methods which permit specific identification and quantitative determination of these compounds in biological materials. This report concerns a spectrophotometric method of analysis for diphenhydramine (Benadryl), [a-(benzohydryloxy)-N, N-dimethylethylamine hydrochloride], and the related compounds bromodiphenhydramine (Xmbodryl) and diphenylpyraline (Diafen). Dimenhydrinate (Dramamine) differs from diphenhydramine only in that it is the 8-chlorotheophyllinate rather than the hydrochloride of the same organic base, and will not be considered separately. Each of the compounds has a structure which includes a diphenylmethyl ether functional group : HE
a
CH-0-R
Many analytical methods have been devised for determining these compounds. Some of the methods are suitable for assaying the drugs in pharmaceutical preparations. Others are sufficiently sensitive for use on biological materials but lack the specificity
necessary for identification of the substance measured. Thus, many nitrogenous bases other than the drugs in question give a positive reaction in the colorimetric method of Dill and Glazko ( 7 ) , in the modification of that method for fluorometric assay ( 6 ) , and in the nonaqueous titration techniques described in several reports (8, 12, 13, 17,18,22,22). Various chromatographic means have been adapted for determining the antihistamines. Paper chromatography (1, 2 , 4, 11, 15, 2 4 , thin layer chromatography (5, Q), and gas chromatography ( 1 O , l 4 , 16) have all been used to separate, identify, and quantitate these compounds. The chromatographic procedures are sensitive, but many of them lack an ability to provide the positive identification from biologic samples which is required in a forensic toxicology laboratory. Diphenhydramine and its related compounds absorb ultraviolet radiations weakly, and for that reason ultraviolet spectrophotometric methods are not sufficiently sensitive for the direct determination of these drugs in therapeutic concentration in biologic materials (3, 23). Further, direct ultraviolet spectrophotometric methods are subject to interference by other alkaline drugs which absorb in the region of 260 mg. The method to be described is rapid, sensitive, and is group specific for diphenhydramine and those related compounds which extract as bases and which have a diphenylmethyl ether functional group. Preliminary separation from other alkaline drugs is unnecessary. I n the determination, acid hydrolysis of the ether linkage is followed by oxidation of the resulting benzhydrol to benzophenone in acid dichromate. The benzophenone is steam distilled from the reaction flask for final spectrophotometric assay. The method and its results will be described for diphenhydramine, with notation of the differences which apply when it is used for bromodiphenhydramine. EXPERIMENTAL
Apparatus. A Beckman DK-2A ratio - recording spectrophotometer with linear presentation, settings and operation routine, was used for the
ultraviolet absorption measurements. The sample path length was 10 mm. throughout. A Beckman IR-4 double beam infrared spectrophotometer was used for infrared spectral characterization of the reaction products. These were purified and prepared for infrared analysis by extraction of an appropriate steam distillate with methylene chloride, evaporation of the organic solvent in a rotary vacuum evaporator, and suspension of the residue in anhydrous potassium bromide. Procedure. Ten-milliliter amounts of whole oxalated blood or serum, or 20- to 50-ml. amounts of urine containing the compound or suspected of containing the compound, are placed in a separatory funnel to which 5 ml. of 1 N NaOH and 200 ml. of ether are added. This mixture is shaken vigorously for 5 minutes, after which the organic and aqueous phases are allowed t o separate. The lower aqueous layer is removed and discarded. The ether layer is transferred from the separatory funnel and is filtered through Whatman No. 541 filter paper into a graduated cylinder. The volume of ether recovered is recorded. Complete recovery of the ether should not be attempted, but a correction for the lost ether must be included in the final calculations. Ten milliliters of 1N H2SOI are added to the filtered ether. This mixture is shaken for 5 minutes. After the ether and aqueous layers have separated, 9 ml. of the aqueous layer are transferred to a 1000-ml. round-bottom flask to which a steam distillation assembly can be connected. Ten milliliters of 0.1N potassium dichromate in lo& HzSOa are added to the flask. Distillation is initiated by allowing steam to pass into the intact assembly. The distillate should appear at a rate of approximately 5 ml. per minute. Fifty milliliters of distillate are collected and read against a blank a t 257 mg except for bromodiphenhydramine, which is read a t 267 mp. The blank should be a distillate obtained by subjecting 10 ml. of blood or 20-50 ml. of urine containing no drug to the same procedure outlined above. If the spectrophotometric reading indicates a concentration of reaction product in the distillate equivalent to less than 2.5 pg. of drug per ml., then the following concentration procedure should be carried out. Fifty milliliters of distillate are extracted by vigorous shaking with 10 ml. of methylene chloride, in which the reaction products are preferentially soluble. After the VOL. 38, NO. 7, JUNE 1966
831
RESULTS AND DISCUSSION
I ,
I
1
a iM
,
220
\
\,-----2.0
I60
1w
280
310
3 . 0
1
360
Y,III*tCIOUO
Figure 1 . Uitraviolet absorption spectra of diphenhydramine 10 pg. per ml. of water and of the steam-distilled diphenhydramine reaction product corresponding to a sample of equivalent concentration
phases have separated, 8 ml. of the methylene chloride are placed in a 125ml. flask which is then attached to a rotary vacuum evaporator. The solvent is evaporated by vacuum while the flask rotates a t room temperature, with no heat applied. The vacuum is released as soon as the methylene chloride is completely removed. The residue is dissolved in 2 , 4,or 8 ml. of water, depending on the absorbance previously observed for the unextracted distillate. The content of diphenhydramine or related drug in the sample is calculated from a standard curve prepared from aqueous solutions of the drugs in appropriate concentrations carried in identical fashion through the procedures B hich have been outlined, including the concentration technique for solutions containing only small amounts of drug.
Table 1. Standard Curve of Steam Distilled Diphenhydramine Reaction Product
Diphenhydramine Absorbance in sample, of steam Absorbance Mg./ml. distillate concentraGZ 20 15 10 5 2
1.37
0.66 0.33 0.12
Table
Diphenhydramine added, p g . / m l . 50.0 25.0 10.0
5.0 2.5
Average recovery
832
0.068 0.067 0.066 0.066 0.060
1,oo
II.
The ultraviolet absorption curve of the steamdistilled reaction product of diphenhydramine is well defined. I t has an absorption maximum a t 257 mp and a minimum a t 228 mp. The high absorbance a t 257 mp, illustrated in Figure I , provides the increase in sensitivity of the present method over that of techniques which rely on direct ultraviolet spectrophotometric analysis of unconverted diphenhydramine. Bromodiphenhydramine yields a steamdistillable product whose ultraviolet absorption curve has a maximum a t 267 mp and a minimum a t 233 mp (Figure 2). The ultraviolet absorption curves of the steamdistilled products of diphenhydramine and of diphenylpyraline are identical except that the
Figure 2. Ultraviolet absorption spectra of bromodiphenhydramine 15 pg. per ml. of water, and of the steam-distilled bromodiphenhydramine reaction product corresponding to a sample of equivalent concentration
curve for diphenylpyraline yields lower values per unit concentration of drug in the original sample. Inability of the method to distinguish between diphenhydramine and diphenylpyraline is considered to be no serious defect. The medical use and pharmacologic effect of the two drugs is similar. Absorbance of the steamdistilled reaction product has a linear relationship, a t the levels investigated, to concentration of drug in the original
Recovery Studies of Diphenhydramine
Recovery, mean f std. dev. ( p g . / m l . ) Whole blood Urine Homogenized liver 45.7 i. 1.1 23.5 + 0 . 8 9.0 f 0 . 2 4 . 4 f 0.2 2 . 2 f 0.1 90.27,
ANALYTICAL CHEMISTRY
48.5 f 0 . 7 23.4 f 0 . 5 9.8 + 0 . 3 4.6 f0.2 2 . 4 Z!C 0.2
46.0 f 1 . 1 2 2 . 5 =I= 0.9 9.0 f0.2 4.7 f0.2 2.3 f0.2
95.270
91.570
SEMlCARBAZONl
ZtO
I40
I(i0
XXI
320
140
0
UILLHIIC"0"S
Figure 3. Ultraviolet absorption spectra of the semicarbarones of steamdistilled reaction products of antaroline and diphenhydramine, each derived from aqueous solutions corresponding to 20 pg. of drug per ml.
sample. Table I shows this relationship for diphenhydramine. Table 11, summarizes the recovery of diphenhydramine which had been added in known amounts to whole blood, urine, and homogenized liver. The average recovery was greater than 90% in each case. I t four adult human males, each of whom received 100 mg. of Benadryl by mouth, the concentration of diphenhydramine in the urine was found to be 0.55 to 2.90 (average 1.36) pg./ml. in samples collected a t the end of the first 4 hours following ingestion of the drug. Samples collected a t the end of the second 4-hour interval contained from 0.56 to 2.09 (average 1.36) pg./ml.; and those from the third 4-hour interval, 0.11 to 2.63 (average 1.73) pg./ml. Samples from the interval 12 to 24 hours after taking the drug contained from 0.37 to 3.51 (average 1.59) pg./ml. While the diphenhydramine concentration in the individual samples varied considerably from person to person, the total amount excreted in the urine by the four subjects during the first 24 hours was less variable: 1.38 to 1.59 (average 1.49) mg. Other drugs, including numerous antihistamines, which might be given to patients together with diphenhydramine or its related compounds were investigated for interference. Each substance was added to whole blood and carried through the procedure outlined. Only antazoline produced a steamdistillable product with significant absorbance a t 257 mp (Table 111). That product had an ultraviolet absorption maximum a t 247 mp and a minimum a t 222 mp. Antazoline (Antistine) is not used internally; it is used topically as eye drops. If, under unusual circumstances, there is a question of the presence of this drug, it can be determined quantitatively, even in the presence of diphenhydramine, by conversion of the steam-distilled antazoline reaction product to a semicarbaaone which has an ultraviolet
W A V E L E N G T H IN MICRONS
W A V E L E N G T H I N MICRONS
KBR WAFER
5000
4000
3000
ZOM) 1800 1600 1400 W A V E N U M B E R IN C Y - I
I200
IO00
100
600
$00
Figure 4. Infrared absorption spectrum of diphenhydramine hydrochloride, 2 rng. in 400 mg. of potassium bromide
absorption maximum at 278 ing. For this specific determination of antazoline, 4 ml. of the steam distillate and 1 ml. of 0.5-11 semicarbazide in phosphate buffer at p H 7.0 are permitted to react a t 100' C.for 10 minutes:, a time sufficient to achieve equilibrium. The reaction niisturt is cooled and read a t 278 mg against a lilank prepared by the addition of 1 inl. of distilled water to 4 nil. of the steam-distilled reaction product. S o significant semicarbazone formation by the diphenhydramine product takes place under these conditions (Figure 3). Sulfuric acid in the acid of choice for extraction and hydro1 cedure. Other acid specificity of the method by permitting hydrolysis of linkages besides that of the dil)henylniethyl ether group concerned here. Compari-on of the infrared absorption spectra of diphenhydramine (Figure 4) and of it.; stcamproduct (Figure 5 ) shoi diff ereiiccs, aniong them a strong carbonyl abborption band at 1660-1680 cm.-l which is present in the product but not in the parent compound. This carbonyl band is in the approximate region reported for the stretching vibrations of carbonyl groups with two aryl group< directly attached (19, 20). Conjugation of the carbonyl group with aromatic rings would account for the increased ultraviolet absorptivity of the steam-distilled Ixoduc ta. The closc similarity of the ultraviolet and infrared absorption spectra of diphenhydramine (Figures 1 and 4) with
Table 111.
4000
3000
2000
1800 1600 ,400 WAVENUMBER IN CM - I
,200
1000
800
6OG
Figure 5. Infrared absorption spectrum of the steamdistilled diphenhydramine reaction product, 2 mg. in 400 mg. of potassium bromide
Compounds Investigated for Interference with the Determination of Diphenhydramine"
Absorbance of steam distillateb
Compound Diphenhydramine Bromodiphenhydramine
1.37 0.Z 0.62 0.01 0.01 0.02 0.01 0.92 0.01 0.01 0.02 0.01 0.02 0.01 0 06 0.05 0 01 0.01 0.09 0.01 0.01 0.04 0.01
(267 mp)
Diphenylpyraline Blank hcetylsalirylic acid Ami trip tyline .4mphetamine Antszoline (247 mp) Atropine Azacyclonal Bucliziiie Caffeine Cap todiamine Carhinoxamirie Chlorcyclizine Chlorothen Chlorpheriiramine Clemizole Cyclizitie Iloxylamine Ephedritie Hydroxyzine Lidocaine
those of benzophenone (Figures 6 and 7 ) suggests that the reaction product is benzophenone. Its formation may be thought to include acid hgdrolysi- of the ether linkage followed by oxidation of the resulting secondary alcohol to a carbonyl grouping. Calculations from the ultraviolet absorption curve4 indicate that 967, of the diphenhydramine in the sample is converted to benzophenone. It is assumed that bromodiphenhydramine is converted to a parabromo derivative of benzophenone.
Compound lleclizitie lleperiditie IIeprohaniate Methyl Plienidate Norphitie Pheiiiiidami tie Phetiiraniitie Phetiyrnniidol Pi lo carpi t i e Pipradol Prochlorperazine Propoxypheiie Theiiyldiiiniiiie Thoiizylamiiie Tripeleriiiamitie Triprolidine Each compoiind was tletermiiied from whole blood, to which it had heeii :idded. The level iti the distillate correywiid. to a conceiitration of 20 pg. (Jf ttte compoiuid per ml. of sample. E:ic.h vnliie is the average of three c1etermiii:it ioirs. Read at 257 mp unless otherwise indicated.
LITERATURE CITED
(1) Baruffiiii, A., Farmaco ( P a ~ z a )Ed. Prat. 13, 459 (19.53). (2) Belle>, Q. S.,Sievert, II. K., J . Lab. Clzn X e d . 46, 624 (1935). ( 3 ) Bratid,tatter-Kithiiert, AI., Hoffman, R., Senti, 31., Jficrochern. J . 7, 357 (1963). ( 4 ) Ciirry, A. S., Powell, I€., .\-due 173, 1143 (1954). ( 5 ) Ilaly, J. IT., Cochin, J., J . Pharrmcol. Exp.?'hemp. 139, 160 (1'163). (6) Dill, \V. A., Ghzko, A . J., Federation Proc. 21, 269 (1962).
WAVELENGTH IN MICRONS
100
E 2
90
80
; C 0 0
'.I
E d
--%o
BENZOPHENONE
70
SO
: 50 40 30
Y '
1;.
' Go '
A 0
20
d o
MILLIMICRON,
,",OD
Figure 6. Ultraviolet adsorption spectrum of benzophenone, 10 pg. per ml. of water
1000
3OW
2000 I800 I600 1400 WAVENUMBER It4 C Y - I
1200
1000
800
630
Figure 7. Infrared absorption spectrum of benzophenone, 2 mg. in 400 mg. of potassium bromide VOL. 38, NO. 7, JUNE 1966
833
(7) Dill, W. A., Glazko, A. J., J. Biol. Chem. 179, 395 (1949). (8) Ekeblad, P., J . Pharm. Pharmacol. 4, 636 (1952). (9) Fike, W. W., Sunshine, I., ANAL. CHEM.37, 127 (1965). (10) Fontan, C. It., Smith, W. C., Kirk, P. L., Ibid., 35, 591 (1963). (11) Goldbaum, L. R., Kazyak, L., Ibid.. 28. 1289 (1956). (12) Jasinski, T., Acta Polon. Pharm. 14, 45 (1957). (13) Kashima, T., Kano, K., J . Pharm. Sac. Japan 76, 931 (1956).
(14) Kazyak, L., Knoblock, E. C., ANAL. CHEM.35, 1448 (1963). (15) Kitamura, M., Okajima, S., Yakkyoku 6, 837 (1955). (16) MacDonald, A., Pflaum, R. T., J. Pharm. Sci. 52, 816 (1963). (17) Mainville, C. A,, Chatten, L. G., Ibid., 53, 154 (1964). (18) Murata, T., Ochiai, T., Kumamoto Pharm. Bull. 4, 15 (1959). (19) Randall, H. AI., Fowler, R. G., Fuson, N., Dangel, J. R., “Infrarei Determination of Organic Structures, p. 16, Van Nostrand, New York, 1949. (20) Rasmussen, R. S., Tunnicliff, D. D.,
Brattain, R. R., J . Am. Chem. SOC. 71, 1068 (1949). (21) Rink, M., Riemhofer, M., Mitt. Deut. Pharm. Ges. 31, 197 (1961). (22) Selles, E., Rodriguez, A. M., Galencia Acta (Madrid)9, 33 (1956). (23) Sunshine, I., Gerber, S. It., “Spectrophotometric Analysis of Drugs Including Atlas of Spectra,” p. 39, Chas. C Thomas, Springfield, Ill., 1963. (24) Vecerkova, J., Sulcova, M., Kacl, K., Pharmazie 17, 22 (1962). RECEIVEDfor review January 5, 1966. Accepted March 21, 1966.
Ultra Sensitive, Specific Method for Cyanide Using p-Nitrobenzaldehyde and o-Dinitrobenzene GEORGE G. GUILBAULT and DAVID N. KRAMER Defensive Research Department, Research laboratories,
b p-Nitrobenzaldehyde reacts with cyanide specifically to give an active reductant, which is capable of effecting the reduction of various compounds to form highly colored products. The intermediate cyanohydrin reacts with o-dinitrobenzene to give a highly colored purple compound, the dianion of o-nitrophenyl hydroxyiamine, or with triphenyl tetrazoiium chloride to produce a red dye, triphenyl formazan. Because cyanide is regenerated, a catalytic reaction occurs, with corresponding increase in sensitivity. Adding isonitroso benzoyl acetone, as little gram of cyanide per as 1.3 X mt. of total solution (3 nanograms total CN-) may be detected. Only p-nitro- and p-cyanobenzaldehyde of over 3 0 aldehydes tested react at an appreciable rate to form the reductive cyanohydrin. Only cyanide of more than 3 5 anions tested is detectable by this procedure.
I
PREVIOUS paper ( 5 ) , we described a specific method for the detection and determination of cyanide using various quinone derivatives. Further experiments were conducted, in attempts both to improve the sensitivity of detection and to develop a water soluble system which might be applicable to enzymatic reactions. Feigl (2) reported that a variety of reductants was capable of effecting the reduction of 1,2-dinitroLenzene(III) in the presence of alkali to o-nitrophenyl hydroxylamine. This reaction served as the basis for qualitative spot tests for such compounds as hydroquinone, ascorbic acid, benzoic acid hydrazide, and reducing sugars, See Reactions (1) and (2). It was thought that a catalytic reaction involving aromatic aldehydes and cyanide would result in a benzoin N A
834
ANALYTICAL CHEMISTRY
U. S. Army Edgewood Arsenal, Md.
condensation, whose products would reduce o-dinitrobenzene. The formation of the characteristic blue color of the o-nitrophenylhydroxylamine dianion (IV) could then be used to monitor the reaction. Aldehydes incapable of undergoing the benzoin condensation due to substitution by electron withdrawing groups (p-Br>p-NO.J, would not, therefore, be expected to participate in the reduction of o-dinitrobenzene (3, 4). Nevertheless, it was found that pnitrobenzaldehyde, in the presence of cyanide ion, rapidly effects this reduction. p-Nitrobenzaldehyde readily forms a cyanohydrin ( 1 ) but does not undergo the benzoin condensation. Preliminary experiments indicated that this reaction can serve as the basis for a sensitive, specific method for the detection of cyanide. The cyanohydrin produced (11) reacts with o-dinitrobenzene(II1) to give the highly colored blue compound(1V). Because cyanide is regenerated, a catalytic reaction occurs, with increased sensitivity. Isonitrosobenzoyl acetone(1BA) serves as a further source for the production of the catalyst cyanide, permitting the determination of as little as 1.3 nanograms of cyanide. Only
p-nitro and p-cyano benzaldehyde of over 30 aldehydes tested react a t a n appreciable rate to form the cyanohydrin, and only cyanide of over 35 anions tested is detectable by this procedure. No other ions interfere in concentrations up to 0.1N. EXPERIMENTAL
Reagents. All experiments were performed with reagent grade chemicals and pure solvents. All anions tested were added in the form of the C.P. sodium or potassium salt. p-Nitrobenzaldehyde and o-dinitrobenzene solutions, 0.1M, were prepared by dissolving the C.P. compounds (Eastman Organics, Rochester, N. Y.) in methyl cellosolve. Triphenyl tetrazolium chloride, 0.1M solution in trir (hydroxymethy1)amino methane buffer, O.liM, p H 8.5, was obtained from the Nutritional Biochemical Corp. Isonitrosobenzoyl acetone solution, 0.02M, was prepared by dissolving the pure compound [prepared in these laboratories by the method of Wolf (IO)] in pH 8.5 borax buffer, 0.1M. Preparation of Pure o-Nitrophenyl Hydroxylamine Monoacetate. Two grams of o-dinitrobenzene (Aldrich Chemical Go.) (0.012M) was dissolved in 50 ml. of tetrahydrofuran with
---
H-C=O
IT 0
I1
XV
