ebook img

Chemical Science PDF

15 Pages·2014·0.96 MB·English
by  
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Chemical Science

Chemical Science EDGE ARTICLE View Article Online View Journal | View Issue – The Houk List transition states for organocatalytic ce. mechanisms revisited† n AM. Lice Citethis:Chem.Sci.,2014,5,2057 53:41 ported AMlaatntAJ.rmHsatrrvoenyg,d,aNRicohboelratsoJA..MBaostoon,bacaPnadulHDeinnrgywSa.llR,azeJpualia*aContreras-Garc´ıa,bc 12:Un 3 0 202n 3. ThetenyearoldHouk–Listmodelforrationalisingtheoriginofstereoselectivityintheorganocatalysed 1/6/utio intermolecular aldol addition is revisited, using a variety of computational techniques that have been d on Attrib idnitsrpoedrusicoendinotreriamcptiroonvsedisssihnocwenthteoboerigcinruacliaslt,uadlyo.nEgvwenithfotrhesuucsheaofreblaastiisvesleytssmwahlelrseysthteems,utpheerprooslietioonf des oaon errorsarelow.AnNCI(non-covalentinteractions)analysisofthetransitionstatesisabletoidentifythe wnlmm noncovalent interactions that influence the selectivity of the reaction, confirming the role of the DoCo electrostatic NCHd+/Od(cid:1) interactions. Simple visual inspection of the NCI surfaces is shown to be a uary 2014. a Creative Received12thDecember2013 uiwnshveoifculvhlinctgooomalpwfuoattreedrthkmeinoedleteiccsuigilsenootoorpftehaeletffeHrenacjaottssiv–aePrearreirnaiscchotamanlttpsea.rtniAballtteievrewn,iatathriveeexshpmoeerwicmnheatnnotis.bmTehsh,eigsAuhmcehrstienarsdeanpmerorgmtoyanan-nirfdeelsaftoyosr, d on 24 Febrensed under ADwcOwcIwe:p.1rt0sec.d1.o02r3g49/tc/hch3Fesemcbi5cru3a4alsr1cy6ieb2n0c1e4 wddaahttiaachtraebepsloepssoitauossrseyesamnthbdelethdpervimniacsicepallvellesstthoaasttshisgecnideenadttaiafiDcDOOdaII.t(adisghitoaul-lodbbjeectc-itidaebnlet,ifiiserfso)lfloowrceadlchuelaretiobnysuhseinldgoinnteardacigtiitvael Article. PublisheThis article is lic ITnhetrporodmuotciotnioofnawiderangeoforganicsyntheticreactions HcaoruAbkoo–nnL–eics-aptrrbmooloinndeeblmoinnedcvhofalovnremissmathtiebonats.reaTdnhsoiisntioiesnnatshmtaeitnereaftoaecr-dtisevttaeetrriemoongin,etinnhigec ss usingnewgenerationsofso-calledorganocatalysts(metalfree step2 of the intermolecular aldol catalytic cycle in which e cc systems)ishighlytopical.Sotooistheincreasingadoptionof the catalytically active enamine attacks an electrophile. The A n computationalmodellingtochartthemostprobablemecha- carboxylic acid group of proline plays a central role in the e p O nistic pathway and to establish the factors responsible for model,directingtheelectrophiletotheRefaceoftheenamine. reactivity, selectivity, and particularly stereoselectivity.1 In Theenaminecanbeeitherantiorsyn,relativetotheacid,and this regard, one report2 in 2003 of computational investiga- theelectrophilecanoffertwoprochiralfacesforattack,ReorSi, tionoftheproline-mediatedasymmetricintermolecularaldol resultinginfourdifferentstereochemicaloutcomes(Scheme2). reactions (Scheme 1) has achieved the rare distinction of Informed by extensive computational studies,2–5 Houk and having the key transition state model being named aer the Bahmanyarsuggestedthattheenergydifferencesbetweenthese two principal authors; Houk and List. This computational transition states, and so the origin and degree of stereo- model has informally become known as the gold standard selectivity displayed by the reaction, depends on two critical for such investigations; since it exemplied the procedures for comparing computational modelling with experimental results for organocatalysed reactions and outlined a meth- odology for predicting the stereoselectivity of new organo- catalysedreactions. aDepartment of Chemistry, Imperial College London, South Kensington Campus, London,SW72AZ,UK.E-mail:[email protected] bSorbonne Universit´es, UPMC Univ Paris 06, UMR 7616, Laboratoire de Chimie Th´eorique,F-75005,Paris,France cCNRS,UMR7616,LCT,F-75005,Paris,France dHighPerformanceComputing,ICTservices,ImperialCollegeLondon,UK Scheme1 Schematicmechanismfortheintermolecularaldolreac- †Electronicsupplementaryinformation(ESI)available:AnNCIgurealongwith tion withprolineas a chiralauxiliary,illustratingthe cyclic transition a detailed analysis of the NCI interactions and their relationship to the D3 stateandkeystereocentrescreatedbybondformationbetweenthe correction.SeeDOI:10.1039/c3sc53416b enamineC]Candthecarbonyl(redbond). Thisjournalis©TheRoyalSocietyofChemistry2014 Chem.Sci.,2014,5,2057–2071 | 2057 View Article Online ChemicalScience EdgeArticle cyclohexanone–benzaldehydesystem.Houkhassincedisclosed that calculations carried out subsequent to publication, in which a greater volume of conformational space was investi- gated,revealedthatstructuresofthetransitionstateswithlower energiesexist;1,13potentiallyalteringthepredictedselectivityof thereaction.Inaddition,repetitionoftheexperimentalresults by List found product selectivities to vary, with the reaction found to be extremely sensitive to both water content and to e. temperature. In the present article, we set out to establish nc whetherthesuccessoftheHouk–Listmodelasaconcept,and M. ce ALi thereportedlyexcellentagreementbetweencomputationaland 53:41 ported eitxpiseirnimdeeendtajlursetsiueltds.,isjustcoincidenceorwhetherthefaithin 12:Un 3 0 1/6/202ution 3. Advancesincomputationalmethods ded on s Attrib Tseheentemnayneyarisntchreamt henavtael,elaanpsdedsosminecteimtheesodrriagminaatlicr,epiomrtprhoavvee- oaon ments to computational methods used. In the present article, wnlmm we re-evaluate the Houk–List model by taking advantages of oo 4 February 2014. Dunder a Creative C Scheme2 Thestereochemicalpossibilitiesfortheasymmetricaldol titctohosemebBslefepe;curseauttshaubutsejileeotgcniftonauellddlslmyutsogostgaudienmbesdlstpatiiiornrtodnuvtstueeitdfrdsonersrlacfefrlualumorcttwtiuhinnsseytgrt,heewsexvyhposeiletxcverphemiem.crsAiaemnncmatepannolotrttoaeeelsnatetscnin.acdulhlirynaagvitnees d on 2ensed reaction.Cahn–Ingold–PrelogconventionsareshownforR¼Ph. >la2r0geatmomolsecaunleds,thteheorlgeavenlocoaftathlyesotrsythademopsteeldvetsoacreonpsotrteuncttiathlley Article. PublisheThis article is lic ssstttraautteicctcusartnaablaeildliesomaptteionantsp:pltrahonevairrdeeleadntaitvomeitdnheeeg,rfeaoenrmdtoitnwhgheiacdlhkeogerxaeicdeheo.trfAaenplselicattniroaonr- Hmbbeaoitsuacebhkbl–eLreutitswontecmieosnmoadtpheutelytpieqniucth2aa0llei0trmy3esoowofdauetsrhlc(eieantshaaavelraoweirlaayasybaosln)nedaatbpotlhreaemtgirommesqaetu)t.(iiacrTecfhmooeumernlpetdvratoeoyl- s enamineallowsforthegreatestpossiblenucleophilicityofthe originallyselectedbyHoukandco-workersiscommonlyiden- ces terminal olen while also reducing the geometric distortion tiedbytheabbreviationB3LYP/6-31G*.Geometries,activation c n A experiencedbytheformingiminiumgroup.Theprotontransfer enthalpies, and free energies (DG298) based on calculating the e p fromthecarboxylicacidtotheformingalkoxidewassuggested normalvibrationalmodesatthislevelwereobtainedforagas O asprovidingthemajorityoftheelectrostaticstabilisationandis phase model. To these energies (and at these geometries) keytotheHouk–Listmodel.2,6Smaller,yetimportant,stabilis- variousfurthercorrectionscouldbeaddedusingthepreviously ing contributions also result from NCHd+/Od(cid:1) interactions computedgeometries,includinganestimationofthedifferen- fromthepyrrolidinering.7 tial solvation energy computed for a polarizable continuum Very good agreement between calculated and experimental solvationmodel.Thiscompoundprocedurewasthenappliedto stereoisomer ratios was observed, with the (S,R)-isomer most an exploration of the conformational space available to the favouredforbothR¼PhandiPr(Scheme2).Thechallengeof system. Because this has many degrees of freedom for even computationalmethodsthenistoestablishalevelofreliability quitesimplereactions,onlythemostlikelyconformationswere which can be used to predict the outcome of further reaction explored.Sohowcanthisbasicapproachbeimproveduponin variations. Any success of such models can then be used to 2014?Wehavefocusedonsixaspectssetoutbelow: build condence for diversication to new reactions and 1. The B3LYP density functional as used has, over the last mechanisms. twenty years, proven to be a very effective one forthe study of TheHouk–Listmodelhasenjoyedahighdegreeofsuccess reaction mechanisms, and it has been subjected to extensive in prediction of the stereochemical outcomes of a number of testing and scrutiny during that period. In recent years, there proline mediated reactions: from the inter-2 and intra- have been attempts to benchmark the performance of B3LYP molecular5 aldol, Mannich,3 a-alkylation,8 and Michael reac- againstmoremodernfunctionals.14–16Asaresult,oneparticular tions,9 among several others,1 to the correct prediction of the problem was identied with this functional; it captures only outcome of reactions mediated by analogues of proline.6,7,10–12 short range dynamic correlation energy and fails to capture However, despite the apparently excellent comparison made longerrangecorrelations.Thistermcanbeidentiedinpartas between computational and experimental selectivities, for the dispersion interactions between non-bonded atoms (also which this research became widely known, two signicant referred to as the non-covalent interactions or NCI). It is now issueshavecometolightregardingthedetailsoftheproline– recognisedthatthemosteffectivewaytocorrectaDFTmethod 2058 | Chem.Sci.,2014,5,2057–2071 Thisjournalis©TheRoyalSocietyofChemistry2014 View Article Online EdgeArticle ChemicalScience for dispersion terms is to add empirical corrections to the and even toapply the larger QZVP basis(1944 basis functions nuclear–nuclear repulsion terms, and three or more genera- forR¼Ph)toselectedpoints. tions of methods have been developed over the last decade to 4. The implementation ofcontinuum solvation modelshas achieve this. Grimme17 in particular has pioneered this also improved considerably since 2003.21 Previously, optimisa- approachandwehaveadoptedhisthirdgenerationprocedure, tionofmoleculargeometrieswithsuchmodelswasunreliable, known as D3, to correct the B3LYP method for these energy sincetherstandsecondderivativesoftheenergyinasolvent terms. Most modern functionals also incorporate such terms, continuumoenresultedinsmallinhomogeneitiesduetothe and the uB97XD method, which incorporates a version of way in which the solvent cavity was constructed. York and e. Grimme'searlierD2correction,wasdevelopedveyearsago18 Karplus21 revolutionised this with the introduction of an algo- nc tospecicallyincludebothdispersionandtheabilitytorepro- rithmforobtainingasmoothedsolventcavityandsinceabout M. ce ALi ducereactionbarriers. 2010 it has been possible to fully optimise geometries within 53:41 ported imp2r.oTvheed,baassiscosemtpquutaelritsyhisaavenoitnhcerreafesaetdurinetshpaetecda,nabsepisrtienagdiolyf stiuocnhaappmroodacehl.22caTnhibsemuenadnesrttahkaetna, wmhoerreebseylft-hceonnsoisrtmenatl msoolvdae- 3 12:0 Un coursetoapracticalcomplete-basis-setlimit(CBS).In2003,the frequencies required to correct for thermal and entropy terms 202n 3. limit using conventional computational resources to the total can be computed for geometries obtained using the solvation 1/6/utio number ofbasis functions that could beused fora modelling model at the relaxed solvation geometry, rather than simply ded on s Attrib bstyudHyowukas(tahbeoumto6d5e0rfnunncottiaotniosn.19wFoourlad6b-3e16G-*31bGa(sdis),sientdaiscautsinedg breeainctgioanncaadnhdoecvceolorrpeccthioanrgaepspelpieadrattoioangdausrpinhgasietsmcooduerls.eT(htoe oaon only d-polarisation functions on non-hydrogen atoms) that formazwitterion),and(continuummodel)solvationofcharge- wnlmm wouldnormallymeanalimitofaround100atoms,orperhaps separated species can in turn have a signicant impact upon oo 4 February 2014. Dunder a Creative C ulbpteiaapvvsreeaitslsolwleosaileftbhtpotrlhsueotetveceer1lies2,cns0asielnalriygfygyloScanoTrwglmOyiet-pth3shuuGetbr.e1esrr9steipoAatfeutcctettithnhnetigtssopcdseoegirzreieeoon,dmweu,daeststhianrtesygapse(maetcchiutoewcnehodHdsoedamrsetsarfitiolahvluneaisrr-, ectthhoxaeer5rimree.cffpgTtleeiheooc,etnmsiHtfeooiotsfrruipsdekoosi–slsLvacsairnisetbdtitloeemcnoaionfndrdsedaeeeinltqeiyruonerenceenaoratglgcinptmeiirossoo,neplbdeetuchrtuttahiilteetesspde.riiffdonednctuohtcteoeafsmlsswoooalditvneeelrcn.latuFsdboaeyr d on 2ensed rceoqmupiruetdedtoinloacbaoteuta1n7d0–c2h0a0rahcoteurrisseoftrealnaspistieodntismtaet.esT)hcisouwlodublde ppraoteduincttthheatmaetclheaasntisomne, eeixtphleicritacwtaivteerlymboyleaccutilnegcoausldapparrottiocni- Article. PublisheThis article is lic nctmhoaueutuclmdhraobllarleyeruglcneiomrdmmeitrmteatmohkneeonrady.veTa(giihmrlaeepbetreiloocitfvhyencdooofnlfor1fogo2mir/c1ma6(cid:3)la-apt4idroGovnacbaenylsctseeeosxtrpooslvo(cid:3)yersr9at4eatimdoGnesbcywatthdietaseht), rbwgeaoaltnateeyddrss.o.OraIncnmtdeienocergoden,apssmeacqesocsuihnevasneenlcycieusbmtiisyvsetfihonparvrmtootltivhnoiengngorkveaeelsyaramyaslldahndfraiyetvieaeosnebfnaoeleuernhrgyydidniosrvfocergtseehttniee- s and faster algorithms for analytically computing the Hessian reaction can decrease. Of course, such a super-molecule can s ce havereducedtheelapsedtimeforsuchacalculationbyafactor itselfbetreatedbyacontinuumsolventmodellayeredontopof c n A ofabout100.Consequently,amuchbetterlevelofbasissetcan theexplicitmodel.Herewereportanexplorationoftheroleof e p now be deployed on molecules of such size. In the present waterasbothprotonrelayandashydrogenbondingsolvent. O study,wehaveusedtheTZVPbasis,20atriple-zqualitysetwith 6.In2003,optimisedgeometricalcoordinates,butnoother polarisation functions on both hydrogen as well as non- derived properties (the full computed wavefunction, an IRC hydrogenatoms.ThisbasiswasselectedsincetheGrimme-D3 pathway, etc.)wereinsertedintopaginatedPDFdocuments as dispersion corrections have been specically parametrised for textandsubmittedastheESI.†Re-usingthisdatabyextraction such basissets. This nowgives anopportunity toevaluate the from the PDF document is non-trivial, as we ourselves experi- errors due to basis-set-superposition that might have been enced in re-using the original Houk–List data. The introduc- incurred using the 6-31G* basis. For the system R ¼ Ph tion23 of open digital chemical data repositories in 2005 (Scheme 1), this results in 556 basis functions for the calcula- hasrevolutionisedthisaspectofdatacurationandcitation.24All tion,comparedwith376forthe6-31G*basisandfurtherallows the results in the present study are presented in the form of selectedintrinsicreactioncoordinates(IRC)tobecomputedin interactive tables, themselves assigned a citable persistent a reasonable time. The IRC procedure requires the DFT-based digital-object-identier(DOI),andinwhichtheDOIassignedto Hessiantobeevaluated20–50times,eachtakingabout1hour individual calculations is used to retrieve and visually present usingtheTZVPbasis,inordertomapafullIRC.Thisrequire- thecalculationloglefromtheappropriatedigitalrepository. mentimposesacurrentpracticallimitof(cid:3)800basisfunctions for such a study. It is worth noting that this size of basis Results and discussion precludesthe use ofother Hamiltoniansbasedonperturbative TheHouk–Listbasemodel expansionssuchasMP2,analternativeprocedureforevaluating dynamiccorrelation.Evenonaverylargememory64-processor The investigation began by locating the Houk–List transition system,itwasnotpossibletolocateasingletransitionstatefor statesforthefourstereochemicaloutcomespresentedinScheme theTZVPbasisinanyreasonabletimeusingthismethod. 2.TheoriginalstudylocatedseventransitionstatesforR ¼ Ph 3. It did prove feasible using the TZVP basis to perform a and eight for R ¼ iPr. Two transition states differing in the moresystematicexplorationofconformationspace(Scheme2), conformation of the cyclohexene ring were located for each Thisjournalis©TheRoyalSocietyofChemistry2014 Chem.Sci.,2014,5,2057–2071 | 2059 View Article Online ChemicalScience EdgeArticle stereoisomer, apart from (S,R) anti R ¼ Ph, where only one thecomputercodesthatwereavailabletenyearsearlierin2003, conformation was located. In the present study, we considered acongruenceof0.01kcalmol(cid:1)1betweentheenergydifferences four conformations for each stereochemical outcome,resulting (DDH ) previously reported and those calculated using the 298 in16transitionstatesforbothR¼PhandR¼iPr.Thisexpanded modern versions of the codes25 assures us that the basic conformationalanalysisincludes“chair”or“twist-boat”confor- methodologyishighlyreproducible. mations for the cyclohexene ring and puckering of the proline We nd that the three additional conformations for the ringawayfromortowardstheprotontransfer(Scheme3). observed (S,R) anti product (conformations 1, 2 and 4) are all A repeat of the Houk–List calculations for R ¼ Ph is pre- lower in free energy than the single conformation previously e. sented in Table 1; calculated at the same theoretical level as reportedforthatdiastereomer(conformation3). nc previously employed2 but including the additional conforma- Theoriginalstudypredictedhighenantioselectivity,butlow AM. Lice tionsdescribedabove.AswellastherelativeDFTenergy(DE), diastereoselectivity, for the reaction with benzaldehyde (based 3 12:53:41 0 Unported awpcohtpiivcuahltaiwtoieonrneesnctaohrnaeslpisdhyeo(rDwedHn2i9bn8a)tshaeneddoorfnrigeietnheaenlefsrrtgeueydoye.fnMseorilgnvyadtf(iuDolnGt;2h9ea8a)tc,hitthoiesf rHoenoowmDeHevr2ei9rc8,).rinaTtchilouisstiwhoanastoiwnfatgshoeoesdxepanegreriwmeeelmonwetanletlynweidrtgehytethtrrmeani(cid:3)ns1eitdi:o1nbdysitaLasittseets-. 202n 3. notpracticaltoreplicatethestudybyusingthesameversionsof changesthepredicteddiastereoselectivityofthereactionsothat 1/6/utio it is highly selective towards the (S,R) anti product and is no d on Attrib longeringoodagreementwithexperiment. des an oo nlm Basissetsuperpositionerror wm oo 4 February 2014. Dunder a Creative C taHcehoffcaecnevofciiountnsrngtmotriofaandditgniudocfcrnooteairrsolpsBneosS.drpSaAaEttichner(ebeg,catateshhpnieseptcasanherreteatnxniscttuglleyeppserairiinropngcrooumistmeiyetditpowhlnteoahtdseeeortrlieooomxgrp)pie,lexoosaprrstnalaotowniroteecenledlthaioonesff d on 2ensed itniocnlusduinsgindgistpheersBio3nLYcPo/r6re-3c1tiGo(nds), fmoretthhoedrm(tohcehesmamicaellceavlecluloaf- Article. PublisheThis article is lic poethfffreoeortceertliysanotuibfvsaBeesldSyeSdiEfnecawatrhteaeloliygHshetotsnu,maktys–apsLiuinicsmatlgesrodtorutugdoapycn)a.oe2nl6cecaCmetaloelomynrttpibsca.ersoeTydfshtaeetrmmoefasmogcreneoti,natulsthidoseert s Scheme 3 The four considered conformational possibilities for the small enoughtoroutinelyignorewhen calculating energiesof s ce asymmetricaldolcondensation. organocatalytic transition states with similar connectivity. To c A n e p O Table1 Transitionstateenergies(R¼Ph)calculatedusingtheHouk–Listmethod Transitionstate B3LYP/6-31G* B3LYP/6-31G*/SCRF¼DMSOf Isomer Conf. DEc Pop.b DDG c Pop.b DDH c Pop.b DOIe DEc Pop.b DOIe 298 298 (S,R)[anti] 1 0.00 98.45(54.15) 0.00 99.51(94.89) 0.00 98.75(64.98) 10042/25001,pj3 0.00 99.40(89.49) 10042/25809,pkn 2 0.07 0.06 0.17 10042/25004,pj4 0.59 10042/25091,pkp 3a 2.36 1.60 2.29[2.29]d 10042/25002,pj5 1.87 10042/25090,pkq 4 1.74 1.22 1.62 10042/25005,pj6 1.87 10042/25092,pkr (S,S)[syn] 1a 2.48 1.52(44.34) 3.44 0.45(4.26) 2.68[2.68]d 1.22(33.59) 10042/25012,pkh 3.24 0.54(8.79) 10042/25162,pk4 2 2.60 3.07 2.72 10042/25015,pkj 3.48 10042/25163,pk5 3a 4.34 4.98 4.66 10042/25013,pkk 4.72 10042/25811,pk6 4 3.70 4.29 3.92 10042/25016,pkm 4.48 10042/25164,pk7 (R,R)[ent-syn] 1 8.10 0.01(0.28) 8.20 0.00(0.09) 8.73 0.00(0.15) 10042/25009,pkc 7.02 0.03(0.81) 10042/25810,pkx 2a 5.48 5.73 5.88[5.89]d 10042/25011,pkd 4.65 10042/25099,pkz 3 9.24 9.89 10.53 10042/25008,pkf 7.67 10042/25097,pk2 4a 7.11 7.39 7.91 10042/25010,pkg 5.81 10042/25098,pk3 (R,S)[ent-anti] 1 7.53 0.02(1.23) 6.70 0.03(0.76) 7.50 0.02(1.27) 10042/25014,pj7 7.22 0.03(0.90) 10042/25094,pks 2a 4.60 4.46 4.62[4.62]d 10042/25007,pj8 4.59 10042/25096,pkt 3 8.59 8.10 8.69 10042/25003,pj9 8.26 10042/25093,pkv 4a 6.17 5.54 6.09 10042/25006,pkb 5.77 10042/25095,pkw aConformationsconsideredintheoriginalstudyfromref.2.b%Populationsbasedupontheconformationsconsideredintheoriginalstudy2in parentheses.cRelativeenergiesinkcalmol(cid:1)1.Originalconformationsfromref.2showninblueshade.dRelativeenergiesfromref.2.ePersistent identiers(DOI)fordigitalrepositoryentry.fSinglepointenergycalculationsforasolvation(CPCM)modelatpreviouslyoptimisedgas-phase geometries(B3LYP/6-31G*).AninteractiveversionofthistableisarchivedatDOI:p9d. 2060 | Chem.Sci.,2014,5,2057–2071 Thisjournalis©TheRoyalSocietyofChemistry2014 View Article Online EdgeArticle ChemicalScience testthisassumption,wehaveevaluatedthedifferenceinBSSE d r bBeotyws–eBenertnhaedianctoiuanntderspyonistreanmsiettihoondstfaoter gfoeoumr ebtariseissusseitns:gtthhee DE(SE) 1.13 0.78 0.19 (cid:1)0.04 ntierfo original Houk–List 6-31G(d), the commonly used rather larger de 6q2-.u93a.11l1i)tG2y7(dQu,psZi)Vn,Pgt.hbTeehnterziacplaldelec-zhu-ylqaduteiaolniatsnydwTeZtrhVeePpcaeynrcfdloorhmtehxeeadnqoiunnaedO–rprucrpaollei(n-Vze-. dE(SE) 7.348.474.675.451.511.700.250.21 ersistenti P enamineasfragments,withoverallgeometriesfullyoptimised c 53:41 AM. ported Licence. admretilffoeaTleat(cid:1)hircv1eehensaecreteerrsrsteophsbreeuseclttotawisvrreiee(geTsilnanetibvaslelltleelrr6eo2e-lf)3oat1icthsiGlvoeee(amodlryr)leyylr.lasesrhvcgeoaelnw(s(cid:3)beth0ele.ao8cttfkeetdcrhraeolinromsrd2od0elu(cid:1)r0e13o)t.fofoT1BrhkStecShsaEeel cDOI 10042/2475310042/2475510042/2476210042/2476610042/2474510042/2474910042/2663110042/26627 ––(cid:1)CPBCPB)E(B).B uary 2014. Downloaded on 1/6/2023 12:a Creative Commons Attribution 3.0 Un lTmtbtmrmahheeZrsoiueacVgnultlae(cid:1)Ptialuimrtt1bpihbs)n6slueaeeo-gams3ollciru1itfsoenha1tlcn(epGoa(cid:3)ooftpu(omBpi0drgvrSt.o,ehmp2ipSpmluyE)karttiictbhascaieataaoeotrilestnomrimiogonsfspoernaowanralslhn(cid:1)checatfdr1edoaas)nTs.eslottlbFZnn.iuoetVeleFyscyrnPereoitodobnmhtumeeoresdgcefitpoh.btethlimerenhtees-oseiotzssi-ecQgaiqhnnblrZuleseaioVaimsswglPuciinsitalcbyeitnasxaitls,bpcspwaial(sop(cid:3)nsrererist0saedes.tfctui2ioi,lcosgurtenkgidvottecehhenoasndeetfl, –cDOICP(CPB) (cid:1)3110042/24754634.15629(cid:1)2310042/24758634.14522(cid:1)5410042/24763634.33119(cid:1)5910042/24767634.32019(cid:1)9210042/24744634.37995(cid:1)1110042/24750634.37134(cid:1)4210042/26630634.44060(cid:1)7710042/26626634.43090 –––¼–(cid:1)CPBCPBCPBE)E(CPB)E(CP)+E(CPCPB d on 24 Febrensed under TTenoheeergxeayffmaecnitndsesotafrnutyhcteeuffdreeis,cptthseeras1ido6intsrpcaeonrrssriietoicnotniocsnotrarteecstiforonmmtihgehetxhpaavnedoend CP(CP) (cid:1)634.151(cid:1)634.139(cid:1)634.328(cid:1)634.316(cid:1)634.378(cid:1)634.370(cid:1)634.440(cid:1)634.430 nerror,E(S Open Access Article. Publishe This article is lic ihgmps(vTaTwcbsSiionerehZyiatcoootnnbnidVehhTortmtfrozPeoehnhoigaawTdmnereeelnbasmsdtodreudbfftatremiuoaraeesblch1awetrnstiate)eysminsulicnlotsdaetsrlt,snheaert3eifieseodeffetaowei(tsrnfflu(eoRiohaRnacwntefwfsia¼tdrlccDi.neot¼oothtuihrhsFasmTPufceelTchheroayPsoiaods)nunte+bhnnnhip.ascas)DcneeTtlnsatiuawBoa3hemhd(lclsStn3neceaciesCciiLudo-orsstrcbh–YabnloegrraoeCbnr4edPoebtmneyaoodx+iubcowdty(fmDaetutiRolneyoriit3pnoterddshgn3h/atn¼d.Tre)rtladoo1oydtZr7iifhuwCseiesVaP0topehtPePwsa.rasf3ryC)/clunotae.dSa3cotMmpcCrehsoTegArp˚ttRemm(seheehlsR-iFrr(eoewplearFe-o¼ledCtraoiivcmoriatgngaPziahtxci.poietPeoCtyiteie1ldnnisrgoeMtdod))thien-.cntizsn¼benhoS-yfcmyqeobBsDmceoFutrvae3nofMteiaianrLgtdmoklttalrSt.YlhhieueenretOsPl1eeesyysr-., a–¼forHoukListtransitionstate,RPh bs) –ccDOIE(CPB)DOI (cid:1)10042/24751345.3432810042/24752(cid:1)10042/24755345.3462910042/24756(cid:1)10042/24760345.4419310042/24761(cid:1)10042/24764345.4423010042/24765(cid:1)10042/24742345.4727310042/24743(cid:1)10042/24747345.4715110042/24748(cid:1)10042/26628345.5084610042/26629(cid:1)10042/26625345.5065510042/26624 berivedenamine.EnergiesinHartree.SuperpositioeversionofthistableisarchivedatDOI:qd7. ctaf11hran.7o1er4dm6b.D69oR/32n1t¼-.oy72clo451isPr.1.8ur2ret(cid:4)b4o(csF7ttt2i/ioro1g.a2.n.t171e26f37.)o27AAc˚r.hAt˚,2thatt,/hh1nea7eegnt6erpnd.as1renot(cid:4)wthshtoiebe(tnRiapoostlnprea¼ta-nsilnmetaasvPrifteihseetleyr)go.dgoefeffoTtoohmhtrmheemeeoetrirtcenyrione,ygrsaifrnCmrefcoo–slimCrpunoRsel1nieo¼f.nd2rngoi4Pntmo2hhgf/ nerrorscalculated Fragment(basi E(B) (cid:1)345.33656(cid:1)345.33879(cid:1)345.43715(cid:1)345.43721(cid:1)345.47136(cid:1)345.47002(cid:1)345.50824(cid:1)345.50635 –hexanoneprolined(cid:1)1mol.Aninteractiv t3bc611vchaah147.a1erlG-3asc5pii.nau(s74rldtgaol/(cid:4)ia,1estocptttes.eneio2)tos/fr6Sosn1ii2sCorsv7craRRt8iiaonl.8nFun5¼d8¼1e/uf1b.siGD1aP7cec5berM6atr1dy.aru1/Stetr1en(cid:4)hObe.2eysmf.2ooi6.nmaWbr1e9m5etitceA˚hlsh9haoteefantrroobnyloparta2gtlosrestae.ygh1insest1teotahshe5frftpaamoiffrtrttorueyo.rsnottwohootbcfhhinuttsieatirochtcifsrnnheeoamaernaeadlmavsnl(afpliatueeinomrlrsraga,iubicna6CntlBgn-ie–e3c9dC.1atM7oLGl1bXaP7u(oMDrd25sng1,/P.,8ped56)2ar)/-,; BasissetsuperpositioTable2 MethodIsomer B3LYP/6-31G(d)antisynB3LYP/6-311G(d,p)antisynB3LYP/TZVPantisynB3LYP/QZVPsynanti ¼¼aBbenzaldehyde,CPcyclodkcaldigitalrepositoryentry. Thisjournalis©TheRoyalSocietyofChemistry2014 Chem.Sci.,2014,5,2057–2071 | 2061 View Article Online ChemicalScience EdgeArticle accepted wisdom tends to the belief that such corrections are onlysignicantformuchlargersystems. For both the R ¼ iPr and R ¼ Ph systems, inclusion of D3 corrections results in an increase in the planarity of the enamine. Examining the enamine geometries for R ¼ Ph, two cleartrendsappear;theenaminenitrogenissignicantly(by5 to 10(cid:4)) less planar in the enantiomer, (R,R) and (R,S), confor- mationsthancomparedtothe(S,R)and(S,S)conformations.In e. addition, enamine planarity is loosely correlated to relative c n stability; the lowest energy structure has the most planar M. ce ALi enamine and vice versa. The lowest energy structure, (S,R) 53:41 ported wcohnefroeramsatthieonhi2g,hheasst aennearlgmyosstrtutcottuarlley,p(Rla,Sn)acroennfaomrminaetio(1n773.,2i(cid:4)s) 1/6/2023 12:ution 3.0 Un S2FiCg(R.ScF1h¼eDmCMeoSmO3)p(ufBot3erLdRYP¼s+eDlPe3hc/t,TesZdhVoPwb/SoinCngdRFth¼leeDnMcgotShnOstr)al(ceA˚tv)ioelnaotifncBno3onLnfYo-Prbm/oTnaZtdVioePdn/ hthigaThthltyheepry(eSrsa,uRml)tissdtfaeolriresoeRids¼o(m1P6eh7r.0fisu(cid:4))lt.lhyecdoonmrimnanthtespeeacrileiesrfocormnceldu;sitohne ded on s Attrib dveisrtsaionnceosftwhhisefingudriespiserasricohnivecdorartecDtOioIn:qids8.included. An interactive nfoeramre,swthalitcehrneamtievregestseareso4is.5o9mkecrailsmcoonl(cid:1)fo1rhmigahtieornin2forfeetheen(eSr,gSy) oaon usingB3LYP/TZVP/SCRF¼DMSOwithoutdispersioncorrection, wnlmm butthedifferenceisreducedto2.97kcalmol(cid:1)1atthisbasisif oo d on 24 February 2014. Densed under a Creative C hscbtAraooniadnmnnEdddispnenwfuenetifrrtrcoegi.ihnrndiTmeesthdsaieacetrrtbmreBiereos3eetnuiwaLdslcYietinatiePsintoon+egns,DsCyfit3nrg–hco/nCcoTemhoZobfrrVrfoodtuPhwnnisn/eodhtSaruCamaftstoeRtoerCFimrdfnr¼eoeafrDmcotboirMooRemmtrnSh2pa¼O9atlhneic(otaadFPenssihpgecros.roefao3fnnetw)afordosirntmrhehmdRotirnwaaptito¼snrmioosCatufnsioe–PmarCnarl. ttdtilamhhoereac9gpWtre8Doe.erai36Qtns%acceZanollVuttserhPosrepeieaofbcrripntaeroirscom(eiaRnosdlf,cie¼icusffrDteleePias3ndchtutec,ileldcvtucnoieondarltnygrenhefdt1eocii0ntoriin0mtsor%aenatanlhnes,tiecswoitinfpnitiovictrorhir2otneoy,catue(hfsSsdtoei,tsurRdair)titn,eseshpy.s(iSstesOht,rfSevmeosm)eir)oornaanduttlth.elme,tlenh,wbdedmeesovrsoweetorsnoeytf s Article. Publishe This article is lic eR(cid:1)cpoxo3p¼rs9rlieotiickProtacrnit.aoilonTnehmorearnosorleer(cid:1)thr1se.eefsotBucrooolutmtsththpiiinntushctTleeucaddobmrerlreeaetlgh3acnttefiiioovteuernffdReeencr,¼taesanrPgngohidifensatgih,nnetdhfpDrieona3mrt-oTtditaciasbu(cid:1)lplleaes3urr64spitfoehotrnoer- aaDbcannaDlscGiiiunsm‡2lc9afpr8uteriinaaoncscnctetriiseococaanfasls(cid:3)leeasf1sprs0opal0mirsgofh(cid:3)ao5tclN5ldhy63,–ffmtrNooor4am1k(g9Ni2en4.n4g9¼e.7trTnhatohlueem3mu.ts2beie7mechrokeafocnatfsaliubskmctaehisncoilas(cid:1)efoxfl1uarp(rnlTtgohcaertebaiboltDeianosF3sniT))s,,. ces variation ((cid:3)3 kcal mol(cid:1)1), may come as a surprise, since the aresultthatmightbeassumedtobeclosetothecompletebasis c A n e p O Table3 CalculatedtransitionstatepropertiesforR¼Ph(Scheme2) Transitionstate B3LYP/TZVP/SCRF¼DMSO B3LYP+D3/TZVP/SCRF¼DMSO Isomer Conf. DE DDG a Pop. DOIb DE D3c DDG a Pop. DOIb 298 298 (S,R)[anti] 1 0.00 0.00 99.94% 10042/24848 0.00 (cid:1)36.52 0.05 99.30% 10042/24862,prz 2 0.44 0.18 10042/24847,prf 0.25 (cid:1)36.79 0.00d,g,i(+20.5)f 10042/24863,pr2 3 1.04 1.68 10042/24849,prg 1.47 (cid:1)35.99 2.20 10042/24864,pr3 4 0.98 1.64 10042/24850,prh 1.07 (cid:1)36.38 1.87 10042/24867,pr4 (S,S)[syn] 1 3.75 4.60 0.05% 10042/24859,prt 1.57 (cid:1)38.93 3.04 0.68% 10042/24875,psf 2 3.99 4.59 10042/24860,prv 1.73 (cid:1)39.06 2.97j 10042/24877,psg 3 4.67 6.34 10042/24861,prw 2.70 (cid:1)38.75 4.78 10042/24876,psh 4 4.35 6.00 10042/24866,prx 1.96 (cid:1)39.19 4.27 10042/24878,psj (R,R)[ent-syn] 1 5.73 6.96 0.01% 10042/24855,prp 5.24 (cid:1)37.11 7.05 0.01% 10042/24871,pr9 2 3.99 5.43 10042/24857,prq 3.19 (cid:1)37.56 5.14 10042/24872,psb 3 6.16 7.72 10042/24856,prr 5.87 (cid:1)37.08 7.59 10042/24873,psc 4 4.70 6.31 10042/24858,prs 4.30 (cid:1)37.22 5.93 10042/24874,psd (R,S)[ent-anti] 1 7.34 7.51 0.00% 10042/24851,prj 7.51 (cid:1)36.32 7.14 0.01% 10042/24868,pr5 2 5.43 6.15 10042/24852,prk 5.35 (cid:1)36.54 5.38 10042/24865,pr6 3 8.06 7.78 10042/24853,prm 8.36 (cid:1)36.24 7.88(7.72)e[8.11]h 10042/24869,pr7 4 6.19 6.52 10042/24854,prn 6.43 (cid:1)36.44 5.81 10042/24870,pr8 akcal mol(cid:1)1. bPersistent identier (digital-object-identier) for digital repository entry. cGrimme's D3 dispersion correction (ref. 17), in kcal mol(cid:1)1. duB97XD/TZVP/SCRF¼DMSO, DOI: n47. euB97XD/TZVP/SCRF¼DMSO, DOI: n48. fHajos–Parrish mechanism, DOI: n49. gMP2(FC)/6- 31(G)/SCRF¼DMSO, DOI: ppd. hMP2(FC)/6-31(G)/SCRF¼DMSO, DOI: ppf. iB3LYP/QZVP/SCRF¼DMSO, DOI: p8h. jDDG 3.27 kcal mol(cid:1)1, 298 DOI:p8gAninteractiveversionofthistableisarchivedatDOI:qcc. 2062 | Chem.Sci.,2014,5,2057–2071 Thisjournalis©TheRoyalSocietyofChemistry2014 View Article Online EdgeArticle ChemicalScience Table4 CalculatedtransitionstatepropertiesforR¼iPr(Scheme2) Transitionstate B3LYP/TZVP/SCRF¼DMSO B3LYP+D3/TZVP/SCRF¼DMSO Isomer Conf. DEa DDG a Pop. DOIb DEa D3c DDG a Pop. DOIb 298 298 (S,S)[anti] 1 0.00 0.00 100.00% 10042/24880,psq 0.00 (cid:1)35.89 0.00 99.99% 10042/24895,ps9 2 0.41 0.24 10042/24881,psr 0.19 (cid:1)36.13 0.38 10042/24896,ptb 3 0.75 0.32 10042/24882,pss 1.20 (cid:1)35.41 0.59 10042/24898,ptc 4 0.66 0.46 10042/24883,pst 0.70 (cid:1)35.81 0.62 10042/25980,ptd e. (S,R)[syn] 1 6.11 6.66 0.00% 10042/24891,ps5 4.99 (cid:1)37.06 5.60 0.01% 10042/24906,ptp nc 2 6.49 6.69 10042/24892,ps6 5.40 (cid:1)37.10 5.59 10042/24908,ptq 3:41 AM. orted Lice (R,S)[ent-syn] 431 777...237451 888...033046 0.00% 111000000444222///222444888989473,,,pppsss8z7 586...851866 (cid:1)(cid:1)(cid:1)333775...310112 679...903282 0.00% 111000000444222///222444999000927,,,ppptttsjr 3 12:50 Unp 32 76..8118 96..1885 1100004422//2244888889,,ppss32 86..4780 (cid:1)(cid:1)3355..1482 97..7379 1100004422//2244990043,,ppttmk d on 1/6/202Attribution 3. (R,R)[ent-anti] 3412 86—6...657817 87—7...150296 0.00% 111—000000444222///222444888898504,,,pppsssw4v 866—...682397 (cid:1)(cid:1)—(cid:1)333556...954548 886—...509686 0.00% 111—000000444222///222444998009059,,,ppptttgnf des 4 6.55 6.55 10042/24886,psx 6.23 (cid:1)36.24 6.41 10042/24901,pth an oo nlm akcal mol(cid:1)1. bPersistent (persistent-object-identiers) for digital repository entry. cGrimme's D3 dispersion correction,17 in kcal mol(cid:1)1. An wm oo interactiveversionofthistableisarchivedatDOI:qcd. DC uary 2014. a Creative 4 Februnder d on 2ensed s Article. Publishe This article is lic Fig. 2 Computed selected bond lengths at B3LYP+D3/TZVP/ es SCRF¼DMSOfor(a)R¼iPrconformation1(Scheme3)inA˚ (original c Ac Houk–Listvalues)and(b)R¼Phconformation2(originalHouk–List n values at the B3LYP/6-31G* level) [MP2(FC)//6-31G(d,p)/SCRF ¼ e Op DMSO],{uB97XD/TZVP/SCRF¼DMSO}. set(CBS)limit.Thisresultisneverthelessstillatoddswiththe reportedapparentexperimentaldiasteromericratioofbetween 1:1and 4:1(DDG‡ < 0.8kcalmol(cid:1)1)butaswe note inthe 298 introduction, the experimental value may be sensitive to both water content and temperature. Similar conclusions can be drawnforR¼iPr(Table4). The maximum difference in energy between the lowest energy conformer of the (S,R) form and the highest energy conformer of the least stable diastereomer (R,S) is 7.88 kcal mol(cid:1)1 (B3LYP+D3/TZVP), 7.72 kcal mol(cid:1)1 (uB97XD/TZVP, which includes an implicit D2-dispersion correction) and 8.11 kcal mol(cid:1)1 (MP2/6-31G(d,p); it proved impractical to use the larger TZVP basis for this calculation), a result which is grati- Fig.3 Computedintrinsicreactioncoordinate(IRC)atthereference fyinglynotoverlysensitivetothemethodused. B3LYP/TZVP+D3/SCRF¼DMSO level for (a) R ¼ Ph, (b) R ¼ iPr. The correspondingdigitalrepositoryidentifiersareDOI:n46andn45. Non-covalentinteractions In general terms, stereoselectivity is controlled by a combina- interactions within the framework. Although the balance of tion of stereoelectronically-induced bond orientations in the theselatterinteractionscanbedifficulttocomputefromtotal transition state, but inuenced by non-bonded weak energies alone, it has been recently shown30 that the reduced Thisjournalis©TheRoyalSocietyofChemistry2014 Chem.Sci.,2014,5,2057–2071 | 2063 View Article Online ChemicalScience EdgeArticle density gradient (RDG) can be used to reveal the interactions interactions, green dispersive interactions appear elongating thatarepresentinthesystem(seeref.28formoredetailsonthe the NCI feature (see for example Fig. 4b) which highlight the method). importance of the planarity of this region (highlighted by the Thisindexenablestheidenticationandcharacterizationof distancesinTable3)andwhichcannotbemerelyexplainedby weak interactions of various strengths as chemically intuitive electrostatics. RDGisosurfacesthatrevealbothstabilizing(hydrogen-bonding 2. An extra region appears in the syn conformer (Fig. 3c), a interactions in blue, van der Waals interactions in green) and greensurfacebetweentheprolineandtheR¼Phgroup.This destabilizing interactions (steric clashes in red, and weaker interactioncanbeidentiedastiltedT-shapeinteractionorasa e. ones to yellow). This approach can be much more effective at p-facial hydrogen bond (a gure for such interactions is nc revealingthenon-bondedinteractionspresentandalsoprovide providedintheESI†forreference).Itisimportanttonotethat M. ce ALi amorerealisticanalysisthanpuretopologicalindicessuchas this newinteraction,which stabilizesthe synconformers, had 53:41 ported QviTsuAaIMl .m31aTnhnuers iitnrewphriecshenttos aansaelmysie-quwahnattitaitsiveoaenndmhiegrhellyy nHootwebveeern,itisdpernetsienecdeebneafobrleesbuystmoeexrpelagineotmheetfraiccttihnastptehcetiosynn. 3 12:0 Un described in a non-quantitative (and oen intuitive) manner conformersaretheoneswiththegreatestdispersioncorrection 202n 3. merely as transition state steric clashes and hydrogen bond/ (Table3). ded on 1/6/s Attributio eTlaebcAltlerlosNs2tCatIaincdgaut3trr.eascIntcioaonnrsdb.eerftoounhdighinligthhtecinerttearianctfiveeatvuerressi,onthsreoef bdoisnpTdehriusnsio,thnae(reciinothgmerbreigniinaotnito)hndeeNtoeCfrmHNidnC+/eHtOdh+d/e(cid:1)oOuodtr(cid:1)coTme-slehecaotpfroet/hsptea-ftraieccasicatliaonHnd-, oaon representative examples from the R ¼ Ph reaction have been withonlyanti andsynasobservable diastereoisomers (seethe wnlmm chosen. Fig. 4 dissects the diastereoisomer with the highest ESI†foramoredetailedanalysis).Itshouldbenotedthatthis oo d on 24 February 2014. Densed under a Creative C e(cicfssaohoononrrvoemltrarawielg)taeceytatnaiitoto(nhgenndirenn.etp(taA-tseatuhtyrenrnearte)cirl.)Rtsdy,iTtDointahgnhGoslesetanesnicrduncuceoireotasfvo,iastasffbatcloeboeemr(n,tierlhtwioez¼ritihinshniotewcg0emh.iar1ttanehhccrtaeoti.itnwuaodhn.niie)tardshsmhhtataehtslishrogeetnhebhosgeeyielssneowtonrwciomtdeolcheseinhstrtfopoihocesefrenermtnsneCrioareo–gnntrCnoys-- bocnKanaoilnlctcaeeubntleicaacegtoiiaboosiotnfnats,eoi,nltpeeheecldetser.e3onff2st,ee3ht3ccaeettssiccsoitaryrnedtcotdieinsnpcelerurgdseietviecdsiisnaptneedrrasgicoetinoomnesffetehrciietgsshliwingohuthtlsde, s Article. Publishe This article is lic sifIanetintatcidetoournranrrceoestsnctipoao-ctnoreornssevd.sahsplFaeootvnonretd.aabsAmemilenoinodnar-glerhlaiwtdnghihegrttehelaiegiitlnhcehtatdeeesrdeaaCsnc–ittnaCoiloytFnfshoii,egsri,.mnfot4-ahbr.tmeeiTotiwmnhne,egoetsCndwt–eocrCeoepplvbeauobvlrnaeleunndleytt. TtrHmrehaaaeenjcnosttusiiao–ttliinPlolyia.nt3ry4aissnoFthafdo–trheEcaediovsxemiatrnmh–pgSaupatatluaiectet,oicroMm–anWnepayyiuleelityerce,lhddiHenfrkvootieunrscrketeei,tgaiccacaonttiiensdodosnttc,ao1onp3-sCwteumogeKrffgokIeeEdescrsettsislnffefogooxrrrptehttthhhraieeet- s ce non-covalentregionsappear: the rate determining step of the proline-mediated intra- c n A 1. The region of around the heteroatoms (C]O/N) shows molecularaldolreactionoccurspriortoC–Cbondformation.35 e p stabilizingfeaturesinalltheconformers. Thisfeature ismost In contrast, for the proline-mediated intermolecular aldol O important in the anti and syn conformers, whereas it is much reaction between acetone and o-chlorobenzaldehyde, Arm- weakerintheent-antiandent-synones.This3Dviewcoincides strong,Blackmond, andco-workersreportedanormal isotope with previous approaches, which locate the relevance in the effect of kH/kD ¼ 1.93–2.26 when using acetone-d .36 Previous 6 NCHd+/Od(cid:1)interaction.However,alongwiththeelectrostatic kineticinvestigationsofthereactionhadledtheresearchersto Fig.4 NCIanalysisofseveralconformersofthecomputedtransitionstate:(a)thelowestenergyisomer,(b)thehighestenergyisomer(c)the highestD3correctionisomer(Table3).Thegradientisosurfaces(s¼0.5a.u.)arecoloredonaBGRscaleaccordingtothesign(l )roverthe 2 range(cid:1)0.03to0.03a.u. 2064 | Chem.Sci.,2014,5,2057–2071 Thisjournalis©TheRoyalSocietyofChemistry2014 View Article Online EdgeArticle ChemicalScience expect an inverse isotope effect. The observation of a small, theory.10 The researchers proposed, that only via the explicit normalisotopeeffectimpliedthatthenalKIEwastheresultof inclusionoftwomethanolmoleculesinacooperativehydrogen- abalancebetweennormalandinverseisotopeeffectsresulting bonding network between the carboxylic acid proton and the respectively from synchronous proton transfer and C–C bond oxygenatomsofanitrogroup,couldthecorrectstereochemical formationinvolvingachangeofC-hybridisationfromsp2.This outcome of the reaction be predicted. This study however is impliesthatsuchisotopeeffectsareindeedasensitivemeasure based on the use of total calculated energies for transition ofhowrealisticacomputedtransitionstateis.Computational states, rather than the more correct thermal and entropy-cor- modelling at the original Houk–List level of B3LYP/6-31G(d)/ rectedrelativefreeenergiesforthereactionprole.Inaddition e. SCRF¼DMSOleveloftheoryhadpredictedaKIE of1.97,thus to molecules of solvent, additives such as triethylamine9 and nc offeringfurthersupportforthishypothesis.Thisevidencealso DBU37 have been shown to play an important role in deter- M. ce ALi allowedthereasonableinferencethatnucleophilicadditionof mining the selectivity of proline-mediated reactions through 53:41 ported tmhienienngamsteinpeotfoththeecaetlaelcyttriocpchyiclelewbayskiinndeteiecdintvheestrigaatetiodnest,era- theHireerxep,liwceitiinncvleusstiioganteinntoratntshiteiornolsetaotef caadldcuitliavteiobnus.t that of 3 12:0 Un processaidedbyamoreorlesssynchronousprotontransferto water in the Houk–List transition state. The rst step of the 202n 3. thecarbonyloxygenfromthecarboxylproton. catalytic cycle, which involves the formation of an enamine 1/6/utio Wefeltitimportanttoestablishthatthevariouscorrections fromthecarbonylcompoundandproline,liberatesonewater ded on s Attrib dKeIEscsri(bTeadbleab5o)veforrettahienptrhoilsinbee-mhaevdiioautre.dWineterrempoorletccualalcrualaldteodl msuoblseecquulee.nPtorteeancttiiaolnly.,Htheirsew, aatesrincgolueldwathteernmpoalreticcuipleatiesiinnttrhoe- oaon reactionbetweencyclohexanoneandbenzaldehydeandalsothe ducedintotheHouk–Listmodel(R¼Ph)asapassivesolvent nlm wm acetone and m-chloro-benzaldehyde system for which previ- foroneofthethreeoxygensviaahydrogenbond,augmenting oo d on 24 February 2014. Densed under a Creative C TrahaeoeffnsuhavdvseeeeelcasyctctlbsaeaamerdlbabecelgunoibasalnsyhapiunteeteerhdrffedfredoeeicrinIsmtvResoatCcetlhatoudenfpetodusehprserahieRenffavssgdi¼myeonctuahbtcPslsehelherfsli,noooneenrwcvxrtoeeehtiruohspistsbneoeie.rtnoetniTennaxedhytgarug.egm3erny6enpiotTnrrroeeheamffdfenfaiestocclchitrtatedemilod(ceRneruueidlts¼saaeotctrfatirtiPootuioehopnmmsn)es,. ktws(toShrmcnaa,eaRtnalet)clshrmloiaetcmnniohotdrloina(cid:1)enln(l1euasS.gctt,uMeuaSimvt)lioeeensdsstteeteooihnlsrlseeevtshorawhcgteshooiieoemimwlcsnenhpocoatudiiiflnntvereilecttdThlaywuade,fadbrycseealheaencmapto6enipomn.glarAeiipennerdugddgwdyt.aieistTdttdiyehiofnffrrtneooemdmeroiieofffnanle2sdoectc.nucee9tucer7bteelheeeotsxooitmspwnw3lelehie.yrce1ariin5aecst Article. PublisheThis article is lic paienffcrcoeeotlcoirtnpneioes-radaan6tedcdaosamicnseptotoolovsnteihnetertereoeafardecaitsimntpignoricrmwetialtcorhoycmamhtpiy-olcdenhrxsol,.ogrsTeoinhn-bucKeesn,IEdzthaeoluedrteioeghbriiynsudeamretvineiindgs niPnrocotrteosantsuerdesilaeraydp,midnelcoyh.tanleiassmtsbecause the stochastic complexity s from the proton transfer and an inverse one originating from ces theisotopeattachedatthecarbonformingtheC–Cbond.The Beyond a passive role, it is possible that a single molecule of c n A agreement between experiment and theory is if anything even water could participate directly in the Houk–List mechanism, e p morecongruent. acting as an active proton transfer relay between the proline O carboxylic acid group and the incoming aldehyde (R ¼ Ph) Passiveexplicitsolvation (Table 7). This results in the expansion of the ring size of the cyclicmodelbytwoatoms,thusallowingsomestereochemical Patil and Sunoj have computationally investigated the role of models previously excluded on the basis of strain to be explicit solvent molecules in the modeling of the proline- included. Proton relay mechanisms in organocatalysis have mediated conjugate addition reaction in polar protic solvents, precedent. One of the initial mechanistic proposals for the at the mPW1PW91/6-31G(d) and B3LYP/6-31G(d) levels of intermolecularaldol reaction,the HPESWreaction, involveda second molecule of proline acting as a proton transfer agent duringC–Cbondformation,basedontheapparentobservation Table5 Calculatedakineticisotopeeffectsat298K,R¼Ph of a non-linear effect.38 This proposal was later discredited R kH/kD k12C/k13C]O k12C/k13C]C k16O/k18O]C through experimental studies conrming the absence of non- linear effects in the proline-mediated aldol reactions.4,39 Patil Phb 4.37c(3.01)d 1.019 1.007 0.993 and Sunoj have also investigated the role of a proton-relay Phe,f 1.18c 1.043 1.027 1.019 mechanism in hemiacetal-, iminium, and enamine formation iPrh 4.05c 1.026 1.0096 1.005 m-Cl-Phi 2.11[1.93–2.26]g — foramodelsystemmimickingtherststepinthecatalyticcycle of proline catalysis, at the mPW1PW91/6-31G(d) level of aB3LYP+D3/TZVP/SCRF¼DMSO,derivedfromthethermallycorrected theory.40It was found that protic additives, such asmethanol, free energy DG based on harmonic uncorrected frequencies, with appropriateisot2o98pemasses.bReactantDOI:pn6.cd onOH.dd on allow for signicantly lower activation energies of these path- 1 4 OHandthethreeC–Hgroupsonthecyclohexanone-derivedenamine. ways. The reported energies did not include the effects of eReactantDOI:pn7.fValuesforprotonrelaymechanisminvolvingan entropy, which would be expected to result in signicantly additionalwatermolecule.gd onOHandtheveC–Hgroupsonthe acetone-derived enamine. hRe6actant DOI: pn8. iReactant DOI: ppb, higherfreeenergiesofactivationcompared tothose basedon transitionstateDOI:pn9forthesystemreportedinref.34. total energies alone. Here we present computed free energy Thisjournalis©TheRoyalSocietyofChemistry2014 Chem.Sci.,2014,5,2057–2071 | 2065 View Article Online ChemicalScience EdgeArticle Table6 PassiveexplicitsolvationoftheHouk–Listtransitionstate,R¼Pha Transitionstate B3LYP+D3/TZVP/SCRF¼DMSO Isomer Conformationd DEb DDG b Pop. DOIc 298 (S,R)[anti] 1(2) 0.37 1.08 99.31% 10042/25146,pk9 1(3) 0.22 0.00 10042/25147,pmb 2(1) 0.00 2.29 10042/25148,pmc 2(2) 0.72 1.55 10042/25149,pmd e. 2(3) 0.48 0.18 10042/25150,pmf c n (S,S)[syn] 1(1) 1.53 4.62 0.69% 10042/25151,pmg AM. Lice 1(2) 1.22 3.52 10042/25818,pmh 3:41 orted 12((31)) 11..6553 33..7747 1100004422//2255115523,,ppmmjk 3 12:50 Unp 22((23)) 11..5807 33..5185 1100004422//2255185149,,ppmmmn 1/6/202ution 3. eanMtroyd.edlAcsomdeprniseindginonSechademdietio3n.aAlnwiantteerrmacotilveecuvleer,shioyndroofgtehnisbotanbdlienigsianrvcahriivoeudsalotcDaOtiIo:nqsc.sb.kcalmol(cid:1)1.cPersistentidentierfordigitalrepository d on Attrib des an nlomo Table7 Protonrelaymechanism,R¼Pha solvatedisomer.Becauseprotontransferreactionsinparticular wm oo can depend on the nature of the density functional used, we uary 2014. Da Creative C ITsroamnseirtionstaCteonf.d BD3ELbYP+DDD3/GT2Z9V8bP/SCPRoFp¼. DMSODOIc uaTlshsieongefrvBeae3luLeaYntPeed+rgDtyh3i,dsbieffunetersregtniylcledciofflefearr6el.yn4ceekxccaaltultdhmeesoult(cid:1)hB1e97iisXnDvsmo/TlvaZelVlmePrelntehtvaeonlf. 4 Februnder (S,R)[anti] 12 34..7291 53..1655 23.03% 1100004422//2255882201,,pppp78 winactoerrreincttshteereporochtoenm-rieclaalyo.uTthcoismme;oidteml wayounlodtaallswoaypsrebdeicsto,thoef d on 2ensed 45 10..5053 20..5689 1100004422//2255116555,,pppq9b cou3r.se,Tfhoerotihnetrrinmsieccharenaiscttiiocnvarciaotoiordnisn.ate (B3LYP+D3/TZVP/ Article. PublisheThis article is lic 11678901 220232......471868308885 230333......895441058352 111111000000000000444444222222//////222222555555111881765226176236,,,,,,ppppppqqqqqqhjcdfg SriinseCltnaReoyrFvwd¼eenaDltamMiyosunScCO).h–,OCsFhnibgoloy.rnt5aed)rfei1nor.9rdtm2ihc0aiasAt˚tieot(srcnaot;nhmistaispttilaoetrhnneegdstptthaorteae2st.e2itsnh7ce3peatA˚rsoaswfneistdthihtedoiooupnetrsssoutttaohctnhee s protontransfercommence.AtavalueoftheIRCindicatedwithan s 12 2.83 4.03 10042/25169,pqk cce 13 2.15 3.96 10042/25172,pqm arrow(Fig.5b),C–Cbondformationislargelycomplete(1.676A˚) A n (S,S)[syn] 1 4.78 6.26 76.97% 10042/25168,pqn andaso-calledhiddenintermediateisrevealedinwhichastrong e Op 2 3.65 3.56 10042/25170,pqp symmetric hydrogen bond between the proton relay and the 3 0.08 0.27 10042/25157,pqq carbonyl group manifests. The signicance of detecting such 4 0.00 1.06 10042/25158,pqr intermediates is that the electronic inuences at this geometry 5 0.08 0.00 10042/25159,pqs 6 0.95 0.93 10042/25173,pqt couldbepotentiallytargetedinordertoinducethesystemtoform 7 1.49 1.33 10042/25824,pqv arealintermediate;therelativetimingofe.g.C–Cbondformation 8 2.56 2.56 10042/25825,pqw andproton-relayscanthereforebeseenasafunctionofstructure 9 3.84 5.73 10042/25826,pqx and substituents. This feature of a potential energy surface is 10 3.00 4.90 10042/25827,pqz relatively insensitive to e.g. the density functional procedure 11 1.07 2.22 10042/25828,pq2 12 0.36 1.42 10042/25832,pq3 employed.ShowninFig.5cisthesamereactionchartedusingthe 13 0.44 0.98 10042/25160,pq4 uB97XDfunctional;thehiddenintermediateoccursattheposi- aBasedonconformationsderivedfromthetransitionstatesdescribed tion in the IRC, but its prole is stronger (the gradient norm inTables1–4.bkcalmol(cid:1)1.cPersistentidentierfordigitalrepository approacheszeromorecloselythanwiththeB3LYPfunctional). entry.dAsdenedintheESI;†seeDOI:rdpAninteractiveversionof 4.Thekineticisotopeeffectscomputedforthismechanism thistableisarchivedatDOI:qc3. (Table5)areverydifferentfromtheoneinvolvingnointerven- tionbywater. We therefore consider it unlikely that an explicit water differencesfortheprotonrelaymechanism,preciselyisomeric molecule is actively involved in the rate-determining step for withthepassivemechanismdescribedintheprevioussection. thisreaction. Fourfeaturesarenoteworthy. 1.The(S,S)diastereomerisnowpredictedasslightlylowerin Hajos–Parrishmechanism freeenergythantheobserved(S,R)products. 2.Thelowestfree-energyforaproton-relaytransitionstateis This mechanistic variation proposed by Hajos and Parrish41 nevertheless 9.1 kcal mol(cid:1)1 higher than that of the passively fortheintramolecularaldolreactioninvolvesaninitialproton 2066 | Chem.Sci.,2014,5,2057–2071 Thisjournalis©TheRoyalSocietyofChemistry2014

Description:
noncovalent interactions that influence the selectivity of the reaction, confirming the role of the electrostatic tion of the proline-mediated asymmetric intermolecular aldol . parallel processing computer of the period, the second deriva- bonding network between the carboxylic acid proton and th
See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.