rowe & machkovech (1977)

Download Rowe & Machkovech (1977)

Post on 13-Apr-2018




0 download

Embed Size (px)


  • 7/27/2019 Rowe & Machkovech (1977)


    Rap id Acetylation of a Dihydroxy Compound by4- Dimethy1amino)pyridineCatalysis: A pplication t o GLCDetermination of Clindamycin Pa lmita te HydrochlorideENGLEBERT L. ROWE and S U S A N M. MACHKOVECH

    Abstract 0 The acetylation of clindamycin palmitate hydrochloride,preparatory to GLC assay, was accelerated greatly by the useof 4-(di-methy1amino)pyridineasa catalyst. The reaction time was reduced from2.5 hr at 100' using pyridine and acetic anhydride to 30min or less atambient temperature upon addition of excess catalyst.Keyphrases Clindamycin palmitate hydrochloride-GLC analysisafter acetylation catalyzed by 4- dimethy1amino)pyridine 0 4-(Di-methy1amino)pyridineatalyst for acetylation of clindamycin palmitatehydrochloride preparatory to GLC assay GLC-analysis, clindamycinpalmitate hydrochloride, after acetylation catalyzed by 4-(dimethyl-amino)pyridine Dihydroxy compounds-clindamycin palmitate hy-drochloride, GLC analysis after acetylation catalyzed by 4-(dimethyl-amino)pyridine Antibacterials-clindamycin palmitate hydrochloride,GLC analysis after acetylation catalyzed by 4-(dimethy1amino)pyri-dine

    The acetylation of clindamycin palmitate hydrochloride(I)for GLC assay is catalyzed by the use of pyridine, whichalso acts as a solvent for the reaction (1).Pyridine formsan intermediate with acetic anhydride, acetylpyridiniumion (21, which then reacts with the hydroxyl groupsat the3-and 4-positions of clindamycin palmitate hydrochloride.The reaction requires 2.5 hr at 100 for completion.4- Dimethy1amino)pyridine (11)recently was reportedto be a much better catalyst than pyridine for various ac-ylations (2, 3). The present study compares the rate ofacetylation of I in pyridine in both the presence and ab-sence of 11.

    EXPERIMENTALMaterials-All materials were analytical reagent grade.GLC Conditions-A gas chromatograph' equipped with a flame-ionization detector and an electronic integrator2 was used. The glassU-shaped column, 3 mm X 60 cm, was packed with 1 SE-54 on GasChromQ (100-120 mesh). The column was conditioned for 1week at 295'under low helium flow. The operating temperatures were: column, 275';

    CH3IH l1H,I



    F M model 402.Infotronics model CRS-104.

    detector, 290'; and flash heater, off. Helium was used as the carrier gasa t about 60mllmin. One microliter of sample was injected, and the at -tenuation was adjusted to give peak heights averaging about 50 of thechart heights.In te rnal Sta ndard Solution-A solution of trilaurin in chloroformwas used at a concentration of 5 mg/ml.Procedure-About 75 mg of I was accurately weighed and dissolvedin 5 ml of double-distilled water. One milliliter of 30 sodium carbonatesolution and 5 ml of internal standard were added, and the solution wasshaken vigorously for 10min and centrifuged to form distinct layers. Theaqueous layer was aspirated off, and 1-ml portions of the chloroform layerwere transferred to 15-ml centrifuge tubes. One milliliter of acetic an-hydride and 1ml of pyridine were added to the tubes, and the tubes werestoppered and allowed to react at t he appropriate temperature.In the standard method, the reaction was carried out in an oven at 95Oand sampled at various intervals. In the modified method,I1was dissolvedin pyridine a t a concentration of 15 mglml. One milliliter of this solution(instead of pure pyridine) was used for the acylation medium. The tubeswere held at ambient temperature (23-26') and sampled for GLC in-jection a t various times.

    RESULTS AND DISCUSSIONThe progress of acetylation of I using I1as a catalyst at room temper-ature is compared to the pyridine-catalyzed reaction at 95O in Fig. 1.TheGLC peak for diacetylated I becomes sharper as he reaction progresses.The disappearing peak on the tailing edge of the main peak is the 3-monoacetate ester of 1. It was reported (4) that the 3-hydroxy group isthe least sterically hindered hydroxyl group in the lincomycin3molecule

    a ) 7 11 10 m i n 30 m i n6

    . 5l-.4c 3 3I 2w

    10 0 2 4 6 81012MINUTES

    I r I 4 5 m i nlr;180 min

    Figure 1-Gas chrom.atograms illustrating the acetylation of I . Key:a, 4- dimethylamino)pyridine added to pyridine, reaction at roomtemperature; and b, pyrid ine alone, reaction at 95 .

    3 Clindamycin is the 7-deoxy-7-(S)-chloro analog of lincomycin.

    Vol 66,N o . 2 February 1977 1273

  • 7/27/2019 Rowe & Machkovech (1977)


    Table I-Compariso n of Two Procedures forAcetylation of I GLC ssay)CH3CO),0 + N@NIGH, , -I1C H ~ C O ~ ~ N ( C H , ) , CH,COO-


    and, therefore, he most reactive. The 4-hydroxy group reacts more slowlyand is the controlling actor in the diacetylationof I.The shoulder on theleading edge, always present even in highly purified lots of I, is the acetateesters of clindamycin B-2-palmitate (the 4'-ethyl analog of I) and clin-damycin 3-palmitate,as ndicated by Brown ( 1 ) . In practice, this latterpeak is not integrated with the main peak.The acetylation reaction, using I1as a catalyst, was virtually completewithin 30 min at ambient temperature (Fig. l a ) . n the absence of 11 thereaction was not complete until well after 2 hr a t 95O (Fig. l b ) .The quantitative GLC data are given in Table I.Clearly, the two pro-cedures differed considerably in their reaction rates. Even after 26 hr,the peak area ratio for the 11-catalyzed reaction was still at the plateauvalue without significant change.The complete acetylation of I is a two-step reaction as already de-scribed. Acetylation of the more labile 3-hydroxy group is fast, so thereaction iscontrolled primarily by the second acetylation of the4-hydroxy



    Jtu 10a0 8Iz 6IzluVu 4un


    1 50 100 150 200MINUTES

    Figure2-Pseudo-first-order plot of acetylation of by acetic anh y-dride. Key: 0 yridine on ly, reaction a t 95O; and 0 4-(dimethyl-amino)pyridine added to pyridine (15 mglm l), reaction at room tem -perature.

    PeakAcetylation AreaProcedu re Minutes Sample Ratio AveragePyridin e, 95 30 1 1.476 1.2492 1.02260 1 1.640 1.5482 1.5122 1.492120 1 1.894 1.9171 1.9632 1.9172 1.895180 1 2.038 2.0242 2.009I1 in pyridine, 10 1 1.694 1.822

    2 1.77130 1 2.050 2.1202 2.214room temperature 2 2.002


    607 590

    1 2 026 hr





    group. The rate of reaction is dependent on the concentrations of I, 11,pyridine, and acetic anhydride. Since the latter three reagents are presentin large excess compared to I,they c nbe considered constant. Therefore,the reaction can be approximated by the simple first-order expression:

    In integrated form, expressing concentration in percent. .where [ o ] is the initial concentration of I, and [I] is the concentration a ttime t .In Fig. 2, the data are plotted according to Eq. 2. The slopes differ bya factor of about nine. In other words, evenwt a temperature differenceof 70 , he room temperature reaction usingI1asa catalyst was about ninetimes faster than the pyridine-catalyzed reaction at 95'. If the temper-atures were equal, the differences would be several orders of magnitude.Connors and Albert (2) found that the rate constant usingI1was 2 X 104greater than th at of the corresponding pyridine-catalyzed acetylationof 2-propanol a t 54O.The intercepts from least-squares regression analysis of the data areconsiderably less than low s can be seen from Fig. 2. This result is dueto the stepwise nature of the acetylation, which is better represented by

    two consecutive first-order reactions rather than by a single first-orderreaction. If sufficient early time points were available, it would be ap-parent that there is a steep slope initially which would intersect the y-axisat about100 .The mechanism of the catalytic action of pyridine involves formationof an intermediate with acetic anhydride. The acetylpyridinium ion thenundergoes nucleophilic attack by the hydroxy group. There is spectro-scopic evidence thqt the acetylpyridinium ion has a very small stabilityconstant because its presence is almost undetectable (5). Connors andAlbert (2) suggested that the intermediate formed by I1 has enhancedstability due to the resonance forms made possible by the dimethylaminogroup (Scheme I).The modified assay was applied to other 2-acyl esters of clindamycinranging in chain length from 10 to 18carbon atoms. The reaction rateswere about equal to the palmitate, requiring no more than 30 min atambient temperature in any case. Compound I1 should prove to be aversatile catalyst for acylation of other hydroxy compounds being pre-pared for GLC analysis.

    274 I Journal of Pharmaceutical Sciences

  • 7/27/2019 Rowe & Machkovech (1977)


    REFERENCES(1) . W. Brown,J . Pharm. Sci., 63,1597 (1974).(2)K.A. Connors and K. S. Albert, ibid., 62,845 (1973).(3)W. Steglich and G. HBfle, Angew. Chem., Int. Ed., 8, 981(1969).(4)W. Morozowich,F.A. MacKellar, and C.Lewis,presented a t APhAAcademy of Pharmaceutical Sciences, Washington meeting, Apr.1970.(5) G. H. Schenk, P. Wines, and C. Mojzis, Anal. Chem., 36, 914


    ACKNOWLEDGMENTS AND ADDRESSESReceived January 21,1976, rom Pharmacy Research, The UpjohnAccepted for publication March 29,1976.The authors are grateful to Mr E. J. Kubiak at The Upjohn Co. forproviding materials and advice on column conditioning and assay pa-rameters and to Dr: w. Morozowich for suggestions and consultations.Present address: School of Pharmacy, University of Michigan, AnnArbor, MI 48104.

    Company, Kalamazoo, MI 49001.

    To whom inquiries should be directed.

    Tissue Distribution of 3H-Canrenoate Potassium in RabbitsA. MICHAELS FINN, R. BROWN, and W. S A D B E .

    Abst ract Plasma and various organ concentrations of canre