Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

High resolution gas chromatography of conjugated estrogens with glass capillary columns Pillai, Gopalakrishna 1981

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1981_A6_7 P54.pdf [ 4.67MB ]
Metadata
JSON: 831-1.0095073.json
JSON-LD: 831-1.0095073-ld.json
RDF/XML (Pretty): 831-1.0095073-rdf.xml
RDF/JSON: 831-1.0095073-rdf.json
Turtle: 831-1.0095073-turtle.txt
N-Triples: 831-1.0095073-rdf-ntriples.txt
Original Record: 831-1.0095073-source.json
Full Text
831-1.0095073-fulltext.txt
Citation
831-1.0095073.ris

Full Text

HIGH RESOLUTION GAS CHROMATOGRAPHY OF CONJUGATED ESTROGENS WITH GLASS CAPILLARY COLUMNS by GOPALAKRISHNA PILLAI Pharm., B i r l a Inst i tute of Technology and Science, 1 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in the Faculty of Graduate Studies Faculty of Pharmaceutical Sciences Div is ion of Pharmaceutical Chemistry We accept th i s thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA January 1981 © Gopalakrishna P i l l a i 1981 I n p r e s e n t i n g t h i s t h e s i s i n p a r t i a l f u l f i l m e n t o f t h e r e q u i r e m e n t s f o r an advanced degree a t the U n i v e r s i t y o f B r i t i s h C o l u m b i a , I a g r e e t h a t t h e L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e and s t u d y . I f u r t h e r a gree t h a t p e r m i s s i o n f o r e x t e n s i v e c o p y i n g o f t h i s t h e s i s f o r s c h o l a r l y p u r p o s e s may be g r a n t e d by t h e head o f my department o r by h i s o r h e r r e p r e s e n t a t i v e s . I t i s u n d e r s t o o d t h a t c o p y i n g o r p u b l i c a t i o n o f t h i s t h e s i s f o r f i n a n c i a l g a i n s h a l l n o t be a l l o w e d w i t h o u t my w r i t t e n p e r m i s s i o n . Department o f P Wcwrm dce.^JnccS ^C\-e,r>ceS The U n i v e r s i t y o f B r i t i s h C o l u m b i a 2075 Wesbrook P l a c e V ancouver, Canada V6T 1W5 Date J ^ n - 2 L 2 L . I °[<Z\ DE-6 (2/79) i i ABSTRACT The pharmaceutical product, conjugated estrogens, is a mixture of ten or more equine estrogen steroid conjugates used as a replacement therapy to t reat estrogen def ic iency symptoms associated with menopause. This complex mixture, containing mainly sodium estrone sulphate and sodium equ i l i n sulphate, is derived from pregnant mare's urine. The United States Pharmacopeia spec i f ies a co lor imetr ic method for the estimation of estrone, equ i l i n and the to ta l estrogen steroid content and includes a gas chromatographic i den t i f i c a t i on test for the minor components. Many attempts have been made to resolve and quantitate each and every component in the mixture, however, most of these methods suffer drawbacks.in terms of e i ther incomplete resolut ion or long e lut ion times. In order to achieve a better reso lut ion, espec ia l ly of the pair estrone/equil in and 3-estradiol/a-dihydroequi1 i n , c ap i l l a r y column gas chromatography was invest igated. A var iety of der ivat ives of estrogens, suitable for gas chromatography, such as t r ime thy l s i l y l ethers, methoxime-trimethyl-s i l y l ethers, te r t .buty l dimethyl s i l y l ethers, heptafluorobutyrates, methoxime-heptafluorobutyrates, and ox ime-tr imethy ls i ly l ethers have been chromatographed on selected columns of varying po la r i t y . A short OV-225 column (7 m x 0.25 mm) did resolve estrone and equ i l i n as the i r t r ime thy l s i l y l der ivat ives but resolut ion of g-estradiol from a -dihydroequi l in could not be achieved. Increasing the column length to 50 meters further improved the resolut ion of estrone and i i i equ i l i n but the two d io l s did not resolve. The l a t t e r pair was well resolved as the i r heptafluorobutyrates and ter t .buty l dimethyl s i l y l ethers on an OV-225 column but estrone overlapped with equ i l i n in the former and equ i l i n overlapped with g-dihydroequil in in the l a t t e r case. The ox ime-tr imethyls i ly l or methoxime-trimethylsilyl der ivat ives of the ketosteroids have been noted to form syn- and anti-isomers as evidenced by two peaks on OV-225 as well as S i l a r 10 C c ap i l l a r y columns. However, a difference in the chromatographic character i s t i c s of the oximes was observed when the two columns were compared. The resolut ion of the methoxime-trimethylsi lyl isomers was far greater than the ox ime-tr imethyls i ly l isomers on a 25 m S i l a r 10 C column. This led to the use of a shorter column (15 m) to further suppress the resolut ion of syn- and anti-isomers without a f fect ing the resolut ion of other steroids. A baseline resolut ion of nine equine steroids as the i r ox ime-tr imethy ls i ly l ethers has been achieved on a 15 m S i l a r 10 C glass c ap i l l a r y column in a chromatographic time of 28 minutes. Employing the methodology developed, a var iety of commercial formulations, such as preparations meant for intravenous administrat ion, tablets of varying strength, vaginal cream and combination products with meprobamate and methyl testosterone have been analyzed. The estrogen conjugates present in these formulations were hydrolyzed to the i r free phenolic form by incubation with sulphatase enzyme under ca re fu l l y contro l led conditions. These steroids were then extracted into chloroform solution containing the internal standard, ethynyl e s t r ad i o l . The mixture of keto and d io l steroids were then der ivat ized to the i r oxime-i v t r i m e t h y l - s i l y l ethers by sequential treatment with hydroxylamine hydrochloride and N , 0 - b i s - ( t r imethy1s i l y l ) tr if luoroacetamide in pyr id ine. For quant itat ive purposes, response factors and l i n e a r i t y were established using at least s ix data points for each of ten stero ids. The high resolut ion achieved with the cap i l l a r y column enabled precise quantitat ion of each steroid as well as a common impurity, equol. The method compares favourably with previously published methods and in add i t ion, has the advantages that i t i s considerably more rap id, completely resolves a l l the known steroids and requires only a s ingle in ject ion of a dual der ivat ive. TABLE OF CONTENTS v Page ABSTRACT i j LIST OF TABLES v \ \ \ LIST OF FIGURES i x LIST OF ABBREVIATIONS xi i I. INTRODUCTION 1 1-1. Therapeutic considerations 2 1-2. Estrogens and cancer 3 1-3. Analyt ical methods 6 A. Colorimetry 6 B. Paper chromatography 9 C. Thin-layer chromatography 9 D. High performance l i q u i d chromatography 1 0 E. Gas l i q u i d chromatography 1 1 I- 4 . Glass c ap i l l a r y gas chromatography 1 6 A. Column technology 1 7 B. Injection techniques 2 0 I I. EXPERIMENTAL 2 2 I I- l . Materials and supplies 2 2 1 1 - 2 . Methods 2 5 A. Glass cap i l l a r y column coated with OV-101 2 5 B. Glass cap i l l a r y column coated with 0V-17 2 7 Page C. Glass cap i l l a r y column coated with OV-225 28 D. Glass c ap i l l a r y column coated with OV-275 and with a mixed phase, 0V-275+0V-225. 29 E. Glass Cap i l la ry column coated with S i l a r 9 CP 30 F. Quartz c ap i l l a r y column coated with S i l a r 9 CP 30 11-3. Deri vat i zat ion 31 A. T r imethy l s i l y l ethers 31 B. Methoxime-trimethylsilyl ethers 31 C. Heptafluorobutyrates 31 D. Methoxime-heptafluorobutyrates 32 E. Tert-butyldimethyl s i l y l ethers 32 F. Oxime-tr imethyls i ly l ethers 32 11-4. A. Preparation of standard solut ion of steroids 33 B. Preparation of standard mixture of steroids 33 C. Cal ibrat ion curves 34 11-5. Enzyme hydrolysis of conjugated estrogens derived from a tab let formulation • 35 11-6. Assay of f in ished dosage forms -• 36 A. Tablet formulations 36 B. Injectable formulation 37 C. Vaginal cream formulations 38 D. Combination products 38 1. Conjugated estrogens with methyl testosterone 38 2. Conjugated estrogens with meprobamate. 39 Page III RESULTS AND DISCUSSION ^ 0 I I I - l . Coating of cap i l l a r y columns in the laboratory ^ A. Glass c ap i l l a r y column coated with 0V-101 ^2 B. Glass c ap i l l a r y column coated with 0V-17 ^ C. Glass cap i l l a r y column coated with 0V-225 ^ 9 D. Glass cap i l l a r y column coated with 0V-275 52 E. Glass cap i l l a r y column coated with -0V-225+0V-275 5^ F. Glass c ap i l l a r y column coated with S i l a r 9 CP 5b G. Quartz c ap i l l a r y column coated with S i l a r 9 CP 5 8 111-2. Der ivat izat ion studies 62 A. S i l y l a t i o n 62 B. Methoxime-trimethyl s i l y l derivatives 6 7 C. Heptafluorobutyrate derivatives 6 7 D. Tert. butyl dimethyl s i l y l derivatives 7 1 E. Methoxime-heptafluoro derivatives 7b F. Oxime-trimethyl s i l y l derivatives 7b I I I-3. Assay of Formulations 8b A. Cal ibrat ion curves for stero id standards 84 B. Sample preparation 8 7 C. Chromatographic analysis of formulations 8 8 D. Quantitation of conjugated estrogen formulations 9 9 E. Precis ion of analysis \0b F. Comparison of three GC methods for a multiple in ject ion v i a l . 104 SUMMARY AND CONCLUSIONS 1 0 8 REFERENCES 1 0 9 LIST OF TABLES Corre lat ion, slope and intercept for ca l i b ra t i on curves of steroid standards Optimization of enzyme hydrolysis of conjugated estrogens derived from a commercial tablet formulation A sample data sheet and ca lculat ions for the assay of an injectable formulation Composition of commercial formulations containing conjugated estrogens Precis ion of analysis of a mult iple in ject ion v i a l containing conjugated estrogens Comparison of three GC methods for a mult ip le in ject ion v i a l containing conjugated estrogens LIST OF FIGURES Structure of equine estrogens Structure of equol Chromatogram of the t r ime thy l s i l y l der ivat ives of estrone and equ i l i n on OV-lOl c ap i l l a r y column (4 m x 0.3 mm) Chromatogram of ' p o l a r i t y mixture' on OV-17 c ap i l l a r y column (7 m x 0.3 mm) Chromatogram of the t r ime thy l s i l y l der ivat ives of a synthetic mixture of equine estrogens on OV-17 c ap i l l a r y column (7 m x 0.3 mm) Chromatogram of the t r ime thy l s i l y l der ivat ives of a synthetic mixture of equine estrogens on OV-225 cap i l l a r y column (7 m x 0.25 mm) Chromatogram of the t r ime thy l s i l y l der ivat ives of a synthetic mixture of equine estrogens on OV-275 cap i l l a r y column (7 m x 0.25 mm) Chromatogram of the t r ime thy l s i l y l der ivat ives of a synthetic mixture of equine estrogens on OV-225+0V-275 cap i l l a r y column (7 m x 0.25 mm) Chromatogram of ' p o l a r i t y mixture' on S i l a r 9 CP cap i l l a r y column (7 m x 0.25 mm) Chromatogram of the methoxime-trimethylsi lyl der ivat ives of a synthetic mixture of equine estrogens on S i l a r 9 CP c ap i l l a r y column (7 m x 0.25 mm) Chromatogram of ' p o l a r i t y mixture' on carbowax Z0 M quartz cap i l l a r y column (50 m x 0.2 mm) Chromatogram of the t r ime thy l s i l y l der ivat ives of a synthetic mixture of equine estrogens on S i l a r 9 CP quartz c ap i l l a r y column (50 m x 0.2 mm) Der ivat izat ion reactions - T r imethy l s i l y l ethers Chromatogram of the t r ime thy l s i l y l der ivat ives of a synthetic mixture of equine estrogens on a commercial OV-225 cap i l l a r y column (7 m x 0.25 mm) Chromatogram of the t r ime thy l s i l y l der ivat ives of a synthetic mixture of equine estrogens on a commercial OV-225 cap i l l a r y column (50 m x 0.25 mm Der ivat izat ion reactions - Methoxime-trimethylsi lyl der ivat ives Chromatogram of the methoxime-trimethylsi lyl der ivat ives of equ i l i n on a commercial OV-225 cap i l l a r y column (7 m x 0.25 mm) Der ivat izat ion reactions - Heptafluorobutyrate derivat ives Chromatogram of the heptafluorobutyrate derivat ives of a synthetic mixture of equine estrogens on a commercial OV-225 cap i l l a r y column (7 m x 0.25 mm) Chromatogram of the heptafluorobutyrate derivat ives of e -e s t rad io l , a -d ihydroequi l in, estrone and equ i l i n on a commercial OV-225 cap i l l a r y column (50 m x 0.25 mm) Der ivat izat ion reactions - Te r t .buty ld imethy l s i l y l der ivat ives Chromatogram of the Ter t .buty ld imethy l s i l y l der ivat ives of a synthetic mixture of equine estrogens on a commercial 0V-225 cap i l l a r y column (7 m x 0.25 mm) Der ivat izat ion reactions - Methoxime-heptafluoro der ivat ives Chromatogram of the methoxime-heptafluoro der ivat ives of estrone and equ i l i n on a commercial OV-225 cap i l l a r y column (7 m x 0.25 mm) Structure of OV-225 and S i l a r 10 C Der ivat izat ion reactions -Oxime-tr imethyl der ivat ives Chromatogram of the methoxime-trimethylsi lyl and ox ime-tr imethy ls i ly l der ivat ives of estrone on a commercial S i l a r 10 C c ap i l l a r y column (25 m x 0.25 mm) Chromatogram of the ox ime-tr imethy ls i ly l der ivat ives of a synthetic mixture of equine estrogens on a commercial S i l a r 10 C c ap i l l a r y column (15 m x 0.25 mm) Ca l ibrat ion curve for a -e s t rad io l Chromatogram of the ox ime-tr imethyls i ly l der ivat ives of equine estrogens derived from a mult ip le in ject ion v i a l on a commercial S i l a r 10 C cap i l l a r y column (15 m x 0.25 mm) Chromatogram of the ox ime-tr imethyls i ly l der ivat ives of equine estrogens derived from a tab let formulation on a commercial S i l a r 10 C c ap i l l a r y column (15 m x 0.25 mm) Der ivat i sat ion reactions - Oxime-tr imethyls i ly l der ivat ives of Methyl testosterone Chromatogram of the ox ime-tr imethyls i ly l der ivat ives of Methyl testosterone on a commercial S i l a r 10 C c ap i l l a r y column (15 m x 0.25 mm) Ca l ibrat ion curve for Methyl testosterone xi i L I S T OF ABBREVIATIONS a - E 0 2 17a-ESTRADIOL 3 - E 0 2 173-ESTRADIOL a-DHEQ 17a-DIHYDROEQUILIN 3-DHEQ 173-DIHYDROEQUILIN.:.. a-DHEQN 17a-DIHYDROEQUILENIN 3-DHEQN 173-DIHYDROEQUILENIN EO ] ESTRONE EQ E Q U I L I N EQN E Q U I L E N I N E E 0 2 ETHYNYL ESTRADIOL WCOT WALL COATED OPEN.TUBULAR TMS T R I M E T H Y L S I L Y L t.BDMS T E R T I A R Y B U T Y L D I M E T H Y L S I L Y L HFBA HEPTAFLUOROBUTYRIC ANHYDRIDE HFBI HEPTAFLUOROBUTYRYL IMIDAZOLE MO-TMS METHOXIME - T R I M E T H Y L S I L Y L MO-HFB METHOXIME - HEPTAFLUOROBUTYRATE B S T F A N.O-BIS ( T R I M E T H Y L S I L Y L ) TRIFLUOROACETAMIDE xiii A C K N O W L E D G E M E N T The author would l i k e to thank Dr. K.M. McErlane for suggesting the problem and for his able guidance and supervision throughout the course of th i s study. The author i s grateful to Dr. Frank Abbott and Dr. Jim Orr for the i r helpful suggestions and encouragement. The author would also l i k e to acknowledge Ayerst Laboratories, Montreal, Canada for the generous g i f t of equine estrogens reference standards, Dr. B. Lodge, Health Protection Branch, Ottawa, Canada for a reference sample of equol and Mr. R. Burton of th i s facu l ty for the mass spectral analysis required in th i s work. The f i nanc ia l support provided by a grant from Natural Applied and Health Science, Univers ity of B r i t i s h Columbia i s g ra te fu l l y acknowledged. 1 I . INTRODUCTION The pharmaceutical product, conjugated estrogens, i s a mixture of nine or more stero ida l substances obtained from the urine of pregnent mares. They ex i s ts i n the form of sodium sa l t s of the sulphate esters. Conjugated estrogens entered the US market 30 years ago and have become the most widely used form of estrogens for the treatment of symptoms associated with menopause, accounting for about 80% of the estrogens administered (1). The normal decline in estrogen l eve l s , fol lowing cessation of menstruation, i s often accompanied by hot f lashes, c h i l l y sensations, sweating and muscle cramps. In more severe cases, an unbearable uneasiness may give r i se to anxiety, overbreathing, dizziness and in rare cases, women may become i nva l i d s , i f l e f t untreated. The physical symptoms of menopause include seni le v a g i n i t i s , u r e t h r i t i s , myalgia, and A r t h r a l g i a . In add i t ion, females lose the i r immunity to coronary atherosclerosis and gout and some of the endometrial changes of the cervix and breast resembles ear ly changes in certain forms of carcinomas (2). In the majority of cases administration of da i l y or c y c l i c doses of conjugated estrogens reverse the symptoms and a .sense of wel l-being has been reported by many users (3). Conjugated estrogens are avai lable in the form of tablets (0.3, 0.625, 1.25 and 2.5 mg per t ab l e t ) , mult iple i n jec t i on v ia l s and vaginal cream (0.625 mg/§). They are also used in combination with anx io ly t i c s (meprobamate), sedatives (phenobarbitone) and 2 androgens (methyltestosterone) to a l l e v i a t e some of the assoc iated symptoms of menopause. 1.1. THERAPEUTIC CONSIDERATIONS Replacement therapy with conjugated estrogens c l e a r l y re l i eves the hot f lashes and other vasomotor symptoms and atrophic v a g i n i t i s . However, the ro le of estrogens in the treatment o f post-menopausal osteoporosis i s not well de f ined. Osteoporosis i s a d i sorder of the skeleton assoc iated with the loss o f both hydroxyapatite (calcium phosphate complexes) and prote in matrix ( c o l l o i d ) r e s u l t i n g in th inning and weakening o f the bones and an increased incidence of f r a c tu re s , p a r t i c u l a r l y compression f ractures of the vertebrae. Postmenopausal osteoporosis occurs in about 25% of postmenopausal women (4). Estrogens cannot reverse es tab l i shed osteoporosis but can temporari ly a r re s t the progress o f the disease. It has been demonstrated (5,6,7) tha t , a f t e r several months o f estrogen replacement in postmenopausal pa t ien t s , the calcium balance became pos i t i ve and bone resorpt ion decreased to normal. The benef i t s o f estrogen replacement, however, may l a s t only fo r 9 to 14 months in sp i te of continued estrogen treatment (4). While some physic ians prescr ibe a small dose o f androgen along with estrogen, many prefer not to administer any androgen. Androgen enhances strength and imparts a sense of we l l -be ing and withdrawal b leeding from estrogen i s sa id to be reduced when androgen i s given concurrent ly . Large doses of androgen in women may increase 3 l i b i d o , but in the small doses used in the menopause, the e f fec t i s seen only in some cases. Excessive dosage leads to undesirable musculinizing e f fec t s , but the usual dose of 5 to 10 mg of methyl-testosterone da i l y , interrupted when the estrogen i s stopped, does not usually cause detectable v i r i l i z a t i o n . Some observers believe that the estrogen counteracts the musculinizing influence (4). Intravaginal administration of conjugated estrogens cream rel ieved vasomotor and vaginal postmenopausal symptoms. Eff icacy of the cream was measured by maturation indices of the vaginal e p i t h e l i a l c e l l s and severety rat ing of vaginal and vasomotor symptoms such as hot f lushes, skin f lu sh , sweating, vaginal i t ch ing and painful coitus (8). Improvement in atrophic v a g i n i t i s , severity of hot flushes and sweating decreased s i g n i f i c an t l y by the eighth day of therapy. However, a corre lat ion was not found between the degree of improvement and estrogen plasma levels and side effects such as breast tenderness and mild abdominal bloating l ikewise were not related to plasma estrogen levels (8). The plasma levels of estrone and p . - estnad tor "were measured in th is study by radio-immunoassay. 1-2. ESTROGENS AND; CANCER: In some species, notably mice, estrogens may cause mammary gland tumors (2) and th i s area of concern has prompted many investigators to assess the incidence of carcinomas of the breast and endometrium in humans. Several authors (9-13) have reported retrospective studies that indicate there i s a s i g n i f i c an t r i s k of endometrial carcinomas in estrogen users compared to non-users. In a recent two year study (14) the investigators noted that estrogen users had a 1-3% chance of endometrial carcinoma-development, while non-estrogen users had a 0.1 to 0.3% chance. Conf l i c t ing data have been obtained by other authors (15),who noted that there was an i n s i gn i f i c an t increase in endometrial carcinomas in estrogen users,and that estrogen therapy led to a reasonable survival rate due to ear ly diagnosis. In add i t ion, the authors reported that c y c l i c administration of estrogens and proges-terone led to an apparent protection from endometrial carcinoma as well as improvement in cardiovascular disease and hypertension. Conf l i c t ing evidence for carcinogenicity of estrogens continues to be reported in the l i t e r a t u r e and several authors have expressed reservation regarding the a b i l i t y of retrospective studies to establ i sh proof of cause and e f fec t (16-18). In a recent ten-year prospective study of a small group of postmenopausal women, no s i gn i f i c an t differences were found between estrogen users and non-users when conjugated estrogens were used on a c y c l i c basis (19). The use of intravaginal cream also has been evaluated by one group (20),who noted that no corre lat ion existed between users and non-users of conjugated estrogens,even though estrogens are well absorbed from the vaginal mucosa. One of the recent prospective studies (19) noted that c y c l i c conjugated estrogen therapy combined with Medroxy-progesterone, even at high doses (2.5 mg estrogens da i l y , 10 mg 5 progesterone) y ie lded a s i g n i f i c an t l y lower incidence of breast cancer, a lower incidence of endometrial carcinoma and an overal l higher survival rate. In addit ion to studies in humans, mutagenicity and implied carcinogenicity has been studied by the use of bacter ia l Ames test and i so lated human tissue cultures by one group of invest igators (21), who reported that no s i gn i f i c an t increase in maturation rate or chromosomal aberrations could be detected af ter exposure to conjugated estrogens. As to which s tero id in the mixture causes cancer, none may be s ingled out. Doubts have been expressed that estrone may be the causative agent (22). This i s based on the fact that estrone sulphate i s the major component of conjugated estrogens, that endometrial cancer was observed in experimental animals when estrone sulphate was used alone, and that estrone i s the major hormone produced outside the ovary during menopause. That th is conclusion may not be va l i d is suggested by metabolic studies (23)> which showed that estrone, estrone sulphate and estradio l are in equi l ibr ium in the c i r cu l a t i on and therefore there i s no conclusive evidence to support the view that estrone i s the dangerous component of conjugated estrogen preparations (24)_. 6 1-3 ANALYTICAL METHODS Conjugated estrogens, in addition to being the subject of a great deal of discussion regarding therapeutic uses and side e f f ec t s , have presented a very d i f f i c u l t problem to the ana ly t i ca l chemist. The complexity of the mixture i s due to the presence of a large number of s t ruc tu ra l l y re lated steroids (Figure 1).Al1 of these steroids have a phenolic hydroxyl group in common at pos it ion 3, but they d i f f e r from each other in the nature of subst i tut ion . as well as stereochemistry at pos it ion 17. In addit ion, they also d i f f e r in the degree of unsaturation in r ing B. B - E s t r a d i o l i s the most potent s tero id in the mixture, being 12-times more potent than estrone and 360 times more potent than equi lenin (25). The introduction of a double bond in r ing B reduces the estrogenic potency. Therefore, the overal l potency of conjugated estrogens i s not only due to tota l s tero id content, but also due to the re la t i ve amount of each component in the mixture. A. Colorimetry . For many years, the accepted method of analysis of conjugated estrogens has been the co lor imetr ic method adopted by the United States Pharmacopei.a(26).This method involves separation of the water soluble s tero id conjugates from free s tero ids , di luents or tablet excipients by adsorption on to a column of chromatographic s i l i c i o u s earth, followed by e lut ion with dicyclohexylamine acetate. The ketosteroids are then separated from the d io l s by the use of C O N J U G A T E D E S T R O G E N S /3-DIHYDR0EQUIUN tt-D I HYDRO EQUI LENIN ESTRONE EQUI LIN FIGURE 1 9H a -D IHYDRO EQUILIN /3-DIH YDROEQUILENIN EQUILENIN 8 Girard ' s reagent (trimethylacethydrazide ammonium ch lor ide) . This reagent forms a water soluble der ivat ive of the ketosteroids. The d io l s are extracted into an organic solvent. The ketosteroids are then recovered from the i r complex with the Girard reagent by ac id i f y i ng and extract ing into benzene. Estrone and e q u i l i n , the two major steroids,are estimated co lo r imet r i ca l l y by comparing the intens i ty of colour produced by Iron-Kober reagent (27) with that of standard estrone or e q u i l i n . This method i s notoriously laborious and time-consuming and would also produce i n te r fe r i ng background colour as a resu l t of the charring e f fec t of sulphuric acid on non-estrogenic impurit ies (27). Presumably, because of a lack of an accurate and s pec i f i c assay method, very broad content ranges are spec i f ied for the major components. On the basis of the co lor imetr ic assay, conjugated estrogens i s required to contain not less than 50% and not more than 65% of sodium estrone sulphate and not less than 20% and not more than 35% of sodium equ i l i n sulphate. Although none of the minor components are quant i f ied, the i r presence in the mixture i s required as evidenced by an i d en t i f i c a t i o n test involv ing gas chromatography with a packedcolumn coated with diethyleneglycol succinate (DEGS). The monograph requires a prominent peak for a -d ihydroequi l in , suggesting that th is s tero id i s the th i rd major component in the mixture. A consequence of the broad and non-specif ic requirements for both major and minor components i s that compendial 9 standards may not assure pharmaceutical equivalence. This suggests the need for more rigorous spec i f i cat ions which are possible only i f an ana lyt ica l technique i s employed that would resolve and quantify each and every component. For many years invest igators have been searching for such a method and have turned to chromatographic methods in order to resolve a l l the components. These include Paper (28-30), th in - layer (31-35), gas - l iqu id (36-43) and high-performance l i q u i d chromatographic procedures (44-47). B. Paper Chromatography Paper pa r t i t i on chromatography has been applied for the resolut ion of a -est rad io l from B-estradiol and equ i l i n from equi lenin (30). These steroids were v i sua l i zed on the paper by spraying with a reagent containing a diazotised amine. C. Thin-layer Chromatography The most se lect ive th in - layer chromatographic separations of synthetic mixtures have been that of Crocker and Lodge (33,34) which were able to resolve seven of the nine constituents of conjugated estrogens. These authors used s i l v e r nitrate impregnated s i l i c a Gel H plates to take advantage of the formation of a TT complex of those steroids containing a double bond in pos it ion 7-8. Such steroids exhibited a lower value on the impregnated plates than on p la in s i l i c a ge l . Other steroids had higher R f values on 10 s i l i c a gel G than on the impregnated s i l i c a gel H plates. A quant itat ive TLC procedure (35) has been reported to be useful for the determination of estrone, e q u i l i n , equilenin and estradio l (presumably the a-isomer) in tab let s . However, other estrogens were not accounted for in th i s study. D. High Performance-Liquid Chromatography The f i r s t high performance l i q u i d chromatographic separation of conjugated estrogens was published by Bu t te r f i e ld et al (44) in 1973. Although th i s manuscript marked the f i r s t complete separation of the known const ituents, i t was a qua l i t a t i ve analysis for standard mixtures of the equine estrogens as the i r free phenols. It was not un t i l 1976 that a quant itat ive HPLC method was reported (45). However, th i s procedure was not capable of resolving B-dihydroequilenin and 9-dehydroestrone from equ i l i n and in addition required a one meter column (ETH-permaphase) and several chromatographic manipulations. The quant i tat ive analysis of estrone, e q u i l i n , equi lenin and B -est rad io l in formulations of e s t e r i f i e d estrogens (which d i f f e r from conjugated estrogens in that they consists mainly of estrone and equ i l i n ) has l ikewise been reported using an HPLC technique (46). This method would not be capable of analyzing the more complex mixture contained in conjugated estrogens. Separation of the glucosiduronates and sulphate sa l t s of B - e s t r ad i o l , e s t r i o l and estrone in b io log ica l f l u id s has also been reported (47), however, the method was concerned with monitoring endogenous steroids and the i r metabolites only. 11 E. Gas-Liquid Chromatography Gas- l iquid chromatography has been r e l a t i v e l y more successful in providing reso lut ion, i den t i f i c a t i on and quantitat ion of most of the estrogens in conjugated estrogen substances. The f i r s t manuscript (38) to report a GLC resolut ion of standard mixtures of equine estrogens employed dual i n jec t ion of t r ime thy l s i l y l and methoxime-trimethylsi lyl der ivat ives. The resu l t ing chromatograms obtained on a polar OV-225 stationary phase exhibited separation of most of the components except e-estradiol,which appeared under the peak of a -d ihydroequi l in . This method was l a te r expanded to provide a quant itat ive procedure suitable for the analysis of the bulk drug, tab let and injectable formulations in a chromatographic time of 55 minutes (39). A combined GLC-Mass spectrometric method (36) i reported the unambiguous i den t i f i c a t i on of the estrogens. The GLC column used in th i s work was a diethyleneglycol succinate (DEGS) l i q u i d phase that provided resolut ion of each s te ro id , except B -estradio l which appeared as a shoulder on the a -d ihydroequi l in peak, as t he i r TMS der ivat ives. An a l ternat ive GLC procedure for the quant itat ive analysis of conjugated estrogen formulations (40) also used a DEGS l i q u i d phase. This method has been incorporated into USP (26) as a new i den t i f i c a t i o n procedure. As such, the monograph requires that the chromatogram exhib i ts major peaks for estrone and equ i l i n and a prominent peak for a -d ihydroequi l in and addit ional minor peaks for a - e s t r a d i o l , 3 -es t rad io l , g-dihydroequi l in, equi lenin and 9-dehydro estrone. 12 The essential differences between the quantitat ive method using the OV-225 phase (39) and the DEGS phase (40) l i e s in the i r chromatographic approach. With the OV-225 column, the tota l analysis time.was55 minutes and required two inject ions of~two der ivat ives. Under the operating condit ions, the column was extremely stable and the peaks were sharp due to the i r r e l a t i v e l y short retention times. The DEGS column, on the other hand, required a.single in ject ion of a s ingle der ivat ive. However, the time of analysis was 72 minutes, which led to broad peaks, espec ia l ly in the case of equi len in, the l a s t component to e lute. In add i t ion, the DEGS column was operated above i t s maximum temperature and was sens i t ive to oxygen.-both conditions leading to short column l i f e t imes . The authors (40) also noted that the DEGS phase must be thoroughly tested to ascertain that the pa r t i cu la r batch i s su itable for the steroid analys is . G. Carrignan et a l . , (41) have noted that an impurity in some formulations overlapped with the peak due to e-dihydroequil in when an OV-225 packed column was used. For th is reason, a benzene pre-wash was included pr io r to the enzyme hydrolysis step to remove this i n te r fe r ing substance. This impurity was subsequently i den t i f i ed as Equol (F ig. 2), an isof lavonoid compound of plant o r i g in (41). Equol (4 1 ,7- isof lavandiol) .was so named because i t was f i r s t i so lated from pregnent mare's urine (48). At that time the plant o r i g in of i t s precursor formononetin (7-hydroxy-4'-methoxyisoflavone) was unknown. It was l a te r reported in the urine of c a t t l e , sheep and goats and in the excreta of laying hens (50) . Although equol FIGURE 2 13 EQUOL 14 was o r i g i n a l l y reported as inact ive as an estrogen (48), l a te r work has shown i t to possess some estrogenic a c t i v i t y (49). This impurity was i den t i f i ed in some commercial preparations and was found at concentrations of up to 6% in some formulations of conjugated estrogens on the Canadian market (41). Neither the USP (26) nor the DEGS method (40) referred to th i s impurity. The reasons for th i s may be that procedures were developed in the United States and that formulations in that country may have lacked th i s impurity. However, i f the DEGS method was used for formulations containing s i gn i f i can t amount of equol, then th i s substance would appear as a major peak overlapping with a- and 6-dihydroequilenin. As such, the sample may not have qua l i f i ed as USP standard since the USP monograph allows only certa in minor steroids to be present in the mixture at th i s locat ion in the chromatogram. One of the GLC procedures (40) has been expanded to an in yi'vd.analysis of conjugated estrogens in the urine of postmenopausal women dosed to a steady-state level (42). The procedure was used to quantify the estrogens both before, and during, therapy with conjugated estrogens in an attempt to study the bioequivalence of conjugated estrogen products. These authors have noted that the urinary estrogen excretion p r o f i l e was s i g n i f i c an t l y d i f fe rent from the composition of the estrogens in the dosage form because of extensive metabolism. However, seven out of nine steroids could s t i l l be measured although i t was noticed that the equ i l i n type 15 compounds were oxidised to the corresponding equilenins up to 12%. Continuation of th i s study (43) in four postmenopausal women revealed s i gn i f i c an t differences in the urinary excretion p r o f i l e when the patients were dosed with d i f fe rent products. This unequal urinary excretion for products of compendial standards c l ea r l y established the bio-inequivalence of the products invest igated. In these studies, i t was not possible to measure 17g-dihydroequilenin and equ i len in , presumably due to a lack of s e n s i t i v i t y . From the foregoing l i t e r a t u r e i t appears that a s ing le , robust and se lect ive ana lyt ica l procedure for conjugated estrogens has yet to be developed. Thus f a r , no method i s ava i lable for routine d i f f e ren t i a t i on of a- and g -est rad io l s , a- and g -d ihydroequi l in, a- and. g-dihydroequilenin, estrone, e q u i l i n , 9-dehydroestrone and equilenin as well as the common impurity, equol. In add i t ion, due to the complexity of the mixture and interference from endogenous substances, no reports have appeared that o f fe r the necessary s e l e c t i v i t y and s e n s i t i v i t y for monitoring both plasma and urinary steroids fol lowing therapy with conjugated estrogens. In recent years, high resolut ion gas chromatography with c ap i l l a r y columns has been frequently employed to resolve very complex mixtures such as drugs and the i r metabolites in body f l u i d s . The high resolut ion capab i l i t y of glass c ap i l l a r y columns i s due to the very high number of theoret ica l plates that can be attained with long columns. Since packed columns could not resolve a l l the 16 components of conjugated estrogens mixtures, a l og i ca l a l ternat ive was to invest igate the s u i t a b i l i t y of glass c ap i l l a r y columns for the assay of conjugated estrogens. 1-4 GLASS CAPILLARY GAS CHROMATOGRAPHY . The invention of the c ap i l l a r y column (51) can be regarded as one of the most important developments in the f i e l d of gas chromato-graphy. In th i s technique, a long c ap i l l a r y column, usually of glass, replaces the conventional packed column used in gas chromatography. The length of such columns may range from 10 meters to 100 meters and the internal diameter ranges from 0.1 to 0.5 mm. The inside wall of the open-tubular column i s coated with a very th in f i l m of a stationary phase. Such columns are ca l led Wall Coated Open Tubular (WC0T). The absence of any s o l i d packing mater ia l , the presence of a th in f i lm of l i q u i d phase, the increased length and small diameter,and the unrestr icted gas flow through the tubular column are some of the factors that contribute to the very high e f f i c i ency of c ap i l l a r y columns. It i s common nowadays to have 150,000 theoret ical plates for a 50 meter column. The high resolut ion capab i l i t y of c ap i l l a r y columns has been demonstrated with very complex mixtures such as steroid metabolites (52, 53), a i r and smoke (54), aromatics and flavours (55) and tobacco smoke (54). Such columns are now being used more frequently to analyse mixtures which would not resolve s u f f i c i e n t l y with conventional packed columns. For example, the use of c ap i l l a r y columns has been reported 1 7 for organic acid p r o f i l i n g of synovial f l u i d (56), ac id i c metabolites in the urine (57), analysis of neurotransmitters (58), s tero id mixtures (59), human urinary s tero ids, urinary sugars, and polyols, drugs and drug metabolites (60), s tero id metabolite rat ios in women with breast lesions (61), separation of prostaglandins (62) and organic acid p ro f i l e s of human tissue biopsies (63). A. Column Technology The pract i ca l appl icat ions of cap i l l a r y GC i s somewhat d i f fe rent from that of conventional gas chromatography. The two main areas where they d i f f e r are in the column technology and sample introduct ion. Toiachieve high separation e f f i c i ency with a cap i l l a r y column, a uniform and homogeneous f i l m of stationary phase must be applied to the inner wall of the c ap i l l a r y . This has lead to problems with reproducible preparation of c ap i l l a r y columns due to poor adhesivity of organic l iqu ids on s o l i d materials. Squalane has been considered as the only l i q u i d phase which y ie lds a perfect f i lm on an unmodified glass surface. (64). Most other stat ionary phases do not create homogeneous layers and condense as droplets and hence these columns f a i l s to give the desired chromatographic performances. Appropriate surface pretreatment of the inner wall of the c ap i l l a r i e s has been reported in many cases to lead to deposition of a homogeneous f i l m of a given stationary phase. In fact, th is step has proven to be the most important factor in the technology of high e f f i c i ency cap i l l a r y columns. 18 Surface Pretreatment Surface pretreatment includes surface corrosion with ammonium hydroxide (65), concentrated hydrochloric acid (66) sodium hydroxide (67) and t r i f l uo roch lo roethy l methylether (etching ether) which cleaves o f f hydrogen f luor ide upon heating at high temperatures (68). Fine par t i c les may also be applied d i r e c t l y to the glass wa l l . Blumer (69) used Silanox 101, a s i l an i sed fumed s i l i c a , which has been applied successful ly in high temperature analysis of b io log ica l samples (70). The properties of Cab-0 -S i l , a non-si lanised s i l i c a , and some diatomaceous supports,have been studied by Deelder and co-workers (71). Other support materials,such as barium carbonate or sulphate (72),and carbon black (73) have also been invest igated. An interest ing,a l tered glass surface was reported by Schieke et al (74), who grew s i l i c a whiskers on the walls of the pyrex column. An a l ternat ive method for formation of s i l i c a whiskers was the use of ammonium hydrogen d i f l uo r ide which dissociates at elevated temperatures to l iberate hydrogen f luor ide (75). Of the several techniques described for modifying inner glass surfaces of c ap i l l a r y columns, gaseous hydrogen f luor ide has proven to be the most extensively used roughening agent, for the formation of s i l i c a whiskers. These whiskers represents filamentary crysta ls which exh ib i t high mechanical strength and an increase in the glass surface areas of up to 1000-fold. These character i s t i c s , along with the a b i l i t y to control the whisker growth rate»provide an excel lent means of obtaining high performance c ap i l l a r y columns. 19 Deactivation of the Column Inner Surface The surface roughening methods described,however, invar iab ly produce act ive s i tes on the glass surface. Unless deactivated, the ca ta l y t i c and adsorptive a c t i v i t y of the column wall would adversely a f fec t the performance of the column. The undesired a c t i v i t y of the column i s thought to be due to surface s i l ano l groups and Lewis acids such as those derived fromboron, aluminum and barium,(80). Of the many methods of deact ivat ion, the best known are s i l an i s a t i on (76), treatment with thermostable ion ic surface .active agents, such as benzyltriphenylphosphoniurn chlor ide (BTPPC) (77) and use of non-extractable th in f i l m deposits of a polar l i q u i d (carbowax) (78). In addit ion Lewis bases of low v o l a t i l i t y , such as N-isopropyl-3- azet id ino l (NIA) and triethanolamine (TEA),; have.been reported to block the act ive s i tes on the column (81). Coating of Columns In general, three methods are used to coat cap i l l a r y columns. These are the dynamic method (82), the mercury plug dynamic method (83) and the s t a t i c method (84). The dynamic method consists of forcing through the column, a solut ion of the stationary phase (5-15%), under care fu l l y contro l led flow conditions. In the mercury plug dynamic method, a concentrated solut ion (10-40%) of the stat ionary phase, i s allowed to flow through the column followed by a small length of mercury plug. The mercury plug i s useful to smoothen the coated 20 surface and to remove any microdroplets of the stationary phase. The s t a t i c method consists of f i l l i n g the cap i l l a r y column with a d i l u te solut ion of the stationary phase (0.2-0.7%), seal ing one end of the column and then evaporating the solvent by applying suction on the other end of the column while care fu l l y cont ro l l i ng the temperature of the column environment. B. Injection Technique. I t has been widely recognised that i n ject ion and evaporation of heat sens i t ive samples are c r i t i c a l points in gas chromatographic analys i s . I t would be of l i t t l e use to have an i ne r t column without preventing the sample from decomposing in the i n jec t i on port. A l l - g la s s systems have therefore been recommended. Cap i l la ry columns have very small sample capacity and therefore a s p l i t t e r i s used in the in jec t ion port to prevent overloading of the column. The s p l i t t e r has been designed in such a way that a major portion of the sample i s vented out,allowing only a small amount to enter the column. S p l i t rat ios ranging from 1:10 to 1:500 have been used rout inely in cap i l l a r y gas chromatography. In highly sens i t ive assay methods,a s p l i t l e s s mode has been incorporated to allow the ent i re sample to reach the detector. In th i s mode, the solvent traverses part of the column as a 'bead' at lower oven temperature. The oven temperature i s then increased to v o l a t i l i s e the solvent. As the rapidly r i s i n g temperature reaches the vaporisation point of the analyte, i t s tarts migrating through the 21 column in the conventional manner. This method has been pa r t i cu l a r l y valuable for the analysis of trace materials at sub-nanogram levels (85). 22 II EXPERIMENTAL 11-1. MATERIALS AND SUPPLIES Steroids and other standards 1 7 a - and 1 7 g - e s t r a d i o l , 1 7 a - and 173 -d ihydroequ. i l in, 1 7 a - and 17p-d ihydroequ i len in , estrone, equ i l i n and equi lenin were obtained as a generous g i f t from Ayerst Laboratories, Montreal, Quebec; equol from Dr. B.A. Lodge, Health Protection "Branch Ottawa, Ontario; Cholesterol from B r i t i s h Drug Houses (Canada) L td . , Toronto, Ontario; ethynyl estradio l and methyl testosterone from Sigma Chemical Co., St. Louis, Missouri, U.S.A. Reagents The fol lowing reagents were obtained from Pierce Chemical Co., Rockford, I l l i n o i s , U.S.A.: N,0-bis ( t r imethy l s i l y l ) tr if luoroacetamide (BSTFA), heptafluorobutyryl imidazole (HFBI), heptafluorobutyric anhydride (HFBA), t e r t . buty1 dimethyl -chlorosi lane (t^BDMS), 2% methoxamine hydrochloride in pyridine (MOX) and pyridine. Hydroxylamine hydrochloride was obtained from B r i t i s h Drug Houses, London, England, and Benzyltriphenylphosphoniurn-chloride (BTPPC) from A ldr ich Chemical Company, Inc., Milwaukee, Wisconsin, U.S.A. A 2% solut ion of hydroxylamine hydrochloride was prepared in 50 ml of r e d i s t i l l e d anhydrous pyridine and was stored in a polytetraf luoroethylene- l ined screw-capped tube. 23 Sulphatase enzyme, type H-2, derived from Helix pomatia was obtained from Sigma Chemical Co., St. Louis, Missour i , U.S.A. A working solut ion was prepared by d i l u t i n g the enzyme concentrate with d i s t i l l e d water to y i e l d a solut ion containing 1000 units of sulphatase a c t i v i t y per ml. The a c t i v i t y stated by the manufacturer using nitrocatechol sulphate as a substrate was used as a basis for preparing d i l u te solut ions. The enzyme and the solut ion were kept at 4°c un t i l required. An aqueous phosphate, buffer(pH 6) was prepared by combining 12.3 ml of 0.2 M disodium hydrogen phosphate with 87.7 ml of 0.2 M sodium dihydrogen phosphate. An aqueous acetate buffer (pH 5.2) was prepared by combining 21.0 ml of 0.2 M acet ic acid with 79.0 ml of 0.2 M sodium acetate so lut ion. The pH of the buffer solut ion was measured and adjusted to 5.2 i f necessary. Miscellaneous Supplies Polytetrafluoroethylene (PTFE)-lined screw-capped conical v i a l s and temperature cont ro l led , dry heating bath (Reacti-therm) were obtained from Pierce Chemical Co., Rockford, I l l i n o i s , U.S.A; PTFE - l ined screw-capped culture tubes (50 ml) from Fischer S c i e n t i f i c Co. L td . , Vancouver, B r i t i s h Columbia; and 10 y l gas chromatographic syringe from Hamilton Company, Reno, Nevada, U.S.A. Si lanox, grade 101, was obtained from Cabot Corporation, Boston, Massachusetts, U.S.A. 2k Capi l la ry Columns A l l boros i l icate ' glass c ap i l l a r y columns were of 0.30 or 0.25 mm l.D. and were obtained in 50 meter lengths to be cut to the desired length. A cap i l l a r y column wall-coated with cyanopropylmethyl phenyl s i l i cone (OV-225) stationary phase was obtained from A l l tech Associates, Arl ington Heights, I l l i n o i s , U.S.A.; a cap i l l a r y column wall-coated with cyanopropylmethyl s i l i cone ( S i l a r 10 C) stat ionary phase from Applied Science Laboratories, State College, Pennsylvania, U.S.A.; a c ap i l l a r y column etched with t r i f luoroch lo roethy l methylether (etching ether) from Dr. Pearce, Simon Fraser Univers i ty, Burnaby, B r i t i s h Columbia and a fused s i l i c a cap i l l a r y column (0.20 mm l.D.) from Hewlett-Packard, Avondale, Pennsylvania, U.S.A. Stationary Phases Cyanopropylmethyl phenyl s i l i cone (0V-275), methyl, s i l i cone (0V-101), methyl phenyl s i l i cone (0V-17), cyanopropylmethyl phenyl s i l i cone ( S i l a r 9 CP) and OV-225 were obtained from Applied Science Laboratories, State College, Pennsylvania, USA and Carbowax 20 M from A l l tech Associates, Arl ington Heights, I l l i n o i s , U.S.A. Po la r i t y Mixture A po la r i t y mixture containing the fol lowing components was obtained from A l l tech Associates, Arl ington Heights, I l l i n o i s , U.S.A.: Nonane, decane, dodecane, dibutyl ketone, tetradodecane, c i s -p ropy l -cyclohexanol, trans-propylcyclohexanol and 2,6-dimethylani l ine. 25 Dosage Forms Premarin intravenous, 25 mg v i a l s , Lot No. 552; Premarin tab le t s , 0.3 mg, Lot No. 868; Premarin Vaginal Cream; Lot No* 874; Premarin with meprobamate, Lot No. 880; and Premarin with Methyl Testosterone, Lot No. 878 were obtained from Ayerst Laboratories, Montreal, Quebec; Oe s t r i l i n tab le t s , 2.5 mg, Lot No. 354 KC from Desbergers L td . , Montreal, Quebec and Conjugated Estrogens tablets 1.25 mg, Lot No. 80766 from ICN Canada L td . , Montreal, Quebec. Instrumentation A Hewlett Packard gas chromatograph, model 5830A with cap i l l a r y i n l e t system and equipped with a flame ion izat ion detector and a GC terminal, Model 18850A was used throughout th i s work. The gas chromatograph was operated with an i n l e t s p l i t t e r . Gas chromatography/mass spectrometry (GC/MS) data were obtained on a Varian Mat 111 mass spectrometer, using a Hewlett Packard 5700A gas chromatograph and a data acqu i s i t ion system based on Varian 620/L computer. II-2. METHODS COATING OF GLASS CAPILLARY COLUMNS A. Glass Cap i l la ry Column Coated with 0V-101 A glass c ap i l l a r y column (4 m x 0.3 mm) was washed sequential ly with 10% HC1, d i s t i l l e d water, acetone and methanol and dried with a stream of nitrogen. 26 S i l an i zat i on The cap i l l a r y was s i l an i sed by forcing a plug of 5% (v/v) solut ion of dimethylchlorosilane in toluene through the column. The column was washed with toluene to remove excess reagent and then with methanol and dried with nitrogen. Deposition of the I n i t i a l Layer This step was carr ied out according to the dynamic method of coating cap i l l a r y columns ( 82 ). 0.1 g of 0V-101 was dissolved in 20 ml of chloroform containing 0.04 g of rbenzyltriphenylphosphoniurn-chloride (BTPPC). To th i s solut ion was added 0.4 g of hydrophobic s i l i c i c acid (Si lanox.W) and mixed w e l l . . . The column was f.i=rs,t-wetted with chloroform. The cap i l l a r y coating reservoir was f i l l e d with the above mixture. One end of the cap i l l a r y column was inserted into the mixture through a septum type seal and the side arm of the reservoir was connected to a nitrogen source. The coating so lut ion was then forced through the column at a rate of 2 cm/sec by nitrogen pressure. When 25% of the column length was f i l l e d with the coating so lut ion, the column end was drawn out of the l i q u i d in the coating reservoir and the l i q u i d plug was allowed to move through the column at a rate of 2 cm/sec. In order to avoid a sudden change in coating ve loc i ty as the stationary phase l e f t the column, the main column was connected to a buffer column of approximately 3 meters. A f ter the plug of l i q u i d l e f t the column, the nitrogen pressure was increased to evaporate the solvent. This step was 27 repeated 2 times to deposit a uniformly white layer of Si lanox. Liquid Phase Coating The cap i l l a r y was coated with addit ional l i q u i d phase using the same experimental arrangements as for the i n i t i a l coating. lg> of OV-101 was dissolved in 20 ml of isooctane and 25% of the column length was f i l l e d with this so lut ion. The plug of stat ionary phase was then moved through the column at a rate of 2 cm/sec. After the plug l e f t the column, the nitrogen pressure was increased to evaporate the solvent. Drying with nitrogen was continued for 3 hrs. Conditioning the Column The column was transferred to the oven of the gas chromatograph and one end was connected to the i n jec t i on port . The oven temperature was raised to 200°C at a rate of l°/min, while maintaining helium ca r r i e r gas flow through the column. The column was held at 200°C overnight, a f te r which period the temperature was raised to 250°C at a rate of l°C/min and held at th is temperature for 2 hours. B. Glass Cap i l la ry Column Coated with 0V-17 The same procedure and experimental arrangements were used to deposit a thin f i l m of 0V-17 stationary phase over the Silanox treated glass surface using a 7 m x 0.3 mm column. 28 Etching of Glass Cap i l l a r ie s in the Laboratory A boros i l i ca te cap i l l a r y (10 m x 0.25 mm) was etched in the laboratory according to a published method (75) as fo l lows: the column was f i l l e d with concentrated hydrochloric ac id , the ends were sealed and heated overnight at 80°C. After cooling to room temperature, the column ends were broken and i t was washed sequential ly with d i s t i l l e d water, acetone and ether,and blown dry with nitrogen. The column was then f i l l e d with a saturated solut ion of ammonium-hydrogen f luor ide in methanol by the same experimental arrangements used for dynamic coating of c ap i l l a r y columns. The f i l l e d column was kept for one hour a f te r which period, the solvent was removed by passing nitrogen unt i l a uniformly white f i l m was observed. The column ends were then sealed and placed in a muffle furnace at 450°C for three hours. A f ter cool ing, the column was opened in a fume hood, washed thoroughly with methanol and then dried with nitrogen. C. Glass Cap i l la ry Column Coated with OV-225 A glass c ap i l l a r y column (7 m x 0.25 mm) that has been previously etched by t r i f luoroch loroethy l methylether was coated by the mercury plug dynamic method (83) as fo l lows: the cap i l l a r y coating reservoir was f i l l e d with a 50% (w/v) solut ion of the stationary phase in chloroform. The reservoir also contained a small pool of mercury. One end of the cap i l l a r y column was inserted into the solut ion of stationary phase through a septum type seal and the side arm of the reservoir was connected to a nitrogen 29 source. The coating solut ion was forced through the column by nitrogen pressure at a ve loc i ty of 2 cm per sec. When 25% of the column length was f i l l e d with the stationary phase, the column end was lowered to the pool of mercury to draw in a 5::cm plug. The stationary phase, along with the mercury plug was allowed to flow through the ent i re length of the column at a constant ve loc i t y . In order to avoid a sudden change i n coating ve loc i ty as the stationary phase l e f t the column, the main column was connected to a buffer col-umn of approximately 10 meters. When the l a s t drop of mercury l e f t the column, the flow of nitrogen was increased to evaporate the solvent residue. After drying under a stream of nitrogen for 3 hours, the column was transferred to the oven.of the gas chromatograph and one end of the column was connected to the in ject ion port. The oven was then temperature programmed from 45°C to 250°C a ta rate of l°Cperminute and the helium flow through the column was maintained at I. rill per minute. The column was held at 250°C for 48 hours. Af ter condit ioning, the performance of the column was evaluated by i n jec t i ng a mixture of nonane, decane, dodecane, dibutyl ketone, tetradodecane, cis-propylcyclohexanol, trans-propylcyclohexanol and" 2,6-dimethylani1ine. D. Glass Cap i l la ry Column Coated with OV-275 Stationary Phase  and with a Mixed Phase The mercury plug dynamic method, as described above was used to prepare a column coated with OV-275 and with a mixed phase, consist ing of equal proportions of OV-225 and OV-275. 30 Both columns were 7 m x 0.25 mm. E. Glass Cap i l la ry Column Coated with S i l a r 9 CP The mercury plug dynamic method as described above was used to coat a glass c ap i l l a r y column (7 m x 0.25 mm) with a cyanopropyl-methylphenyl s i l i cone ( S i l a r 9 CP) stationary phase. COATING OF FUSED SILICA COLUMNS -F. Quartz Cap i l l a ry Column Coated with S i l a r 9 CP Stationary Phase. Column Deactivation with Carbowax 20 M (49). A f l e x i b l e , fused s i l i c a c ap i l l a r y column (50 x 0.2 mm) was f i r s t coated with a 2% solut ion of Carbowax 20 M in methylene chloride by the dynamic method as described above. This column was conditioned at 220° for 48 hours and the preliminary performance was tested at th i s point with the po la r i t y mixture previously described. For deactivation of the column surface, both ends of the column were sealed by a flame and the column was heated in an oven at 280° for 1:2 hours. The sealed ends were then broken, the column was flushed with methylene:chloride and dried with nitrogen. Column Coating The column that had been deactivated with Carbowax 20 M was coated with a 50% w/v solut ion of S i l a r 9 CP in chloroform by the method described above. 31 11-3 DERIVATIZATION^: A. Tr imethy l s i l y l ethers. To 1 mg of the s tero id mixture (50-150 ug of each steroid) contained in a 5 ml PTFE-lined, screw-capped conical vial was added 150y l of the s i l y l a t i n g reagent, BSTFA.and 50 u l of dry pyr id ine. The v i a l was closed and heated at 70°C for 10 minutes. B. Methoxime-trimethylsi lyl Ethers To 1 mg of the s tero id mixture (50 - 150yg of each stero id) contained in a 5 ml PTFE-lined screw-capped conical vial was added 200 p i of a 27o solut ion of methoxyamine hydrochloride in pyr id ine. The v i a l was closed and heated at 70° for 3 hours. Most of the pyridine was removed from the v i a l by a gentle stream of clean nitrogen and to the residue was added 150 u l of BSTFA and 50 ...u.l of dry pyr idine. The v i a l was again heated at 70° for 10 minutes. C. Heptafluorobutyrates. Y To 1 mg of the s tero id mixture (50 - 150 ug of each steroid) contained in a 5 ml PTFE-lined, screw-capped conixal v i a l was added 0.5 ml of dry benzene and 0.2 ml of a 0.1 M t r i e t h y l amine i n benzene, followed by 50 u1 of heptafluorobutyric anhydride (HFBA). The mixture was shaken well for 30 seconds and kept at room temperature for 15 minutes. To this mixture was added 0.5 ml of phosphate buffer (pH=6)and the v i a l was shaken for 30 seconds. The benzene layer containing the heptafluorobutyrates was separated and dried with anhydrous sodium sulphate. 32 D. Methoxime-heptafluorobutyrates To 1 mg of the s tero id mixture (50-150 yg of each steroid) contained in a 5 ml PTFE-lined, screw-capped conical v i a l was added 200 y l of a 2% solut ion of methoxamine hydrochloride in pyr id ine. The v i a l was closed and heated at 70° C for 3 hours. One mi H i l i t r e of d i s t i l l e d water was added to the v ia l and the contents were extracted three times with 0.5 ml portions of benzene. The combined benzene extract was washed successively with IN . hydroch lo r i cac id and 1% sodium bicarbonate so lut ion. The benzene solut ion containing the methoxime der ivat ive was dried with anhydrous sodium sulphate and concentrated under a gentle stream of nitrogen. To the residue was added 200 y1 of 0.1 M triethylamine in benzene followed by 50 y1 of HFBA. The v ia l was shaken well for 30 seconds and kept at room temperature for 15 minutes. To this mixture was added 0.5 ml of phosphate buffer (pH = 6) and the mixture was shaken for 30 seconds. The benzene layer containing the heptafluorobutyrates was separated and dried with anhydrous sodium sulphate. E. Tert. Buty ld imethy l s i l y l ethers To 1 mg of the s tero id mixture (50-150 yg of each steroid) contained in a 5 ml PTFE-lined, screw-capped conical v ia l was added 150 y l of t e r t . butyl dimethylchiorosilane and 50 y l of dry pyr id ine. The v ia l was closed and heated at 70°C for 10 minutes. 33 F. Oxi me-tri methyl s i l y l e t he r s : To 1 mg of the s tero id mixture (50-150 ug of each steroid) contained in a 5 ml PTFE-lined, screw-capped conical v i a l was added 200 u l of a 2% (w/v)solution of hydroxylami hydrochloride in pyr id ine. The v i a l was closed and heated at 70°C for 30 minutes. The pyridine in the v i a l was removed by evaporation under a gentle stream of clean nitrogen. To the residue was added 150 ul of BSTFA and 50 u1 of dry pyridine and the v i a l was again heated at 70°C for 10 minutes. II-4. ESTROGEN STANDARDS A. Preparation of Standard Solutions of Steroids The fol lowing stock solutions of the steroids were prepared in methanol: 1 7 a - est rad io l (100ug/ml), 17g-estradiol (100 ug/ml), 17a -d ihyd roequ i l i n (1000 u g/ml), 170-dihydroequilin (100ug/ml), estrone (1000 ug/ml), equ i l i n (1000 ug/ml) 17a-d ihydroequi len in (100 ug/ml), 17B-dihydroequilenin (100 ug/ml), equilenin (100 ug/ml), ethynyl estradio l (1000 ug/ml), cholesterol (1000 ug/ml) and methyl testosterone (1000 u g/ml). B. Preparation of Standard Mixture of Steroids A standard mixture of equine estrogens was prepared so as to contain in each ml, approximately the same proportion of each stero id as present in formulations as fol lows: 20 ug 1 7 a - e s t r a d i o l , 10ug 1 7 g - e s t r a d i o l , 100ug 17a -d i h yd roequ i l i n , 5 u g 173-dihydroequil in, 300 ug estrone, 150 ug e q u i l i n , 20 u g 3b 17a -d ihydroequi len in, 10 yg 173 -d ihydroequi len in, 50 y g equ i len in, and 200 yg ethynyl est rad io l ( internal standard). C. Cal ibrat ion Curves The fol lowing concentrations of standard solutions (A) were used to plibt 6 point ca l ib ra t ion curves for each s tero id : Ua - e s t r a d i o l - 5, 1 0 , 2 0 , 3 0 , 4 0 , 50 yg 173-estradiol - 4, 6, 8, 1 0 , 1 2 , 16 yg 17a -d ihydroequ i l in - 6 0 , 8 0 , 1 0 0 , 1 2 0 , 1 4 0 , 160 yg 173 -d ihydroequ i l in - 4, 6, 8, 1 0 , 1 2 , 20 yg estrone - 1 2 0 , 1 4 0 , 1 6 0 , 1 8 0 , 2 0 0 , 2 2 0 yg equ i l i n - 1 2 0 , 1 4 0 , 1 6 0 , 1 8 0 , 2 0 0 , 2 2 0 yg 17a -d ihydroequi lenin - 6, 8, 1 0 , 1 2 , 1 4 , 18 yg 173 -d ihydroequi lenin - 1, 2, 4, 8, 1 6 , 20 yg equilenin - 2 0 , 3 0 , 4 0 , 5 0 , 6 0 , 70 yg The amount of internal standard added to each sample was 200 yg. Representative Procedure for One Sample Preparation of Each Steroid To a 5 ml PTFE-lined, screw-capped conical v i a l was added methanolic solutions of the steroids so as to contain 5 yg 17a - e s t r a d i o l , 4 yg 173 - e s t r a d i o l , 60 yg 17a -d i hyd roequ i l i n , 4 yg 173 - d i hyd roequ i l i n , 200 yg estrone, 120 yg e q u i l i n , 6 yg 17a -d ihydroequi len in, 1 yg 1 7 3 -dihydroequilenin, 20 yg equilenin and 200 yg of ethynyl e s t r ad i o l . The methanol was removed from the v ia l under a gentle stream of clean nitrogen. To the residue was added 200 y l of a 2% so lut ion of hydroxylami hydrochloride in pyridine the v i a l was closed and heated on an:.aluminum 35 block at 70 C for 30 minutes. The pyridine was removed by a stream of nitrogen and to the residue was added 150 y l of s i l y l a t i n g reagent, BSTFA,and 50 y l of dry pyr id ine. The v i a l was again heated at 70°C for 10 minutes. From the resu l t ing solut ion containing the ox ime-t r imethy l s i l y l der ivat ives , 2 y l were injected into the gas chromatograph. Area rat ios of each steroid and internal standard were plotted against the corresponding weight ra t io s . The fol lowing chromatographic conditions were maintained throughout: glass c ap i l l a r y column coated with S i l a r 10C (15 m x 0.25 mm),injection temperature 250°C, detector temperature 250°C, column temperature, programmed from 170°C for 8 minutes to 220°C at a rate of 2.3°C/min., column i n l e t pressure 10 p s i , helium ca r r i e r gas flow through the column 0.8 ml/min, s p l i t vent 40 ml/min, make-up gas 50 ml/min, s p l i t r a t i o 1:50, attenuation 3. This procedure was repeated for each of the d i f fe rent concentrations of the pure steroids to y i e l d s ix ca l i b ra t i on points for each s tero id . II-5 ENZYME HYDROLYSIS OF CONJUGATED ESTROGENS DERIVED"FROM  A"TABLET FORMULATION: Al iquots of a tablet powder ( O e s t r i l i n , 2.5 mg) equivalent to 1 mg of conjugated estrogens were weighed into four 50 ml PTFE-l i ned , screw-capped culture tubes. To each tube were added 15 ml of acetate buffer (0.02 M, pH=5.2) and the tubes were shaken for 36 30 minutes on a mechanical shaker. To these mixtures were added 1000, 2000, 3000 and 4000 units/ml of sulphatase enzyme and the tubes were incubated in a water-bath at 45°C for 30 minutes. To each tube, 0.2 ml of a methanolic solut ion of ethynyl estradio l ( internal standard, 1 mg/ml) were added followed by 10 ml of chloroform. The contents of the tubes were extracted into chloroform by shaking the tubes on a mechanical shaker for 30 minutes. Af ter centr i fugat ion at 2000 r.p.m. for 15 minutes, the chloroform layer was separated and f i l t e r e d through a pledget of glass wool covered with a bed of anhydrous sodium sulphate. The chloroform so lut ion, containing the equine estrogens, was evaporated, in small port ions, in a 5 ml conial v i a l under a gentle stream of clean nitrogen and the residue was der ivat ised to i t s oxime-trimethyl s i l y l ethers by the procedure described before. A 2 y l a l iquot of the derivatives were injected onto a S i l a r 10 C glass cap i l l a r y column (15 m x 0.2 mm). The chromatographic conditions were the same as described under ' c a l i b r a t i on curves ' . II-6. ASSAY OF FINISHED DOSAGE FORMS A. Tablets Formulations The average weight of 20 tablets was determined accurately and an .aliquot of the tablet powder equivalent to 1 mg of conjugated estrogens was weighed d i r e c t l y into a 50 ml PTFE-lined, screw-capped centrifuge tube. To the tablet powder was added 15 ml of acetate buffer (pH = 5.2) and the contents were shaken for 30 minutes on a mechanical shaker. Af ter th i s time, 2 ml of sulphatase enzyme 37 (1000 units/ml)were' added and the tube was incubated for 30 minutes in a water-bath maintained at 45°C. To the hydrolysate was added 0.2 mg of ethynyl .est rad io l ( internal standard, 1 .mg/ml): followed by 10 ml of chloroform. The contents of the tube were extracted into chloroform by shaking the tube for 30 minutes on a mechanical shaker. After centr i fuging the tubes at 2000 r.p.m. for 15 minutes, the chloroform layer was separated as completely as possible and f i l t e r e d through a pledget of glass wool covered with a bed of anhydrous sodium sulphate. The chloroform solut ion was evaporated to dryness, in small port ions, in a 5 ml PTFE-lined, screw-capped conical v i a l under a gentle stream of nitrogen. The residue was der ivat i sed to i t s ox ime-tr imethy l s i ly l ethers as previously described. A 2 y l a l iquot of thes^ederivatives were injected onto a S i l a r l O ' C . c a p i l l a r y column (15 m x 0.2 mm). The chromatographic conditions were the same as described under ca l ib rat ion curves. B. Injectable Formulations Injectable formulations, containing the l yoph i l i sed material in a dry, buffer were dissolved in acetate buffer (pH =.5.2), transferred to a 50 ml volumetric f lask and the volume made up to 50 ml. An a l iquot equivalent to 1 mg of conjugated estrogens was quant i tat ive ly transferred into a 50 ml centrifuge tube,15 ml of acetate buffer and 2000 units of sulphatase enzyme were added and the sample was treated as before. 38 C. Vaginal Cream Formulations A sample of the cream equivalent to 1 mg of conjugated estrogens was weighed into a 50 ml PTFE-lined, screw-capped centrifuge tube, 10 ml of a 10% w/v solut ion of sodium chloride was added and the tube was shaken thoroughly to disperse the cream. The aqueous solut ion was washed with 25 ml portions of methylene chloride by shaking for 20 minutes on a mechanical shaker. A f ter centr i fuging at 2000 r.p.m. for 15 minutes, the methylene chloride phase was separated and the aqueous phase was processed as before. D. Combination Products 1. Conjugated estrogens with methyl testosterone: To an a l iquot of the tablet powder equivalent to 1 mg of conjugated estrogens contained in a 50 ml PTFE-Tined, screw-capped centr i fuge tube, was added 15 ml of acetate buffer and the contents were shaken for 30 minutes on a mechanical shaker. The contents were then extracted with three 25 ml portions of chloroform by shaking for 30 minutes on a mechanical shaker. Af ter each ext ract ion , the tube was centrifuged at 2000 r.p.m. for 15 minutes and the chloroform layer was separated. The chloroform layer was processed for methyl testosterone and the aqueous phase was processed for conjugated estrogens. Assay of Methyl testosterone: An a l iquot from the chloroform extract , equivalent to 0.5 mg of methyl testosterone, was evaporated to dryness 39 in 1 a 5 ml PTFE-lined, screw-capped conical v i a l under a gentle stream of nitrogen. To the residue was added 0.2 ml of a methanolic solut ion of ethynyl estradiol (0.2 mg) and the methanol was removed by a stream of nitrogen. The residue was der ivat ised to i t s oxime-t r ime thy l s i l y l ethers as described before. Ca l ibrat ion curve for methyl testosterone: From a stock solut ion of methyl testosterone (1 mg/ml in methanol)0.1 ml, 0.2 ml, 0.3 ml and 0.4 ml were dispensed into four 5 ml PTFE-lined, screw-capped conical v ia l s and to each v i a l was added 0.2 ml of a methanolic solut ion of ethynyl estradiol ( internal standard, 1 mg/ml). The solvent was removed by a gentle stream of nitrogen and the residue was der ivat i sed to i t s ox ime-tr imethyls i ly l ethers as described before. A 2 y l a l iquot of this solut ion were injected onto a S i l a r , 10 C glass cap i l l a r y column. The chromatographic conditions were the same as described under ' c a l i b r a t i o n curves ' . Assay for conjugated estrogens: The aqueous phase remaining a f ter removal of methyl testosterone was assayed for conjugated estrogens as previously described. 2. Conjugated estrogens with meprobamate The assay of conjugated estrogens in th is product was carr ied out by the same method as described under conjugated estrogens with methyl testosterone: The quantit ies of meprobamate were not determined. ho I I I . RESULTS AND DISCUSSION The gas chromatographic analysis of complex mixtures of b io log ic o r i g i n , such as conjugated estrogens derived from pregnant mare's urine, is best carr ied out with high resolution glass cap i l l a r y columns rather than with conventional packed columns. Ea r l i e r attempts to analyse conjugated estrogens by gas chromatography with packed columns revealed that polar stationary phases, such as diethyleneglycol succinate (DEGS) or a cyanopropylmethyl phenyl s i l i cone (OV-225) would-be required for the i r reso lut ion. At the i n i t i a l stages of the present work, cap i l l a r y columns coated with polar phases were not commercially avai lable nor had experimental methods been reported that could lead to t he i r production in the laboratory. Therefore, attempts were made to prepare cap i l l a r y columns coated with stationary phases of varying po la r i t y and these were evaluated for the i r s u i t a b i l i t y in the analysis of steroids contained in conjugated estrogen mixtures. I I I - l . COATING OF CAPILLARY COLUMNS IN THE LABORATORY In order to achieve high resolut ion and column e f f i c i ency , a uniform and homogeneous f i lm of the stationary phase must be coated on to the inner wall of the c ap i l l a r y . With a few notable exceptions, l i q u i d phases deposited on untreated glass surface usually produce unsatisfactory coatings; the cohesive forces of the l iqu ids are usually superior to the adhesive or wetting forces between the l i q u i d and the glass surface and the l i q u i d phase draws up into discrete globules and beads. Such columns would y i e l d unsatisfactory re su l t s , because the concept of a thin uniform f i l m of the l i q u i d phase, essent ial to high e f f i c i ency separations has been v io lated. This problem has led to a number of suggestions for changing the wetting character i s t ic s of the glass surface, e i ther by the addit ion of a surface act ive or wetting agents, or by changing the character i s t i c s of the surface i t s e l f . The wetting agent usually plays a dual ro le : i t lowersthe surface tension of the l i q u i d phase and decreases the contact angle between the glass surface and the l i q u i d phase, thereby encouraging wetting and spreading; i t also deactivates adsorptive s i tes and therefore reduce ' t a i l i n g ' of polar compounds. However, there are problems associated with the use of such agents because of temperature i n s t a b i l i t y and the i r e f fec t on retention character i s t i c s . Silanox pre-treatment One of the ea r l i e s t methods for surface modif ication of glass cap i l l a r y columns involved the deposition of a uniform layer of hydrophobic fumed s i l i c i c acid (Silanox ® ) 0 n the innerwall of the cap i l l a r y (70). The presence of such f i ne l y dispersed par t i c le s ( 6 - 10 u) in the walls of the c ap i l l a r y aids in the maintenance of a mechanically stable f i l m of stationary phase that w i l l not break up on heating. Thermal i n s t a b i l i t y of the stationary phase has been a major problem in preparing e f f i c i e n t cap i l l a r y columns. Investigations on the s u i t a b i l i t y of surface pretreatment with S i lanox, as i n t r o -duced by Horning et a l . , (70) formed the i n i t i a l part of th i s study 42 in the search for an e f f i c i e n t method of coating cap i l l a r y columns in the laboratory. With an apparatus assembled in the laboratory, a non-polar column (OV-101) was f i r s t prepared and i t s e f f i c iency was evaluated for estrogens stero id analys is. Although i t . was known from previous, reports employing packed columns (39, 40) that polar phases such as QV-225 and diethyleneglycol succinate resolved most of the components of conjugated estrogen mixtures, i t was decided to employ a non-polar column f i r s t because of the re l a t i ve ease with which they could be prepared and to test the e f fec t of the increased number of theoret ical plates that could be attained on cap i l l a r y columns. A. Cap i l la ry column coated with OV-101 The column preparation method consisted of an i n i t i a l s i l an i s a t i on of the glass surface to remove act ive s i t e s , which i f not removed, would cause t a i l i n g of polar substance, followed by coating the surface with a uniform layer of Si lanox. Silanox, as such,would not adhere to a smooth glass surface but i f dispersed in a viscose solvent containing the stationary phase and a surface active agent, the stationary phase w i l l act as a binder. Therefore a mixture containing Si lanox, stationary phase (OV-101) and a surface act ive agent, benzyltriphenylphosphoniurn - chloride (BTPPC) in chloroform was used to deposit a layer of Silanox on the glass surface. Addit ional l i q u i d phase was deposited by passing a solut ion of the l i q u i d phase in isooctane, through the column fol lowing b3 the dynamic method of coating c ap i l l a r y columns ( 8 2 ) . Polar solvents, required to dissolve polar stationary phases, would not be useful in th i s step as they would displace the Silanox layer that had i n i t i a l l y been deposited. This problem i s a serious l im i ta t i on to the overal l usefulness of the method. After the second step of coating, the column was conditioned and i t s performance was evaluated by in ject ing a mixture of estrone and equ i l i n as the i r t r ime thy l s i l y l ethers. This pa i r was chosen as they represented the two major steroids contained in conjugated estrogens preparations and had h i s t o r i c a l l y been one of the d i f f i c u l t pairs to resolve. As shown in f igure 3, only a pa r t i a l resolut ion of estrone could be achieved from equ i l i n and therefore no further study was done with th i s non-polar phase. B. Cap i l l a ry column coated with OV-17 In order to study the e f fec t of increasing po la r i t y of the stationary phase on resolut ion of equine estrogens, attempts were made to prepare a c ap i l l a r y column coated with a methyl phenyl s i l i cone (OV-17). Pre-treatment of the column inner surface with Silanox was repeated four times un t i l a uniformly white layer was produced. OV-17, dissolved in isooctane (5% w/v), was coated over the Silanox treated surface by the dynamic method previously described. Performance of the OV-17 column was evaluated by in ject ing a test probe mixture (po lar i ty mixture) containing n-alkanes, a lcohols, ketones, and amines, and studying the i r chromatographic characte r i s t i c s . This mixture i s designed to provide information about the inertness FIGURE 3 CHROMATOGRAM OF THE TRIMETHYLSILYL DERIVATIVES OF ESTRONE AND EQUILIN ON OV-101 GLASS CAPILLARY COLUMN 1. ESTRONE 2. EQUILIN 1 0 M I N U T E S Chromatographic Condit ions: Column, OV-lOl (4 mx 0.3 mm); Inject ion temperature, 260°C; Detector (F.I.D.) temperature, 260 Column temperature, 200°C; Car r ie r gas (Helium flow) l ml/min; S p l i t vent flow, 50 ml/min; Make-up gas (Helium) flow, 50 ml/min Inlet pressure, k p . s . i ; Chart speed, 0.5 cm/min. 45 of the column and i t s e f f i c i ency . Adsorption or t a i l i n g of the alcohols would indicate presence of active s i l o x y l groups, the peak shape of amines would indicate the presence or absence of ac id i c s i te s and the shape of the alkane peaks would indicate column and system e f f i c i e n c i e s . As shown in f igure 4, some components of th is mixture were only p a r t i a l l y resolved and there was considerable t a i l i n g of the alcohol peaks. Figure 5 shows the chromatographic character i s t i c s of the t r ime thy l s i l y l derivatives of a synthetic mixture of equine estrogens on the OV-17 column. Lack of resolut ion of three pairs of equine estrogens may be due to the poor e f f i c i ency of the column as well as poor s e l e c t i v i t y of the stationary phase. Pre-treatment of the glass surface with Si lanox therefore proved to be a non-productive exercise because of the extreme d i f f i c u l t i e s in coating a uniform layer of the stat ionary phase. Several columns were found to plug during passage of the suspension of Si lanox resu l t ing in pa r t i a l or complete obstruction and therefore such columns had to be discarded. Etching of glass c ap i l l a r y columns Recognising that an etched glass surface would have d i f fe rent wetting character i s t i c s compared to a smooth glass surface, several invest igators explored surface pre-treatment that included etching with hydrochloric ac id , hydrofluoric ac id , sodium hydroxide and ammonium hydroxide. Among these methods, gaseous hydrogen f luor ide has been the most extensively used and two methods have been described FIGURE 4 CHROMATOGRAM OF POLARITY MIXTURE ON OV-17 GLASS CAPILLARY COLUMN 46 1 2 3 1. NONANE 2. DECANE 3. UNDECANE k. DODECANE 5. DIBUTYLKETONE 6. TETRADECANE 7. CIS-PROPYICYCLOHEXANOL 8. TRANS-PROPYLCYCLOHEXANOL 9. 2,6-DIMETHYLANI LINE M I N U T E S 1 0 Chromatographic Condit ions: Column, OV-l7 (7 m x 0.25 mm); Inject ion temperature, 250°C; Detector (F.I.D.) temperature, 250°C; Column temperature, 40°C (1.5 min) to 200°C at a rate of 20°C/min; Ca r r ie r gas (Nitrogen) f low, l ml/min; S p l i t vent flow, 100 ml/min; Make-up gas (Nitrogen) f low, 50 ml/min; Inlet pressure, k p.s.i ; Chart speed, 0.5 cm/min. FIGURE 5 CHROMATOGRAM OF THE TRIMETHYLSILYL DERIVATIVES OF A SYNTHETIC MIXTURE OF EQUINE ESTROGENS ON OV-17 CAPILLARY COLUMN. 1 2 , 3 5 , 6 7,8 0 1. 17orESTRADIOL 2. 17&-ESTRADIOL 3. 17a~DIHYDROEQUILIN 4. 1 7 3-D I HYDROEQUILIN 5. 17«-DIHYDROEQUILENIN 6. 173-DIHYDROEQUILENIN 7. ESTRONE 8. EQUILIN 9. EQUILENIN 0 5 1 0 M I N U T E S Chromatographic Cond i t ions : Column, OV- l7 (7 m x 0.25 mm); Inject ion temperature, 270°C; Detector (F.I.D.) temperature, 270°C; Column temperature, 2kO°C (l min) to 260OC at a rate of l ° / m i n ; Ca r r i e r gas (Helium) flow, l ml /min;Sp l i t vent flow, l30 ml/min; Make-up gas (Helium) flow, 50 ml/min; I n l e t pressure, 10 p.s. i . . ; Chart speed, 0.5 cm/min. 48 (74, 75 ) to generate hydrogen f luor ide in s i t u . One of these methods (75) consisted of passing a saturated so lut ion of ammonium hydrogen f luor ide in methanol through the column and subsequent removal of the solvent to leave behind a f ine layer of ammonium hydrogen f luor ide . Sealing the column and heating i t at 450°C l iberated hydrogen f luor ide. An a l ternat ive method (74), which has become more popular because of the uniform etching, consisted of f i l l i n g the cap i l l a r y column with the vapours of t r i f l u o r o -chloroethyl methyl ether (etching ether) under high vacuum and heating the sealed column at 400°C. The hydrogen f luor ide gas l iberated would react with the glass surface to form s i l i c a whiskers. I t i s believed that hydrogen f luor ide probably reacts with the glass to form s i l i c o n te t ra f luor ide which i s then converted to s i l i c o n dioxide and deposited in the form of whiskers (74). Whisker columns have a number of advantages as the s i l i c a whiskers are reported to be ideal for retention of any stationary phase (92) and the increase of surface area allows for a higher sample capacity as compared to a smooth, w a l l -coated open tubular column (WCOT). Attempts to etch cap i l l a r y columns in the laboratory by the ammonium hydrogen f luor ide method were not successful because i t was not possible to control the temperature of the muffle furnace that was used to obtain the high temperature required for the etching process. Temperature f luctuat ions during etching resulted in columns that were unevenly etched and therefore exhibited poor e f f i c i ency when coated with a polar stationary phase. Since i t was also not bS possible to carry out the etching process with the use of etching ether in the laboratory because of lack of a high vacuum source (10~^ tor r ) to vapourise the etching ether, a 32 meter column that had been etched using th i s method was obtained from Dr. H. Pearce of the Simon Fraser Univers i ty, Burnaby, B.C. This column was broken into 4 short columns and the fol lowing polar stationary phases were coated: OV-225, OV-275,mixture of OV-225 and OV-275, and S i l a r 9 CP. C. Cap i l la ry column coated with OV-225 The mercury plug dynamic method (83) was used to coat the whisker-walled cap i l l a r y column (7 m x o.25 mm). This method consisted of forc ing a concentrated solut ion of the l i q u i d phase (50% w/.v) in acetone" through the column, followed by a short plug of mercury (10 cms). The plug of mercury employed in th i s technique helped to smoothen the coated surface and to remove any microdroplets of the l i q u i d phase that may be l e f t behind. The presence of any droplet in the cap i l l a r y would adversely a f fect the performance of the column. OV-225, a cyanopropylmethylphenyl s i l i c one -stationary phase was selected because of the reported (39) advantages of th i s l i q u i d phase over diethyleneglycol succinate l i q u i d phase, both of which were used for analysing equine estrogens employing packed columns. Figure 6 i s representative of the resolut ion of t r ime thy l s i l y l der ivat ives of a synthetic composite of equine FIGURE 6 50 CHROMATOGRAM OF THE TRIMETHYLSILYL DERIVATIVES OF A SYNTHETIC MIXTURE OF EQUINE ESTROGENS ON OV-225 GLASS CAPILLARY COLUMN 17a-ESTRADIOL 17B-ESTRADI0L 1 7a-DIHYDROEQUI LI N 1 7B-DIHYDR0EQ.UI LIN 1 7a-DIHYDR0EQ_UI LENI N 1 73-D IHYDROEQU I LENIN ESTRONE EQUILENIN 9 A 5 i 1 0 M I N U T E S - i — 1 5 2 0 - | — 2 5 Chromatographic Condit ions: Column, OV-225 (7 m x 0.25 mm); Injection Temperature, 260°C; Detector (F.I.D.) temperature, 260°C; Column temperature, 250°C; Car r ie r gas (Helium) flow, I ml/min; S p l i t vent flow, 130 ml/min; Make-up gas (Helium) flow, 50 ml/min; Inlet pressure, 7 p.s.i ; Chart speed, 0.5 cm/min. 5 1 estrogens on a 7 meter OV-225 column. Most of the steroids were resolved except B -estradiol and a -d ihydroequi l in . The tota l number of theoret ical plates calculated for the equilenin peak according to the formula: n - 16 £ j 2 where t = retention time of equilenin w = width of the peak at baseline n = tota l number of theoret ica l plates was 8590,and the number of theoret ica l plates per meter was 1227. Resolution obtained for the estrone/equil in pair according to the formula: R - 2 (VA " VB) a b where R = degree of separation of two components (resolution) t r A = retention time of equ i l i n t g = retention time of estrone W + W. = width of the two peaks at baseline a b ^ was 1.04. To achieve a baseline resolut ion of th is pair, defined as R = 1.5, the to ta l number of theoret ical plates required was calculated according to the formula: n „ p n = 16 R 2 ( - V ) 2 req a-1 t. 'rA where a = , R = 1.5 52 and was found to be 12912. This would predict a column; of 10.5 meters. No such calculat ions could be made for the pair B-estradio l/a-dihydro-equ i l i n since retention times for both appeared i d e n t i c a l . D. Capi l lary Column Coated with OV-275 As an a l ternat ive to a longer column coated with OV-225, i t was considered advisable to employ a more polar stat ionary phase such as OV-275, which i s also a cyanopropylphenyl methyl s i l i cone with a higher content of cyano moieties. As shown in f igure 7, i t was possible to obtain a pa r t i a l resolut ion of 3-estradiol from a-d ihydroequi l in , but the ketosteroids, estrone, equ i l i n and equilenin were not eluted up to 50 minutes. The non-elution of the ke to -steroids was considered to be due to the presence of e i ther act ive s i tes on the glass surface, or was due to the very high a f f i n i t y of th i s phase towards the ketones. (The suppl ier of th i s phase l i s t e d the retention index value, A l , for 2-pentanone as 885 and that for OV-225 as 338). Considering the high McReynold's constant l i s t e d for the ketone, 2-pentanone, i t was not unexpected that the ketosteroids would exh ib i t long retention times on the OV-275 column. However, a chromato-graphic e lut ion time in excess of 50 minutes, using a 7 meter column, was considered impractical for the present work. The fact that a pa r t i a l resolut ion of 3 - e s t r ad i o l from a-dihydroequi l in was achieved with the use of OV-275 cap i l l a r y column, however, indicated that a stationary phase, more polar than OV-225, but less than OV-275 might favour better resolut ion of the FIGURE 7 CHROMATOGRAM OF THE TRIMETHYLSILYL DERIVATIVES OF A SYNTHETIC MIXTURE OF EQUINE ESTROGENS ON OV-275 GLASS CAPILLARY COLUMN 1. 17«-ESTRADIOL 2. 173-ESTRADIOL 3. 17a_DIHYDROEQUILIN k. 173-DIHYDROEQUILIN Chromatographic Condit ions: Column, OV-275 (7 m x 0.25 mm); Inject ion temperature, 260°C; Detector (F.I.D.) temperature, 260 Column temperature, 250°C; Ca r r i e r gas (Helium) flow, 1 ml/min; S p l i t vent flow, 130 ml/min; Inlet pressure, 7 p . s . i ; Make-up gas (Helium) flow, 50 ml/min; Chart speed, 0.5 cm/min. 5k dio ls while f a c i l i t a t i n g e a r l i e r e lut ion of the ketosteroids. E. Cap i l la ry Column Coated with Mixed Phase, OV-225 + OV-275 (1:1) As a th i rd a l ternat ive i t was decided to combine the two polar phases in an attempt to make use of the r e l a t i ve merits of both the types. With the mixed phase column, i t was possible to elute estrone and equ i l i n with good resolution (R =1.2); but equilenin was unfortunately not eluted (Figure 8 ) . The resolut ion of p -e s t rad io l from a-dihydroequi l in was s l i g h t l y reduced as compared to that obtained with the OV-275 column, and was apparently due to a s l i g h t decrease in overal l po la r i t y of the mixed phase. Hence the optimum po la r i t y required for maximum resolution of a l l the components was..not achieved by mixing equal proportions of OV-225 and OV-275 stationary phases. This study indicated that inclus ion of OV-225 did favour e lut ion of the two ketosteroids, estrone and e q u i l i n . F. Glass Cap i l la ry Column Coated with S i l a r 9 CP S i l a r 9 CP, a cyanopropylphenylmethyl s i l i c one l i q u i d phase, was selected since th i s phase has a po la r i ty intermediate between OV-225 and OV-275. The po la r i t y of th i s phase was considered su itable since OV-275 was found to be too polar and OV-225 less polar for equine stero id analys is. Column performance evaluation by i n jec t ing a test probe mixture revealed poor e f f i c i ency and t a i l i n g of peaks (Figure 9). Figure 10 shows the resolut ion obtained on this 55 FIGURE 8 CHROMATOGRAM OF THE TRIMETHYLSILYL DERIVATIVES OF A SYNTHETIC MIXTURE OF EQUINE ESTROGENS ON MIXED PHASE (0V-225+0V-275) GLASS CAPILLARY COLUMN 1 1. 17O-ESTRADI0L 2. 173-ESTRADIOL 3. 1 7 a _DI HYDROEQUI LI N k. 173-DIHYDROEQUILIN 5. 17a-DIHYDROEQUI LENIN 6. 17 3-D I HYDROEQU I LENIN 7. ESTRONE 8. EQUILIN Chromatographic Condit ions: Column, OV-225+0V-275(1•1) (7 m x 0.25 mm); Injection temperature, 260°C; Detector (F.I.D.) temperature, 260°C; Column temperature, 250°C; Ca r r i e r gas (Helium) flow, 1 ml/min; S p l i t vent flow, 130 ml/min; Inlet pressure, 7 p . s . i ; Make-up gas (Helium) flow, 50 ml/min. Chart speed, 0.5 cm/min. FIGURE 9 56 CHROMATOGRAM OF 'POLARITY MIXTURE' ON SILAR 9 CP GLASS CAPILLARY COLUMN 4 6 1. 2. 3. k. 5. 6. 7. 8. 9. 8 9 NONANE DECANE UNDECANE DODECANE DIBUTYLKETONE TETRADECANE CIS-PROPYLCYCLOHEXANOL TRANS-PROPYLCYCLOHEXANOL 2,6-DIMETHYLANILINE 1 0 0 5 M I N U T E S Chromatographic Condit ions: Column, S i l a r 9 CP (7 m x 0.25 mm); l.D.) temperature, 270°C; Inject ion temperature, 270°C; Detector (F Column temperature, 80°C (2 min) to 200°C at a rate of lO°C/min; Ca r r i e r gas (Helium) flow, l ml/min; S p l i t vent f low, 130 ml/min; make up gas (Helium) flow, 50 ml/min; Inlet pressure, 10 p . s . i ; Chart speed, 0.5 cm/min. FIGURE 10 CHROMATOGRAM OF THE METHOXIME-TRIMETHYLSILYL DERIVATIVES OF A SYNTHETIC MIXTURE OF EQUINE ESTROGENS ON SILAR 9 CP GLASS CAPILLARY COLUMN. 1. 17ct-ESTRADI0L 2. 17B-ESTRADI0L 3- 17a~DIHYDROEQUILIN k. 17 3"DI HYDROEQUI LIN 1 0 M I N U T E S Chromatographic Condit ions: Column, S i l a r 9 CP (7 m x 0.25 mm); Inject ion temperature, 260°C; Detector (F.I.D.) temperature, 260°C; Column temperature, 230°C; C a r r i e r gas (Helium) f low, l ml/min; S p l i t vent flow, 50 ml/min; Inlet pressure, 7 P - s . i ; Chart speed, 0.5 cm/min; Make-up gas (Helium) f low, 50 ml/min. 58 column for the t r ime thy l s i l y l ethers of a synthetic mixture of equine estrogens. Although B -est rad io l was resolved from a-dihydro-equil i n , the ketosteriods did not elute from the column, perhaps due to act ive s i tes on the column. However, a study of the chromatographic character i s t i c s of the d io l s on th i s column indicated that S i l a r phases with a po la r i ty s l i g h t l y more than OV-225 would favour the resolut ion of the d io l s as the i r t r ime thy l s i l y l ethers. Therefore th i s phase was used again on a quartz cap i l l a r y column. G. Quartz Cap i l la ry Column Coated with S i l a r 9 CP F lex ib le fused s i l i c a cap i l l a r y columns were introduced into the market within the l a s t year. These columns were reported to be completely i ne r t , very e f f i c i e n t , thermally s tab le, mechanically strong, easy to use and su i table for analysis of a lcohols, amines, phenols and acids without der ivat i zat ion (86). The introduction of these columns also appeared to el iminate many of the problems and c r i t i c i sms associated with cap i l l a r y gas chromatography. Unfortunately, such columns coated with polar phases, considered necessary for s tero id analys i s , have not yet been made ava i l ab le . Therefore, attempts were made to coat a polar stationary phase on a fused s i l i c a surface. Because of the poor ' w e t t a b i l i t y ' of s i l icone-based polar phases such as OV-225, OV-275, S i l a r 9 CP S i l a r 10' C etc.on such a surface, i t was not feas ib le t odepo s i t a uniform f i l m of a polar stationary phase d i r e c t l y on the untreated surface. Therefore, a technique of coating with a thin f i lm of 59 non-extractable polymer matrix (79), followed by re-coating with the required polar stationary phase was adopted for the present study. The method consisted of f i r s t coating the column with carbowax 20'* M and then decomposing the carbowax by heating at an elevated temperature,. This process leaves a non-extractable polymeric material on the column surface over which polar phases could be coated. I t was also of interest to evaluate the performance of the carbowax 20M column before i t was destroyed by heating. As shown in f igure 11, a l l components of the test probe mixture were well resolved and the peaks were symmetrical, ind icat ing an e f f i c i e n t column. The e f f i c i ency of the column was further supported by the baseline resolut ion of the two isomers, cis-propylcyclohexanol and trans-propylcyclohexanol. Equine Estrogens TMS Ethers on S i l a r 9 CP Cap i l la ry Column. The chromatographic character i s t i c s of the TMS ethers of a synthetic mixture of equine steroids i s shown in f igure 12. A l l the nine steroids were p a r t i a l l y resolved on th i s column, but the resolut ion obtained for? the pair p -est rad io l/a -d ihydroequi l in was not considered sat i s factory for quantitat ion purpose. Although the e f f i c i ency was poor (n = 16,610 for a 50 meter column), the s e l e c t i v i t y of th i s phase appeared to be far greater than OV-225 phase. Since a baseline resolut ion of a l l the known equine stero ids, su itable for quant itat ive analys i s , was not achieved by employing FIGURE 11 CHROMATOGRAM OF 'POLARITY MIXTURE' CAPILLARY COLUMN ON CARBOWAX 20 M QUARTZ * K 1. NONANE 2. DECANE 3. UNDECANE k. DODECANE 5. DIBUTYLKETONE 6. TETRADECANE 7. CIS-PROPYLCYCLOHEXANOL 8. TRANS-PROPYLCYCLOHEXANOL 9. 2,6-DI METHYLAN ILINE 17 MINUTES Chromatographic Cond i t ions : Column, Carbowax 20 M (50 m x 0.2 mm); Inject ion temperature, 260°C; Detector (F . l .D. ) temperature, 260°C; Column temperature, 70°C (5 min) to 220°C at a rate o f 10°C/min; Ca r r ie r gas (Helium) flow, 1 ml/min; S p l i t vent flow, 130 ml/min; Inlet pressure, 25 p . s . i ; Make-up gas (Helium) f low, 50 ml/min; Chart speed, 0.3 cm/min. FIGURE 12 61 CHROMATOGRAM OF THE TRIMETHYLSILYL DERIVATIVES OF A SYNTHETIC MIXTURE OF EQUINE ESTROGENS ON SILAR 9 CP QUARTZ CAPILLARY COLUMN 1. 17a-ESTRAD10L 2. 17iS-ESTRADlOL 3. 1 7a_DI HYDROEQUI LIN 4. 1 73"DIHYDROEQUI LIN 5. 17a-DIHYDROEQUILENIN 6. 173-DIHYDROEQUILENIN 7. ESTRONE 8. EQUILIN 9. EQUILENIN 1 — 1 0 To— M I N U T E S To" Chromatographic Condit ions: Column, S i l a r 9 CP (50 m x 0.2 mm); Inject ion temperature, 260°C; Detector (F.I.D.) temperature, 260°C; Column temperature, 220°C; (12 min) to 250°C at a rate of 15° c /min; Car r ie r gas (Helium) flow, 1 ml/min; S p l i t vent flow, 130 ml/min; Inlet pressure, 25 p . s . j ; Make-up gas (Helium) flow, 50 ml/min; Chart speed, 0.3 cm/min. 62 cap i l l a r y columns coated in the laboratory, i t was considered necessary to invest igate commercial cap i l l a r y columns for the i r s u i t a b i l i t y in stero id analys is. In add i t ion, the studies were expanded to evaluate the s u i t a b i l i t y of alternate derivatives of equine estrogens other than t r imethy l s i l y l ethers. I l l - 2 DERIVATIZATION STUDIES Gas chromatography i s a valuable technique for analysing steroids. However, many steroidal compounds cannot be subjected to the conditions required for analysis by gas chromatography unmodified, since the functional groups would make the compound very polar and reduce the i r v o l a t i l i t y . Such a compound would exh ib i t long retention time or would suf fer thermal breakdown before e lut ing from the column. By der i vat i s ing some or a l l of the functional groups, the po la r i t y of the molecule would be decreased, the v o l a t i l i t y increased and the compound would be eluted in a reasonable time. The steroids found in conjugated estrogen mixtures are polar compounds containing e i ther two hydroxyl groups or one hydroxyl and one keto group (Figure 1). I t would be necessary, therefore, to a l t e r these functional groups to a less polar der ivat ive. A. S i l y l a t i o n A systematic study of formation of s i l y l der ivat ives of hydroxy and ketosteroids was reported by. Chambaz and Horning (87). 63 The e f fect of the nature of the s i l y l a t i n g reagent, the reaction conditions and the influence of s t e r i c ef fects on the rate of reaction and structure of the resu l t ing product was investigated in order to better define the conditions needed for sat i s factory quantitat ion of a given s tero id . Based on these resu l t s , hydroxylgroups were ranked in the order of degree of d i f f i c u l t y to form s i ly letheirs and reaction conditions to s i l y l a t e each hydroxyl were recommended. For unhindered as well as moderately hindered hydroxyl» groups, N,0-bis ( t r imethy l s i l y l ) tr if luoroacetamide (BSTFA) was found to be adequate for quant itat ive conversion at room temperature or at 60°C. In the present study , BSTFA was found to be the most su itable reagent since the by-products, N - t r i m e t h y l s i l y l t r i f l u o r o -acetamide and trif1uoroacetamide, are more v o l a t i l e than that of other s i l y l a t i n g reagents and therefore.jwould elute with the solvent. These by-products would not cause any interference in the chromatogram. Pyr id ine, used as a solvent, also acts as a cata lyst in th is reaction (Figure 13). The chromatographic character i s t i c s of the TMS ethers of a synthetic mixture of equine estrogens on a commercial OV-225 cap i l l a r y column (7m x 0.25 mm) i s shown in Figure 14. A l l steroids except g-estradio l anda-dihydroequil in were well resolved on this column. Increasing the length of the column to 50 meters, thus increasing the to ta l number of theoret ica l p lates, did not help in resolving these two stero ids, but a better separation of estrone/ equ i l i n pair was achieved. (Figure 15). FIGURE 13 DERIVATIZATION REACTIONS: TRIMETHYLSILYL ETHERS HO ESTRADIOL ESTRONE 1.3STFA 2. PYRIDINE 70 C, 10 MINUTES TMSO' OTMS TMSO ESTRADIOL DI-TMS ESTRONE TMS FIGURE 14 65 CHROMATOGRAM OF THE TRIMETHYLSILYL DERIVATIVES OF A SYNTHETIC MIXTURE OF EQUINE ESTROGENS ON COMMERCIAL OV-225. GLASS CAPILLARY COLUMN. 6 1 2 ,3 1. I7a-ESTRADIOL 2. 173-ESTRADI0L 3. I7a-DIHYDR0EQ.UILIN k. I73"DIHYDROEQUILIN 5. I7a-DIHYDROEQUILENIN 6. 17f3-DI HYDROEQUI LEN IN 7. ESTRONE 8. EQUILIN 9. EQUILENIN Chromatographic Condit ions: Column, OV-225 (7 m x 0.25 mm) Inject ion temperature, 260°C; Detector (F.I.D.) temperature, 260°C; Column temperature, 250°C; Ca r r i e r gas (Helium) f low, 1 ml/min; Inlet pressure, 10 p.s.i ; S p l i t vent flow, 100 ml/min; Make-up gas (Helium) flow, 50 ml/min; chart speed, 0.5 cm/min. 66 FIGURE 15 CHROMATOGRAM OF THE TRIMETHYLSILYL DERIVATIVES OF A SYNTHETIC MIXTURE OF EQUINE ESTROGENS ON OV-225 GLASS CAPILLARY COLUMN 1 2.3 1. 1 7ct- ESTRADIOL 2. 173-ESTRADIOL 3- 1 7a-DI HYDROEQ.U I LI N k. 173-DIHYDROEQUILIN 5. 17cr DIHYDROEQUILENIN 6. 173-DIHYDROEQUILENIN 7- CHOLESTEROL 8. ESTRONE 9. EQUILIN 10. EQUILENIN 10 SO r-o to MINUTES Chromatographic £ o n d i t i o n s : Column OV-225 (50 m x 0.25 mm); Inject ion temperature, 260°C; Detector (F. l .D. ) temperature, 260°C; Column temperature, 235°C; Ca r r i e r gas (Helium) flow, 1 ml/min; S p l i t vent f low, 80 ml/min; Inlet pressure, 23 p.s.i ; Make-up gas (Helium) flow, 50 ml/min; Chart speed, 0.5 cm/min. 67 B. Methoxime-trimethylsi lyl Derivatives (MO-TMS) This dual der i vat i za t ion technique (88) was used to block the r eac t i v i t y of both the hydroxyl groups as well as ketogroups in the stero id molecule. (Figure 16). The MO-TMS der ivat ives of steroids usually have a shorter retention time than the corresponding TMS der ivat ives(35). However, the methoximes have been reported to form syn- and anti-isomers which could resu l t in two chromatographic peaks when a se lect ive and e f f i c i e n t column i s used (88). With packed column gas chromatography, isomer formation was not observed for the 17-and 20-ketosteroids. But in the present study with c ap i l l a r y columns, two peaks could always be detected, presumably due to isomer formation (Figure 17). C. Heptafluorobutyrate Derivatives (HFB esters) Heptafluorobutyrate derivat ives were prepared with the aim of providing not only a l i k e l y der ivat ive that would f a c i l i t a t e resolut ion of the s tero ids, but would also open the p o s s i b i l i t y of extremely sens i t ive detection using electron capture detectors. The method (Figure 18) consisted of acylat ing the stero id with heptafluorobutyric anhydride in the presence of trimethylamine (TMA) as a cata ly s t . The excess anhydride and the acid formed during the reaction were removed by par t i t i on ing between an aqueous phosphate buffer,(pH *6)and benzene. HFB esters of phenols are stable in the presence of water i f the aqueous phase i s at pH of 68 FIGURE 16 DERIVATIZATION REACTIONS: METHOXIME-TRIMETHYLSILYL ETHERS ESTRADIOL ESTRONE METHOXI ME ESTRADIOL DI-TMS ESTRONE METHOXI ME TMS FIGURE 17 CHROMATOGRAM OF THE METHOXIME-TRIMETHYLSILYL DERIVATIVES OF EQUILIN ON OV-225 GLASS CAPILLARY COLUMN. r— 1 r-0 5 1 0 M I N U T E S Chromatographic Condi t ions : column, OV-225 (7 m x 0.25 mm); Inject ion temperature, 260°C; Detector (F.I.D.) temperature, 260°C; Column temperature, 250°C; Ca r r i e r gas (Helium) flow, l ml/min; S p l i t Vent flow, 80 ml/min; Inlet pressure, 10 p . s . i . ; Make-up gas (Helium) f low, 50 ml/min; Chart speed, 0.5 cm/min. FIGURE 18 DERIVATIZATION REACTIONS: HEPTAFLUOROBUTYRATES H F B O ESTRADIOL ESTRONE ESTRADIOL DI-HFB 1. TRIMETHYLAMINE IN BENZENE 2. HEPTAFLUOROBUTYRIC ANHYDRIDE ROOM TEMP. 15 MINUTES 3. PHOSPHATE BUFFER, pH * 6 O H F B H F B O ESTRONE HFB 7 1 6 or less . A pH less than 6 maintains TMA in the protonated form. If l e f t in the unprotonated form, TMA w i l l catalyse decomposition of the esters. The heptafluorobutyrates have been found to have a shorter retention time than the corresponding TMS der ivat ives and therefore the tota l analysis time would be reduced. An example of the chromatographic character i s t i c s of the HFB derivat ives of equine steroids on a commercial OV-225 column (7m x 0.25 mm) i s shown in f igure 19. Excel lent resolut ion of g-estradiol from a-dihydroequil in was obtained, although the i r pos it ion i s now reversed as compared to the corresponding TMS ethers on the same column. Unfortunately, the two ketosteroids estrone and e q u i l i n , did not separate. Increasing the column length to 50 meters did not resolve these two ketones (Figure 20) showing that i t was not an increased number of theoret ica l plates that was important for resolut ion in th i s case. D. Ter t . buty ld imethy l s i l y l Derivatives (t.BDMS ethers) Tert. butyl dimethyl s i l y l e the r s have been reported (89) to be su i table for structure analysis of steroids using mass spectro-metry or for sens i t ive detection using single iion monitoring. The t.BDMS ethers have much simpler mass spectra than the corresponding TMS ethers and almost a l l show a strong M-57 peak which i s very useful in quant itat ive work (90). These derivat ives also exh ib i t 72 FIGURE 19 CHROMATOGRAM OF THE HEPTAFLUOROBUTYRATE DERIVATIVES OF A SYNTHETIC MIXTURE OF EQUINE ESTROGENS ON OV-225 GLASS CAPILLARY COLUMN 1. 17<x-ESTRADIOL 2. 1 7c-DIHYDROEQUI LIN 3. 17S-ESTRADIOL k. 173-DIHYDROEQUILIN 5. 17a-DIHYDROEQUILENIN 6. 176_DIHYDROEQUILENIN 7. ESTRONE 8. EQUILIN 9. EQUILENIN 7.8 0 5 10 MINUTES Chromatographic Condit ions: Column, OV-225 (7 m x 0.25 mm); Inject ion temperature, 260°C; Detector (F.I.D.) temperature, 260°C; Column temperature, 220°C (7 min) to 235°C at a rate of 15°C/min; Ca r r i e r gas (Helium) f low, 1 ml/min; S p l i t vent f low, 60 ml/min; Inlet pressure, 10 p . s . i ; Make-up gas (Helium) flow, 50 ml/min; Chart speed, 0.5 cm/min. 73 FIGURE 20 CHROMATOGRAM OF THE HEPTAFLUORODERIVATIVES OF A MIXTURE OF 4 EQUINE ESTROGENS ON OV-225 GLASS CAPILLARY COLUMN 1. 17a-DIHYDROEQUILIN 2. 17B-ESTRADI0L 3. ESTRONE it. EQUILIN Chromatographic Condit ions: Column, OV-225 (50 m x 0.25 mm); Inject ion temperature, 260°C; Detector (F.I.D.) temperature, 260°C; Column temperature, 220°C (12 min) to 235°C at a rate of 15°C/min; Ca r r ie r gas (Helium) flow, 1 ml/min; S p l i t vent f low, 60 ml/min; Inlet pressure, 23 p s i Make-up gas (Helium) flow, 50 ml/min; Chart speed, 0.5 cm/mm. 7 4 a s t a b i l i t y towards moisture not usually found in TMS der ivat ives. The preparation of t^ BDMS ethers i s s im i l a r to that of TMS ethers (Figure 21). The chromatographic character i s t i c s of the t-BDMS ethers of equine steroids on a commercial OV-225 column i s shown in f igure 22. The two d i o l s , g-estradiol and a-dihydroequi l in were well resolved, but the peak due to equ i l i n now overlapped with that of B-dihydroequi l in . Increasing the length of the column to 50 meters increased the chromatographic time to 98 minutes but did not resolve th i s l a t t e r pa i r . E. Methoxime-Heptafluorobutyrates (MO-HFB esters) This der ivat ive had the desired electron capturing property for sens i t ive detection and the keto group was der ivat i sed (Figure 23) but they could not be used for analysis due to i n t e r -ference from addit ional peaks, presumably due to syn- and anti-isomerism (Figure 24). F. Oxime-tr imethyls i ly l Ethers Since a baseline resolut ion of a l l the nine equine steroids was not achieved by any combination of the above derivatives on OV-225 cap i l l a r y column, i t was decided to study the i r chromatographic character i s t ic s on a more polar stat ionary phase. S i l a r 10 C was the next log ica l choice since a par t i a l resolut ion of estrone and equ i l i n was achieved by the use of th is phase (91). S i l a r l O C i s also a cyanopropylmethyl s i l icone, but unlike OV-225, there are no FIGURE 21 75 DERIVATIZATION REACTIONS: TERT. BUTYLDI METHYLS ILYL ETHERS ESTRADIOL ESTRONE TERT.BUTYLDIMETHYLCHLOROSI LANE , PYRIDINE, 6 0°C, 10 MINUTES CH3CH3 Jo HjC-C-Si-O 1 1 CH 3CH 3 0-S i -C-CH 3 CH3fcH3 CH CH 3 3 ESTRADIOL DI-BDMS CH 3CH 3 H 3 C-C-Si-0 CH 3CH 3 ESTRONE BDMS 76 FIGURE 22 • CHROMATOGRAM OF THE TERT. BUTYLDIMETHYLSILYL DERIVATIVES OF A SYNTHETIC MIXTURE OF EQUINE ESTROGENS ON OV-225 GLASS CAPILLARY COLUMN 1. 17a-ESTRADIOL 2. 17a_DIHYDROEQ.UI LIN 3. 173-ESTRADIOL k. ESTRONE 5. EQUILIN 6. 17B-DIHYDROEQUILIN 7. 17a-DIHYDROEQUILENIN 8. EQUILENIN 9. 1 76-DIHYDROEQUILENIN i i Vo G MINUTES Chromatographic Condit ions: Column, OV-225 (7 m x 0.25 mm); Inject ion temperature, 260°C; Detector (F.I.D.) temperature, 260°C; Column temperature, 250°C; Ca r r ie r gas (Helium) f low, 1 ml/min; S p l i t vent flow, 100 ml/min; Inlet pressure, 7 p.s.i ; Make-up gas (Helium) flow, 50 ml/min; Chart speed, 0.5 cm/min. 77 FIGURE 23 DERIVATIZATION REACTIONS: METHOXIME-HEPTAFLUOROBUTYRATES H F B O ESTRADIOL I 1. METHOXYAMINE HYDROCHLORIDE V ESTRONE IN PYRIDINE 70°C, 30 MINUTES N O C H , ESTRADIOL ^ ® ~ ESTRONE METHOXIME 1. TRIMETHYLAMINE IN BENZENE 2. HEPTAFLUOROBUTYRIC ANHYDRIDE ROOM TEMP,. 15 MINUTES PHOSPHATE BUFFER, pH = 6 H F B ESTRADIOL DI-HFB ESTRONE MOX-HFB 78 FIGURE 24 CHROMATOGRAM OF THE METHOXIME-HEPTAFLUOROBUTYRATE DERIVATIVES OF ESTRONE(A) AND EQUILIN (B) ON OV-225 GLASS CAPILLARY COLUMN eo o m u. z o B CO Chromatographic Condit ions: Column, OV-225 (7 m x 0.25 mm); Inject ion temperature, 260°C; Detector (F.I.D.) temperature, 260°C; Column temperature, 235°C; Car r ie r gas (Heliurn)flow, 1 ml/min; S p l i t vent flow 100 ml/min; Inlet pressure, 7 P - s . i ; Make-up gas (Helium) flow, 50 ml/m Chart speed, 0.5 cm/min. 79 phenyl groups in th is molecule. I t i s more polar than OV-225 because of the presence of a high percentage of cyanopropyl moiety (Figure 25). Attempts to separate the TMS ethers on this phase was not successful because of the excessive t a i l i n g of estrone and equ i l i n peaks. Equilenin apparently could not be eluted with in 50 minutes.(McReynold.'s. constant,retention index value, A l , given by the manufacturer, fo r 2-pentanone i s 659). Der ivat izat ion of the keto group to i t s methoxime lead to resolut ion of the syn- and anti-isomers and therefore was not useful for analysis. An a l ternat ive der ivat ive for the ketosteroids was the oxime formed by reaction with hydroxylamine hydrochloride, which, on further s i l y l a t i o n y ie lded ox ime-tr imethy ls i ly l ethers (Figure 26). The chromatogram obtained for these derivat ives on S i l a r 10 'Ccapi l la ry column showed evidence of isomerism but the peak due to one isomer was not well resolved from the mainpeak (figure 27). This observation suggested that reducing the column length would further suppress the resolut ion of the syn- and anti-isomers. As shown in f igure 28, a l l the nine s tero ids , along with an internal standard, ethynyl estradio l , were well resolved on a 15 meter column without any addit ional peaks due to syn- and ant.irisomer separation. The e lut ion time was also reduced compared to the corresponding MO-TMS der ivat ives. Thus i t appeared that the most suitable der ivat ives, for c ap i l l a r y gas chromatography of equine steroids on S i l a r 10 C column were the ox ime-tr imethy ls i ly l ethers. Having developed a FIGURE 25 O V - 2 2 5 CH, CH 3 C=N (CH2)3 (CH2)3 I • C=N CH 3 I Si - O t S i - C H 3 I CH 3 -•x S ILAR 10C 8 1 FIGURE 26 DERIVATIZATION REACTIONS: OXIME-TRIMETHYLSILYL ETHERS ESTRADIOL 1 . NH20H.HC1/PYRIDINE 70°C, 30 MINUTES 2. BSTFA/PYRIDINE 70°C, 10 MINUTES ESTRONE TMSO TMS TMSO ESTRADIOL DI-TMS NOTMS ESTRONE OXIME TMS fl 82 FIGURE 27 CHROMATOGRAM OF THE METHOXIME-TRIMETHYLSILYL (A) AND OXIME-* TRIMETHYLSILYL(B) DERIVATIVES OF ESTRONE ON SILAR IOC GLASS CAPILLARY COLUMN I ' Vo • K — i • r o — r ~ MINUTES Chromatographic Condi t ions : Column, S i l a r 10 C(25 m x 0.25 mm);Injection temperature, 260°C; Detector (F.I.D.) temperature, 260°C; Column temperature, 220°C, Ca r r ie r gas (Helium) flow, 0.8 ml/min; S p l i t vent flow, kO ml/min; Inlet pressure, 15 p . s . i ; Make-up gas (Helium) flow, 50 ml/min; Chart speed, 0.3 cm/min. FIGURE 28 CHROMATOGRAM OF THE OXIME-TRIMETHYLSILYL DERIVATIVES OF A SYNTHETIC MIXTURE OF EQUINE ESTROGENS ON SILAR 10 C GLASS CAPILLARY COLUMN i To To MIN UTES Chromatographic Condit ions: Column, S i l a r 10 C (15 m x 0.25 mm) Inject ion temperature, 260°C; Detector (F.I.D.) temperature, 260°C; Column temperature, 170°C (6 min) to 220°C at a rate of 2.3°C/min; Ca r r i e r gas (Helium) flow, 0.8 ml/min; S p l i t vent flow, kO ml/min; Inlet pressure, 10 p . s . j ; Make-up gas (Helium) flow, 50 ml/min; Chart speed 0.3 cm/min; At tenuat ion, 3 84 chromatographic method that would resolve a l l the known components in equine estrogens, the analysis of dosage forms of these steroids was undertaken. II1-3. ASSAY OF FORMULATIONS A. Cal ibrat ion Curves for Steroid Standards L inear i ty of detector response (FID) over a wide range of concentrations of a l l the nine equine steroids was established with an internal standard, ethynyl e s t r ad i o l . Among the potential internal standards invest igated, such as cho lestero l , testosterone, methyl testosterone and ethynyl e s t r ad i o l , the l a t t e r was found to be most su i table because of i t s close s i m i l a r i t y in structure with that of equine estrogens, the lack of any addit ional peaks a f te r de r i va t i sa t i on , and an appropriate retention time. Six d i f fe rent concentrations of the pure steroids were selected for ca l i b ra t ion purpose in such a way as to bracket the amount expected in commercial formulations. Over the concentration ranges studied, the flame ion i sat ion detector response was found to be l inear as represented by a -es t rad io l in f igure 29, and as l i s t e d in table T for a l l s tero ids. 8 5 FIGURE 29 CALIBRATION CURVE FOR 17a -ESTRADIOL 0 . 2 5 WEIGHT RATIO 86 TABLE 1 Ste roici Correlation Slope Intercept a - E 0 2 * 0.9995 1.0166 0.0060 B-EO2 0.9996 1.0083 0 . 0 0 0 8 a-DHEQ 0.9992 1.0578 0.0403 (3-DHEQ 0.9994 0.9037 0.0043 E 0 1 0.9984 0 . 9 4 0 3 0.0690 EQ 0.9999 0 . 7 8 0 3 0.0330 aDHEQN 0.9992 0.6374 0.0131 BDHEQN 0.9994 0. 3 9 5 8 0.0049 EQN 0.9936 0.2319 0.0241 * Refer to page x i i for l i s t . o f abbreviations. 87 B. Sample Preparation Equine estrogens are present in f in i shed dosage forms as the i r sodium sa l t s of the sulphate esters. In order to analyse these steroids by gas chromatography, the esters must be hydrolysed to l i be ra te the free phenolic steroids. Two methods of hydrolysis have been employed in the past; one involv ing the use of hydrochloric acid at elevated temperature (36), and the other employing sulphatase enzyme derived from Hel ix pomatia(40). I t has been shown that acid hydrolysis either f a i l s to cleave stero id sulphates completely or leads to destruction of a considerable portion of the molecule. (93). In the case of androsterone sulphate, hydrochloric acid hydrolysis has been reported to lead to a loss of 90% of the s tero id structure and to lead to the production of several degradation products, (93). Since the aim of th i s study was to determine the or ig ina l components of the mixture without creating a r t i f a c t s during work-up, acid hydrolysis was not used. Enzyme hydrolysis,using sulphatase enzyme>has been the preferred method of hydrolysis of conjugated estrogens in pharmaceutical products, and optimum conditions of hydrolysis such as the pH, molarity of the buffer, temperature and duration of incubation have already been established (39, 40). The use of sulphatase enzyme to hydrolyse s tero id conjugates i s not without disadvantages, as i t has been shown that phosphates greatly reduce the effectiveness of the enzyme(94). Although the optimum conditions of hydrolysis have been reported, i t was considered necessary to ascertain the quantity of enzyme to 88 be used for a new batch of formulations, as many of these may contain varying amount of i nh ib i to r s such as phosphates. In the present study, Q e s t r i l i n tablets were selected as a model to establ i sh the optimum quantity of enzyme, as th i s formulation was suspected to contain phosphates (39). The level of enzyme used varied between 1000 to 4000 units to hydrolyse an a l iquot of the tab let powder equivalent to 1 mg of conjugated estrogens. For the sample invest igated, 2000 units were found to be adequate to accomplish complete hydrolysis. Above this l e v e l , no increase in quantity of the tota l steroids recovered was observed (Table 2). TABLE 2 Sample Erizyme (units/ml) Oe s t r i l i n tablets 1000 2000 3000 4000 Amount of estrogen, found (mg) 0.9874 1.0053 1.0050 1.0020 Amount as per label claim (mg)1.0043 1.0043 1.0043 1.0043; C. Chromatographic Analysis of the Formulations It was found that the estrogenic components could be resolved as the i r ox ime-tr imethy ls i ly l ethers using temperature programmed gas chromatography on a 15 meter S i l a r 10C column. The conditions required for chromatography, such as the temperature of the in ject ion port, detector, oven, gas flow, i n l e t •pr^ -suire^ -.-a-nd 89 s p l i t ra t io were a l l kept constant throughout the assay procedure, and these conditions were the same as used for establ i sh ing l i n e a r i t y of the detector with standard samples. Determination of the optimum s p l i t r a t i o , i e . the ra t i o of rate of gas flow through the column (adjusted by i n l e t pressure) to the rate of gas.flow at the s p l i t vent was important, since at high s p l i t rat ios the component present in the smallest amount, 3-dihydroequilenin, would not be detected and at very low s p l i t r a t i o s , the column would be overloaded with those components present in large quantity, estrone and e q u i l i n . This also had to be balanced with the r ight select ion of instrument attenuation. The purpose of employing an i n l e t s p l i t t e r was not only to l i m i t the amount of sample introduced into the cap i l l a r y column (sample capacity i s very low for c ap i l l a r y columns compared to packed columns) but also to avoid the deposition of broad bands on the column. S p l i t t i n g the flow stream permited the use of high flow rate through the vapourisation chamber so that i t i s rap id ly flushed clean while s t i l l maintaining the r e s t r i c ted flow through the column. The s p l i t rat io,attenuat ion and the volume of the sample injected were adjusted in such a way as to get f u l l scale def lect ion of the detector response on the chart paper for the major component, estrone, and a detectable response for B-dihydroequilenin. The complete baseline resolut ion of a l l the known steroids of conjugated estrogen mixtures in two formulations marketed by d i f fe rent manufacturers i s shown in figures30 and 31. While the 2 .. metre packed column gas chromatographic methods required 55 minutes(39) FIGURE 30 90 CHROMATOGRAM OF THE OXIME-TRIMETHYLSILYL DERIVATIVES OF EQUINE ESTROGENS DERIVED FROM A MULTIPLE INJECTION VIAL ON SILAR 10 C GLASS CAPILLARY COLUMN 1. 17a-ESTRADI0L 2. 17B-ESTRADIOL 3. I7a-DIHYDR0EQUILIN-4. 173-DIHYDROEQUILIN 5. ETHYNYL ESTRADIOL 6. ESTRONE 7. EQUILIN 8. 17a-DIHYDR0EQUILENIN 9. 17B-DIHYDR0EQUILENIN 10. EQUILENIN i 1 1 1 0 1 0 2 0 3 0 M I N U T E S Chromatographic Condit ions: Column, S i l a r IOC (15 m x 0.25 mm); Inject ion temperature, 260°C; Detector (F. l .D. ) temperature, 260 C; Column temperature, 170°C (8 min) to 220°C at a rate of 2 .3° c/min; Ca r r i e r gas (Helium) f low, 0.8 ml/min; S p l i t vent f low, 40 ml/min; Inlet pressure, 10 p . s . i ; Make-up gas (Helium) f low, 50 ml/min; C h a r t s p e e d , 0.3 cm/min; At tenuat ion, 3 91 FIGURE 31 CHROMATOGRAM OF THE OXIME-TRIMETHYLSILYL DERIVATIVES OF EQUINE ESTROGENS DERIVED FROM A TABLET FORMULATION ON SILAR IOC GLASS CAPILLARY COLUMN 1 3 1. 17a-ESTRADIOL 2. 17B-ESTRADIOL 3. 17a-DIHYDROEQUI LIN k. 176-DI HYDROEQ.UI LIN 5. ETHYNYL ESTRADIOL 6. ESTRONE 7. EQUILIN 8. 17a-DIHYDROEQUILENIN 9. 173-DIHYDROEQUILENIN 10. EQUILENIN 11. EQUOL 12. DIETHYL PHTHALATE 1 0 — I — 2 0 — r 3 0 M I N U T E S Chromatographic Condit ions: Column, S i l a r IOC (15 m x 0.25 mm); Inject ion temperature, 260°C; Detector (F.I.D.) temperature, 260°C; Column temperature, 170°C (8 min) to 220°C at a rate of 2 .3°c/min ; Ca r r i e r gas (Helium) f low, 0.8 ml/min; S p l i t vent f low, kO ml/min; Inlet pressure, 10 p . s . i ; Make-up gas (Helium) f low, 50 ml/min; Chart speed, 0.3 cm/min; At tenuat ion, 3 92 and 72 minutes (40) to achieve a par t i a l reso lut ion, c ap i l l a r y gas chromatography with a 15 m S i l a r 10 C column required only 28 minutes to achieve a baseline resolut ion. I t i s also s i gn i f i c an t to note that chromatograms obtained with packed columns did not exh ib i t peaks other than that of 10 known stero ids, while the present chromatogram revealed the presence of addit ional minor peaks. These extraneous peaks did not or ig inate from the reagent used for der ivat i zat ion nor they were derived from degradation products, since the chromatogram of the blank (reagent) did not reveal any peaks other than the solvent peak. Chromatography of the ind iv idual steroids also did not give addit ional peaks. Thus, the addit ional peaks possibly indicate the presence of yet unident i f ied steroids or s im i l a r other substances in these formulations. The presence of 9-dehydroestrone, which has been reported ea r l i e r (36) could not be confirmed due to the unava i l ab i l i t y of a pure sample of th is s te ro id . In Oe s t r i l i n tablets (Figure 31), i t was possible to i dent i f y two non-steroidal compounds. One prominent peak with a retention time of 7.32 minutes was i den t i f i ed as diethyl phthalate, by comparison with retention time of pure diethyl phthalateas well as mass spectral data (m/e, 222, 178, 177, 149). The occurrence of this material i s probably due to i t s use as a p l a s t i c i s e r in the f i lm-coat ing technique of the tab lets . Another extraneous peak with a retention time of 22.07 minutes (Figure 31) was i den t i f i ed as equol, by comparison with the retention time of a reference sample as well as by mass spectral data of the di-TMS derivat ive (m/e, 386, 267,193, 192, 178, 177). Equol, a non-steroidal compound 93 f i r s t i so lated from pregnant mare's urine (48), was e a r l i e r i d en t i f i ed as an impurity in some formulations ava i lable in the Canadian market (41). I t has been shown (39) that i f equol i s not removed pr io r to chromatographic analys i s , i t would in ter fere with the quantitat ion by overlapping with the peak due to B - dihydroequi1in. In order to remove this substance from conjugated estrogens and allow quant itat ive analys i s , repeated washings with benzene were required (39). As shown in f igure 31, equ'ol was resolved from equine stero ids, thus o f fe r ing a d i rect method of simultaneous determination of equine steroids as well as the impurity, equol. The non-interference of this impurity with quantitat ion of equine steroids in commercial formulations has . resulted in reducing the tota l time of analysis by el iminat ing the r e l a t i v e l y lengthy benzene pre-wash required by the e a r l i e r procedure( Another s i gn i f i can t d i f ference, as compared to previous methods,was that a baseline resolut ion of B -estradio l and a -d ihydroequi l in was achieved by employing a cap i l l a r y column of high s e l e c t i v i t y . B -estradio l can now be accurately measured in commercial formulation of conjugated estrogens. This i s also true of equ i len in, the l a s t peak to elute from the column. This compound was eluted from the c ap i l l a r y column as a sharp and symmetrical peak with a retention time of 26.22 minutes while the packed column methods required e i ther 55 minutes (39) or 72 minutes (40) and exhibited broad, i l l - d e f i n e d peaks, espec ia l ly in the l a t t e r case. Determination of Conjugated Estrogens in Combination Products Conjugated estrogens are often combined with meprobamate or methyl testosterone. These compounds were separated pr ior to enzyme hydrolysis by pa r t i t i on ing between an acetate buffer and chloroform. Here, advantage is taken of the fact that conjugated estrogens are soluble in water while meprobamate and methyl testosterone are not. Extraction of the sample three times with 25 ml portions of chloroform was found to be adequate to completely remove both meprobamate as well as methyl testosterone. No attempt was made to quantitate meprobamate by a gas chromatographic method since GLC analyses of the carbamates have been reported to be prone to serious error due to thermal i n s t a b i l i t y , resu l t ing in pyrolys is in the in ject ion port. Attempts at using a lower in ject ion port-temperatures helped to a l l e v i a te the problems considerably but did not t o t a l l y el iminate the formation of two major pyrolys is products, propanediol-monocarbamate and propanediol (95). I t was possible to determine methyl testosterone i n the combination product by the same der ivat i zat ion technique (figure 32) and using the same cap i l l a r y column under the same chromatographic conditions. However, methyl testosterone had to be chromatographed separately because of the large amount of th i s s tero id present in an a l iquot of the tablet, powder equi valentv.to ;lmg of conjugated estrogen . This quantity had been found to be optimum for determination of conjugated estrogens ( i e . 0.3446 gms of the tab let powder i s equivalent to 1 mg FIGURE 32 D E R I V A T I Z A T I O N REACTIONS: O X I M E - T R I M E T H Y L S I L Y L ETHERS OF TZa-METHYL TESTOSTERONE. 1 7 a - METHYLTESTOSTERONE HYDROXYLAMINE HYDROCHLORIDE PYRIDINE, 70°C, 30 MINUTES HON-1 7 a - METHYLTESTOSTERONE OX I ME (CH3)SiON BSTFA, PYRIDINE 70°C, 10 MINUTES 17a-METHYLTEST0STER0NE OXIME-TMS 96 of conjugated estrogens and 8 mg of methyl testosterone). If th i s mixture i s chromatographed d i r e c t l y , the column would be overloaded with methyl testosterone, and the l a t t e r s tero id would inter fere with B -d ihydroequi l in. If the extract ion of methyl testosterone was not complete, addit ional peaks in the v i c i n i t y of B-dihydroequi l in would be observed when conjugated estrogens are chromatographed. Extraction of the unhydrolysed sample three times with 25 ml portions of chloroform did remove methyl testosterone completely as evidenced by the absence of an addit ional peak in the chromatogram of conjugated estrogens. Chromatographic character i s t i c s of the ox ime-tr imethyls i ly l der ivat ive of methyl testosterone, with an internal standard ethynyl e s t r ad i o l , are shown in f igure 33. As expected, the methyl testosterone oxime exhibited syn- and anti-isomerism which was qua l i t a t i ve l y and quant i tat ive ly d i f fe rent from that of isomerism of 17-ketosteroid oximes. The resolut ion of two isomers of the 3-ketosteroid, methyl testosterone, was far greater as compared to that of a 17-ketosteroid, estrone. The re l a t i ve peak heights of the two isomers were also d i f fe rent (f igure 33). This difference in peak heights can be explained on the basis of the fact that, the ketogroup at pos it ion 3 of methyl testosterone i s not s t e r i c a l l y hindered, and therefore i s free to form both the isomers in approximately equal proportions. In the case of a 17-ketosteroid, estrone, the s t e r i c hindrance from the neighbouring angular methyl group at posit ion 13 would be expected to r e s t r i c t the formation of one of the isomers. For quantitat ion purposes, the area of the two peaks 97 FIGURE 33 CHROMATOGRAM OF THE OXIME-TRIMETHYLSILYL DERIVATIVE OF 17a-METHYL-TESTOSTERONE ON SILAR IOC GLASS CAPILLARY COLUMN 1 , 2 . SYN-AND ANTI-ISOMERS 3. ETHYNYL ESTRADIOL 1 2 r-0 — I — 1 0 — I — 2 0 M I N U T E S Chromatographic Condit ions: Column, S i l a r IOC (15 m x 0.25 mm); Inject ion temperature, 260°C; Detector (F.I.D.) temperature, 260°; Column temperature, 170°C (8 min) to 220°C at a rate o f 2.3°C/min; Car r ie r gas (Helium) f low, 0.8 ml/min; S p l i t vent flow, 40 ml/min; Inlet pressure, 10 p . s . i ; Make-up gas (Helium) f low, 50 ml/min; Chart speed, 0.3 cm/min; At tenuat ion, 3-98 FIGURE 34 CALIBRATION CURVE FOR METHYLTESTOSTERONE Cor re la t i on c o e f f i c i e n t 0.9992 Slope 0.9515 Intercept -0.0731 0 5 1.0 1.5 2 0 W E I G H T R A T I O 99 were added together and the tota l area ra t i o was referred to the standard curve for methyl testosterone prepared by p lo t t i ng the area ra t io of methyl testosterone to ethynyl estrad io l against the corresponding weight ra t io s . (Figure 34) Corre lat ion, slope and intercept for th is least, square l i ne .ware 0.9992, 0.9515 and 0.0731, ,. respect ively. The percent label c la im, as determined from the resu l t of th i s assay, was 101.08%. D. Quantitation of Conjugated Estrogen Formulations Standard curves, constructed for each stero id by p lo t t ing the area rat ios of the s tero id and internal standard against weight rat ios of these two substances were used to quantitate steroids present in commercial formulations. The reporting gas chromatograph employed in th i s study recorded the area of each peak in the chromatogram. The ra t i o of area of a stero id peak to area of internal standard was then calculated and by extrapolation from the standard curves, the amount of each stero id was obtained. These amounts, added together, represented the tota l " f r ee " estrogenic steroids present in an a l iquot of the sample taken for analys is. Mu l t i p l i ca t i on by 1.38 converts the quantity of free phenolic steroids to the i r sodium sa l t s of the sulphate esters. The percentage of each stero id with respect to the tota l conjugated estrogens found and the percent of label claim was then calculated. A sample data sheet and calculat ions are shown in table 3, and the TOO TABLE 3 ASSAY OF CONJUGATED ESTROGENS IN A MULTIPLE INJECTION VIAL Sample - Premarin Intravenous Manufacturer - Ayerst Laboratories, Montreal, Canada Lot No - 552, Labelled strength - 25 mg Conjugated estrogens in the al iquot taken for analysis - 1 mg. Quantity of internal standard, 0.2 mg. Steroid Peak .area Area ra t i o weight. r a t i o amount % found(mg) a - E 0 2 * 1030 0.1265 0.1025 0.0204 2.74 B-E0 2 345 0.0424 0.0429 0.0085 1.14 a-DHEQ 4426 0.5440 0.6047 0.1204 16.27 B-DHEQ 373 0.0458 0.0440 0.0088 1.18 E 01 13180 1.6199 1.7967 0.3543 48.35 EQ 5047 0.6203 0.8372 0.1674 22.53 a-DHEQN 401 0.0492 0.0565 0.0113 1.52 B-DHEQN 44 0.0054 0.0011 0.0002 0.03 EQN 241 0.0296 0.2316 0.0463 6.23 E E 0 2 ( internal " . 8136 - - - -standard) Total (phenolic steroids) 0.7431 mg Total x 1.38 (conjugated steroids) 1.0254 mg Total amount in the v i a l 25.5 mg Percent label claim 102.0% * Refer to page x i i for l i s t of abbreviations 101 TABLE 4 COMPOSITION OF COMMERCIAL FORMULATIONS CONTAINING CONJUGATED  ESTROGENS: (These results are the percentages of each conjugated estrogen and is..;the average of two assays). COMPONENT SAMPLE PERCENTAGES A B C [ ) E 1 F G a - E 0 2 * 2. 82 2. 72 3. 14 0. 06 0. 36 3, .15 2. .34 3-E0 2 1. 06 0. 21 0. 86 0. 12 0. 36 1. .03 2, .04 a-DHEQ 16. 23 21. 66 15. 07 2. 87 2. 41 17 .84 16. .48 3-DHEQ 1. 11 0. 89 1. 18 0. 23 0. 19 0. .45 0. .65 E 01 48. 07 44. 10 49. 99 62. 01 68. 54 48. .04 55. .29 EQ 22. 45 21. 73 23. 97 26. 95 16. 29 21. .78 15. .47 a-DHEQN 1. 65 1. 33 1. 81 1. .1.2 4. 29 1. .65 2. .13 3-DHEQN 0. 11 3. 46 0. 24 0. 12 0. 07 0. .03 0. ,06 EQN 6. 49 3. 90 3. 73 6. 51 7. 47 6. .02 5. .53 EQUOL 3. 85 4. 97 TOTAL 101.40 112.66 99.92 99.19 99.02 99.55 84.5 A = Premarin intravenous, 25 mg v i a l B = Premarin tab le t s , 0.3 mg C = Premarin Vaginal Cream, 0.625 mg/g D = Oe s t r i l i n tab le t s , 2.5 mg E = Conjugated Estrogens tab le t s , (ICN), 1.25 mg. F = Premarin (0.625 mg) with methyl testosterone (5 mg) G = Premarin (0.4 mg) with meprobamate (200 mg) * Refer to page x i i for l i s t o f abbreviations and to page 105 for "precision of the assay. 102 assay results of d i f fe rent formulations areshown in Table 4. Employing the methodology developed, a var iety of commercial formulations, such as preparations meant for intravenous administrat ion, tablets of varying strength, vaginal cream and combination products with meprobamate and methyl testosterone have been analysed. There has been no attempt made in the past to analyse combination products of conjugated estrogens with androgens or anx io ly t i c s ,o r for conjugated estrogens present in the vaginal cream.Since these estrogens are readi ly absorbed into the blood by intravaginal administration (4) i t was f e l t necessary to develop a qua l i ty control method for such products as we l l . It would be of interest to compare the composition of d i f fe rent products made by the same manufacturer and also the products of d i f fe rent manufacturers. As shown in Table 4, close s i m i l a r i t y i s evident for products from the same manufacturer, whether they are tab let s , in jectab les , vaginal cream or combination products. For example, the percentage of a -d ihydroequi l in in products marketed by the same company i s f a i r l y consistent (16.23, 16.48, 15.07, 17.84%)., but in the case ofa d i f fe rent trade name product, the percentage of th is s tero id i s only 2.84%. S i gn i f i cant var iat ions are also noted for;the two major components, estrone (44-68%) and equ i l i n (15-26%). Since the USP allows a var iat ion of ± 10% of tota l conjugated estrogens, most of these products would part ly s a t i s f y compendial standards, but the requirement of 50-65% estrone and 20-35% equ i l i n would not be met by most of these products, except perhaps, Oe s t r i l i n 2.5 mg tablets and Premarin vaginal cream. It must be rea l i sed , however, that the 103 l a t t e r requirements (50-65% and 20-35%) are based on a co lor imetr ic method and may not be d i r e c t l y compared with the more se lect ive GLC method, espec ia l ly where only subtle differences are noted. The apparent low percentage (84.5%) for tota l conjugated estrogens in the combination product, premarin and meprobamate, i s probably due to the fact that the par t i cu la r l o t analysed was several years older than the other samples. 104 E. Precis ion of Analysis In order to assess the reproduc ib i l i t y of the assay developed with glass c ap i l l a r y column, the assay was repeated four times using an intravenous preparation. . This formulation, when reconstituted with a buffer, y ie lded a very homogeneous solut ion and therefore was found to be most suitable for a l iquot determination. Enzyme hydrolys i s , extract ion of the s tero ids, der i vat i zat ion etc. were carr ied out on separate a l iquots . The precis ion of quantitat ion of each stero id as well as the tota l amount is shown in Table 5. F. Comparison of Three GC Methods for a Mult ip le Inject ion V ia l Table 6 shows the assay of the same formulation, meant for intravenous use, by three d i f fe rent gas chromatographic methods. S ign i f i cant differences can be observed for B - e s t rad io l , B -dihydroequi lenin, equ i len in, as well as in the to ta l amount.;of conjugated estrogens. B-estradiol could not be measured accurately when DEGS phase was used because of poor resolut ion from a -d ihydroequi l in . This s tero id could not be measured at a l l when OV-225 phase was used since i t could not be resolved from a -d ihydroequ i l in . B- Dihydroequilenin, the s tero id present in the smallest amount, was not detectable in some formulations when packed columns were ...used and equi len in, the l a s t peak toelute:exhib ited i l l - d e f i n e d peak. While packed column gas chromatography with DEGS stationary phase reported a label claim of 108.5% for to ta l conjugated estrogens and OV-225 TABLE 5 ASSAY OF MULTIPLE INJECTION VIAL Amount i s expressed as Mil 1igrams of each steroid in a 1.0 mg al iquot Steroid Mean ±S.D. Percentage aE0 2 0.0210 0.0204 0.0209 0.0206 0.0207±0.0002 2. 82 BE0 2 0.0072 0.0085 0.0143 0.0144 0.0111+0.0037 1. 51 aDHEQ 0.1177 0.1209 0.1157 0.1244 0.1196±0.0038 16. 29 BDHEQ 0.0076 0.0084 0.0084 0.0083 0.0081±0.0003 1. 10 E 01 0.3472 0.3593 0.3519 0.3478 0.3515±0.0055 47. .87 EQ ; 0.1626 0.1674 0.1675 0.1673 0.1662± .0024 22. 64 aDHEQN 0.0130 0.0113 0.0096 0.0083 0.0106± .0020 1. 44 BDHEQN 0.0014 0.0011 0.0005 0.0003 0.0008±0.0005 0. .11 EQN 0.1491 0.0463 0.0433 0.0431 0.0454±0.0028 6. 18 Total (mg) 0.7268 0.7431 0.7323 0.7345 0.7342 Total x 1.38 (mg) 1.0029 1.0254 1.0105 1.0136 1.0131 Amount of conjugated estrogens found = 1 .0131 ± 0.0093 mg * •••Refei?-(.feo. page x i i for l i s t of abbreviations. o 1 0 6 TABLE 6 COMPARISON OF THREE GC METHODS FOR ONE OF THE FORMULATIONS  MEANT FOR INTRAVENOUS USE (25..mq v i a l ) Amount i s expressed as milligrams of conjugated estrogen in each v i a l . Steroid DEGSa 0V-225 b S i l a r IOC packed co l . packed c o l . c ap i l l a r y c a-E0 2* 0.85 0.81 0.71 e-E02 0.14 - 0.27 a-DHEQ 3.43 4.14 4.12 3-DHEQ 0.37 0.97 0.28 a-DHEQN 0.49 0.36 0.42 B-DHEQN 0.13 0.06 0.03 E 01 13.74 12.66 12.18 EQ 7.17 6.64 5.69 EQN 0.82 0.68 1.64 Total 27.14 mg 26.90 mg 25.35 mg Label Claim 25.00 mg 25.00 mg 25.00 mg Percent Label Claim 108.50 % 107.00 % 101.40 % a - Data taken from Johnson et a l . , (40) b - Data taken fromMcErlane et a l . , (39) * Refer to page x i i for l i s t of abbreviations. 107 packed column method reported 107.0%, the cap i l l a r y gas chromatographic method claimed only 101.4%. The apparent high percent label claim when packed columns were employed, i s perhaps due to the fact that some minor components, other than the known stero ids, did not resolve, and therefore may have contributed to the peak area measured for the known steroids. Since the glass c ap i l l a r y gas chromatographic method resolved these addit ional components, the peak area measured can be assumed to represent the amount of a s ingle component. The high resolut ion that i s attained with a c ap i l l a r y column, within a short period of time i s a s i g n i f i c an t factor that has contributed to the accuracy of th i s method. 108 SUMMARY AND CONCLUSIONS A highly se lect ive and rapid technique was developed for the simultaneous determination of conjugated estrogens in a var iety of formulations such as tablets of varying strength, preparations meant for intravenous administrat ion, vaginal cream and combination products with androgens (methyltestosterone) and anx io ly t i c s (meprobamate). The method consisted of hydrolyt ic cleavage of the stero id sulphates with sulphatase enzyme, extract ion of the free phenolic steroids into an organic solvent, der i vat i zat ion to ox ime-tr imethy ls i ly l ethers and resolut ion of these derivatives on a short glass c ap i l l a r y column coated with a cyanopropylmethyl s i l i cone stat ionary phase. Baseline resolution of a l l the steroids ., has been achieved, along with a previously reported impurity, equol, and therefore the method can be used to quantify th i s impurity as we l l . The tota l chromatographic time was only 28 minutes whereas the previous methods required e i ther 55 minutes or 72 minutes. The high resolut ion that was attained by the use of cap i l l a r y columns within a short period of time has contributed to the increased accuracy of th i s method compared to the ex i s t ing procedures. 109 REFERENCES 1. J.R. Jones and E. Kemmann in 'The Menopause, Comprehensive Management' ed. , B.A. Eskin, p. 173, Masson Publishing USA Inc., New York, N.Y., USA, (1980). 2. E.B. Atwood in 'Pharmacological Basis of Therapeutics' eds., L.S. Goodman and A. Gilman, 4th ed., p. 1538, The MacMillan Co., London, England, (1970). 3. W.H. Ut ian, S. Af r . Med. J . , 46, 1079 (1972). 4. F. Murad and R.C. Haynes in .'Pharmacological Basis of Therapeutics'• eds., L.S. Goodman and A. Gilman, 6th ed. , p. 1431, The MacMillan Publishing Co. Inc., New York, N.Y., USA, (1980). 5. B.L. Riggs, J . Jowsey, R.S. Goldsmith, P.J. Ke l l y , D.L. Hoffmann and CD . Arnaud, J . Clin.. Invest., 51_, 1659 (1972). 6. R. Lindsay, J.M. A i t k i n , J.B. Anderson, D.M. Hart, E.B. McDonald and A.C. Clark, Lancet, 1 , 1038 (1976). 7. R.R. Recker, P.D. Sav i l l e and R.P. Heaney, Ann. Intern. Med. 87, 649 (1977). 8. J . Dickerson, R. Bressler and CD . Chr i s t i an , C l i n . Pharmacol. Ther., 26, 502 (1979). 9. M.B. L ip se t t , J . Amer. Med. Ass., 237, 1112 (1977). 10. T.W. McDonald, J.F. Annegers, W.M. O'Fallow, M.B. Dockerty and G.D. Malkasian Amer. J . Obstet. Gynecol., 127, 572 (1977). 11. D.C. Smith, R. Prent ice, D.J. Thompson and W.L. Hermann, N. Engl. J . Med., 293, 1164 (1975). 12. H.K. Z ie l and W.D. F ink le , N. Engl. J . Med., 293, 1167 (1975). 13. T.M. Mack, M.C Pike, B.E. Henderson, R.I. P fe f fe r , V.R. Gerkins, M. Arthur and S.E. Brown, N. Engl. J . Med., 294, 1262 (1976). 14. H. J i c k , R.N. Watkins, J.R. Hunter, B.J. Denan, S. Madsen, K.J. Rothman and A.M. Walker, N. Engl. J . Med., 300, 218 (1979). 15. C.B. Hammond, F.R. Jelovsek, K.L. Lu, W.T. Creasman and R.T. Parker, Amer. Obstet. Cynecol., 133, 525 (1979). 110 16. G.S. Gordon and B.C. Greenberg, Postgrad. Med., 59, 66 (1976). 17. R.W. Kistner, Obstet. Gynecol., 48, 479 (1976). 18. A. Debancens, S. Fuensabda, 0. Gonzalez and J . Zanartu, Cancer, 43, 658 (1979). 19. L.E. Nacht iga l l , R.M. Nacht iga l l , R.D. Nachtigall and E.M. Beckman, Obstet. Gynecol., 54, 74 (1979). 20. R.I. Horwitz and A.R. Fe inste in, J . Amer. Med. Ass., 241, 1266 (1979). 21. S.R. Ser ianni , P. Hale, M. Brown, C C . Huang, B. Paigin and J.L. Ambrus, J . Med., 9, 423 (1978). 22. P.K. S i i t e r i , B.E. Schwartz, P.C. McDonald, Gynecol. Oncol.,2, 228 (1974). 23. S.S.C. Yen, J . C l i n . Endocrinol. Metab. 40, 578 (1975). 24. H.J. Ruder, L. Lauriax, M.B. L ippsett, J . C l i n . Invest., 5^, 1020 (1972).. 25. Ole Gisvold in 'Textbook of Organic Medicinal and Pharmaceutical Chemistry' eds. CO. Wilson and 0. Gisvold. 3rd. ed., p. 578, J.B. L ippincott Company, Ph i ladelphia, Pa. USA, (1956). 26. The United States Pharmacopea, 19th rev i s i on , Mack Publishing Company, Easton, Pa., USA, (1975). 27. E.O. Haenni, J . Pharm. S c i . , 39, 544 (1950). 28. E. Heftman, Science ITU 571 (1950). 29. J.E. Cox, J . Chromatogr., 61_, 193 (1971). 30. E. Heftmann, J . Amer. Chem. S o c , 73_, 851 (1951). 31. G.J. K ro l , G.R. Boyden, R.H. Moody, J . C Cameau and B.T. Kho, J . Chromatogr., 61_, 187 (1971 ). 32. K.M. Rajkowski, G.D. Broadhead, J . Chromatogr., 89_, 374 (1974). 33. L.E. Crocker and B.A. Lodge, J . Chromatogr., 62_, 158 (1971 ). 34. L.E. Crocker and B.A. Lodge, J . Chromatogr., 69^ , 419 (1978). 111 35. I. Schroeder, G. Lopez-Sanchez, J.C. Medina-Acevedo, M. Del Carmen Espinosa, J . Chromatogr. S c i . , 1 3 , 37 (1975). 36. R. Roman, C H . Yates, J.F. M i l l e r and W.J.A. VandenHeuvel, Can. J . Pharm. S c i . , 1_0, 8 (1974). 37. I. Shroeder, J.C. Medina-Acevedo and G. Lopez-Sanchez, J . Chromatogr. S c i . , 1TJ, 183 (1972). 38. K.M. McErlane, J . Chromatogr. S c i . , 1_2, 97 (1974). 39. K.M. McErlane and N.M. Curran, J . Pharm. Sci.,'66, 523 (1977). 40. R. Johnson, R. Masserano, R. Haring, B. Kho and G. S c h i l l i n g , J . Pharm. S c i . , 64, 1007 (1975). 41. G. Carrignan, N.M. Curran, B.A. Lodge and K.M. McErlane, Can. J . Pharm. S c i . , 1_3, 73 (1978). 42. R.N. Johnson, R.P.:-Masserano, B.T. Kho and W.P. Adams, J . Pharm. S c i . , 67_, 1218 (1978). 43. W.P. Adams, J . Hasegawa, R.N. Johnson and R.C. Haring, J . Pharm. S c i . , 68, 986 (1979). 44. A.G. Bu t t e r f i e l d , B.A. Lodge and N.J. Pound, J . Chromatogr j. S c i . , Jl, 401 (1973). 45. R.W. Roos, J . Chromatogr. S c i . , 1_4, 505 (1976). 46. G. Capitano and R. Tscherne, J . Pharm. S c i . , 68_, 311 (1979). 47. P.I. Musey, D.C. Co l l i n s , and J.R.K. Prudy, Steroids, 31, 583 (1978). 48. G.F. Marrian and G.A. Haselwood, Biochem. J . , 26_, 1227 (1932). 49. D.A. Shutt, Endeavour, 35_, 110 (1976). 50. W. Klyne and A.A. Wright, Biochem. J . , 66, 92 (1957). 51. M.J.E. Golay in 'Gas chromatography' p . l , Academic press, New York, N.Y., USA,(1958). 52. M. Novotny, M.P. Muscarinek, A.T.G. Steverink and R. Farlow, Anal. Chem., 48, 468 (1976). 112 53. C. Madani, J . Chromatogr., 126_, 161 (1976). 54. U. Rapp and U. Shroeder, Chromatographic,8, 474 (1975). 55. H.T. Badings, Chromatographia, 8, 440 (1975). 56. J . Henninger, E. Munthe, J . Pable and E. Jel lum, J . Chromatogr., 158, 297 (1978). 57. B.A. Knights, M. Legendre, J.L. Laseter and J.S. Storer, C l i n . Chem., 21_, 888 (1975). 58. J . J . Fran ken and M.F. T r i j ebe l s , J . Chromatogr., 91_, 425 (1974). 59. M. Novotny and A. Z l a t k i s , J . Chromatogr. S c i . , 8 346 (1970). 60. E.C. Horning, M.G. Horning, J . Szafranek, P. Vanhout, A.L. German, J.P. Thenot and CD . Pfaffenberger, J . Chromatogr., 91 , 367 (1974). 61. CD. Pfaffenberger, L.R. Malinak and E.C.Horning, J . Chromatogr., 158, 313 (1978). 62. F.A. F i t zpa t r i c k , D.A. Str ingfe l low, J . Maclouf and M. Rigaud, J . Chromatogr., 177, 57 (1979). 63. S.I. Goodman, P. Helland, 0. Stokke, A. Flatmark and E. Jel lum, J . Chromatogr., 142, 497 (1977). 64. D.H. Desty, Adv. Chromatogr., 1_, 199 (1965). 65. M. Mohnke and W. Saf fer t in 'Gas Chromatography' ed. M. Vansway, p. 216,Butterworths, Washington, D.C. USA (1962). 66. A.V. Kiseley in 'Gas Chromatography' ed. M. Vansway, p.3, Butterworths, Washington, D.C. USA (1962). 67. F.A. Brunner and C P . Cartoni, Anal. Chem., 36, 1522 (1964). 68. M. Novotny and K. Tesarik, Chromatographia. 1_, 332 (1968). 69. M. Blumer, Anal. Chem., 45;,.980 (1973). 70. A.L. German and E.C Horning, J . Chromatogr. S c i . , 11_, 76 (1973). 71. R.S. Deelder, P.J.H. Hendricks and M.G.F. K r o l l , J . Chromatogr., 57_, 67 (1971). 113 72. K. Grob and E. Grob, J . Chromatogr., J25_, 471 (1976). 73. G. Nota, G.C. Goret t i , A. Armanante and G. Marino, J . Chromatogr., 95, 229 (1974). 74. J.D. Schieke, N.R. Comins and V. Pretor ius, J . Chromatogr., 112, 97 (1975). 75. F.I. Onuska, M.E. Comba, T. B i s t r i c k i and R.T. Wilkinson, J . Chromatogr., 142, 117 (1977). 76. A.L. German and E.C. Horning, J . Chromatogr. S c i . , 1J_, 76 (1973) . 77. G.A.F.M. Rutten, J.A. Luyten, J . Chromatogr., 74, 177 (1972). 78. R.C.M. DeNijs, J.F. Franken, R.P.M. Dooper and J.A. R i jk s , J . Chromatogr., 167, 231; (1978). 79. D.A. Cronin, J . Chromatogr., 97, 263 (1974). 80. C. Hista and J . Bornstein, Adv. Chromatogr., 9_, 215 (1970). 81. P. Sandra and M. Verzele, Chromatographia, 1_0, 419 (1977). 82. G. Dykstra and J . DeGoey in 'Gas Chromatography' ed. N. H. Desty, p. 5, Butterworths, London, England, (1958) 83. G. Schomburg, H. Husmann and F. Weeke, J . Chromatogr., 99, 63 (1974) . 84. J . Bouche and M. Verzele, J . Gas Chromatogr., 6_, 501 (1968). 85. K. Grob and G. Grob, J . Chromatogr. S c i . , 7_, 584 (1969). 86. R. Dandeneau, P. Bente, T. Rooney and R. Hiskes, Amer. Lab. 11_, 61 (1979). 87. E.M. Chambaz and E.C. Horning, Anal. Biochem., 30, 7 (1969). 88. E.C. Horning, Anal. Biochem., 22, 284 (1968). 89. G. P h i l l i p o u , D.A. Bigham and R.F. Seamark, Steroids, 26, 516 (1975). 90. R.W. Kel ly and P.L. Taylor , Anal. Chem., 48, 465 (1976). 91. J.S. Zwig, R. Roman, W.B. Hagerman and W.J.A. VandenHeuvel, Journal of HRC and CC., 3, 169 (1980). 92. P. Sandra, M. Verstappe and M. Verzele, Journal of HRC & CC, 1 , 28, (1978). 93. CA . Muehlbaecher and E.K. Smith, C l i n . Chem., 16_, 158 (1970). 94. K.S. Dodgson and B. Spencer, Biochem. J . , 55, 315 (1953). 95. K. Holch and J.G. Gjaldback, Dansk. Tidsskr. Farm., 45, 32 (1971); through "GLC and HPLC determination of Therapeutic agents ' , Chromatographic Science Series, Volume 9, part 2, ed. Kiyoshi T su j i , p;.-.610,Marcel Dekker, Inc. New York.', N.Y., USA, (1978). 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0095073/manifest

Comment

Related Items