Theoretical Studies of the Tautomers of Pyridinethiones

Transcription

1 J. Phys. Chem. A 2008, 112, Theoreticl Studies of the Tutomers of Pyridinethiones Yu Adm Zhng, Vishkh Mong, Chris Orvig,* nd Yn Alexnder Wng* Deprtment of Chemistry, UniVersity of British Columbi, VncouVer, BC V6T 1Z1, Cnd ReceiVed: June 7, 2007; In Finl Form: October 29, 2007 Pyridinethiones re importnt lignd precursors of coordintion complexes of therpeutic vlue. In queous solution, pyridinethiones cn dimerize nd tutomerize to the corresponding thiols. However, the tutomerism of pyridinethiones, which cn impct on therpeutic performnce, is yet not fully understood. To resolve this importnt issue, we hve crried out b initio nd DFT clcultions to compute the geometries, energies, dipole moments, nd NMR, IR, nd UV-vis spectroscopic properties of ll possible tutomers of pyridinethiones nd compred our theoreticl results with the existing experimentl dt. We found tht the thione form of the tutomer is dominnt for monomers of the pyridinethiones studied here. This work cn serve s reference for exploring other similr orgnosulfur compounds. Introduction There hs been continued interest in orgnosulfur compounds with thiocrbonyl groups becuse of their diverse chemistry, 1 vrious biologicl pplictions, 2,3 nd the tendency of the CdS bond to oxidize esily to form the corresponding thiols nd disulfides. 4 Pyrnthiones nd pyridinethiones re two importnt clsses of such orgnosulfur compounds. 5 Four yers go, pyrnthiones were explored s lignds becuse of their trnsinfluencing bility. 6 A number of pyridinethiones hve lso been ptented for their therpeutic ntioxidnt properties 7 nd effectiveness ginst crcinom. 8 Pyridinethiones nd their oxygen nlogues, pyridinones, cn tutomerize to the corresponding thiol/enol form (Figure 1). 9,10 It hs been shown, however, tht the previling form of these compounds is the thione/ketone form. 11 For pyridinethiones, the S-H group in the thiol form cn esily oxidize to bridge two molecules through disulfide bond in polr solvents in ir (Scheme 1). 10,12 Interestingly, Stoynov et l. reported tht pyridinethiones convert to the thiol tutomer when left in queous or lcoholic solutions for 24 h, even though the thione form is more dominnt. 10 Evidence for this tutomerism ws lso observed recently in the syntheses of pyridinethiones studied by our group. 5 We recently reported 5 the preprtion of 3-hydroxy-2-methyl- 4-p-pyridinethione (Hmppt) nd 3-hydroxy-1,2-dimethyl-4-ppyridinethione (Hdppt) by recting thiomltol with mmoni nd methylmine (Scheme 2), respectively. The resulting pyridinethiones were thoroughly chrcterized by elementl nlysis, electron impct mss spectrometry, melting point determintion, s well s UV-visible bsorption, IR, nd NMR spectr. X-ry crystl structures were lso obtined for some of these compounds. Although the thiol nd thione forms of the tutomer could not be seprtely isolted, some experimentl chrcteriztion dt suggested tht the thione form is the dominnt tutomer. A strong ν CdS stretch in the IR spectr confirmed the thione form of the tutomer to be the solid-stte chemicl structure of the pyridinethiones. The solution-stte chrcteriztion (vi NMR nd UV-vis spectr) of these * Corresponding uthors. Phone: ; Fx: E-mil: ywng@chem.ubc.c (Y.A.W.). Figure 1. Tutomeric forms of pyridinones nd pyridinethiones. TABLE 1: Selected Bond Lengths (in Å) nd Bond Angles (in deg) of Hdppt Dimer bond X-ry /theory b bond ngles X-ry /theory b S(1)-S(1*) (7)/ S(1)-S(1*)-C(1) (5)/ S(1)-C(1) 1.763(1)/1.771 S(1)-C(1)-C(2) 112.1(1)/119.8 O(1)-C(2) 1.278(2)/1.245 S(1)-C(1)-C(5) 126.0(1)/119.3 N(1)-C(3) 1.362(2)/1.355 C(3)-N(1)-C(4) 123.0(1)/122.8 N(1)-C(4) 1.351(2)/1.364 C(3)-N(1)-C(7) 119.4(1)/119.3 N(1)-C(7) 1.482(2)/1.472 C(4)-N(1)-C(7) 117.5(1)/117.9 C(1)-C(2) 1.424(2)/1.457 N(1)-C(3)-C(2) 119.9(1)/121.2 C(1)-C(5) 1.383(2)/1.390 N(1)-C(3)-C(6) 120.1(1)/120.7 C(2)-C(3) 1.427(2)/1.462 N(1)-C(4)-C(5) 119.9(1)/119.4 C(4)-C(5) 1.377(2)/1.385 O(1)-C(2)-C(1) 120.8(1)/124.4 C(3)-C(6) 1.489(2)/1.497 O(1)-C(2)-C(3) 123.4(1)/121.0 C(2)-C(1)-C(5) 121.9(1)/120.6 C(1)-C(2)-C(3) 115.8(1)/114.6 C(1)-C(5)-C(4) 119.4(1)/121.3 C(2)-C(3)-C(6) 120.0(1)/118.1 From ref 5, with experimentl uncertinty shown in the prentheses. b Optimized geometry from B3LYP/6-31G(d,p) clcultions. compounds lso suggests the presence of the thione tutomeric forms. Nonetheless, the dt collected could not be unmbiguously ssigned to one tutomeric form over the other. As mentioned before, when left in queous or lcoholic solutions, the monomeric (thiol) forms of these compounds were found to convert to their dimeric (disulfide) forms. Evidence for the formtion of n S-S bridge between two molecules ws obtined with IR nd NMR spectr. Unfortuntely, the S-H or N-H proton could not be ssigned distinctly due to probble exchnge of the proton between S nd N in solution (Figure 1). This process hs been studied before, nd the structures hve been verified by compring the experimentl dt with theoreticl clcultions for the IR nd NMR dt. 10 Hence, more extensive comprisons between the theoreticl nd experimentl /jp CCC: $ Americn Chemicl Society Published on Web 03/15/2008

2 3232 J. Phys. Chem. A, Vol. 112, No. 14, 2008 Zhng et l. SCHEME 1. Proposed Mechnism for the Dimeriztion of the Two Pyridinethiones Studied in this Work: 3-Hydroxy- 2-methyl-4-p-pyridinethione (Hmppt) nd 3-Hydroxy-1,2-dimethyl-4-p-pyridinethione (Hdppt) SCHEME 2. Synthesis of the Pyridinethiones () excess NH 4OH, H 2O/EtOH, 34 C, 36 h; (b) excess 40% MeNH 2,H 2O, 75 C, 12 h. Figure 2. ORTEP digrm of Hdppt dimer (50% therml ellipsoids), dpted from ref 5. dt should llow the ssignment of the dominnt tutomeric form of these pyridinethiones. Herein, we report b initio nd DFT clcultions of Hmppt, its methyl nlog (Hdppt), their thiol tutomers, nd their TABLE 2: Theoreticl Predictions of the Dipole Moments (in D) of All Tutomers in Different Medi (vcuum, DMSO, or H 2 O), Clculted t the B3LYP/ G(2df,p)// B3LYP/6-31G(d,p) Level of Theory tutomer vcuum DMSO H 2O Hmppt Hmppt Hmppt-d Hmppt-d dimeric forms. Our results reported here provide comprehensive librry of dt for ny future comprisons of these nd other such similr compounds. Detils of Computtionl Methods We hve performed b initio nd DFT clcultions to obtin the NMR, IR, nd UV-vis dt for Hmppt, Hdppt, nd their dimers. We optimized Hmppt-1 nd Hmppt-2 in the gs phse t the B3LYP 13 /6-31G(d,p) nd B3LYP/ G(2df,p) levels of theory nd found only negligible differences between the two types of geometries. So we will use the B3LYP/6-31G- (d,p) geometries for ll species throughout this pper, unless specilly mentioned. Geometries of Hmppt-1 nd Hmppt-2 obtined on different levels re included in the Supporting Informtion for reference. The totl energies nd the Gibbs free energies (t 298 K) of molecules of interest were computed t the B3LYP/6-31G(d,p) nd MP2/6-31G(d,p) levels of theory. Prtil chrges were computed from the nturl bond orbitl (NBO) nlysis. 14 The 13 C nd 1 H NMR chemicl shifts of ll tutomers were predicted with the guge-independent tomic orbitl (GIAO) 15 nd the continuous set of guge trnsformtions (CSGT) 16 pproches t the HF/ G(2df,p), MP2/ G(2df,p), B3LYP/ G(2df,p), nd PBEPBE/6-

3 Studies of the Tutomers of Pyridinethiones J. Phys. Chem. A, Vol. 112, No. 14, Figure 3. Atomic prtil chrges of ll species (in vcuum/dmso/h 2O). () Hmppt-1, (b) Hmppt-2, (c) Hmppt-d-1 (left prt), (d) Hmppt-d-2 (left prt), (e) Hdppt, (f) Hdppt-d (left prt) G(2df,p) levels of theory. The polrizble continuum model 17 (PCM) ws employed to ccount for the solvtion effects. All geometries were reoptimized t the B3LYP/6-31G- (d,p) level in the solution environment. Vibrtionl frequencies were clculted t the HF, MP2, nd B3LYP levels of theory with the bsis set 6-31G(d,p). To improve the greement between theory nd experiment, we took the usul prctice of employing scling fctors to bring our clculted frequencies closer to the existing experimentl dt. 18 Since the scling fctor for B3LYP/6-31G(d,p) clcultions is not vilble, we insted

4 3234 J. Phys. Chem. A, Vol. 112, No. 14, 2008 Zhng et l. TABLE 3: Totl Energy nd Gibbs Free Energy (t 298 K) Differences with the Zero-Point Energy Correction (in kcl/mol) of All Tutomers in Different Medi (vcuum, DMSO, or H 2 O), Clculted t the B3LYP/6-31G(d,p) nd MP2/6-31G(d,p)// B3LYP/6-31G(d,p) Levels of Theory vcuum DMSO H 2O energy difference method E G E G E G E Hmppt-2 - E Hmppt-1 B3LYP MP E Hmppt-d-2 - E Hmppt-d-1 B3LYP MP TABLE 4: Theoreticl IR Pek Positions (in cm -1 ) nd Intensities of the Importnt Peks (in prentheses) Clculted t the HF/6-31G(d,p), MP2/6-31G(d,p), nd B3LYP/6-31G(d,p) Levels of Theory, Compred with the Experimentl Results HF MP2 B3LYP expt ssignment b scling fctor c Hmppt (171.5) 3161(98.0) 3192(160.4) ν NH (ν OH) ν NH (ν OH) ν NH (ν OH) 3329(182.5) 3455(177.0) 3499(119.5) 3427 ν OH (ν NH) ring nd C-N-C bnds ring nd C-N-C bnds ring nd C-N-C bnds 1136(468.5) 1154(119.1) 1162(104.9) 1174 ν CdS ring nd C-N-C bnds ring nd C-N-C bnds Hmppt (128.1) 3488(104.2) 3527(104.2) 3427 (ν OH) 2437(5.9) 2507(4.1) 2553(9.1) (ν SH) ring nd C-N-C bnds ring nd C-N-C bnds ring nd C-N-C bnds 1157(12.9) 1124(53.5) 1134(55.9) 1174 (ν C-S) ring nd C-N-C bnds ring nd C-N-C bnds Hmppt-d (133.7) 3505(103.3) 3544(126.6) 3508 ν H2O (ν OH) ring nd C-N-C bnds ring nd C-N-C bnds ring nd C-N-C bnds ring nd C-N-C bnds ring nd C-N-C bnds Hmppt-d (165.7) 3398(115.9) 3437(93.6) (ν NH) (ν NH) (ν NH) ring nd C-N-C bnds ring nd C-N-C bnds ring nd C-N-C bnds ring nd C-N-C bnds ring nd C-N-C bnds See the Supporting Informtion of ref 5 (with (4cm -1 uncertinty). b See ref 5. New ssignments from this work re shown in prentheses. c See ref 18. used the scling fctor for B3LYP/6-31G(d), which should not bring ny significnt numericl error. Time-dependent density functionl theory 19 (TDDFT) with different density functionls (B3LYP nd PBEPBE 20 ) ws utilized to clculte the UV-vis spectr. All clcultions were performed with the quntum chemistry pckge of Gussin Results nd Discussions Both Hmppt nd Hdppt cn dimerize in solution. While in solution, it is very difficult to crystllize the Hmppt monomer/ dimer mixture, so X-ry dt re vilble only for Hdppt dimer. 5 X-ry-qulity crystls for Hdppt dimer were obtined by slow evportion in ethyl cette. The structure is shown in Figure 2, nd the selected bond lengths nd bond ngles re listed in Tble 1. For comprison, we lso provide the dt of the optimized geometry from our DFT clcultions. As cn be seen in Tble 1, our theoreticl results gree quite well with the experimentl X-ry dt. Becuse the S-S bond is very flexible, we imposed C 2 symmetry on the dimer to ccelerte the clcultion nd to reduce the demnd on computtionl resources (especilly in the property clcultions). In the end, we found tht the optimized geometry without the C 2 symmetry constrint is virtully identicl to tht obtined with the constrint. Thus, it is legitimte for us to impose the sme C 2 symmetry constrint in dditionl clcultions whenever possible. We crried out NBO prtil chrge nlysis for Hmppt-1 nd other molecules shown in Scheme 1 (Hmppt-1 is resonnce form of Hmppt-1). The prtil chrge results re shown in Figure 3. For Hmppt-1, we found tht both in vcuum nd in solution (H 2 O or DMSO), the N tom lwys crries lrge negtive chrge (> -0.5), nd the S tom lwys crries reltively smll negtive chrge (<-0.5). This might suggest tht Hmppt- 1 poorly represents the ctul electronic structure, since in Hmppt-1 the N tom should crry less negtive chrge thn usul nd the S tom should crry more negtive chrge. Moreover, the bond length of C1-C2 (see the numbering in Figure 3) in Hmppt-1 is clculted to be 1.43 Å, which is much longer thn the bond length of C2-C3 (1.37 Å), indicting double bond structure between C2 nd C3. Thus, it is sfe to

5 Studies of the Tutomers of Pyridinethiones J. Phys. Chem. A, Vol. 112, No. 14, TABLE 5: Theoreticl 13 C nd 1 H NMR Chemicl Shifts (in ppm) of Hmppt-1 in DMSO, Clculted with the GIAO Method with the G(2df,p) Bsis Set t the HF, MP2, B3LYP, nd PBEPBE Levels of Theory, Compred with Experimentl Results Obtined in DMSO HF MP2 B3LYP PBEPBE expt c C C C C C δ b H H H H See Figure 2 for the numbering of the toms. b Averge bsolute error with respect to the 13 C NMR experimentl dt. c From the Supporting Informtion of ref 5. TABLE 6: Theoreticl 13 C nd 1 H NMR Chemicl Shifts (in ppm) of Hmppt-2 in DMSO, Clculted with the GIAO Method with the G(2df,p) Bsis Set t the HF, MP2, B3LYP, nd PBEPBE Levels of Theory, Compred with Experimentl Results Obtined in DMSO HF MP2 B3LYP PBEPBE expt c C C C C C δ b H H H H See Figure 2 for the numbering of the toms. b Averge bsolute error with respect to the 13 C NMR experimentl dt. c From the Supporting Informtion of ref 5. TABLE 7: Theoreticl 13 C nd 1 H NMR Chemicl Shifts (in ppm) of Hmppt-d-1 in DMSO, Clculted with the GIAO Method with the G(2df,p) Bsis Set t the HF, B3LYP, nd PBEPBE Levels of Theory, Compred with Experimentl Results Obtined in DMSO HF B3LYP PBEPBE expt c C C C C C δ b H H H See Figure 2 for the numbering of the toms. b Averge bsolute error with respect to the 13 C NMR experimentl dt. c From the Supporting Informtion of ref 5. TABLE 8: Theoreticl 13 C nd 1 H NMR Chemicl Shifts (in ppm) of Hmppt-d-2 in DMSO, Clculted with the GIAO Method with the G(2df,p) Bsis Set t the HF, B3LYP, nd PBEPBE Levels of Theory, Compred with Experimentl Results Obtined in DMSO HF B3LYP PBEPBE expt c C C C C C δ b H H H See Figure 2 for the numbering of the toms. b Averge bsolute error with respect to the 13 C NMR experimentl dt. c From the Supporting Informtion of ref 5. conclude tht Hmppt-1 does not represent the dominnt form of the rel electronic structure. We lso found similr sitution in the chrge nd structurl nlyses of Hdppt nd Hdppt : Hdppt is not good representtion of the ctul electronic structure of Hdppt. Hdppt dimer might be somewht different, since it cn only exist in the formlly chrged form (see Scheme 1). Its N tom hs chrge of c. -0.3, smller thn those in Hmppt-1 nd Hmppt-2 nd their dimers, wheres its O tom crries lrger negtive chrge (> -0.7). These results, compred to chrge distributions of Hmppt dimers, support the N-forml-positivechrged structure of Hdppt dimer, in greement with the X-ry crystl structure dt. It ws lso observed tht different solvents ffect the chrge distributions slightly. The dipole moments of ech species nd totl/free energy differences between them were lso studied with smll nd lrge bsis sets, both in vcuum nd in solution (see Tbles 2 nd 3 nd Supporting Informtion for detils). We found tht the bsis set ffects dipole moments only slightly. Dipole moments clculted from the smll bsis set re included in Supporting Informtion for reference. Dt shown in Tble 3 clerly indicte tht Hmppt-1 is more stble thn is Hmppt-2 in DMSO or H 2 O, lthough the MP2 clcultions predict tht Hmppt-2 is just slightly more stble thn is Hmppt-1 in vcuum. Such irregulrity should not pose ny rel problem, becuse we lredy hve coherent results (from different methods) for the reltive stbility in DMSO or H 2 O, bsed on which we cn compre our theoreticl results with experimentl dt. As expected from those positive energy differences in Tble 3, the more polr thione form of Hmppt is fvored in the polr solution environment. In ddition, the polr solvents even increse the energy difference between Hmppt-2 nd Hmppt-1. The dimers show nother story. In the tutomerism of Hmppt, the proton migrtes between the N nd S toms, but the proton migrtes between the N nd O toms in Hmppt dimer. Hmpptd-2 hs structure with forml discrete positive/negtive chrge distribution tht usully hs n higher energy in vcuum, nd its dipole moment is very lrge ( 10 D), compred to the smll dipole moment of Hmppt-d-1 (only 1.7 D in vcuum). Although Hmppt-d-1 is fvored in terms of totl energy, the energy difference between Hmppt-d-2 nd Hmppt-d-1 is smller thn the corresponding vlues of the Hmppt monomeric tutomers in solutions. This is, of course, becuse polr solvents cn stbilize the more polr Hmppt-d-2. It is interesting to consider how the S nd H toms re bonded in these molecules, so we crried out vibrtionl frequency clcultions nd compred the results with the experimentl dt. The finl results re shown in Tble 4. Experimentl IR dt were collected from smples in solid phse, but brod ν OH pek for H 2 O ws seen in the IR spectrum of Hmppt dimer becuse the smple ws not dry. In ddition, O-H groups in Hmppt or in H 2 O cn form H-bonds with other O-H nd N-H groups, which mkes the IR spectrum bove 2700 cm -1 very complicted. Hence, it is very difficult to mtch the clculted results with the experimentl dt precisely. Theoreticlly, we found tht the vibrtionl frequencies of O-H nd N-H bonds re nd cm -1, respectively, in direct contrdiction to previous experimentl ssignments: cm -1

6 3236 J. Phys. Chem. A, Vol. 112, No. 14, 2008 Zhng et l. TABLE 9: Theoreticl UV-vis Absorption Positions (in nm) nd Oscilltor Strengths (in Prentheses) for Single Excittions of Thiol nd Thione Tutomers of Hmppt nd Hmppt Dimer in Vcuum nd H 2 O, Clculted t B3LYP Levels, Compred with Experimentl Dt Obtined in H 2 O e tutomers vcuum H 2O expt tutomers vcuum H 2O expt Hmppt-1 c 206 (0.2072) 206 (0.2621) 210 (4.13) Hmppt-d-1 d 270 (0.2546) 264 (0.2938) 270 (3.71) H b -3 f L b 0.29 H-3 f L H-5 f L 0.67 H-5 f L 0.67 H-2 f L H-2 f L (0.0002) 323 (0.0029) 336 (4.32) H-1 f L H-1 f L H-5 f L 0.10 H-4 f L H-1 f L H f L H-2 f L 0.68 H-3 f L 0.65 H f L H f L (0.0975) 249 (0.1454) 245 (3.92) 364 (0.1262) 354 (0.1581) H-2 f L 0.63 H-3 f L H f L 0.68 H-3 f L H-3 f L H-2 f L 0.61 H f L 0.67 H f L Hmppt-d-2 d 201 (0.2213) 203 (0.3136) 281 (0.0037) 280 (0.0001) 276 (3.43) H-10 f L H-10 f L H-1 f L H-2 f L H-9 f L 0.11 H-7 f L 0.11 H f L H-7 f L 0.19 H-6 f L (0.3166) 319 (0.3693) 326 (4.21) H-6 f L H-5 f L H-2 f L H-2 f L H-5 f L H-4 f L H f L 0.60 H f L 0.63 H-4 f L H-3 f L Hmppt-2 c 199 (0.2005) 208 (0.0583) H-3 f L H-2 f L H-3 f L 0.41 H-3 f L H-3 f L H-1 f L H-3 f L H-2 f L H-2 f L H f L H-2 f L H-1 f L H-1 f L H-2 f L H-1 f L H f L H-1 f L H f L (0.1801) 221 (0.1271) H f L H f L H-7 f L H-6 f L 0.11 H f L H-6 f L 0.23 H-5 f L H f L H-5 f L H-3 f L (0.0038) H-4 f L H-1 f L H-2 f L H-3 f L H-1 f L H-2 f L (0.1235) H-1 f L H-1 f L H-7 f L H-1 f L H-1 f L H-6 f L (0.0013) 245 (0.0290) H f L H-5 f L H-3 f L H-3 f L 0.11 H f L H-1 f L H-2 f L 0.68 H-2 f L 0.49 H f L H-1 f L (0.0777) H f L H-6 f L (0.0257) 269 (0.0223) 266 (0.1388) H-5 f L H-4 f L H f L H-3 f L H-4 f L H-4 f L H f L 0.30 H-2 f L H-3 f L H-3 f L H-1 f L H-2 f L H-3 f L H f L 0.63 H f L H-2 f L Hmppt-d-1 d 206 (0.0747) 206 (0.0426) 201 (4.27) 271 (0.0002) 269 (0.0008) H-7 f L H-7 f L H-5 f L H-2 f L H-5 f L H-5 f L H-1 f L H-1 f L H-3 f L H-1 f L H f L H-1 f L H f L (0.1134) 318 (0.0844) H f L H f L H-3 f L H-5 f L H f L H-1 f L H-1 f L H f L (0.0018) 337 (0.0005) 206 (0.0264) 206 (0.0279) H-4 f L 0.29 H-4 f L 0.62 H-5 f L H-5 f L H-3 f L 0.21 H-3 f L H-1 f L H-1 f L H-2 f L H f L H-1 f L H-1 f L H-1 f L H f L H f L (0.1157) 246 (0.1393) 238 (4.43) H-7 f L H-7 f L H-5 f L H-1 f L H-1 f L See the Supporting Informtion of ref 5. b H nd L denote the HOMO nd the LUMO, respectively. H-m nd L+n denote the mth orbitl below the HOMO nd the nth orbitl bove the LUMO, respectively. c Clculte t B3LYP/ G(2df,p)//B3LYP/6-31G(d,p) level of theory. d Clculte t B3LYP/6-31G(d,p) level of theory. e Trnsitions with lrge CI coefficients re shown below ech pek position. for O-H nd 3061 cm -1 for N-H vibrtions. Even fter we incresed the bsis set from 6-31G(d,p) to G(2df,p) nd reoptimized the geometries t B3LYP/ G(2df,p), we still got 3179 cm -1 for O-H nd 3484 cm -1 for N-H vibrtions. Since ll different theoreticl methods employed here consistently predict the vibrtionl frequencies of O-H nd N-H bonds, further computtionl studies employing more ccurte (more computtionlly demnding nd more timeconsuming) b initio methods should be crried out to fully resolve this discrepncy. Of course, n experimentl revisit might lso be necessry to settle this issue completely. For the crbon-sulfur bond, the ν CdS pek is observed in Hmppt but not in Hmppt dimer, which strongly suggests the formtion of the S-S bond. Due to symmetry, we were unble to see the vibrtion of the S-S bond stretching in the IR spectr. It is cler tht the clculted CdS stretching frequencies of Hmppt-1 gree better with the experimentl dt thn do those of Hmppt-2 (the corresponding bsolute verge errors re 24

7 Studies of the Tutomers of Pyridinethiones J. Phys. Chem. A, Vol. 112, No. 14, nd 36 cm -1, respectively), which fvors Hmppt-1 in the tutomerism. According to our clcultion, the bsence of the ν SH peks round cm -1 in the experiment lso disfvors the existence of Hmppt-2 in the smple. The GIAO results of the 13 C nd 1 H NMR chemicl shifts of ll tutomers re compred with the experimentl dt in Tbles 5-8. Becuse the CSGT results re bsiclly consistent with the GIAO results, we hve collected the CSGT results in Tbles 5S-8S of the Supporting Informtion. From the verge errors of the clcultions, we found tht the 13 C NMR chemicl shifts of Hmppt-1 mtch better with the experimentl dt thn do those of Hmppt-2, nd the sme is true for those of Hmppt-d-1 nd Hmppt-d-2. These results support the conclusion tht Hmppt-1 nd Hmppt-d-1 re fvored in solution, specificlly in DMSO. From the clculted 1 H NMR results, we cn see tht Hmppt hs n ctive H (from N-H) with the chemicl shift rnging from 9.2 to 10.7 ppm. For Hmppt-2, we found tht the chemicl shift for H (from S-H) rnges from 4.8 to 5.3 ppm. The experimentl vlue is ppm, much closer to the theoreticl prediction of Hmppt-1. This fct thus strongly suggests tht the Hmppt-1 form is fvored. The solvent effects for the NMR properties of both tutomers were not considerble, since the verge errors of the chemicl shifts predicted in the gs nd solution phses re roughly the sme. In ddition, for different levels of theory (b initio nd DFT) nd methods computing NMR prmeters (GIAO nd CSGT), there re no significnt differences in the results. Similrly, the clculted 13 C NMR chemicl shifts lso indicte tht Hmppt-d-1 is fvored (see Tbles 7 nd 8). Moreover, we notice tht the theoreticl chemicl shifts for the ctive H (from O-H) in Hmppt-d-1 nd for the ctive H (from N-H) in Hmppt-d-2 re nd ppm, respectively, wheres the experimentl vlue is bout 9.8 ppm, roughly between these two rnges. This indictes tht both tutomeric forms contribute to the NMR spectrum of Hmppt dimer nd tht Hmppt-d-1 is only slightly more fvored thn Hmppt-d-2. UV-vis bsorption spectr of Hmppts nd its dimers clculted with TDDFT on different levels re compred with the vilble experimentl dt in Tble 9. For Hmppt-1 nd Hmppt-2, we found tht clcultions with the lrge bsis set, G(2df,p), give better results thn do those with the smll bsis set, 6-31G(d,p), but they show the sme tendency. So, results of the monomers from the smll bsis set re shown only in the Supporting Informtion for comprison. Becuse of the sizes of the dimers, our computtionl resources only permitted us to crry out clcultions bsed on the smll bsis set. Clcultions of the monomers suggest tht such smll bsis set results re still cceptble s compromise between performnce nd ccurcy. As different density functionls re concerned, we found tht PBEPBE gives similr results s B3LYP does, so we only list the B3LYP results in Tble 9 nd keep the PBEPBE results in the Supporting Informtion for reference. The mximum UV-vis bsorption wvelength, λ mx, of Hmppt ws determined experimentlly to be 326 nm. Clcultions for Hmppt-1 yield reltively ccurte estimtion of these bsorption pek positions (see Tble 9), wheres clcultions for Hmppt-2 do not support the existence of such pek in the proximity of 326 nm. This strongly disfvors Hmppt-2 in the tutomerism. We cn lso rech similr conclusion for Hmppt dimers. Our theoreticl prediction (246 nm) mtches the experimentl λ mx vlue 238 nm well for Hmppt-d-1, while Hmppt-d-2 hs only very wek peks round 238 nm (t 234 nm), which certinly cnnot be considered s cndidtes for λ mx. On the other hnd, Hmppt-d-2 does show very strong peks bove 400 nm in our clcultions (444 nm, not shown in Tble 9), which were not observed in our previous experiments. Comprison of the clcultions in the gs phse nd in H 2 O shows tht the solvtion effects re negligible. Hence, the chrcteristics of the corresponding UV-vis spectr strongly infer tht the structure of Hmppt-d-1 represents the ctul structure better thn does Hmppt-d-2. Conclusions We hve presented here comprehensive theoreticl study on the tutomers of Hmppt nd its dimer in the vcuum nd in polr solutions (DMSO nd H 2 O). Comprisons were lso mde with Hdppt nd its dimeric form to substntite our conclusions bout the dominnt tutomeric form observed in our previous experiments. Clcultions t rnge of levels of theory consistently support preference for the thione form, lthough the degrees of the preferences my vry. This study cn be used s reference for investigting other similr molecules. Acknowledgment. Finncil support from the Nturl Sciences nd Engineering Reserch Council (NSERC) of Cnd, Cndin Institutes of Helth Reserch (CIHR), nd Cnd Foundtion for Innovtion (CFI) is grtefully cknowledged. WestGrid nd C-HORSE (CFI) hve provided us the necessry computtionl resources. Y.A.Z. would like to thnk Prof. Wei Qun Tin of Jilin University nd Dr. Wei (Dvid) Deng of Gussin Inc. for help with some of the clcultions. Supporting Informtion Avilble: Additionl tbles nd figures s noted in the text. This mteril is vilble free of chrge vi the Internet t References nd Notes (1) Metzner, P.; Hogg, D. R.; Wlter, W.; Voss, J. In Orgnic Compounds of Sulphur, Selenium nd Tellurium; Hogg, D. R., Ed.; The Chemicl Society: London, 1977; Vol. 4, p 125; nd 1979; Vol. 5, p 118. (2) () Block, E. Rections of Orgnosulfur Compounds; Acdemic: New York, 1978; p 26. (b) Cremlyn, R. J. An Introduction to Orgnosulfur Chemistry; Wiley: New York, 1996; Chpters 10 nd 11. (3) Mw, G. A. In Sulfur in Orgnic nd Inorgnic Chemistry; Senning, A., Ed.; Mrcel Dekker: New York 1972; Vol. 2, p 113. (4) Steliou, K.; Mrni, M. J. Am. Chem. Soc. 1982, 104, 3104 nd references therein. (5) () Mong, V.; Ptrick, B. O.; Orvig, C. Inorg. Chem. 2005, 44, (b) Mong, V.; Thompson, K. H.; Yuen, V. G.; Shrm, V.; Ptrick, B. O.; McNeill, J. H.; Orvig, C. Inorg. Chem. 2005, 44, (6) Lewis, J. A.; Puert, D. T.; Cohen, S. M. Inorg. Chem. 2003, 42, (7) Tilbrook, G.; Hider, R.; Moridni, M. Ptent WO 98/25905, (8) Hoyoku, N.; Yukihiro, K.; Toshihiko, S.; Susumu, Y.; Hiromichi, M.; Yoichi, I. U.S. Ptent , (9) Albert, A. Heterocyclic Chemistry n Introduction; Athlone: London, 1959; p 424. (10) Stoynov, S.; Petkov, I.; Antonov, L.; Stoynov, T. Cn. J. Chem. 1990, 68, (11) Besso, H.; Imfuku, K.; Mtsumur, H. Bull. Chem. Soc. Jpn. 1977, 50, 3295 nd references therein. (12) Roberts, J. D.; Cserio, M. C. Bsic Principles of Orgnic Chemistry; W. A. Benjmin: New York, 1964; p 744. (13) () Becke, A. D. Phys. ReV. A1988, 38, (b) Lee, C.; Yng, W.; Prr, R. G. Phys. ReV. B1988, 37, 785. (14) Weinhold, F. Nturl Bond Orbitl Methods. In Encyclopedi of Computtionl Chemistry, Schleyer, P. V. R., Allinger, N. L., Clrk, T., Gsteiger, J., Kollmn, P. A., Schefer, H. F., III, Schreiner, P. R., Eds.; Wiley: Chichester, UK, 1998; Vol. 3, pp (15) () London, F. J. Phys. Rdium. 1937, 8, (b) McWeeny, R. Phys. ReV. 1962, 126, (c) Ditchfield, R. Mol. Phys. 1974, 27, 789. (d) Dodds, J. L.; McWeeny, R.; Sdlej, A. J. Mol. Phys. 1980, 41, (e) Wolinski, K.; Hilton, J. F.; Puly, P. J. Am. Chem. Soc. 1990, 112, (16) () Keith, T. A.; Bder, R. F. W. Chem. Phys. Lett. 1992, 194, 1. (b) Keith, T. A.; Bder, R. F. W. Chem. Phys. Lett. 1993, 210, 223. (c)

8 3238 J. Phys. Chem. A, Vol. 112, No. 14, 2008 Zhng et l. Cheesemn, J. R.; Frisch, M. J.; Trucks, G. W.; Keith, T. A. J. Chem. Phys. 1996, 104, (17) () Cncès, M. T.; Mennucci, B.; Tomsi, J. J. Chem. Phys. 1997, 107, (b) Cossi, M.; Brone, V.; Mennucci, B.; Tomsi, J. Chem. Phys. Lett. 1998, 286, 253. (c) Mennucci, B.; Tomsi, J. J. Chem. Phys. 1997, 106, (18) Scott, A. P.; Rdom, L. J. Phys. Chem. 1996, 100, (19) () Runge, E.; Gross, E. K. U. Phys. ReV. Lett. 1984, 52, 997. (b) Gross, E. K. U.; Kohn, W. Phys. ReV. Lett. 1985, 55, (20) () Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. ReV. Lett. 1996, 77, (b) Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. ReV. Lett. 1997, 78, (21) Frisch M. J.; et l. Gussin 03, Revision C.02, Gussin, Inc.: Wllingford, CT, 2004.