A

KI

a

ARR1A

KPTCDE

1

dabate

eit

armtsapmv

1d

Spectrochimica Acta Part A 75 (2010) 901–907

Contents lists available at ScienceDirect

Spectrochimica Acta Part A: Molecular andBiomolecular Spectroscopy

journa l homepage: www.e lsev ier .com/ locate /saa

theoretical and spectroscopic study of conformational structures of piroxicam

ely Ferreira de Souza, José A. Martins, Francisco B.T. Pessine, Rogério Custodio ∗

nstituto de Química, Universidade Estadual de Campinas - UNICAMP, P.O. Box. 6154, 13083-970 Campinas, São Paulo, Brazil

r t i c l e i n f o

rticle history:eceived 7 April 2009eceived in revised form2 November 2009ccepted 8 December 2009

eywords:iroxicam

a b s t r a c t

Piroxicam (PRX) has been widely studied in an attempt to elucidate the causes and mechanisms of itsside effects, mainly the photo-toxicity. In this paper fluorescence spectra in non-protic solvents anddifferent polarities were carried out along with theoretical calculations. Preliminary potential surfacesof the keto and enol forms were obtained at AM1 level of theory providing the most stable conformers,which had their structure re-optimized through the B3LYP/CEP-31G(d,p) method. From the optimizedstructures, the electronic spectra were calculated using the TD-DFT method in vacuum and including thesolvent effect through the PCM method and a single water molecule near PRX. A new potential surface

otal fluorescenceonformational analysisensity functional theorylectronic spectra

was constructed to the enol tautomer at DFT level and the most stable conformers were submitted tothe QST2 calculations. The experimental data showed that in apolar media, the solution fluorescence israised. Based on conformational analysis for the two tautomers, keto and enol, the results indicated thatthe PRX-enol is the main tautomer related to the drug fluorescence, which is reinforced by the spectraresults, as well as the interconvertion barrier obtained from the QST2 calculations. The results suggestthat the PRX one of the enol conformers presents great possibility of involvement in the photo-toxicity

mechanisms.

. Introduction

Piroxicam (PRX), Fig. 1, is a non-steroidal anti-inflammatoryrug (NSAID). NSAID is between the most utilized therapeuticgents. There are more than 50 different kinds of substanceselonging to this class [1]. Such agents present three main effects:nti-inflammatory, analgesic and antipyretic. However, none ofhese substances acts on organism without provoking undesirableffects.

Actuating as therapeutic agents, the NSAIDs inhibit the COX-2nzyme, promoting the desired effects. On the other hand, they alsonhibit the COX-1, an isoenzyme of COX-2, which probably causeshe side effects on organism [1].

PRX is recognized by its efficiency as anti-inflammatory andnalgesic. Its utilizations include treatment of osteoarthritis andheumatoid arthritis, fractures, post-operative, acute pains in theuscular skeletal system, etc. [1–3]. It presents a long life-time in

he organism compared to similar drugs [1] resulting in the pos-ibility to manage only one daily dose to the patient. It is quickly

bsorbed by the organism and its anti-inflammatory activity sur-asses the Indometacine, Naproxen and Phenylbutazone [1]. Itsolecular structure presents fast tautomeric equilibria presenting

arious isomers [4]. Furthermore, the existence of intramolecu-

∗ Corresponding author. Tel.: +55 19 35213104; fax: +55 19 35213023.E-mail address: roger@iqm.unicamp.br (R. Custodio).

386-1425/$ – see front matter © 2009 Elsevier B.V. All rights reserved.oi:10.1016/j.saa.2009.12.031

© 2009 Elsevier B.V. All rights reserved.

lar hydrogen bonds forming six member rings leads to differentconformers in the fundamental state [5]. The excited state protontransfer, leading to the interconversion between the keto and enolforms [6], is also an important feature of PRX.

The photochemical properties of PRX are sensitive to themedium. Solvent parameters like polarity, pH, hydrogen bonds orviscosity can affect the conformation, as well as the drug fluores-cence [7].

Among the side effects of PRX, photo-allergy is observed in 1% ofthe patients [8]. There are not enough experimental data to affirmif it is triggered by metabolites or photoproducts (stable or not)or even by the drug itself in some stable conformational structure.Westerm et al. from in vitro studies of photo-toxicity indicated aPRX metabolite as the responsible agent [9]. A possible mecha-nism involving the PRX-enol and the molecular singlet oxygen wassuggested by Lemp et al. [10] from kinetic studies and semiempiri-cal calculations. Alencastro and Neto [11] optimized the molecularstructure of three anti-inflammatory non-steroids, including PRXand a metabolite, using AM1 method and carried out calculationsof electronic spectra using the INDO/S method. They concludedthat the photo-toxicity mechanism could not be understood andattributed to the action of a single specie or only one of the possible

tautomers.

In order to improve the level of knowledge on structural andspectroscopic properties of PRX, in this paper we are investigatingthe photophysical processes of PRX in solvents with very differentpolarities through different mixtures of two non-protic solvents

902 K.F.d. Souza et al. / Spectrochimica Acta Part A 75 (2010) 901–907

to–eno

(te

2

oeet

iwts(r−waii

Ftt

ooft�fe2eTfp

TSc

water molecules and preserving the PCM model were also car-ried out. The re-optimized structures at the B3LYP/CEP-31G(d,p)level of theory were submitted to the calculation of the six firstelectronic transitions in vacuum and in the presence of water

Fig. 1. PRX-ke

CH3NO2 and CCl4). The PRX conformational structures and its elec-ronic spectra were obtained by theoretical calculations in order tostablish correlations with the experimental data.

. Experimental and computational methods

The PRX total fluorescence emission was obtained in mixturesf solvents with different polarities (CH3NO2 and CCl4). The totalmission consists of the emission spectra collected over the wholexcitation spectral range. The solutions were prepared accordingo the scheme in Table 1.

Reference solutions were made with CH3NO2 and CCl4 solvents,n the absence of PRX. The total emission surfaces for all solutions

ere recorded on a SPF-500C, SLM AMINCO conventional spec-rofluorimeter. Excitation spectra from 260 nm to 390 nm (2 nmteps) were monitored with emission from 400 nm to 600 nm10 nm steps). Emission and excitation slits were 4 nm and 10 nm,espectively, and the reference and fluorescence PMT voltages were525 V and −990 V, in this order. The fluorescence data examinedere the difference between the surfaces of solution total emission

nd its respective references, because Raman and Rayleigh scatter-ngs and low residual fluorescence (from 220 nm to 280 nm) couldmpair the analysis of results.

The keto–enol equilibrium provides two possible tautomers.ig. 1 shows that the two main structures differ in the localiza-ion of the C(11)–C(14) double bond. These species were taken ashe primary isomers for the structural search.

An initial exploratory analysis of the conformational structuref PRX was carried out at the AM1 semiempirical level of the-ry. The keto and enol most stable structures were determinedrom an energy surface for each isomer calculated with respect tohe dihedral angles defined by �1 = C(11)–C(14)–C(23)–O(24) and2 = O(24)–C(23)–N(25)–C(27). Both dihedral angles were changed

rom 0◦ to 360◦ with steps of 15◦, totalizing 625 optimizations forach conformer, using the AM1 method included in the Gaussian003 package [12]. The energy surface of each isomer (keto and

nol) provided six stable structures being two of them equivalent.he corresponding conformers were re-optimized using densityunctional theory with B3LYP functional and a compact effectiveseudopotential (CEP) along with the respective basis set CEP-

able 1olution composition for total emission data registration (PRX concentration wasonstant at 1 × 10−5 mol L−1).

Solution [CH3NO2] (mmol L−1) Molar fraction CH3NO2 (in CCl4)

a 0.0 0b 37.1 0.0020c 74.2 0.0041d 111 0.0061e 222 0.0123f 445 0.0246g 890 0.0492h 1780 0.0984

l equilibrium.

31G(d,p). The 31G(d,p) basis set used along with CEP is a doublezeta valence basis equivalent to 6-31G(d,p) where the 31G(d,p)structure is preserved and the six inner Gaussians are changed to apseudopotential. The final result is comparable to an all-electron 6-31G(d,p) calculation, but using a smaller number of Gaussians andconsequently lesser computational effort. Fig. 2 is an illustrationof the potential surface obtained for the enol isomer through 625optimizations using B3LYP/CEP-31G(d,p) and reproduces the sixminima observed by the semiempirical method. Structures A andB are equivalent as well as C and D. The rotation of both dihedralangles provides small differences in the equivalent structures as isthe case of C and D forms, which yield small changes in the equilib-rium energy of both structures. However, they will be consideredessentially as the same structures.

The intra- and intermolecular hydrogen bonds are considerablyimportant in the drug behavior. These interactions take place inthe neighborhood of the oxygen atoms, O(12) and O(24) of PRX.The compound presents little solubility in water. However, in orderto consider solvent effect in the calculations, water was chosenbecause it provides stronger dipolar interactions and hydrogenbonds than several solvents. If water causes no significant effecteither by local interaction or field effect on PRX, other solventswill probably provide weaker interactions. Otherwise, in the caseof evidences of strong interactions, some particular solvents mustbe studied. The water effect on the electronic structure of PRXwas taken into account through a hybrid method used to calcu-late the absorption UV–VIS electronic spectra. Explicit presenceof water molecule was considered near O(12) and O(24) of PRXand the solvent effect was improved through the polarized con-tinuum model (PCM) [13]. Some tests excluding the presence of

Fig. 2. Total energy system as a function of �1 and �2 (in degrees) enol PRXform—B3LYP/CEP-31G(d,p) method.

ica Acta Part A 75 (2010) 901–907 903

u[

tnB

3

(tpSd[ftorrsmenbi

nsgidfmstaplr

tm0tsiect

K.F.d. Souza et al. / Spectrochim

sing the time dependent density functional (TD-DFT) method14].

After identification of the most stable isomer and the elec-ronic spectrum, the interconversion barriers among the nearesteighbors were calculated using the QST2 method [16] at the3LYP/CEP-31G(d,p) level of theory.

. Fluorescence analysis

Several works on the excited state proton transfer phenomenonESPT) as a function of the solvent polarity and pH can be found inhe literature [15,17–24]. Significant influence in the photophysicalroperties of some substances has been investigated [15,17–21].ubstances, such as PRX, present very little fluorescence or eveno not fluoresce when dissolved in protic (as water) solvents25,26]. Specific solute–solvent interactions or a distinct stable con-ormer that cannot provide intramolecular hydrogen bonding ishe explanation generally accepted for such behavior. The absencef hydrogen bonds in the ground and excited states provides lessigid structures. Therefore the molecule can be deactivated by non-adiative processes. When the solvent interference is small andpecific structures provide an intramolecular hydrogen bond, theolecule keeps the absorbed energy and can emit intensely in sev-

ral spectral ranges [21,27]. For PRX, when the molecule is in aon-protic solvent, the proton transfer of H(13) (Fig. 1) takes placeetween O(12) and O(24) without hindrance [28,29] because of the

ntramolecular hydrogen bonding.Some investigators have used HPLC, RMN 1H, and other tech-

iques to study the solvent influence on keto–enol equilibrium foreveral compounds [30–35]. The major interest in these investi-ations is related to solute–solvent interactions that depend onntrinsic characteristic of each component in the solution, such asipole moment, acidity, basicity, possibility of a hydrogen bondormation, dielectric constant, etc. The solute–solvent interactions

odify in a significant way the keto–enol equilibrium, affecting thepectroscopic behavior of PRX and consequently, PRX can comeo enol or keto forms. In most cases [34], the enol form, having

smaller dipole moment, is more stable in non-polar solventsroviding intramolecular hydrogen bonds that turn the molecu-

ar structure more rigid, hindering deactivation through vibrationalelaxation and increasing the fluorescence quantum yield.

The excitation spectral bands (Fig. 3) show a red shift andhe emission spectral range does not suffer significant displace-

ent, as the concentration of nitromethane is 0.000, 0.111 and.445 mol L−1, respectively. Also, the larger the solvent polarity (ashe concentration of nitromethane increase), the smaller the emis-

ion intensity. These facts suggest that solute–solvent interactionsnduce modifications in the excited state energy levels. Even so, themission state is the same because there is no shift of the fluores-ence band. Furthermore, the more polar the solvent is, the largerhe excitation band red shifts. There is an equilibrium between the

Fig. 3. Total emission surfaces for three PRX solutions in CCl4 and CH3NO2:

Fig. 4. Relative energies between the conformers after re-optimization—B3LYP/CEP-31G(d,p) method.

PRX-enol (that emits) and -keto (that does not emit) forms as afunction of the solvent polarity. There is a strong decrease in theemission when the polarity increases because the keto form is morestable than the enol form in a polar solvent.

4. Conformational analysis

Fig. 4 shows the relative energies obtained from the equilib-rium geometry at the six minima for each isomer calculated at theB3LYP/CEP-31G(d,p) level of theory. The total energies indicate thatthe enol A (or B) structure is more stable than all the other enoland keto forms. The enol A conformer is approximately 51 kJ mol−1

more stable than enol C (or D) and 41 kJ mol−1 than keto A(or B) andC(or D). The comparison between the keto A and keto C, the ketomost stable structures, and enol C shows that the first structure is∼=10 kJ mol−1 more stable than the last one. All the other structures(enol E and enol F or keto E and keto F) are considerably higherin energy with respect to enol A and keto A (see Fig. 4). Therefore,enol A is energetically one of the dominant structures of the sys-tem. Another very important characteristic is the planarity of theenol structures regarding the keto forms. The planarity is identifiedas one of the responsible by the fluorescence associated with theenol form.

The fluorescence of the PRX-enol is related to its conformationalstructure that should obey two basic conditions: larger planarityand presence of an intramolecular hydrogen bond. The first rein-forces that the enol form is more stable than the keto structurebecause the electronic delocalization is raised. The second turns the

structure more rigid, decreasing deactivation by non-radiative pro-cesses. The keto form does not fluoresce because its conformationalstructure is neither planar nor rigid and non-radiative processestake place, reducing the molecular fluorescence quantum yield.

(a) 0.000 mol L−1, (b) 0.111 mol L−1 and (c) 0.445 mol L−1 of CH3NO2.

904 K.F.d. Souza et al. / Spectrochimica Acta Part A 75 (2010) 901–907

Table 2Atomic distances obtained from the geometry of the more stable conformers after re-optimization – B3LYP/CEP-31G(d,p) – keto and enol forms.

Conformer H(26)–N(15) O(12)–O(24) Conformer H(26)–N(15) H(13)–O(24)

sgbts

tat

ihP

5

ewitstissceea(3(i

fftnptf

TE

O

Keto A 2.92 3.42Keto B 4.00 4.30Keto C 4.24 4.03Keto D 4.50 3.61

Some interatomic distances in the conformational structureshown in Table 2 indicate the presence of the intramolecular hydro-en bond interactions and carbonyl group repulsion. The distancesetween the atoms O(24) and H(13) (1.58 and 1.56 Å) are smallerhan a typical hydrogen bond (1.7 Å in water [36]) for the mosttable structure in PRX-enol.

Furthermore, H(26)–N(15) distances for enol form is shorterhan the keto ones, suggesting stronger interaction. These inter-ctions promote molecular rigidity in all enol conformers becausehese atoms can build an intramolecular hydrogen bond.

The repulsion between O(12) and O(24), which is possible onlyn the keto form, as well as the methyl-pyridinic group stericindrance, precludes planarity and resonance stabilization in allRX-keto conformers.

. Electronic spectra

In order to investigate the nature of the electronic structure, themission spectra should be calculated in order to compare directlyith experimental data available and to investigate the sensitiv-

ty of the electronic structure with the solvent effect. However,he difficulty in obtaining the molecular structures at the excitedtates for large molecules like PRX and accurately describe the elec-ronic emission transitions suggested a simpler approach in thenterpretation of the electronic effects. The time dependent den-ity functional theory (TD-DFT) for the calculation of the absorptionpectra has been used with acceptable accuracy and computationalosts accessible to large structures like the present one. The UV–VISxperimental absorption spectra of PRX is also available in the lit-rature in absolute ethanol giving four maximum absorption bandst 3.45 eV (360 nm), 4.26 eV (291 nm), 4.85 eV (256 nm) and 6.05 eV205 nm) [37]. From pH 5 to 12, the maximum bands are shifted to.52 eV (352 nm), 4.34 eV (286 nm), 4.94 eV (251 nm) and 6.05 eV205 nm) [37]. The maximum absorption band at 3.45 eV (360 nm)s the same observed in water solution [38].

The application of TD-DFT yielded the results shown in Table 3or the most intense transitions of the electronic spectra in vacuumor enol conformers, according to the great oscillator strength to

hem. On the other hand, the keto forms present oscillator strengthear to zero in all cases. The most stable enol conformer – A – alsoresents the most intense transition in 3.55 eV, considerably closeo the experimental data mentioned above. Except for the con-ormer keto C, the transition energies of the enol form are smaller

able 3lectronic spectra – most intense transitions to enol tautomer – TD-DFT B3LYP/CEP-31G(

Conformer Most intense transition Excited state

Enol A HOMO → LUMO 1st

Enol C HOMO (−2) → LUMO 1stHOMO → LUMO

Enol E HOMO (−2) → LUMO 1stHOMO → LUMO

Enol F HOMO (−2) → LUMO2ndHOMO (−1) → LUMO

HOMO → LUMO

bs: Enol B and enol D are equivalent with enol A and enol C, respectively.

Enol A 2.29 1.58Enol B 2.29 1.58Enol C 3.91 1.56Enol D 3.91 1.56

than the keto forms (results not shown), suggesting greater possi-bility of absorption in the UV–VIS region for the PRX-enol. The E andF enol conformers present for all structures studied the largest tran-sition energies even for keto forms and the results will be shownas an example only for enol.

The results including the solvent effect through a continuummodel (Table 4) show greater strength oscillator values comparedwith those obtained in vacuum for the enol and keto forms. Further-more, the same previous tendency is observed: the enol speciespresent higher values than the keto ones. The most stable enolconformer – A – showed, once more, the most intense transitions.Despite the change in the intensities, no significant changes wereobserved for any tautomer in the transition energies in the presenceof the solvent effect through the PCM model.

In an attempt to analyze some local solvent effect, calculationswere carried out for all the keto and enol conformers with one watermolecule next to the atom O12. Then the electronic spectra in TD-DFT were obtained again, using the PCM method. The results areshown in Table 5. The local solvent effect also showed changes,compared to the field effect. The conformers keto A and keto C pre-sented smaller oscillator strength than those obtained only withthe field effect. The transition energies also presented very smallchanges. The largest changes in energies reached 0.1 eV for the ketoC.

The enol form once again presented no significant changes inthe oscillator strength, keeping the same trend: a high intensity forthe first transition (enol A) and a lower value (enol C). Moreover,all the transition energy values showed no significant changes withthe combined solvent effect.

Fig. 5 summarizes the absorption energies obtained using TD-DFT and the experimental data for the most intense transition forthe four stable structures in vacuum and considering the solventeffect either using only PCM or PCM and one explicit water moleculeinteracting with the solute. Fig. 5 shows clearly that the solventeffect for almost all structures tends to reduce the transition ener-gies. The only exception is the keto F using the PCM model. The enolA is not so susceptible by the solvent effect compared with enol C,enol E and enol F. Enol E is strongly affected by the explicit inter-

action with the water molecule. The absorption energy changesfrom 3.74 eV to 3.53 eV, a difference of 0.21 eV, while for enol A thechange is not significant and presents a difference of only 0.02 eV.

Enol A, the most stable conformer, is seen as the structure pre-senting the absorption energy closest to the experimental data.

d,p).

Oscillator strength Coefficient Transition energy (eV)

0.4767 0.66344 3.55

0.2649 0.12687 3.650.62964

0.3348 −0.11834 3.740.65012

0.11630.16606

3.790.595520.22754

K.F.d. Souza et al. / Spectrochimica Acta Part A 75 (2010) 901–907 905

Table 4Electronic spectra – most intense transitions to each tautomer – TD-DFT B3LYP/CEP-31G(d,p) and PCM.

Conformer Most intense transition Excited state Oscillator strength Coefficient Transition energy (eV)

Enol A HOMO → LUMO 1st 0.6253 0.66816 3.53Enol C HOMO → LUMO 1st 0.4015 0.64997 3.65Enol E HOMO → LUMO 1st 0.4541 0.66948 3.69

Enol F HOMO(−2) → LUMO2nd 0.2045

0.106473.83HOMO(−1) → LUMO 0.63430

HOMO → LUMO 0.13737

Keto A HOMO(−7) → LUMO

5th 0.2606

0.14553

4.60HOMO(−4) → LUMO 0.52786HOMO(−2) → LUMO −0.31464HOMO(−1) → LUMO(+1) −0.11996HOMO → LUMO(+2) −0.11376

Keto C HOMO(−7) → LUMO

6th 0.1271

−0.12222

4.55HOMO(−6) → LUMO −0.12117HOMO(−5) → LUMO −0.23309HOMO(−4) → LUMO 0.54197HOMO(−2) → LUMO −0.24021

Obs: Enol B and enol D are equivalent with enol A and enol C, respectively and the other keto structures had oscillator strength near to zero.

Table 5Electronic spectra – most intense transitions to each tautomer – TD-DFT B3LYP/CEP-31G(d,p)—PCM and local effect.

Conformer Most intense transition Excited state Oscillator strength Coefficient Transition energy (eV)

Enol A HOMO → LUMO 1st 0.6525 0.66727 3.53Enol C HOMO → LUMO 1st 0.4293 0.65647 3.61Enol E HOMO → LUMO 1st 0.4910 0.66441 3.53

Enol F HOMO(−1) → LUMO1st 0.1761

−0.104133.75HOMO → LUMO 0.64357

Keto A HOMO(−8) → LUMO(+1)

6th 0.2557

0.15092

4.55

HOMO(−4) → LUMO(+1) 0.53971HOMO(−3) → LUMO(+1) −0.27109HOMO(−2) → LUMO(+1) 0.12264HOMO(−1) → LUMO(+2) 0.11122HOMO → LUMO(+3) 0.10176

Keto C HOMO(−8) → LUMO(+1)

6th 0.1006

−0.12190

4.44

HOMO(−7) → LUMO(+1) 0.10816HOMO(−5) → LUMO(+1) 0.31045HOMO(−4) → LUMO(+1) 0.53836HOMO(−3) → LUMO(+1)HOMO(−2) → LUMO(+1)

Obs: Enol B and enol D are equivalent with enol A and enol C, respectively and the other

Fig. 5. Absorption energies of the enol conformers for the most intense transitionobtained from TD-DFT and the experimental data. The symbol (a) indicates calcula-tions using the PCM model and (b) indicates the use of PCM and one explicit watermolecule.

0.14649−0.14008

keto structures had oscillator strength near to zero.

The maximum difference between the calculated and experimen-tal transition is lower than 0.1 eV. On the other hand, the secondmost stable structure, enol C, is not so close to the experimentaldata and the energy difference reaches up to 0.2 eV. The two otherenol structures – enol E and enol F – usually present higher differ-ences compared to the experimental data reaching a difference upto 0.38 eV. The only exception is the transition for enol E with PCMand explicit interaction with water shown as E(b) in Fig. 5, whichis as close to the experimental transition as enol A. However, anyof the E and F structures are unlikely predominant because of theirstability and the low transition barrier with respect to enol A andenol C.

The keto structures often present much smaller oscillatorstrength than the enol forms and the most intense absorption tran-sitions are considerably different than experimental and are notdiscussed further.

Comparing the results for both tautomers, the oscillatorstrengths obtained for the enol species are higher than thatobtained for the keto ones and the transition energies showed much

higher values for the keto than the enol forms either including ornot the solvent effect. The better agreement of the calculated tran-sition energies with the experimental absorption spectra reinforcesthe predominance of the enol A as most important structure eitherin vacuum or in solution.

906 K.F.d. Souza et al. / Spectrochimica A

Fs

6

prbdpreb

ihbbvt

eHC

7

flTrtga

icttceecsscedw

[

[[

[[[[[[

ig. 6. Relative energies obtained to the interconversion barriers between the mosttable conformers of the DFT surface—QST2 method.

. DFT surface

The previous analysis and information from the literature alsoointed out to the importance of the enol form in the PRX fluo-escence. The theoretical evidences of the enol form as responsibley the fluorescence require a further analysis identifying the pre-ominant conformer and the interconvertion barrier among theossible structures. The QST2 method was applied and yielded theesults shown in Fig. 6. The barrier between the two most stablenol structures A and C shows a great difficulty in interconvertingoth structures since the barrier is about 81 kJ mol−1.

The enol F, the most energetic structure, is not expected signif-cantly in the system, since the barrier about 18 kJ mol−1 is not soigh and will be easily interconverted to the enol C. The same cane verified with respect to the enol E, which present a very smallarrier of only 6 kJ mol−1 to be interconverted in enol A. The con-ersion of enol A in E is as difficult as from enol A in enol C, sincehe barrier is 74 kJ mol−1.

Finally, the interconversion from enol C to enol A is not soasy because the barrier presents a reasonable value of 46 kJ mol−1.owever, the enol A is about 35 kJ mol−1 more stable than the enol, becoming the most probable structure present in the system.

. Conclusion

The experimental data show that in non-polar solvents the PRXuorescence increases and it is attributed mainly to the enol form.he theoretical structures calculated using B3LYP/CEP-31G(d,p)einforce the importance of the hydrogen bonds in the stabiliza-ion of this tautomer. This interaction leads the enol form to presentreater rigidity and planarity than the keto form, becoming it morept to participate of radiative processes.

PRX-keto and PRX-enol conformers do not exist independentlyn solution, there are an equilibria among them. Although thesehemical species can absorb light, only some of them seem to fillhe necessary requirements to present fluorescent emission. Forhat reason, when the solvent polarity varies one or more of thoseonformers will be formed preferentially in detriment to the oth-rs, because the total concentration of PRX is not modified. Thexperimental data are clear in the sense that the total fluores-ence quantum yield of all fluorescent chemical species in solutionuffers variation. Therefore, it is evident that among the present

pecies, the quantum yields are different. This suggests a relativeoncentration variation of those species (with different fluorescentmissions) that causes the alterations observed in the experimentalata. The agreement between the theoretical absorption spectrumith experimental data in solution indicates that one of the enol

[[[

[[

cta Part A 75 (2010) 901–907

conformers is predominant in solution or in vacuum. Therefore, itis possible to affirm that the PRX-enol has important role in thephoto-toxicity observed.

The results involving the electronic spectra showed distinctbehavior to the two PRX tautomers. In vacuum, the enol formspresent stronger transitions. So, more apt to participate in mecha-nisms involving the drug photo-toxicity. The calculation of solventeffect on the lower and most intense transitions using a simple con-tinuum model (PCM) or combined with explicit presence of watermolecules provided no significant change in the transition energieswith respect to the vacuum spectra. The only significant change wasin the intensities of absorption and once again the enol structureswere more susceptible to the effect than the keto forms.

Therefore, all the results indicate that the enol species is thekey to elucidate the mechanism of PRX photo-toxicity. The appli-cation of the QST2 method to predict the barriers of interconvertionamong the enol conformers obtained from the DFT surface pointedout to a particular structure referred to as enol A in this work.This structure probably is present predominantly either in vacuumor in solution, regarding the other possible structures. The greatplanarity of enol A suggests a significant contribution to the drugfluorescence.

Acknowledgments

The authors are grateful to CNPq and Fapesp for the financialsupport and to CENAPAD (Campinas, Brazil) for availability of com-putational time. F.B.T. Pessine expresses his appreciation to PfizerCo. for a free sample of PRX. The authors are also grateful to Dr. FredY. Fujiwara for helpful suggestions.

References

[1] H.P. Rang, M.M. Dale, J.M. Ritter, Pharmacology, C. Livingstone, New York, 1995.[2] M. Mihalic, H. Hofman, J. Kuftinec, B. Krile, V. Capler, F. Kajfez, N. Blazevic, in:

K. Florey (Ed.), Analytical Profiles of Drug Substances, vol. 15, Academic Press,New York, 1986, p. 511.

[3] I.E. Kochevar, W.L. Morrison, J.L. Lamm, D.J. McAuliffe, A. Western, A.F. Hood,Arch. Dermatol. 122 (1986) 1283.

[4] J. Bordner, P.D. Hammen, E.B. Whipple, J. Am. Chem. Soc. 111 (1989) 6572.[5] Y.K. Kim, D.W. Cho, S.G. Kang, M. Yoon, D. Kim, J. Lumin. 59 (1994) 209.[6] S.M. Andrade, S.M.B. Costa, Phys. Chem. Chem. Phys. 1 (1999) 4213.[7] R. Banerjee, H. Chakraborty, M. Sarkar, Spectrochim. Acta Part A 59 (2003) 1213.[8] G. Serrano, J. Bonillo, A. Aliaga, E. Garallo, C. Pelufo, J. Am. Acad. Dermatol. 11

(1984) 113.[9] A. Westerm, J.R. Van Camp, R. Bensasson, E.J. Land, I.E. Kochevar, Photochem.

Photobiol. 46 (1987) 469.10] E. Lemp, A.L. Zanocco, G. Günther, J. Photochem. Photobiol. B: Biol. 65 (2001)

165.11] R.B. Alencastro, J.D.M. Neto, J. Comp. Chem. 22 (1995) 123.12] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheese-

man, J.A. Montgomery Jr., T. Vreven, K.N. Kudin, J.C. Burant, J.M. Millam, S.S.Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G.A.Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J.E. Knox, H.P.Hratchian, J.B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R.E. Strat-mann, O. Yazyev, A.J. Austin, R. Cammi, C. Pomelli, J.W. Ochterski, P.Y. Ayala, K.Morokuma, G.A. Voth, P. Salvador, J.J. Dannenberg, V.G. Zakrzewski, S. Dapprich,A.D. Daniels, M.C. Strain, O. Farkas, D.K. Malick, A.D. Rabuck, K. Raghavachari,J.B. Foresman, J.V. Ortiz, Q. Cui, A.G. Baboul, S. Clifford, J. Cioslowski, B.B. Ste-fanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R.L. Martin, D.J. Fox, T. Keith,M.A. Al-Laham, C.Y. Peng, A. Nanayakkara, M. Challacombe, P.M.W. Gill, B. John-son, W. Chen, M.W. Wong, C. Gonzalez, J.A. Pople, Gaussian 03, in: Revision C.02, Gaussian, Inc., Wallingford, CT, 2004.

13] S. Miertus, J. Tomasi, Chem. Phys. 65 (1982) 239.14] E. Runge, E.K.U. Gross, Phys. Rev. Lett. 52 (1984) 997.15] A. Mordinski, K.H. Grellmann, J. Phys. Chem. 90 (1986) 5503.16] C. Peng, H.B. Schlegel, Isr. J. Chem. 33 (1993) 449.17] G.A. Brucker, D.F. Kelley, J. Phys. Chem. 92 (1988) 3805.18] G.A. Brucker, T.C. Swinney, D.F. Kelley, J. Phys. Chem. 95 (1991) 3190.

19] P.F. Barbara, P.K. Walsh, L.E. Brus, J. Phys. Chem. 93 (1989) 29.20] R.S. Moog, S.C. Bovino, J.D. Simon, J. Phys. Chem. 92 (1988) 6545.21] S. Nagaoka, U. Nagashima, N. Ohta, M. Fujita, T. Takemura, J. Phys. Chem. 92

(1988) 166.22] B.J. Schwartz, L.A. Peteanu, C.B. Harris, J. Phys. Chem. 96 (1992) 3591.23] L. Lavtchieva, V. Enchev, Z. Smedarchina, J. Phys. Chem. 97 (1993) 306.

ica A

[[[[[[

[

[[[

K.F.d. Souza et al. / Spectrochim

24] P.-T. Chou, M.L. Martinez, S.L. Studer, J. Phys. Chem. 95 (1991) 10306.25] M.H. Yoon, N. Choi, H.W. Kwon, K.H. Park, Bull. Korean Chem. Soc. 9 (1988) 171.26] R.S. Becker, S. Chakravorti, M. Yoon, Photochem. Photobiol. 51 (1990) 151.

27] F. Laermer, T. Elsaesser, W. Kaiser, Chem. Phys. Lett. 148 (1988) 119.28] M. Yoon, Y.H. Kim, Bull. Korean Chem. Soc. 10 (1989) 434.29] D.W. Cho, Y.H. Kim, M. Yoon, S.C. Jeoung, D. Kim, Chem. Phys. Lett. 226 (1994)

275.30] J.N. Spencer, E.S. Holmboe, M.R. Kirshenbaum, D.W. Firth, P.B. Pinto, Can. J.

Chem. 60 (1982) 1178.

[[[[[

cta Part A 75 (2010) 901–907 907

31] G. Allen, G.R.A. Donek, J. Chem. Soc. B: Phys. Org. (1966) 161.32] J.L. Burdett, M.T. Rogers, J. Am. Chem. Soc. 86 (1964) 2105.33] P. Svoboda, O. Pytela, M. Vecera, Coll. Czech. Chem. Commun. 48 (1983)

3287.34] S.G. Mills, P.J. Beak, Org. Chem. 50 (1985) 1216.35] V. Aviyente, T. Varnali, M.F. Ruiz-López, J. Phys. Org. Chem. 9 (1996) 119.36] L.C. Allen, J. Am. Chem. Soc. 97 (1975) 6921.37] G.M. Gehad, E.A.E. Nadia, Vib. Spectrosc. 36 (2004) 97.38] R. Banerjee, M.J. Sarkar, J. Lumin. 99 (2002) 255.