Selectivity of Poly(styrene-divinylbenzene) Columns - ACS Publications

5 Selectivity of Poly(styrene-divinylbenzene) Columns John M. Joseph

Downloaded by UNIV LAVAL on May 1, 2016 | http://pubs.acs.org Publication Date: January 27, 1986 | doi: 10.1021/bk-1986-0297.ch005

Squibb Institute for Medical Research, New Brunswick, NJ 08903

Macroporous poly(styrene-divinylbenzene) copolymer, PRP-1, columns were used as the stationary phase in the reverse-phase HPLC of the syntheticβ-lactamantibiotic aztreonam [2S-[2α,3β(Z)]]-3-[[(2-Amino-4-thiazolyl)[(1carboxy-1-methylethoxy)-imino]acetyl]amino]-2-methyl-4oxo-1-azetidinesulfonic acid and related compounds. Aztreonam was separated better from its precursors and therefore could be assayed more accurately. In most cases, the elution order of compounds tested on a PRP-1 column followed that in conventional reverse-phase, suggesting a similar separation mechanism. For various separations investigated, PRP-1 was found to be more suitable for our applications than the silica-based reversed-phase columns. Reverse-phase packings of o c t y l s i l a n e (C%) n d octadecylsilane (Ci8) are by f a r the most popular and widely used stationary phases i n modern l i q u i d chromatography. In spite of their wide popularity as universal l i q u i d chromatography phases, these a l k y l bonded phases ( s i l i c a packings, i n general) have several shortcomings, the major one being their i n s t a b i l i t y i n aqueous phases (1,2). The p o s s i b i l i t i e s f o r the regulation of retention and s e l e c t i v i t y by ion suppression and complexation are limited i n the reverse-phase mode with functionalized s i l i c a , as degradation of the stationary phase occurs outside the pH range 2-8 (3). Column-to-column v a r i a t i o n i s another major disadvantage with bonded phase columns. S i g n i f i c a n t variations i n s e l e c t i v i t y and retention c h a r a c t e r i s t i c s exist among bonded phase columns from manufacturer to manufacturer, and even within l o t to l o t from the same manufacturer. Several reasons are attributed to t h i s problem. A few are worth mentioning as they i n herently are associated with silica-based supports. Due to s t e r i c hindrance, the bonding reaction i s always incomplete leaving unreacted s i l a n o l groups (Si-OH) on the s i l i c a surface (4). These residual s i l a n o l groups can cause mixed retention mechanisms, especially with basic compounds, leading to non-reproducibility i n retention from batch-to-batch of columns (5). In addition, residual a

0097-6156/86/0297-0083$06.00/0 © 1986 American Chemical Society

Ahuja; Chromatography and Separation Chemistry ACS Symposium Series; American Chemical Society: Washington, DC, 1986.


84

C H R O M A T O G R A P H Y A N D SEPARATION CHEMISTRY

Si-OH groups are also f o c a l points of attack for water and other "reagents i n the mobile phase, causing d i s s o l u t i o n of underlying s i l i c a with gradual deterioration i n column performance and eventual reduction i n column l i f e t i m e (5,6). The pH of the s i l i c a surface also varies from a c i d i c , neutral, to basic, and i s believed to i n fluence the preparation and properties of bonded phases (7). The inclusion of traces of t r a n s i t i o n metals i n s i l i c a matrix can further influence the retention of acids, bases or neutral compounds that can undergo complexation reactions (1). Thus the s e l e c t i v i t y and reten tion c h a r a c t e r i s t i c s of bonded phase columns also depend on the quality of s i l i c a used for the bonding reactions. The use of macroporous poly(styrene-divinylbenzene) copolymer, PRP-1, as reverse-phase adsorbent has been previously reported i n the analysis of pharmaceuticals (8,9), organic acids (8), nucleosides (10), plant hormones (11), and phenolic compounds (12). The major advantage of t h i s resin-based packing material i s i t s extreme s t a b i l i t y over a wide pH range of 1-13 (10). The chemical s t a b i l i t y of PRP-1 at elevated pH l e v e l s has enabled new applications. The increased d e t e c t a b i l i t y for barbiturates (13) and thiamine deriva tives (14) was reported using PRP-1 column with mobile phase, at elevated pH l e v e l s (pH 8-9). S i m i l a r l y , PRP-1 functionalized with trimethyl amine (PRP-X100 anion-exchanger) has been used with basic eluents i n ion-chromatography using conductivity detection (15). This paper describes the use of poly(styrene-divinylbenzene) copolymer, PRP-1, as a reverse-phase adsorbent i n the assay of the a n t i b i o t i c aztreonam and related compounds. Comparisons are also made f o r similar assays using silica-based columns. None of the shortcomings described e a r l i e r , associated with bonded phase columns, i s observed. In addition to the reverse-phase mode, the PRP-1 columns are tested i n ion-pair as well as i n size exclusion modes of separation. Superior resolutions are obtained i n the reverse-phase chromatography of i o n i c compounds without the use of ion-pairing agents. In addition to the normal adsorption and/or p a r t i t i o n i n g , TT-TT interaction i s also seen to play a dominant role i n the separa tion of aromatic compounds using a PRP-1 column. 9

Experimental The modular HPLC system used i n this study included the following components: a Model 110A Altex pump (Beckman S c i e n t i f i c Instruments) or a System 4 Series programmable pump (Perkin-Elmer Corporation), an i n - l i n e f i l t e r , a Model 773 variable-wavelength UV detector (Kratos S c i e n t i f i c Instruments) f i t t e d with a 12 uL flow c e l l , a Model 600 Series autosampler with a nominal volume of 20 uL (Perkin Elmer Corporation) and a two channel recorder (Kipp and Zonen). The chromatographic data were processed with the aid of a Model 3357 laboratory computer (Hewlett-Packard). The chromatographic separations were carried out either on a 15 or 25 cm x 4.6 mm I.D. s t a i n l e s s - s t e e l column packed with 10 ym poly(styrene-divinylbenzene) copolymer (PRP-1, Hamilton). The con ventional reverse-phase chromatography was performed using either a 10 um Whatman P a r t i s i l ODS-3 (25 cm x 4.6 mm I.D.) or a Waters 10 ym yBondapak C- (30 cm x 3.9 mm I.D.) column. The a n a l y t i c a l columns were kept at 31° using a column heater (Bioanalytical Systems, Inc., g


5.

JOSEPH

Poly(styrene-divinylbenzene) Columns

85

LC-23A) unless otherwise s p e c i f i e d . Since PRP-1 i s a non-silaceous p a r t i c l e , no precolumn or saturator column was used. The s i l i c a based a n a l y t i c a l columns were used i n conjunction with a precolumn packed with 37 ym s i l i c a . The solvents used were: (a) HPLC grade methanol and acetonit r i l e (Fisher S c i e n t i f i c ) (b) water, d o u b l e - d i s t i l l e d and stored i n glass. Inorganic s a l t s and acid (ammonium sulfate, potassium phos phate monobasic, potassium phosphate dibasic and phosphoric acid, 85%) were a n a l y t i c a l reagent grade (Mallinckrodt Chemical Works). The mobile phase was delivered at a flow rate of 1-1.5 mL/min. The pH of the mobile phase was adjusted before the addition of organic modifiers.


Results and Discussion Aztreonam i s a (3-lactam a n t i b i o t i c with a wide spectrum of a c t i v i t y against a v a r i e t y of gram-negative organisms including Pseudomonas aeruginosa (16). The synthesis of this compound starting from the amino acid, L-threonine has been previously reported (17). Figure 1 shows the structure of aztreonam and related compounds. Aztreonam, the (Z)-isomer, i s the b i o l o g i c a l l y active form. The a l t e r n a t i v e configuration, the (E)-isomer, was synthesized. The open ring form (Z) obtained by the cleavage of the 3-lactam moiety i s the p r i n c i p a l b i o l o g i c a l metabolite. Other possible, related compounds include the ethyl ester of aztreonam, which could be formed i n trace amounts during r e c r y s t a l l i z a t i o n of aztreonam from ethanol, desulfonated aztreonam, the open-ring form (E) and the desulfonated open*-ring form. Several HPLC systems capable of resolving aztreonam from possible related compounds, mainly the (E)-isomer and the open-ring form (Z) were examined. As shown i n Figure 2, the conventional reverse-phase system u t i l i z i n g a Waters yBondapak C±Q column was found to be suitable f o r carrying out these separations. The mobile phase, here, ± composed of 82% phosphate buffer (0.05M, pH 3.0)and 18% methanol. The desulfonated aztreonam almost coeluTed with the open-ring form (Z) i n t h i s chromatographic system. As expected,the aztreonam ethyl ester was retained longer (RT *\> 24 minutes) than the rest of the compounds due to i t s increased hydrophobicity. An inherent problem and perhaps, one of the biggest disadvan tages of silica-based columns, i n general, i s the column-to-column v a r i a t i o n (1). This was c l e a r l y manifested during an attempt to reproduce the separation p r o f i l e obtained on a Waters yBondapak C^g column using other conventional Ci8 columns from diverse manufact urers (Whatman, E. Merck, IBM, etc.) under similar and other chromotographic conditions. Even though the separation of the open-ring form (Z) and the ethyl ester from aztreonam was accomplished, these reverse-phase columns lacked the s e l e c t i v i t y to effect adequate resolution of the (Z)- and (E)- isomers of aztreonam. S i m i l a r l y , amine columns ( s i l i c a bonded to n-propyl amine) from various manufacturers (IBM, Whatman, DuPont, and Waters) were also examined in the same capacity using a mobile phase of methanolic phosphate buffer (pH 3.5) containing small amounts of tetrabutylammonium hydrogen sulfate (18). Here also, the major problem was column-tocolumn v a r i a t i o n , which once again, proved to be r e a l rather than s




AZTREONAM ETHYL ESTER

DE8ULFONATED OPEN-RING FORM

OPEN-RING FORM (Z)

OPEN-RING FORM (E)

DESULFONATED AZTREONAM Figure 1.

Structure of aztreonam and related compounds.



5.

JOSEPH


87

apparent with s i l i c a based columns. Only the Zorbax amine column (DuPont) provided adequate separation of aztreonam from related compounds including the (E)-isomer. In t h i s instance, the mode of separation presumably i s ion-exchange (weak anion-exchanger). The elution order of aztreonam ethyl ester and open-ring form (Z) were found to be the reverse of that found i n conventional reverse-phase HPLC (Figure 3). The less polar aztreonam ethyl ester was eluted f i r s t , and the r e l a t i v e l y more polar open-ring form (Z) became the l a s t eluting component i n t h i s chromatographic system. No change was observed i n the elution order of the (Z)- and (E)- isomers of aztreonam. One d i f f i c u l t y with an amine column i s the longer e q u i l i b r a t i o n time required as opposed to reverse-phase (bonded phase and PRP-1). Figure 4 shows the separation of aztreonam and related compounds accomplished on a PRP-1 column. The mobile phase was e s s e n t i a l l y the same as that used i n conventional reverse-phase except that a small amount of a c e t o n i t r i l e (4%) was incorporated to sharpen the peaks. S e l e c t i v i t y was found to be much better with t h i s resin-based stationary phase compared to a bonded phase, as i t afforded superior and otherwise d i f f i c u l t separations of these very c l o s e l y related compounds. The improved resolution between the desulfonated openring form and open-ring form (Z) would enable accurate quantitation of both of these compounds i f they were present as impurities i n bulk material or formulations of aztreonam. The most noteworthy was the excellent resolution obtained between the (Z)- and (E)- isomers of the open ring form of aztreonam. I t was found that the (Z)isomer always eluted before the corresponding (E)- isomer i n both Ci8 and PRP-1 columns (Figures 2 and 4). Based on t h i s , i t was presumed that the peak at RT 3.9 minutes (Figure 4) might be the (E)-isomer of open-ring desulfonated aztreonam. The aztreonam ethyl ester was retained longer on a PRP-1 column (RT ^ 35 min.) as opposed to conventional C^g columns. The polystyrene matrix i s believed to be extremely hydrophobic (6), which accounts for the prolonged retention of r e l a t i v e l y non-polar e n t i t i e s l i k e aztreonam ethyl ester. Except desulfonated aztreonam which was unavailable at the time of t h i s experiment, the elution order of a l l the other compounds from a PRP-1 column followed that i n conventional reversephase, suggesting a somewhat similar retention mechanism under i d e n t i c a l chromatographic conditions. The desulfonated aztreonam was retained longer (^ 16 minutes), and i t was eluted after the E isomer. Another type of interaction which i s unique to PRP-1 i s Tr-Tr interaction, especially with aromatic solutes (19) . Figure 5 shows the separation obtained on a 15-cm PRP-1 column under the same chromatographic conditions. No loss of resolution was noted. The retention time was lowered with shorter columns. Concomitantly, the peaks appeared to be much sharper. The PRP-1 column also has been used successfully i n the HPLC assay of several key aztreonam intermediates. The carbobenzoxy-Lthreoninamide (CBZ-L-threoninamide) was one. I t was the f i r s t major intermediate isolated i n the synthesis of aztreonam starting from L-threonine. The synthetic sequence i s depicted i n Figure 6. Two major impurities were benzyl alcohol and benzyl carbamate; the former produced by the hydrolysis of benzyl chloroformate and the l a t t e r by the reaction between benzyl chloroformate and ammonia. The purpose T


f

88



(2) 5.9

RT IN MINUTES Figure 2. Conventional reverse-phase HPLC of aztreonam and r e lated compounds. Column: Waters ybondapak C-^g (30 x 3.9 mm i.d.), Mobile phase: 82% 0.05M phosphate buffer (pH 3.0)-18% methanol. Flow rate: 1.5 mL/min. Detection: UV 254 nm. Peaks: (1) = openring form (Z), (2) = aztreonam (Z) , (3) = aztreonam (E), (4)= aztreonam ethyl ester.

(2) 6.5

RT IN MINUTES Figure 3. Separation of aztreonam and related compounds using a Zorbax amine column (25 x 4.6 mm i . d . , 30°C). Mobile phase: 97.5% 0.1M phosphate buffer-5 mM TBAHS (pH 3.5) - 2.5% methanol. Flow rate: 1 mL/min. Detection: UV 220 nm. Peaks: (1) = aztreonam ethyl ester, (2) = aztreonam (Z), (3) = aztreonam (E), (4) = open-ring form (Z).


5.

JOSEPH

Poly(styrene-divinylbenzene)

(2) 2.7

89

Columns

(4) 6.9

(5) 11.0

(3) 5.2


(1) 2.2

RT IN MINUTES Figure 4. Separation of aztrenam and related compounds usifjg a PRP-1 column (25 x 4.1 nm i.d., 31 C). Mobile phase: 83% 0.05 M phosphate buffer (pH 3.0)- 13% methanol 4% acetonitrile. Flow rate: 1 mL/min. Detection: UV 254 nm. Peaks: (1) = open-ring desulfonated form, (2) = open-ring form (Z), (3) = open-ring form (E), (4) = aztreonam (Z), (5) = aztreonam (E). (4) 5.3

(2) 2.4

(5) 8.3

(3) 4.1

(1) 2.0

Ul RT IN MINUTES Figure 5. Separation of aztreonam and related compounds on a 15-cm PRP-1 column. Chromatographic conditions and elution order are the same as i n Figure 4.



90

C H R O M A T O G R A P H Y A N D SEPARATION C H E M I S T R Y

of the HPLC system was to resolve these two possible impurities and the two synthetic precursors, L-threonine and CBZ-L-threonine, from CB Z-L-threoninamid e. Figure 7 shows the separation of these compounds on a 15-cm ( i . d . 4.6 mm) PRP-1 column. The mobile-phase consisted of methanolwater (60:40) containing 0.05M ammonium sulfate. The chromatography provided excellent resolution of a l l compounds of interest with benzyl carbamate as the l a s t eluting component (RT ^8.9 minutes). The use of ammonium sulfate i n the mobile phase helped to reduce peak t a i l i n g . This HPLC system has now been used f o r the routine assay of CBZ-L-threoninamide. This method i s f a s t , rugged, sensi t i v e and reproducible. Column s t a b i l i t y was found excellent with minimum column-to-column v a r i a t i o n . Figure 8 shows the chromatogram of a batch of CBZ-L-threoninamide produced by an overseas manufact urer. The sample was assigned a purity of 100% based on a reversephase HPLC assay (Ci8 column) performed by a d i f f e r e n t laboratory. When the same sample was assayed i n the author s laboratory using a PRP-1 column, i t was found to contain a s i g n i f i c a n t amount (^ 5%) of benzyl carbamate which coeluted with the main peak when the ODS column was used. Consequently, the laboratory using the common ODS column showed erroneously high purity r e s u l t s . f

In the meantime, attempts to develop a reverse-phase HPLC assay for CBZ-L-threoninamide were vigorously being made i n other labora t o r i e s . Eventually, a reverse-phase HPLC assay was developed for this compound using a Waters Radial Pak Ci8 cartridge (10 cm x 8 mm i.d.) and a mobile phase composed of 0.1M phosphate buffer-methanol (82:18). Figure 9 represents the chromatographic p r o f i l e . Even though the most d i f f i c u l t separation of benzyl carbamate from CBZ-L-threoninamide was accomplished, the greatest shortcoming of this method was the prolonged assay time: approximately 55 minutes, six times longer than when a PRP-1 column was used. Such a lengthy assay time i s less a t t r a c t i v e f o r p r a c t i c a l purposes. Styrene-divinylbenzene XAD copolymer was shown to give higher K values for amino acids containing aromatic residues (e.g.,phenyl alanine, tyrosine, etc.) as opposed to those with simple a l i p h a t i c moieties; f o r example, v a l i n e , leucine, etc. (20). This increased retention was attributed to the greater interaction of the aromatic moiety of the amino acid with the r e s i n matrix by TT-TT interaction (20) . With an ODS column such unique interaction (TT-TT) i s non existent. The a b i l i t y of the phenyl group of benzyl carbamate to exert t h i s type of interaction with the PRP-1 matrix might have caused i t s easy separation on t h i s column (Figure 10). Other e f f i c i e n t separations accomplished using a PRP-1 column were of compounds involved i n the synthesis of carbobenzoxy-L-serinamide mesylate starting from the amino acid, L-serine (Figure 11). The purpose of the chromatographic system was to aid organic chemists at various synthetic steps by HPLC monitoring of the reaction i n t e r mediates. Fast and simple HPLC assays are c r u c i a l to t h i s type of work, and i n most cases, as i n the current example, e f f i c i e n t HPLC assays were developed using PRP-1 columns. A l l three synthetic pre cursors (L-serine, CBZ-L-serine and CBZ-L-serinamide) were presumed to be possible impurities i n the f i n a l mesylate intermediate. As shown i n Figure 12, these synthetic intermediates were completely resolved from CBZ-L-serinamide mesylate i n less than eight minutes. 1


5.


JOSEPH

H N,

OH

φ χ ^ ο ^ Ο Ι

3

91

C B Z - N

CBZ-CI

Ί

^ C H ^ — O H

3

CBZ — L—THREONINE

L—THREONINE

Η C B Z - N U

OH J

CIC0 Et

*CH

2

NH

OH

NH

3

3

2


C B Z - L - THREONIN AMIDE

^

.CH OH 2

^

C

H £ H - - 00 2

2

C - N H

2

Ο

BENZYL ALCOHOL

Figure 6.

BENZYL CARBAMATE

Synthesis of CBZ-L-threoninamide.

(3) 4.1

(5) 8.9

(4) 6.1

(2) 2.5

(1) 1.3

RT IN MINUTES

Figure 7. HPLC of CBZ-L-threoninamide. Column: PRP-1 (15 χ 4.1 mm i . d . , 30°C). Mobile phase: 60% methanol - 40% water - 0.05M ammonium s u l f a t e . Flow rate: 1 mL/min. Detection: UV 215 nm. Peaks: (1) = L-threonine, (2) = CBΖ-L-threonine, (3) = CBZ-Lthreoninamide, (4) = benzyl alcohol, (5) = benzyl carbamate.



92 REVERSE-PHASE

PRP-1

5.3


3.4

BENZYL CARBAMATE

7.0 2.2

R T IN M I N U T E S


Figure 8. Chromatograms of a batch of CBZ-L-threoninamide using Ci8 and PRP-1 columns.

(5) 46.2

Ο)

2.5

(2) 15.6 Λ . RT IN M I N U T E S

Figure 9. HPLC of CBZ-L-threoninamide using a Waters Radial Pak C Cartridge (10 χ 8 mm i . d . ) . Mobile phase: 82% 0.1M phosphate buffer (pH 3.0) - 18% methanol. Flow rate: 2 mL/min. Detection: UV 204 nm. Peaks: (1) = L-threonine, (2) = benzyl alcohol, (3)= CBZ-L-threonine, (4) = benzyl carbamate, (5) = CBZ-L-threonina mide. 1 8


5.

JOSEPH

Poly(styrene-divinylbenzene)

Columns

93


BENZYL CARBAMATE

Figure 10. Hypothetical π-π interaction of phenyl group of benzyl carbamate with PRP-1 matrix.

HN 2

0

OH

Z-N

J—OH

Γ

OH

Z

_N

0

OH

J — N H

2

o'—

Z

_H

0

OMeS

J ^ - r NH

2

OH

L—SERINE

CBZ-L-SERINE

Figure 11.

C B Z — L — S E R I N AMIDE

C B Z - L - S E R I N AMIDE MESYLATE

Synthesis of CBZ-L-serinamide mesylate.

RT IN MINUTES

Figure 12. HPLC of CBZ-L-serinamide mesylate. Column: PRP-1 (15 χ 4.1 mm i . d . , 30°C). Mobile phase: 60% methanol - 40% water - ammonium s u l f a t e (1 gm/liter). Flow rate: 1 mL/minute. Detection: UV 210 nm. Peaks: (1) = L-serine, (2) = CBZ-L-serine, (3) = CBZ-L-serinamide, (4) = CBZ-L-serinamide mesylate.



94


The chromatography was performed on a 15-cm PRP-1 column with a simple mobile phase of methanol-water (60:40) containing ammonium sulfate (1 gm/liter). The same HPLC system was also used to monitor the next step of the synthesis where the mesylate underwent ring closure to form i t s tetrabutylammonium derivative (CBZ-L-azetidinone-TBA salt) which was f i n a l l y reduced to the corresponding azetidinone (Figure 13). Another possible impurity besides the precursor, CBZ-L-azetidinone-TBA s a l t , was l u t i d i n e which was used as a reagent during c a t a l y t i c reduction. Figure 14 shows the resolution obtained on a 15-cm PRP-1 column as compared to that obtained on 25-cm Whatman 0DS-1 and ODS-3 columns under i d e n t i c a l chromatographic conditions. Further modification of the mobile phase would be required to achieve similar separations on ODS-1 and ODS-3 columns. The elution order of these compounds from PRP-1 was i d e n t i c a l to that from ODS with the most polar TBA deriva t i v e as the f i r s t eluting component, followed by the less polar CBZ-L-azetidinone. Lutidine being least polar among the three was the l a s t to elute from a l l three columns. Ion-pair chromatography has dramatically extended the scope of bonded reverse-phase HPLC, as i t successfully separates both i o n i c and ionizable compounds, and i s considered an a l t e r n a t i v e to ionexchange chromatography (21, 22, 23). In reverse-phase ion-pair chromatography a surfactant i s added to the aqueous mobile phase as a counter-ion to effect increased resolution of oppositely charged sample ions. Thus, tetrabutylammonium hydrogen s u l f a t e (TBAHS) i s used as a cationic counter-ion for the separation of organic acids, whereas sodium dodecyl s u l f a t e (or, more commonly, sodium heptane sulfonate) i s used as an anionic counter-ion for the separation of organic bases. Various theories have been proposed to explain the mechanism of the ion-pair phenomenon. A complete discussion of the mechanistic aspects of ion-pair chromatography has been given by Bidlingmeyer (22). The objective, here,was to examine the retention c h a r a c t e r i s t i c s of a PRP-1 column as opposed to ODS columns in ion-pair chromatogra phy using TBAHS as the counter-ion. The compound I (Figure 15) i s CBZ-L-azetidinone-TBA s a l t derived from L-threonine, and i t i s another important intermediate i n the synthesis of aztreonam. Compounds I I and I I I are possible impurities associated with I. A reverse-phase ion-pair chromatographic method was o r i g i n a l l y developed f o r I using a Whatman ODS-1 column and a mobile phase com posed of methanol-water (70:30) containing 0.1M ammonium s u l f a t e and 5 mM TBAHS (pH 5.0). Figure 16 shows the separation obtained for I, II and I I I . The chromatography took ^ 11 minutes with CBZ-L-azetidinione-TBA s a l t (compound I) as the l a s t eluting component. When the separation was carried out under s i m i l a r conditions using a 25-cm PRP-1 column, no peak was v i s i b l e even after a period of ^ 50 minutes. A l l three test compounds were believed to be strongly retained on the PRP-1 column when TBAHS was used as the counter-ion. A fresh batch of the same mobile phase was prepared without TBAHS and the chromatography was repeated with another 25-cm PRP-1 column. The test compounds eluted t h i s time with wide separa tion i n less than 30 minutes (Figure 17). On the contrary, loss of resolution for the test compounds I, I I and I I I was observed when the chromatography was repeated using the same Whatman ODS-1 column with


5.

95


JOSEPH

H

H

Z-N

Z-N

o*Q v

SO,"TBA

+


CBZ-L-AZETIDINONE TBA SALT (i>

LUTIDINE

CBZ-L-AZETIDINONE

(3)

(2)

Figure 13. Synthesis of CBZ-L-aztetidinone.

ODS-3

PRP-1

ODS

(2) 4.2

(1) 3.0 (2) 7.8

(3)" 5.5

(1) 3.3

L-L RT IN MINUTES

Figure 14. Comparison of separations on PRP-1 vs C^g columns. Mobile phase: 60% methanol - 40% water - ammonium sulfate (1 gm/ l i t e r ) . Flow rate: 1 mL/min. Detection: UV 215 nm. Peaks: (1) = CBZ-L-azetidinone- TBA s a l t , (2) = CBZ-L-azetidinone, (3) = lutidine.



Ο P h ^ O ^ N *

/

C

H

3

ΤΊ O^

Ο

N

^S0

3

(QBu) N 4

Ο

* Ο ^ Ν

^

H

N

^S0

Ρ Η ^ Ο ^ Ν # [ ^ Ο Η

3

(QBu) N 4

3

C0 H


2

HN

^S0 "(nBu) N 3

π

4

in

Figure 15. Structure of CBZ-L-azetidinone-TBA s a l t and related compounds derived from L,-threonine.

RT IN MINUTES

Figure 16. HPLC of CBZ-L-azetidinone- TBA salt derived from L-threonine using bonded reverse-phase i o n - p a i r chromatography. Column: Cig (25 χ 4.6 mm i . d . , 30°C). Mobile phase: 70% water 5 mM TBAHS (pH 5.0) - 30% methanol - 0.1M ammonium sulfate. Flow r a t e : 1 mL/min. Detection: UV 215 nm.



5.

JOSEPH


97

the modified mobile phase without TBAHS. Evidently, TBAHS helped to enhance the retention of ionic species on a bonded reverse-phase, thereby e f f e c t i n g t h e i r separation. Unlike C^g, poly(styrenedivinylbenzene) matrix, by v i r t u e of i t s polymer structure, possesses strong adsorbent properties (24) . I t i s now evident that PRP-1 could r e t a i n moderately i o n i c species (here, anions) better than bonded reverse-phase columns without further modifications of i t s polymer surface with large organic counter-ions l i k e TBAHS. A small amount of TBAHS could be used to improve separation i n the case of f a i r l y ionic compounds which show poor retention on PRP-1 columns. As shown i n Figure 18, the marginal resolution obtained between aminoxyisobutyric acid and phthalic acid was improved by adding 1 mM TBAHS to the same mobile phase of acetonitrile-water (10:90) containing ammonium s u l f a t e (4 g m / l i t e r ) . TBAHS seemed to have no effect on the retention of aminoxyisobutyric a c i d , but i t enhanced the retention of phthalic acid considerably (from 2.9 minutes to 25.9 minutes). Exclusion effects are also possible with PRP-1 column depending on the size of the solute molecule. This was evident i n the reten tion behavior of the macrocyclic heptaene a n t i b i o t i c , amphotericin B, on a PRP-1 column. The reverse-phase fast LC method recently developed i n the author s laboratory using a 5-cm long Ci8 column (3 micron p a r t i c l e size) gave a retention time of 5.5 minutes for t h i s compound. This method also provided excellent r e s o l u t i o n of amphotericin Β from the cofermented X component (25), another heptaene closely related to i t . The mobile phase consisted of 0. 05M sodium acetate buffer (pH 5.0)-methanol-acetonitrile(45:35:30) containing 3 mM EDTA. Figure 19 shows the separation obtained when a sample of amphotericin Β was chromatographed on a 25-cm PRP-1 column under the same chromatographic conditions. The amphotericin Β was retained even less (4.6 min.) than that on a shorter C^g column (RT 5.5 min.) with marginal separation from the X component. Using a C^g column of approximately equal dimensions(25 cm χ 4.6 mm 1. d.) as the PRP-1 column, a retention time of ^ 20 minutes was obtained f o r amphotericin B. The poor retention of macrocyclic structures l i k e amphotericin Β on PRP-1, compared to bonded reversephase, could be explained only by size exclusion. Here, obviously, the s e l e c t i v i t y and retention were not as good as that obtained with other modes of separation (adsorption, p a r t i t i o n i n g , π-π i n t e r a c t i o n , Η-bonding etc.) which are prevalent i n poly(styrene-divinylbenzene), PRP-1, stationary phase. This was further exemplified by the poor resolution obtained between amphotericin Β and i t s X component. 1

f

f

T

T

f

f

Acknowledgment s The author i s grateful to Dr. G. Brewer, Dr. B. K l i n e , and Dr. J. Kirschbaum for t h e i r encouragement and many h e l p f u l suggestions. The author also wishes to thank Ms. C. Saloom f o r her conscientious typing of t h i s manuscript and Mr. J . Alcantara for the artwork.



ODS 5.6

III 5.2

5.1


3.9

U 4 R T IN M I N U T E S

Figure 17. Comparison of separations on PRP-1 and ODS columns with mobile phase containing no TBAHS.

HCI >

/

2

0

\

(1) A M I N O X Y I S O B U T Y R I C ACID WITHOUT T B A H S

^ ^ C O O H

(2) P H T H A L I C ACID WITH T B A H S

(1) 1.5

(1) 1.6

(2) 2.9l


Figure 18. Effect of TBAHS i n the separation of f a i r l y ionic compounds using a PRP-1 column. Peaks: (1) = aminoxyisobutyric acid, (2) = phthalic acid.


5.

JOSEPH


99

yOH ΌΟΗ

\^0»

AMPHOTERICIN Β


CH

ODS (FAST LC) 5.5

3

PRP-1

(2)

4.6

(2)

(1) 2.1

RT IN MINUTES

Figure 19. Retention of amphotericin Β on PRP-1 and Cl8 columns. Mobile phase: 45% 0.05M sodium acetate buffer - 3 mM EDTA- 35% a c e t o n i t r i l e - 20% methanol (pH 5.0). Flow rate: 1 mL/min. Detection: 405 nm. Peaks: (1) = amphotericin X, (2) = ampho t e r i c i n B.


CHROMATOGRAPHY AND SEPARATION CHEMISTRY


100

Literature Cited 1. Snyder, L.R.; Kirkland, J. J. "Introduction to Modern Liquid Chromatography"; 2nd Edition. John Wiley and Sons: New York, 1979; Chap. 7-9. 2. Majors, R.E. J. Chromatogr. Sci. 1980, 18,488-11. 3. Kroeff, E.P.; Pietrzyk, D.J. Anal. Chem. 1978, 50, 502-11. 4. Roueliotis, P.; Unger, K.K. J. Chromatogr. 1978, 149,211-24. 5. Cooke, N.H.C.; Olsen, K. J. Chromatogr. Sci. 1980, 18, 512-24. 6. Benson, J.R.; Woo, D.J. J. Chromatogr. Sci. 1984, 22, 386-99. 7. Engelhardt, H.; Muller, H. J. Chromatogr. 1981, 218, 395-07. 8. Lee, D.P. J. Chromatogr. Sci. 1982, 20, 203-08. 9. Alton, K.B.; Lietz, F.; Bariletto, S.; Jaworsky, L.; Desrivieres, D.; Patrick, J. J. Chromatogr. 1984, 311, 319-28. 10. Lee, D.P.; Kindsvater, J.H. Anal. Chem. 1980, 52, 2425-28. 11. Greyson, R.; Patch , A. M. J. Chromatogr. 1982, 242, 349-52. 12. Buta, J. G. J. Chromatogr. 1984, 295, 506-09. 13. Gupta, R.N.; Smith, P.T.; Eng, F. Clin. Chem. 1982, 28, 1772-74. 14. Bontemps, J.; Bettendorft, L.; Lombet, J.; Grandfils, C.; Dandrifosse, G.; Schoeffeniels, E. J. Chromatogr. 1984, 295, 486-91. 15. Lee, D.P. J. Chromatogr. Sci. 1984, 22, 327-31. 16. Sykes, R.B.; Bonner, D.P.; Bush, K.; Georgoadakou, N.H. Antimicrob. Agents Chemother. 1982, 21, 82-92. 17. Cimarusti, C.M.; Applegate, H.E.; Chang, H.W.; Floyd, D.M.; Koster, W.H.; Slusarchyk, W.A.; Young, M.G. J. Org. Chem. 1982, 47, 180-82. 18. Joseph, J.M.; Kirschbaum, J.J. Abstr. 131st Apha Annual Meeting. 1984, P.92. 19. Cantwell, F.F.; Puon, Su. Anal. Chem. 1979, 51, 623-32. 20. Kroeff, E.P.; Pietrzyk, D.J. Anal. Chem. 1978, 50, 502-11. 21. Wood, R.; Jupille, T. J. Chromatogr. Sci. 1980, 18, 551-58. 22. Bidlingmeyer, Β.A. J. Chromatogr. Sci. 1980, 18, 525-39. 23. Horvath, C.; Melander, W.; Molnar, I. Anal. Chem. 1977, 49, 142-54. 24. Rotsch, T.D.; Cahill, Jr., W.R.; Pietrzyk, D.J.; Cantwell, F.F. Can. J. Chem. 1981, 59, 2179-83. 25. Margosis, M.; Aszalos, A. J. Pharm. Sci. 1984, 73, 835-38. RECEIVED April 19, 1985


Selectivity of Poly(styrene-divinylbenzene) Columns - ACS Publications

Recommend Documents