Spectral and Electrochemical Investigations of Ketorolac Tromethamine

Indian Journal of Chemistry Vol. 42B, January 2003, pp. 159-165

Spectral and electrochemical investigations of ketorolac tromethamine

G S Moses t, M Sunil Kumar *, A Ramachandraiah* and K M Rao§ * Department of Chemistry, Regional Engineering College, Warangal 506 004, India t Department of Chemistry, Government College, Rajahmundry-533 105, India * School of Chemistry, University of Hyderabad, Hyderabad - 500 046., Indi a §Department of Chemistry, Regional Engineering College, Silchar-7330 10, Assam, Indi a Received 20 March 2001; accepted (revised) 12 April 2002

The spectral and electrochemical studies of ketorolac tromethamine, 1 are presented. 1 undergoes a sin gle step two­ electron transfer reduction in media of pH < 6.00 and in a two step one-electron reduction in basic media to give a second ary alcohol. 1 is also found to act as a bidentate ligand to give 3. Electrochemical investi gati ons of 1 in non-aqueous media are O presented. Relevant electrochemical data such as n, Epc. ipc ' ana, k h, Do, etc. are reported.

Ketorolac Tromethamine, 1 is structurally related to pure aqueous standard solutions of the drug were pre­ Tolmetin and Zomepirac. It is prepared by neutrali sa­ pared from vials only for the investigations. 5 vials, tion of (±)2-benzoyl-dihydro-pyrrolizine-5-carboxylic each of a 30 mg/mL aqueous supply, were carefully acid with 2-amino-2-hydroxymethyl-1 ,3-propanediol'. emptied into a 50 mL standard fl ask equipped with a It is marketed through several trade names such as funnel. The empty vials were thoroughly washed :,ev­ Dolac, Ketanov, etc. eral times with water and the washings were also 1 is basically a non-steroidal analgesic drug having transferred into the standard flask. The funnel was similar pharmaco-kinetic actions as are other non­ removed after cleaning it to drain the washings into steroidal anti-inflammatory drugs (NSAID's) in inhib­ the flask and the contents of the flask were then .made iting the prostaglandin synthesis. Unlike other up to mark. From thi s solution, a second stage aque­ NSAID's, ketorolac has a stronger analgesic activity ous stock solution of the drug was prepared in another besides a moderate anti-inflammatory activity. It may 25 mL standard flask to get 1mM of the compound. cause gastrointestinal side effects, probably by reduc­ Buffers of different pH's (ionic strength = 0.1 M) 2 ing the synthesis and activity of prostaglandins. Spec­ were prepared according to literature procedures . tral and electrochemical studies of the drug in aque­ Their pH was measured by an A TI Orion Model 902 ous media at several pH's would provide useful in­ Ion Meter under thermos tatted conditions (25±0.1Q C). formation about the electron transfer characteri stics of The UV -Vi sible spectra of the solutions were re­ the drug which should provide an insight into its corded on a Shimadzu Model UV-160A Ratio Re­ pharmaco-kinetics. No detailed electrochemical inves­ cording Spectrophotometer fitted with thermos tatted tigations of this important drug molecule are reported 1 cm-path-Iength cuvette holder. The electrochemi cal yet in the literature. Further, its ability to serve as a measurements such as cyclic voltammetry and differ­ bidentate ligand through its carbonyl oxygen and pyr­ ential pul se polarography were recorded on an EG&G rolidic nitrogen, is not yet demonstrated. Prompted by PARC Model 264 A3 Polarographic Ana­ its medicinal importance, its structure and dearth of lyser/Stripping Voltammeter whereas the coulometry reports on its electrochemistry and coordination was on a BAS Model CV-27 Voltammograph, ac­ chemi stry , we studied the spectral and electrochemi­ cessed with an Omnigraphic 2001 x-Y /t Recorder of cal characteristics of the drug in considerable detail Di gital Electronics. A Carlo-Erba (presently Fisons) and the results of these studies are presented here. Model EA 1108 CHNS-O Elemental Analyser linked with a Sartorius Model MC5 Microbalance, was used Materials and Methods for the elemental analysis of the compounds. All the chemicals used were of AnalaR grade. Ke­ All the spectra were recorded using double wall ed torolac tromethamjne is available in the market as cuvette holders of the spectrophotometer thermostat­ both tablets and vials. As the tablets contain additives, ted by an Insref model cryostatic circul ating liquid 160 INDIAN J. CHEM., SEC B. JANUARY 2003 bath with a temperature stability of ± 0.1 DC. The cy­ the wavelength region 200-11 00 nm, suggesting either clic voltammograms and differential pulse polaro­ that the lone pair of electrons on the pyrrole nitrogen grams were run on a 6 mL sample solution transferred does not participate in the electronic transitions in the into the electrochemical cell. Before each run, nitro­ wavelength region studied because it becomes part of gen gas was purged for about 4 minutes. Buffer (5 the aromatic sextet of the pyrrole ring or that the pyr­ mL) and 1 mL of water were mixed for the blank run role nitrogen is hardly protonated because pyrrole is while buffer (5 mL) and 1 mL of the sample were an extremely poor base (pKa = OAO). The basicity of mixed for the sample run. For no~-aqueous measure­ the pyrrole moiety might be further reduced due to the ments, TEAP in acetonitrile is used as the supporting presence of the electron-withdrawing carbonyl and electrolyte. carboxy groups nearby. The drug is expected to pos­ sess three transitions, viz., the n ~ n* and n~ n* Results and Discussion transitions of the benzoyl group and the n ~ n* tran­ Spectral studies sition from the pyrrole ring. The structures of the compound and its protonated form are shown as 1 and 1H+ in Scheme I. Molecular energy minimisation studies using the software, The highest energy band below 200 nm (which PCWIN, reveals that the protonated form is stabilised could not be read due the limitation of wavelength by hydrogen bonding. The aqueous electronic spec­ range for the instrument) should be from the benzene trum of 1, recorded at pH=7 is shown in Figure 1. ring. The low intense (£ = - 8500 lil" mol -, cm-') The spectral features are near independent of pH in band at - 250 nm with a negligible shift in position as pH varies, is attributed to the n ~ n' transition of the carbonyl group. The moderately intense band (£ = - e CH;PH 24000 lil" mol -, cm-') at - 320 nm exhibits a small o 61 I red-shift in the pH range 0.16 < pH < 4.00. This band H:/'J-C-CH;PH I may be originating from the n ~ n * transition of the + CH;PH conjugated three double bonds of the pyrroyl moiety 1H (which includes the carbonyl's double bond). The spectral data of 1, in the wavelength region 1100-200 nm, are collected in Table I.

CH;PH Table I - Electronic spectral data of ketorolac 61 I romethamine* Hj'J-C-CH;PH I pH ')...'(d CH;PH 0.16 317.0 (22,763) 250.0 (10,630) Scheme I 4.20 321.6 (23,515) 249.4 (8,875) 6.50 322.4 (27,576) 1 Absorbance 249.4 (9,326) 0.8 10.0 322.6 (23,816) 0.6 249.4 (8,473) a, in nm; b, lie' mor'cm- '; * in methanol at 25°C 0.4 ._-----

0.2 Electrochemical studies

0 Organic compounds containing carbonyl group 200 250 300 350 400 have evinced considerable research interest among Wavelength (nm) electrochemists for the group's versatile electro­ chemical behaviour with a mechanism that varies with Figure I-Electronic spectrum of 1 (lxlO-.5 M) atpH=7.02 the solvent, pH, nature of the electrode material and MOSES et ClI.: SPECTRAL & ELECTROCHEMICAL STUDIES OF KETOROLAC TROMETHAMINE 161

5 312 ip = 2.69 X 10 n A D"2 yll2 c where ip is the peak current, A the area of the elec­ trode, D the diffusion coefficient, y the scan rate and c t­ Z the concentration of the electroactive species'. The W value of n calculated by this method for 1 was 2 in 0::: 0::: media of pH < 6.00 and 1 for each of the two reduc­ ;::J U tion steps in basic media. Hence it is assumed that the only reducible site i.e., carbonyl group of 1 undergoes electrochemical reduction through different mechanis­ tic pathways in acid and basic media. Assigning a mechanism to the two-electron single step electro­ -1.0 -1.1. -1.8 chemical reduction in acid medium involving two H+ POTENTIAL ions is rather straight forward because it is reported V vs Ag/AgCI that keto groups undergo facile 2 e', 2 H+ reduction to Figure 2-Cyclic voltammograms of 1 (1.6 X 10-4 M) on SMDE give secondary alcohols6. In strongly acidic media, at pH of (a) 2.20 and (b) 10.02 (scan rate, 50 mY/sec) the tertiary nitrogen is protonated, even if it is to a negligible extent. Upon protonation, the -NH- hydro­ 3 6 gen is likely to be engaged in hydrogen bonding with other electrochemical experimental conditions - . Cy­ the carbonyl oxygen as shown in IH+. The carbonyl clic voltammograms of 1 are exhibited in Figure 2 at group is more polarised in IH+ than in Reduction two representative pH's. It is clear from these polaro­ 1. of IH+ is more facile in this medium because of thi s grams that the electrochemical behaviour of the drug enhanced polarisation and high concentration of H+ depends on the pH of the medium. It seems to un­ ions. The proposed mechanism is presented in dergo a single step reversible electron transfer in me­ Scheme Plots of Ep vs pH were linear with a near­ dia of pH below 6.00 and in a two step successive II. uniform slope of 0.1162 in buffers of pH 6.00 and irreversible transfer in those above 6.00. It is further < in those of pH The number of H+ ions evident from this figure that the reduction peak poten­ 0.05982 > 6.00. calculated from these slopes was 2 in acid range and I tials shift cathodically as the pH increases. This trend suggests the involvement of H+ ions in the electro­ for each of the two peaks in the basic media, further supporting the mechanism of Scheme II. chemical reduction step. Usually, the reduction of carbonyl group is irre­ The peak potential, Ep varies with pH at 25° Cas, versible and is complicated with catalytic currents. The free radical intermediate undergoes a chemical Ep = EOp - m (0.05916/n ) pH where EOp is the peak dimerisation before the potential for the second elec­ potential at a reference pH (activity of H+ I or 8 = tron addition is arrived . One of the best ways of un­ pH=O). Hence, a plot of Ep vs pH should give a derstanding the electrochemkal behaviour of carbonyl straight line from the slope of which one can calculate group is to study the effect of scan rate, y and concen­ the number of H+ ions involved in the electrochemkal tration, c of the compound on its cyclic voltammetric reduction step, provided, one knows the value of n by peak current, ip. Plots of ip vs yll2 are linear in acid coulometry or by other means. Coulometric studies of range and slightly non-linear in basic range. As the 1 in several buffers of a wide pH range gave a value scan rate Increases,. th' e lp vs Y 112 P1 ot become s more of 2 for n when electrolysis was carried out at poten­ linear in basic media. At high scan rates, the interme­ tials cathodic to the cyclic voltammetric response. diate is quickly brought to the second electron transfer Attempts to measure the number of electrons involved potential leaving hardly any time for dimerisation. in each of the two reduction peaks in the media of pH Relevant electrochemical data obtained from these > 6.00 were unsuccessful. This suggests that the two studies are collected in Table I. peaks correspond to the electrochemical reduction of Even though the electronic spectral behaviour of 1 the same site but in consecutive electron transfer is similar in aqueous and non-aqueous media, its mechanism. The number of electrons, n for any diffu­ electrochemical response cannot be expected to be so sion-controlled peak of cyclic voltammetry can be because most of the electrochemically important non­ evaluated from the application of Randles-Sevcik aqueous solvent are aprotic in nature and hence can equation, not yield H+ ions for reduction. The electrochemical 162 INDIAN J. CHEM., SEC B, JANUARY 2003

Acidic media

COOY COOY 2H+ , 2e- / O------H

1H+ 2H+ Basic media

o ~/COOY H+, e- COOY ~~~ --~---.. o 1

COOY H+ COOY o , ..

Scheme II redox behavior of 1 was examined in non-aqueous solvents such as acetonitrile, dimethyi formamide and dimethyl sulphoxide on stationery mercury drop electrode (SMDE) in the potential region, +0.200 to - 2.00 V Ys Ag IAgCl . A typicai cyclic voltammogram of 1 in acetonitrile medium on SMDE is shown in Figure 3. Electrochemical reduction of organic compounds in non-aqueous media usually involve I­ participation of the solvent molecules if no other Z 9 lLl reactive solute is available in the medium . If the non­ c::: c::: aqueous medium is very dry, electrochemical ~ u mechanism at mercury electrode passes through formation of radical ions. In the case of 1, an irreversible reduction peak is observed at - -0.85 V vs Ag I Age!. In the reverse scan a sudden adsorption -01}- -0.8 -1.2. complicated cathodic spike is observed at - -0.575 V, POTENTIAL (VOL1S) followed by a normal anodic peak at - -0.40 Y. When the cyclic voltammetric cycle are continued, there Figure 3--Cyclic voltammograills of 1 (1.6 X 10·'1M) on SM DE are two reduction peaks, one at - -0.625 V and the in acetonitrile with TEAP (2 x 10':; M) as the su pporting electro­ other at - -0.750 V from the second cycle onwards. lyte (scan ra te, 50 mY/sec) MOSES et ai.: SPECTRAL & ELECTROCH EM ICAL STUDIES OF KETOROLAC TROMETHAMINE 163

This spike is repeated in all the subsequent cycles. contents of the supernatant acetonitrile solution gave When a trace quantity of water is added to this an additional TLC spot below the one expected for 1. medium and the cyclic voltammogram recorded, no According to Scheme III, electrolys is at this potential such spikes are observed. A plausible mechanism, as should result in only the l,2-diol dianion into solution shown in Scheme III is proposed. A partial proof for through the reduction process (1) and the mercuryl this scheme is obtained from controlled potential anion species on the electrode surface through the electrolysis of 1 in acetonitrile on a mercury pool surface reaction (2). The other products of working electrode at --1.00 V vs Ag I Agel. The Scheme III can form only by electrochemical means, QjCOOY COOY COOY kd ... (1 ) o - -0.85 V dimerisation i{jCOOY

08

COOY COOY n e- ---:--.~ Reduction Products ... (2) surface - -0.525 V

COOY COOY 2 - 2e- + 2 HgO .. .(3) --O.40V • ~ YOOC ~

COOY

.. . (4) HgO Surface YOOC - -0.60 V YOOC

COOY

COOY HgO Surface .. . (5) - -0.75 V

YOOC

Scheme III 164 INDIAN J. CHEM., SEC B, JANUARY 2003 triggered by the oxidation of the mercuryl anion 0.6 species only at (--0.400 V) a potential anodic to the set controlled potential, -1.000 V. Since th is can not 0.5 happen in controlled potential electrolysis, the other 0.4 products do not form into solution. Subsequently, the ..

Table III - Spectrophotometric and differential pulse polarographic assaying data of ketorolac tromethamine in aqueous buffers

Photometrt Polarographl Sample Amount Amount Recovery Amount Recovery labeled recovered (%) recovered (%) (mg) (mg) (mg)

1 30 28.95 96.50 29.83 99.40 2 30 29.02 96.73 29.72 99.06 3 30 29.12 97 .07 29.85 99.50 4 30 28 .99 96.63 29.82 99.43 5 30 29.42 98.06 29.96 99.87

a, in aqueous media of pH= 6.50 at 322 nm t b, in aqueous media of pH= 10.00 at SMOE at scan rate 5 mYs· , pulse height 100 mY (fi rst peak) and drop time at 1 sec access to their Molecular Modeling facilities. One of the authors (MSK) is thankful to the authorities of AICTE for the financial assistance (F.No. 802- I1RDIIIR&OI Rec. 167 dated 24th January 1995), in the form of a Project Assistant.

References 1 Technical pamphlet on ' Oolac' from Candila Chemicals Lim­ ited, Ahmadabad. 2 Ju Lurie, Hand book of Analytical Chemistry, (Mir Publishers, Moscow), 1978. 3 3 Clarke S & Harrison J A, J Electroallal Chern , 36, 1972, 9. 4 Fung K & Grosjean G, Anal Ch em, 53, 1981 , 168. polarography in basic aqueous buffers. The assaying 5 Petropavlovski, Zh Allal Khim , 30, 1975, 7. data are presented in Table III. 6 Yijay Kumar K& Jayarama Reddy S, fndian J Chel1l. 25. 1986, 223. Acknowledgement 7 Bard A J & Faulkner L R, Electrochemical Methods Funda­ mentals and Applications (John Wiley, New York), 1980. The authors are very thankful to the Indian Institute 8 Oolores M & Francisco M, J Ch em Edu, 65, 1988,916. of Chemical Technology, Hyderabad for providing 9 Rakoutz M, Ch em Abstr, 79, 1973, 73006 x.