Chiral recognition by mass-resolved laser spectroscopy

CHIRAL RECOGNITION BY MASS-RESOLVEDLASER SPECTROSCOPY*

Maurizio Speranza,1{ Mauro Satta,2 Susanna Piccirillo,3 Flaminia Rondino,4

Alessandra Paladini,4 Anna Giardini,2,4 Antonello Filippi,1 and Daniele Catone41Facolta di Farmacia, Dipartimento di Studi di Chimica e Tecnologia delleSostanze Biologicamente Attive, Universita di Roma ‘‘La Sapienza’’,pl. A. Moro 5, I-00185 Roma, Italy2CNR-IMIP (sezione Istituto Materiali Speciali),I-85050 Tito Scalo (Pz), Italy3Universita di Roma ‘‘Tor Vergata’’Dipartimento di Scienze e TecnologieChimiche, Via della Ricerca Scientifica, I-00133 Rome, Italy4Dipartimento di Chimica, Universita di Roma ‘‘La Sapienza’’,pl. A. Moro 5, I-00185 Roma, Italy

Received 17 March 2004; received (revised) 19 June 2004; accepted 19 June 2004

Published online 8 November 2004 in Wiley InterScience (www.interscience.wiley.com) DOI 10.1002/mas.20040

Chiral recognition is a fundamental phenomenon in life sciences,based on the enantioselective complexation of a chiral moleculewith a chiral selector. The diastereomeric aggregates, formed bycomplexation, are held together by a different combination ofintermolecular forces and are therefore endowed with differentstability and reactivity. Determination of these forces, which arenormally affected in the condensed phase by solvent andsupramolecular interactions, requires the generation of thediastereomeric complexes in the isolated state and theirspectroscopic investigation. This review deals with chiralrecognition in the gas phase through the application of laser-resolved mass spectrometric techniques (R2PI-TOF and RET-MS). The measurement of the fragmentation thresholds ofdiastereomeric clusters by these techniques allows the determi-nation of the nature of the intrinsic interactions, which controltheir formation and affect their stability and reactivity. # 2004Wiley Periodicals, Inc., Mass Spec Rev 24:588–610, 2005Keywords: gas-phase clusters; supersonic beam; laser spec-troscopy; chirality; mass spectrometry

I. INTRODUCTION

Molecular clusters are noncovalent aggregates widely recog-nized as a new state of matter with neither the properties of theirindividual constituents nor those of their respective bulkcondensed phases. Indeed, the formation of a cluster does affectproperties of its components compared to both the isolatedmolecules and the bulk solution. The stronger are the noncovalentinteractions in the cluster, the larger are the changes in the

properties of the components. These changes are important formonitoring the formation itself of clusters and for evaluating theirstability and reactivity.

Tailor-made molecular clusters have proved to be idealsystems for modeling molecular recognition phenomena. In thelast few years, it has become possible to study the singleinteractions acting between the individual components ofclusters in the isolated state, without any interference from theenvironment. Their structural arrangement can provide a bench-mark against which the influence of any dynamic environmentmight be assessed (Rizzo et al., 1985, 1986; Philips & Levy,1986; Rizzo, Park, & Levy, 1986; Nir et al., 2001; Robertson &Simons, 2001; Plutzer et al., 2001). These studies were carriedout with the use of specifically designed spectroscopic meth-odologies, capable of characterizing molecular clusters in theisolated state and probing their chemical bond breaking andforming on an extremely short time scale. Their applicationfostered important breakthroughs in many scientific fields, whichsurpass the purely chemical threshold by interlocking thephysical and life sciences.

Natural and synthetic enzymes are invariably characterizedby asymmetric structures with a cavity of appropriate shape andsize holding suitable functionalities in specific positions. Theremarkable catalytic proficiency and the exceptional selectivityof enzymes towards biomolecules are ascribed to a combinationof: (i) shape-specific intermolecular interactions betweenfunctionalities located on the host/guest complementary sur-faces, which severely limit their translational and (overall)rotational motion, and (ii) the rate acceleration because of partialdesolvation of the functionalities themselves in the host cavity(Hollfelder, Kirby, & Tawfik, 1996). Thus, solvation/desolvationphenomena may strongly affect chiral recognition and rateacceleration of enzymes and complicate the understanding of theunderlying principles.

As biological function and morphology are stronglycorrelated, knowledge of the structure of a molecule/receptorpair is expected to shed light on its biological activity. Structuredetermination is the primary and probably the most importantstep in the elucidation of molecular cluster properties. A clear

Mass Spectrometry Reviews, 2005, 24, 588– 610# 2004 by Wiley Periodicals, Inc.

————*Dedicated to the memory of Fulvio Cacace, a truly exceptional

scientist, mentor, and friend.

Contract grant sponsor: Ministero dell’Istruzione, della Universita, e

della Ricerca (MIU); Contract grant sponsor: Consiglio Nazionale

delle Ricerche (CNR): P.F. MSTAII.{Correspondence to: Maurizio Speranza, Dipartimento SCTSBA (No.

64), Universita di Roma ‘‘La Sapienza’’, P.le A. Moro 5, I-00185,

Roma, Italy. E-mail: [email protected]

cluster structure permits to face the study of its potential energysurface (PES) and energy distribution processes and finally toexplain chemical reactions occurring in clusters. Structuredetermination of noncovalent clusters is often a difficult task,because of both large amplitude motion and cluster flexibility.The latter effect occurs when an aggregate in a given electronicstate has more than one energetically accessible potentialminimum on its electronic PES. Large amplitude motion forintermolecular cluster modes can involve only one potentialminimum, nevertheless it plays a crucial role in the dynamicalprocesses connecting different potential minima on PES.Describing the relative orientation of the components of aflexible cluster is a daunting challenge from both the experi-mental and theoretical standpoints: a weakly bound clustershould be thought as a dynamic system, rather than one with awell-defined structure.

The state-of-art of the generation, spectroscopic and massspectrometric characterization, and dynamics of molecular chiralclusters in the gas phase are illustrated in the present review,together with a brief description of the experimental methodol-ogies employed.

II. CHIRAL CLUSTERS

The enantiomers of a chiral molecule can have a very differentbehavior when acting in a chiral environment. A chiralenvironment can be represented by the circularly polarized lightor an asymmetric medium. For instance, when a chiral druginteracts with its chiral receptor, one of the enantiomers of thedrug usually displays the desired biological activity, while theother is useless or even poisonous. For this reason, thediscrimination and separation of enantiomeric pairs are in thefocus of intense research and technological developments.Following the notion that the R and S enantiomers of a chiralmolecule have the same properties, except when interacting witha chiral environment, the most commonly employed method forseparating enantiomeric pairs is based on their association with aselector of precisely defined configuration, e.g., CR, to yield thecorresponding [R .CR] and [S .CR] diastereomeric adducts.These are not any longer mirror images, have different physicaland chemical properties and, therefore, can be identified andseparated. The most used separation techniques are liquidchromatography (HPLC), gas chromatography (GC), andcapillary electrophoresis (CE) using chiral stationary phasesacting as selector (Schreier, Bernreuther, & Huffler, 1995;Koppenhoefer, Epperlein, & Schwierskott, 1997; Desiderio &Fanali, 1998). Pirkle & House (1979) first gave to chiralchromatography a systematic basis with the introduction of the‘‘three-point rule,’’ i.e., there must be at least three simultaneousinteractions between a chiral selector and a chiral molecule andone of these must be stereochemically dependent, if chiralresolution is to be effected. After Pirkle and House, researchersdevoted a huge effort in defining the principles underlyingenantiomeric resolution in chiral chromatography, in particularthe specific interactions between the chiral sites of the stationaryphase and those of the analyte (Guebitz, 1990; Taylor & Maher,1992; Agnew-Heard et al., 1997; Dzygiel et al., 2000).

Other techniques such as UV-vis absorbance spectrometry(Kubo et al., 1996), circular dichroism (Fleischhauer et al.(1997), infrared transmission spectrometry (Chappell, 1997), andNMR spectroscopy (Kram & Lurie, 1992) have also been used forenantiomeric discrimination in liquid phase. In absorbancespectrometry, chiral discrimination of two enantiomers occursthrough the measurement of the absorption spectrum of theircomplexes with a chiral reference (host-guest systems). The twodiastereomers show different spectral shift with respect to theabsorption spectrum of the bare reference molecule. Also inNMR technique, a chiral reagent is added to obtain differentchemical shifts from the proton of two optical isomers. This ispossible because the presence of the reagent creates a chiralenvironment, which makes nonequivalent the protons in the twoenantiomers. Fluorescence techniques have often been used tostudy the interaction between enantiomers and receptors (James,Sandanayake, & Shinkai, 1995; Parker, Townshend, & Bale,1995, 1996; Takeuchi et al., 1997). Receptor systems (Schuetteet al., 1994; Grady et al., 1996; Yang & Bohne, 1995), such ascyclodextrins and calixarenes, have been observed to selectivelybind some enantiomers and produce a complex that can bedetected by fluorescence techniques. The centrality of chiralrecognition in biology has prompted these investigators toexplore the use of biomolecules as receptors for a fluorescence-based determination of enantiomeric purity.

The isolated environment of the supersonic beam and thehigh resolution of the laser spectroscopy have demonstrated to bevery useful tools for enantiodiscriminating van der Waalscomplexes in the gas phase, where the undesired leveling effectsof the solvent are excluded and the intrinsic factors governingenantioselective complexation determinable. The diastereoiso-meric [R .CR] and [S .CR] complexes can be distinguishedthrough fluorescence (Lahmani, Le Barbu, & Zehnacker-Rentien, 1999) and resonance enhanced multiphoton ionization(REMPI) spectra (Latini et al., 1999a), hole-burning (Al-Rabaaet al., 1997a) techniques, binding energy measurements,fragmentation thresholds, beam populations, and total dipolemoments measurements. The observed chiral discriminations arebecause of subtle balances between different effects such as stericfactors and additive dispersive interactions, which are moreimportant in the excited state.

III. EXPERIMENTAL METHODOLOGIES

A. Supersonic Expansion and Cluster Formation

The molecular clusters described in the present review have beengenerated by supersonic expansion of their components (Hagena& Obert, 1972; Ryali & Fenn, 1984). Supersonic expansioncluster sources normally operate either in the continuous or thepulsed mode, both providing very effective cooling of the formedclusters. Pre-existing beams of neutral clusters, well beyond theregion of the expansion jet, have been brought into contact withphotons or electrons to form the corresponding cluster ions,which were eventually mass discriminated.

The low temperatures in a supersonic beam provide a uniqueenvironment for the formation of weakly bound molecularcomplexes. In fact, at the translational temperatures available

CHIRAL RECOGNITION BY MASS-RESOLVED LASER SPECTROSCOPY &

589

with supersonic jets, van der Waals clusters are stable becausetheir binding energies can be larger than kT. Their formationproceeds through a sequence of association reaction steps withconcomitant collisional stabilization (Bauchert & Hagena, 1965;Hagena & Henkes, 1965; Milne & Greene, 1968; Anderson,1974; Hagena, 1974; Breen et al., 1984). This process can bedescribed by a kinetic model (Veenstra, Jonkman, & Komman-deur, 1994), which considers a diluted binary mixture of weaklyinteracting molecules expanding in a supersonic beam. The firststep, the formation of a dimer, is the critical step in the process. Infact, the formation of a trimer in a three-body collision is toounlikely for dilute systems. The van der Waals condensationenergy is dumped into the bond joining the two molecules and anactivated complex, containing excess energy, is formed. Ingeneral, the complex bond does not couple strongly to the internalvibrations of the molecule, so if the energy is not carried offduring the collision, the collision complex will dissociate again.Instead, the complex can be stabilized by a collision with an atomof the monatomic carrier gas or with an incoming third molecule.Further growth of the cluster, which mainly involves cluster-monomer collisions, is easier than dimer formation, since thereare more modes into which the ‘‘condensation’’ energy can betemporarily stored.

The rate constants involved in the formation of largerclusters are described in terms of the RRK theory (Robinson &Holbrook, 1972), which states that the substitution reaction ratefor the addition of the strongly bound component is much fasterthan for the addition of the more weakly bound component. Thisgives rise to the experimental observation that the composition ofthe clusters does not reflect the composition of the vapor phasefrom which they are formed. Instead, during the formation stageof the clusters, a nonstatistical enrichment toward the morestrongly bound species occurs (Jonkman, Even, & Kommandeur,1985).

B. Spectroscopic Characterization

Measuring physical-chemical properties of the clusters, such asionization energy (IE), binding energy (BE), electron (EA) andproton affinity (PA), fragmentation channels, electronic structureand so on, provide a basis for the comprehension of the intrinsicforces acting in the clusters and governing their dynamics.Theoretical computation of these quantities may provide afeedback to evaluate the quality of the calculations and theaccuracy of the experimental determinations. Recently, a numberof reviews on the experimental methods applied in the molecularcluster field have been published (Castleman & Bowen, 1996;Muller-Dethlefs & Hobza, 2000; Desfrancois, Carles, &Schermann, 2000). Among the many methodologies employed,those providing valuable information on the electronic groundand excited states of clusters are laser-induced fluorescence (LIF)and resonance enhanced multiphoton ionization (REMPI)spectroscopies (Smalley, Levy, & Wharton, 1976; Smalley,Wharton, & Levy, 1977; Kenny et al., 1980; Amirav, Even, &Jortner, 1981; Babbitt, Ho, & Topp, 1988; Dao et al., 1989;Wittmeyer & Topp, 1989; Leutwyler & Bosiger, 1990; Weberet al., 1991; Scherzer, Selzle, & Schlag, 1992). This lattertechnique is particularly promising since it enables massselection of cluster species and their spectral and thermochemical

characterization (Brutschy, 1990; Courty et al., 1997; Helm et al.,1998; Latini et al., 1999b; Mons et al., 2000). In resonant two-photon ionisation (R2PI), the complex is excited from itselectronic ground state by a photon and then ionized by a secondphoton of equal or different frequency. Very recently, R2PIspectroscopy has been applied to the study of biomolecules, suchas amino acids (Lubman & Li, 1990; Snoek et al., 2000; Snoek,Kroemer, & Simons, 2002), peptides (Cohen et al., 2000), DNAbases and derivatives (Nir et al., 2000, 2002), sugars (Talbot &Simons, 2002), and others (Weinkauf et al., 1995; Nir et al.,1999). LIF and R2PI spectroscopies can be coupled withultraviolet or infrared hole-burning (HB) techniques (Muller-Dethlefs & Hobza, 2000; Robertson & Simons, 2001) tocharacterize different isomers of clusters. An intense laser beamscans through the wavelength region of interest (the pump), whilea counter-propagating laser (the probe), delayed in time, is fixedon a selected resonance of the probed ground-state level. Whenboth lasers excite transitions, which arise from the same ground-state species, the pump beam induced depopulation manifestsitself by a decrease in the intensity of the signal arising from theprobe (spectral hole). If the lasers excite different ground-statelevels, no spectral hole is observed.

Rydberg electron transfer (RET) spectroscopy consists oftransferring electrons from highly excited atoms into diffuseorbitals of polar systems and provides a useful method for thediscrimination between different geometrical configurations(Desfrancois, 1995; Zheng & Ornstein, 1996; Carles et al.,2000b, 2001a). This methodology, together with LIF and R2PI,has been successfully applied to the study of the physical-chemical properties of chiral molecular clusters.

C. Resonant Two-Photon Ionization Spectroscopy

In a two-photon (1þ 1) process, namely resonant two-photonionization (R2PI), the species C is first promoted from itselectronic ground state S0 to the electronic excited state S1 via aresonant absorption step. Then, the absorption of a second photontakes the species into the ionization continuum. Since theionization cross-section is larger for resonant than for non-resonant processes, an increase in the ionization yield ofCwill beobtained each time the energy hn1 excites C from its ground stateto a precise vibronic level of S1. If the frequencies of theexcitation and ionization photons are equal, the process is namedone color R2PI (1cR2PI), otherwise two colors R2PI (2cR2PI)(Fig. 1).

A large number of R2PI experiments on clusters are reportedin literature: these studies provide detailed information on thenature of the intra- and intermolecular forces acting in molecularclusters. The vast majority of the studies involve simple modelsconsisting in organic species, particularly aromatics, because ofthe strong electronic spectral transitions (typically the S1 S0

transitions) available in regions accessible to tunable dye laserused in R2PI spectroscopy. For what concerns biologicalmolecules, many of them exhibit strong absorption bands in thenear-ultraviolet region of the spectrum (200–400 nm). These aregenerally associated with small conjugated ring systems, botharomatic and heterocyclic, whereas the large conjugated systemsabsorb in the visible region. Absorption of proteins in the 280 nmregion prevalently results from the presence of aromatic amino

& SPERANZA ET AL.

590

acidic residues (such as tyrosine, phenylalanine, and tryptophan),which mostly absorb in this range. Similarly, all of the purine andpyrimidines bases absorb strongly in the near ultraviolet.Generally, these absorption bands correspond to transitions ofp electrons in the ring to antibonding p* orbitals, the so-called p-p* transitions.

D. One-Color Resonant Two-Photon Ionization(1cR2PI) Spectroscopy

One color (R2PI) measurements consist in tuning the frequencyof the excitation and ionization photons n1. By monitoring theionic signal Cþ as a function of the laser frequency, thevibroelectronic spectrum of C is recorded.

The speciesC is either a simple molecule or its cluster with amolecule M (henceforth denoted as [C .M]). In this latter case,each component of the cluster electronically perturbs the other.This can be seen by comparing the 1cR2PI spectrum of the barechromophore C with that of the cluster. The spectral shift of theS1 S0 electronic transition band origin of [C .M], compared tothat of the isolated C, provides a measure of the difference in thestabilization energy of the excited and the ground electronic stateof the chromophore induced by the solvent. Two oppositesituations can be observed: (i) a spectral blue-shift of the S1 S0

electronic transition in [C .M] with respect to that in the isolatedC (‘‘ipsochromic shift’’). The complexation of the chromophoreresults in a larger stabilization of the ground state with respect tothe excited state and, thus, an increase in the energy differencebetween the S0 and S1 states; (ii) a spectral red-shift of theS1 S0 electronic transition in [C .M] with respect to thatin the isolated C (‘‘bathochromic shift’’). Following the cluster

formation, a lower stabilization of the ground state with respect tothe excited state takes place and, thus, a reduction of the energydifference between the ground and excited state.

The magnitude and direction of the spectral shift arise from acombination of effects that depends on the nature of theinteraction of the molecule with its solvating partners and theeffects of excitation on molecular properties including polariz-abilities and dipole moments (Amirav, Even, & Jortner, 1981;Amirav et al., 1983; Keese & Castleman, 1990; Muller-Dethlefs,1995; Lakin et al., 1998). A red-shift is most typically observedwhen dispersive forces give rise to long-range interactions. Polarsolvent molecules introduce complications in interpretingspectral shifts. Polar interactions produce red- or blue-shifts,depending on whether the excited state has a larger or smallerpermanent dipole moment than the ground state, and whetherpolar forces contributions are attractive or repulsive in the groundstate complex. If an interaction is largely polar and excitationincreases the polarity of the chromophore C along the interactionaxis, substantial red-shifts are observed (Bombach, Honnegger,& Leutwyler, 1985). If polar forces do not dominate the bondingof the complex in the ground state, then excitation to a more polarstate may lead to a blue-shift (Even & Jortner, 1983). Zero pointenergy (ZPE) differences must be considered as well; these canarise from unequal perturbations to ZPE of the two states of themoleculeC or unequal contributions to the ZPE by the respectiveintermolecular (van der Waals) modes in the ground and excitedcluster (Amirav, Even, & Jortner, 1981).

Analysis of 1cR2PI spectra has provided evidence for theexistence of various conformers (Amirav, Even, & Jortner, 1982;Dao, Morgan, & Castleman, 1984; Schauer & Bernstein, 1985;Leutwyler & Jortner, 1987; Bahatt et al., 1991) in complexmolecules, containing multi-ring systems or long side-chains.Biological compounds generally belong to this type of complexmolecules. The number of aromatic rings plays also an importantrole in the spectral shifts observed and in the frequencies of thevan der Waals modes. It is often possible to assign the lowerfrequency modes observed in the 1cR2PI spectra. Thermalbroadening of resonant line widths is often observed in largeclusters. This is because of the fact that large clusters typicallyform late in an expansion and retain a considerable amount oftheir heat of condensation. This leads to a higher internaltemperature with a concomitant broader rotational envelope.

The isolated enantiomers S (CS) and R (CR) of a chiralchromophore C exhibit the same spectral features since theirphysical properties are indistinguishable. However, their aggre-gation with a chiral solvent molecule of defined configuration(MR/S) leads to the formation of two diastereomeric com-plexes with different spectral properties, i.e., [CR .MR/S] and[CS .MR/S]. The 1cR2PI spectroscopy is able to discriminatebetween MR and MS by measuring the spectral shift of thediastereomeric [CR/S .MS] and [CR/S .MR] complexes with res-pect to that of the bare chromophore CR/S. It is convenient todefine the diastereomeric clusters as ‘‘homochiral’’ when thechromophore and the solvent have the same configuration, and‘‘heterochiral’’ in the opposite case.

The coupling of the 1cR2PI and supersonic beam techniquesallows to obtain good spectra, not congested by the presence ofhot bands, being negligible the population of the high vibrationallevels of the ground state. This comes from the cooling of the

FIGURE 1. Schematic representation of the 1cR2PI (hn1þ hn1) and

2cR2PI (hn1þ hn2) ionization processes for species C. [Color figure can

be viewed in the online issue, which is available at www.interscience.

wiley.com.]


591

translational, rotational, and vibrational degrees of freedom ofthe species produced in the supersonic expansion. Obviously, the1cR2PI methodology presents also some constraints, mostlybecause of the fact that the method does not permit to choose thefrequency of the second ionizing photon. For example, there isthe possibility that the ionization of the species comes out fromautoionization and, thus, the cross-section of this process shouldbe also considered. Autoionization occurs following a resonantabsorption of the ionization photon in a vibrational level of theRydberg states of the neutral (Hager, Smith, & Wallace, 1986a,b).

Furthermore, to ensure ionization of the species and avoid itsfragmentation, the excited states of the species should have highabsorption cross-sections, lifetimes comparable to the coherencetime of the laser and the excitation energy should be somewhatlarger than one-half of the ionization energy (IE) of the species. Infact, if the excited state is much higher in energy than one-half ofthe IE, the ion will be produced with a not negligible excessenergy (2hn1�IE) which is then transferred into vibrationalexcitation of the ionized species. Sometimes, this excitationcauses the formed ion to dissociate: if the species is a molecularion, fragmentation can occur; if the species is a cluster ion, it canalso dissociate into its original components or give rise to morecomplex fragmentation patterns. In general, the fragmentationpattern is that reported in Scheme 1.

If the overflow energy causes the fragmentation or thedissociation of the ionic species, fragments will appear in themass spectrum. These fragments keep memory of their parentions, it is as to say that they have the same vibroelectronic spectralpatterns. This is an important characteristic of the 1cR2PImethod, which in turn can allow the study of the fragmentationand dissociation processes occurring in molecules and clusters(Brutschy, 1990; Piccirillo et al., 1995).

A complication in the 1cR2PI spectrum can come fromthe fragmentation of large clusters, [CR/S.Mn] (Scheme 2). Themeasured spectral features of a cluster can arise also from thedissociation of larger cluster ions, which eventually could notbe detected in the spectrum because of high fragmentationefficiency. In this case, an unambiguous assignment of thevibronic pattern is no longer possible. As been yet mentioned,another complication in the analysis of the 1cR2PI spectrum cancome from the presence of various isomers of the studied species,which cannot be always distinguished in the absorption spectrum.

E. Two-Color Resonant Two-Photon Ionization(2cR2PI) Spectroscopy

All the problems associated with the 1cR2PI laser spectroscopycan be reduced by means of the 2cR2PI laser spectroscopy, a

powerful methodology allowing the measurement of physico-chemical properties of the species of interest, such as theionization and fragmentation potentials and binding energies. Aspreviously described, in the 2cR2PI, a fixed frequency laser (n1)pumps a vibroelectronic transition of a specific molecule orcluster, while a second tunable photon n2 promotes the speciesinto the ionization continuum.

The ionization and fragmentation thresholds can be obtainedby the photoionization efficiency curves through the followingsequence: (i) the first exciting laser (hn1) is fixed on the S1 S0

transition of the species of interest; (ii) a second laser is scannedthrough the cluster ionization (hn2) and dissociation thresholds(hn3) regions (Fig. 2).

The binding energy of the [CR/S .M] adduct in the ground,excited ([CR/S .M]*), and ionized state ([CR/S .M]þ), respec-tively, D0

00, D0*, and D0þ, are computed from the following

relations:

D000 ¼ h1 þ h3 � h01 þ h02� �

ð1Þ

D�0 ¼ D000 � h1 � h01� �

¼ D000 � D ð2Þ

Dþ0 ¼ h3 � h2 ð3Þ

SCHEME 1.

SCHEME 2.

FIGURE 2. Schematic representation of the energy levels of the bare

CR/S (denoted as CR/SþM) and of its complexes with M (denoted as

[CR/S .M]). D000, D0

0, and D0þ as the binding energies of the adducts in

the ground, excited, and ionized state, respectively. [Color figure can be

viewed in the online issue, which is available at www.interscience.

wiley.com.]

& SPERANZA ET AL.

592

The topic of ionization thresholds and cross-sections of clustershas been intensely explored (Jortner, 1984; Maerk & Castleman,1984). A general finding is that the ionization energy ofmolecular clusters (hn1þ hn2) tends to decrease with its size,the change being more pronounced for smaller clusters. In thecase of weakly bound systems, the decreases in ionizationenergies are found to be considerably larger than the spectralshifts of the electronic transitions. Ionization enhances theintermolecular forces by introducing the influence of a chargedcenter (Dao, Morgan, & Castleman, 1985). This is found, forexample, in the case of benzene and its derivatives where theirionization potential decreases upon association with argon byapproximately 100–200 cm�1 compared to spectral shifts for theS0 S1 transitions around 30 cm�1 (Fung et al., 1981; Gonoheet al., 1985). This large difference is an indication of themagnitude of the ion-induced dipole effect. Another conse-quence of the stronger interactions upon ionization is that theequilibrium geometry of the ionized complex may differsignificantly from that of the neutral states. Broadened ionizationonsets are frequently attributed to the spectral superposition ofionization into several vibrational levels for which Franck–Condon factors are more favourable (Fuke et al., 1984). As aresult, the adiabatic ionization potential may be considerablylower than the vertical potential, and the observed ionizationonsets may occur above the adiabatic potential. Another factor tobe considered is the conformation-dependent effect (Brutschy,1990; Mons et al., 2000), because of the different conformationsof the solvent molecules. The most populated form of a complexmay involve a less stable form of the solvent. After photoioniza-tion, the lowest-energy dissociation channel in the complex ionleads to the most stable form of isolated solvent, which has to betaken into account for the estimate of the binding energy.

The molecular or cluster radical cation, arising from R2PI,can give rise to different fragmentation patterns as shown inScheme 1 and Scheme 2. The excess energy in the vibrationalmodes of the cation formed may causes extensive fragmentationeven at threshold ionization (Yao et al., 2000). This happens whenthe vertical ionisation energy is much higher than the adiabaticionisation energy. Alternatively the fragmentation thresholds canbe measured by scanning the energy of the ionising/fragmenting(hn4) photon and looking at the appearance thresholds of theproduced ionic fragment. This allows the evaluation of theactivation barriersDE# towards different type of reactions (eq. 4).

DE# ¼ h4 � hv2 ð4ÞA laser technique, complementary to R2PI, is laser-induced

fluorescence (LIF) spectroscopy. Similarly to R2PI, LIF is basedon the electronic excitation of a supersonically expanded moleculeor cluster by absorption of light quanta from a tunable dye laser.The emitted fluorescence is collected at right angles to both theexcitation laser pulse and the beam axis and recorded. Thefluorescence decay times are measured with a photomultiplier.

In the LIF experiments (Smalley, Levy, & Wharton, 1976),the resulting fluorescence can be detected without wavelengthdispersion, and the spectrum is then a plot of the total detectedfluorescence versus exciting wavelength. Since the molecule canonly fluorescence if it is excited, and it can only be excited if thelaser is tuned to an absorption frequency, the fluorescenceexcitation spectrum is very similar to the absorption spectrum.

One important difference arises because some states may decayvia radiationless processes. In these cases, the molecule may havea well-developed high-resolution absorption spectrum and nofluorescence spectrum. Recently, additional structural informa-tion has come from the deconvolution of the rotational envelopesassociated with the electronic transitions (Muller-Dethlefs, 1995;Lakin et al., 1998).

When different isomers are present in the jet, the applicationof hole-burning (HB) spectroscopy is of particular advantage(Muller-Dethlefs & Hobza, 2000). The population of one isomercan be burned out by employing an additional tunable UV (or IR)laser. The LIF or R2PI spectrum then shows the depletion of thatisomer and the spectral transitions associated with the isomer canbe identified.

F. Rydberg Electron Transfer (RET) Spectroscopy

Rydberg electron transfer spectroscopy is based on the transfer ofelectrons from highly excited atoms into diffuse orbitals of polarsystems and provides a useful method for the discriminationbetween different geometrical configurations.

When an excess electron is attached to a neutral noncovalentcomplex of polar molecules with negative valence electronaffinities, no stable valence negative ions of this complex can beformed. However, the presence of a large enough total permanentelectric dipole moment (Desfrancois et al., 1994) and of aquadrupole moment and polarizability (Abdoul-Carime &Desfrancois, 1998), let the electrostatic field to accommodate avery weakly bound (in the meV range) excess electron in a verydiffuse orbital (in the nanometer range), mostly located outsidethe molecular frame. The formation of a stable anion, calleddipole-bound or multipole-bound, only releases very littleinternal energy into the complex (Desfrancois, Abdoul-Carime,& Schermann, 1996). As a consequence, fragmentation pro-cesses are very unlikely to occur, even for weakly boundnoncovalent complexes, and the structure of the dipole-boundanion is generally very similar to that of its neutral parent.Therefore, information about the original structure can bededuced. This is not always true, and, if a reorganization of theneutral geometry with a low energy expense take places, a higherdipole moment structure, more favourable to dipole-bound anionformation, can be formed (Dao, Morgan, & Castleman, 1985;Carles et al., 2000b).

The dipole-bound anions are formed by electron transfercollisions between cold molecular complexes and laser-excitedRydberg atoms, in a crossed beam experiment under singlecollision conditions. The use of Rydberg electrons allows for avery efficient electron transfer and for the stabilization of thecreated anions, leading to cold negatively charged species.Furthermore, RET spectroscopy is selective with respect to theexcess electron binding energy in the dipole-bound anion andelectron transfer is efficient only if there is an appropriatematching between the initial Rydberg orbital and the final dipole-bound anion orbital. This peculiarity allows the precisedetermination of the excess electron binding energy in thecreated dipole-bound anion, EADB. When rigid complex anionsare produced, the Rydberg n-dependencies are sharply peaked,while more peaks appear when a number of nearly isoenergeticisomers of the neutral complexes are simultaneously present.


593

Each peak corresponds to the production of different anionsarising from electron attachment to either large or small resultingdipole configurations. If a neutral complex configuration isfloppy, there is no well-defined resulting dipole but rather a broaddistribution and the Rydberg n-dependencies for anion produc-tion are then sums of different contributions. As an example, theexperimental data obtained by Schermann and co-workers for theformation, respectively, of monomer dipole-bound anions ofimidazole and adenine and for their complex are shown inFigure 3 (Carles et al., 2000a). Anion creation rates for the twomonomers are clearly peaked around the respective optimumRydberg quantum numbers nmax¼ 12 and nmax¼ 15, while theircomplex has mainly one peak centered around nmax¼ 8–9.

From these data, the excess electron binding energy valuescan be obtained from the EADB¼ 23/nmax

2.8 empirical relation-ship, which has been verified with a very good precision for manypreviously studied dipole-bound species (Desfrancois, 1995;Desfrancois, Abdoul-Carime, & Schermann, 1996). Since excesselectron binding energy is strongly related to the electrostaticcharacteristics (dipole, quadrupole, polarizability, etc.) and thusto the geometry of the complex, it is possible to determine whichneutral configuration is likely to have given birth to the observedanions. For a given cluster size, the resulting dipole moment ofthe different configurations is, in a first approximation, the vectorsum of the individual dipoles.

IV. CHIRAL RECOGNITIONIN MOLECULAR CLUSTERS

A. Structure of Molecular Complexes

The LIF method combines the formation of a molecular complexcontaining an apt fluorescent moiety (F) by supersonic expansionand the measurement of the laser-induced emitted fluorescence.The F molecule is laser excited to the fluorescent S1 state.Relative to the excitation spectrum of the bare F (hn),

complexation with a solvent molecule (M) usually results inthe appearance of new spectral features (hn0) falling at differentwavelengths (microscopic solvent shift: Dn¼ hn0-hn). The shiftmay be toward the red (negative Dn value) or toward the blue(positive Dn value). A red-shift indicates the decrease of theS1 S0 energy gap in the [F .M] complex relative to the bare F,while a blue-shift indicates the opposite. In general, diaster-eomeric pairs from complexation of a chiral M with chiral F arecharacterized by the nonequivalence of their interaction energy inboth the ground and the excited states and, therefore, are expectedto exhibit different microscopic solvent shifts Dn. In addition,complexation normally results in a drastic variation of thefluorescence lifetime t of the chromophore because of change ofthe radiative lifetime or to a decrease of the intersystem-crossingrate constant. This is the basis of the LIF application to chiralrecognition in the isolated state.

The LIF spectral shifts Dn and the fluorescence lifetimes tconcerning the [F .M] complexes between (R)- (1R) or (S)-2-naphthyl-1-ethanol (1S) and a variety of primary (2–7) andsecondary aliphatic alcohols (8–11) and terpenes (12–15) (Chart1) are listed in Table 1 (Al-Rabaa et al., 1995; Latini et al., 1999a;Lahmani, Le Barbu, & Zehnacker-Rentien, 1999).

The origin bands of the [F .M] complexes are red-shiftedwith respect to that of bare F (negative Dn in Table 1). Thissuggests that the major contribution to the intracomplex forces isbecause of hydrogen bonding between M and F, with the latteracting as the hydrogen donor. In fact, when the hydrogen donor isM, as in the case of the acidic trifluoroethanol 4, a blue-shiftis observed (Dn¼þ24 cm�1). Similar band shifts are observedfor the diastereomeric complexes with chiral M¼ 2-methyl-1-butanols (16R and 16S) or 2-chloro-1-propanols (7R and 7S).Here, the band origin of the heterochiral complex falls moretoward the blue (from 37 to 50 cm�1) relative to that of thehomochiral adduct. The trend is fully reversed for the complexeswith chiral secondary alcohols (8–11). In this case, are theheterochiral complexes, which exhibit larger red-shifts (from 11to 63 cm�1).

FIGURE 3. Relative dipole-bound anion formation rates in RET collisions between Rydberg Xe(nf) atoms

with (a) adenine (circles) or imidazole (squares) molecules and (b) adenine-imidazole complex produced in

a supersonic beam. Experimental data are fitted to curve-crossing model calculations, which lead to the

experimental determination of EADBexp values, equal to 11 meV for adenine, 23 meV for imidazole, and 54

meV for adenine-imidazole complex. Reproduced from Carles et al., 2000a with permission from American

Chemical Society, copyright 2000.

& SPERANZA ET AL.

594

In general, the excitation spectra of the [F .M] complexesshow intense features, accompanied by smaller signals placed atthe low energy side. These signals are mainly because of severalconformational isomers of the 1:1 complexes, as demonstratedby HB experiments (Le Barbu et al., 1998).

For instance, LIF experiments, carried out on the hetero-chiral [1R . 8S] complex, indicate the presence of three new bandsat �136, �114, and �73 cm�1 from the band origin of the bare1R. Similar experiments on the homochiral [1R . 8R] complex

exhibit two new bands at �125 and �69 cm�1 from the bandorigin of the bare 1R. HB experiments show that four differentisomers coexist in the jet, two of them corresponding to theheterochiral complex (denoted as Rs1 (�136 cm�1) and Rs2(�73 cm�1) in Fig. 4) and two others to the homochiral one(denoted as Ss1 (�125 cm�1) and Ss2 (�69 cm�1) in Fig. 4). Therelevant interaction energies have been partitioned into theelectrostatic, dispersion, polarization, and repulsion contribu-tions by using the method developed by Claverie (Claverie,


595

1978). Accordingly, it is assumed that the complex whichexhibits the most important dispersion term corresponds to themost red-shifted isomer.

Therefore, the most stable heterochiral anti-gauche struc-ture, observed in the HB experiments, is assigned to the more red-shifted Rs1 isomer (Sr(ag) in Fig. 5). In it, the ethyl chaininteracts strongly with the aromatic ring of 1S. The H atom of thechiral center of 8R points to the aromatic ring and interacts with itrepulsively. The more red-shifted Ss1 diastereomer can also beassigned to the most stable homochiral anti-gauche structureSs(ag) in Figure 5. In it, the increased repulsive interactionbetween the methyl group of 8S and the aromatic ring leads to aswitch of the ethyl chain on the edge of the ring and, therefore, to

a decrease in dispersive interactions (smaller red-shift). Theminor red-shift of the Rs2 and Ss2 diastereomers is insteadassigned to the corresponding gauche-anti structures. Thisexample clearly shows that, in this case, the chiral discriminationis achieved by means of the repulsion-dispersion part of theinteraction energy.

The most intense bands observed in the excitation spectra ofthe complexes often exhibit a decay time t longer than that of bareF (Table 1). However, some exceptions are observed and t doesnot decrease monotonically with the interaction energy as couldbe expected. The t variations can be rather associated with theeffect of complexation on the photophysics of the chromophoreF. Thus, no clear-cut relationship can be established between t

& SPERANZA ET AL.

596

and the structure and configuration ofM. Nonetheless, because ofthe peculiar sensitivity of the excited state dynamics to the subtleperturbation induced by complexation, the decay time t can beseen as a phenomenological tool for enantiodifferentiating chiralmolecules in the isolated state. For instance, the diastereomericcomplexes between 1R and 15 exhibit the same LIF spectra,characterized by an intense single band red-shifted by 89 cm�1

from the origin of bare 1R (Al-Rabaa et al., 1997b; Lahmani, LeBarbu, & Zehnacker-Rentien, 1999). Thus, no discrimination canbe achieved between the two diastereoisomeric complexes on the

exclusive basis of the excitation energy. However, in contrast tothe LIF spectra, the lifetimes of the diastereoisomeric complexesdiffer markedly (25� 3 ns (homochiral [1R . 15R] complex);42� 3 ns (heterochiral [1R . 15S] complex)). In addition, both areshorter than that of bare 1R (45� 3 ns). This observation isattributed to a strong chirality effect on the electronic overlapbetween the excited states of the components of the complexes,which reduces the fluorescence decay time.

The structure of weakly bound diastereomeric complexesbetween 1S and the enantiomers of 2-amino-1-propanol

TABLE 1. Main solvent shifts (Dn) and fluorescence lifetimes (t) of molecular complexes [F .M]


597

(alaninol; 17R and 17S) has been interrogated by means of LIFand IR fluorescence dip spectroscopy (Robertson & Simons,2001; Le Barbu, Lahmani, & Zehnacker-Rentien, 2002). Thehomochiral [1S . 17S] complex exhibits two new bands red-shifted by 94 and 28 cm�1 from the origin of bare 1S, while theheterochiral [1S . 17R] one only one new signal red-shifted by 59cm�1 from the origin of bare 1S. The structures associated withthese signals are determined by fluorescence dip spectroscopy,coupled with ab initio calculations of the most stable structures ofthe complexes and of their associated harmonic frequencies.Accordingly, the fluorescence dip spectrum recorded with theprobe set on the transition located at�59 cm�1 from the origin ofbare 1S is consistent with a chainlike structure of the heterochiral[1S . 17R] complex involving the oxygen center of the chromo-phore (henceforth denoted as O) and the two n-type centers of

17R, i.e. OH . . .OH . . .N. A similar chainlike structure isassociated with the homochiral [1S . 17S] complex with thetransition located at �28 cm�1 from the origin of bare 1S. Acompletely different IR spectrum is associated with thehomochiral [1S . 17S] complex with the transition located at�94 cm�1 from the origin of bare 1S, which is more consistentwith a bridged structure characterized by an intenseOH . . .NH2 intermolecular bond coupled with a weakerOH . . . p interaction between the OH group of 17S and thearomatic ring of 1S.

Further insights into the structural features of supersonicallyexpanded molecular complexes is obtained using the R2PIspectroscopy (Giardini Guidoni & Piccirillo, 1997; Piccirilloet al., 1997; Latini et al., 1999a, 2000). The supersonicallyexpanded species under investigation is ionized through absorp-tion of two laser photons of adequate energy and mass selected bya time-of-flight (TOF) mass spectrometer.

The mass resolved 1cR2PI spectrum of the bare chiralchromophore (R)-(þ)-1-phenyl-1-propanol (18R) shows threeintense signals at 37577 (A), 37618 (B), and 37624 cm�1 (C) inthe electronic S1 S0 band origin region. A similar triplet falls at38106, 38148, and 38155 cm�1. This pattern is common tosubstituted arenes and is interpreted as due to three stableconformers. Quantum chemical calculations at the RHF/3-21Gand B3LYP/6-31G levels of theory confirm this hypothesis.

The 1cR2PI absorption spectra of the complexes between18R and primary alcohols exhibit major characteristic peaks,which somewhat reproduce the pattern of the bare chromophore(Latini et al., 2000) but shifted toward the red by an extentdenoted as Dna and Dnb, respectively (Table 2).

The Dna and Dnb shifts from the complexes with primaryalcohols are linearly correlated to the proton affinity (PA) ofprimary M (PA(M) (kcal mol�1)¼ (180� 0.6)–(0.076�0.006)Dn(cm�1); r2¼ 0.951), which in turn determines thestrength of the OH . . .O hydrogen bond between the 18Rdonor and the M acceptor (Fig. 6).

These features allow to assign thea andb spectral signaturesof Figure 6, two different sets of [18R .M] complexes, where 18Ris in a given conformation and acts as the hydrogen-bond donor toM and where the alkyl group of M maintains a specific spatialorientation toward the aromatic ring of 18R. In this case, both theelectrostatic and the dispersive (polarization, charge exchange,etc.) interactions cooperate in stabilizing the adducts in theground and excited states.

The signals of the complexes of 18R with secondary alcoholsand amines display Dn values, which are less negative thanexpected on the grounds of the linear correlation obtained for theprimary alcohols. These deviations suggest that the relativecontributions of electrostatic and dispersive forces in thesesystems are substantially different from those operating in thecorresponding complexes with primary alcohols.

Comparison of the Dn values of diastereomeric complexeswith 8, 9, and 22 indicates that relative extent of electrostatic anddispersive forces depends upon the nature, the bulkiness, and theconfiguration of M. The D(Dn)¼ ((Dn)homo-(Dn)hetero)¼þ13cm�1 difference between the red-shifts of the homochiral[18R . 8R] and the heterochiral [18R . 8S] complexes finds aclose analogy with LIF red-shift difference of the correspond-ing [1R . 8R] and [1R . 8S] adducts (D(Dn)¼ ((Dn)homo�

FIGURE 5. Calculated structure of the 1:1 diastereomeric complexes of

(S)-2-naphthyl-1-ethanol with either (R)-2-butanol (Sr(ag)) and (S)-2-

butanol (Ss(ag)). [Color figure can be viewed in the online issue, which is

available at www.interscience.wiley.com.]

FIGURE 4. Bottom-up: (i) fluorescence excitation spectrum of the 1:1

diastereomeric complexes between (S)-2-naphthyl-1-ethanol (2-

NEtOH) and 2-butanol (MR/MS); (ii) hole-burning spectrum obtained

with the probe tuned on the transition located at �136 cm�1 ([2-

NEtOH .MR] complex); (iii) hole-burning spectrum obtained with the

probe tuned on the transition located at �69 cm�1 ([2-NEtOH .MS]

complex); (iv) hole-burning spectrum obtained with the probe tuned on

the transition located at �73 cm�1 ([2-NEtOH .MR] complex). The

probed band is denoted by D. The bands of the bare chromophore are

denoted by 2-NetOH. Reproduced from Le Barbu et al., 1998, with

permission from American Chemical Society, copyright 1998.

& SPERANZA ET AL.

598

TABLE 2. REMPI band shifts relative to jet-cooled 1:1 complexes between 18R and primary and secondary alcohols and

amines (M)

aBand shifts of thea peaks, relative to the band origin A of bare 18R; bBand shifts of theb peaks, relative to the band origin B of

bare 18R; cLias, S.G., Bartmess, J.E., Liebman, J.F., Holmes, J.L., Levin, R.D., and Mallard, W.G. (1988) J. Phys. Chem. Ref. Data 17,

Suppl. 1. Values in italic estimated from the PA limits of primary and secondary alcohols (Long, J. and Munson, B. (1977) J. Am.

Chem. Soc. 99, 6822), using the group additivity method (Benson. S.W. Thermochemical Kinetics, Wiley, New York, 1968).

FIGURE 6. Diagram of the band origin shiftsDn of the molecular complexes between (R)-(þ)-1-phenyl-1-

propanol (CR) and primary and secondary alcohols and amines. The circles refer to primary alcohols

(open circles¼Dna; full circles¼Dnb); the diamonds refer to secondary alcohols and amines (open

diamonds¼Dna; full diamonds¼Dnb).


599

(Dn)hetero)¼þ11 cm�1; Table 1). However, while LIF resultsfrom the diastereomeric [1R . 9R] and [1R . 9S] complexes arequalitatively similar to those with the other secondary alcohols,the homochiral [18R . 9R] and the heterochiral [18R . 9S]complexes display a negative spectral shift difference(D(Dn)¼ ca.�89 cm�1), which strikingly contrasts with thepositive one observed for the [18R . 8R] and [18R . 8S] pair. Thismarked spectral diversity can be attributed to the shorter sidechain and the wider ‘‘p-electron bed’’ of 1R, relative to 18R,which allows establishment of a synergy between attractiveelectrostatic and dispersive forces in [1R . 9R] and [1R . 9S] inboth the S0 and S1 states. In the [18R . 9R] and [18R . 9S]complexes, a similar cooperation is hindered by the greater stericcongestion because of the relatively bulky side chain andrelatively small p-system of the chromophore. This is particu-larly evident in the heterochiral [18R . 9S] complex, whose smallred-shift is rationalized in terms of a predominant OH . . . pelectrostatic interaction (or even a changeover in the hydrogen-bond donor/acceptor pair) determined by its high stericcongestion. Indeed, a similar blue-shift is observed in the 1cR2PIabsorption spectrum of the 1:1 cluster between 18R and water,where the solvent can establish with the aromatic ring of thechromophore only electrostatic OH . . . p interactions (Sattaet al., 2000).

Insight into the forces intervening in the supersonicallyexpanded isomeric complexes of (2R,3R)-(23RR), (2S,3S)-(23SS), and (2R,3S)-butanediols (23RS) with 18R is achieved by1cR2PI-TOF experiments (Giardini Guidoni et al., 2000).Figure 7 illustrates the corresponding excitation spectra togetherwith that of bare 18R.

Their spectral patterns are characterized by the presence ofseveral bands red-shifted relative to the 00

0 electronic S1 S0

origin of the bare chromophore 18R at 37618 cm�1 (peak B). Themost intense one is separated from the peak B of the chromophoreby the Dn values reported in the Figure. Since a red-shiftedS1 S0 origin denotes an increase of attractive interactions in thecomplex by the electronic transition, theDnvalues of Figures 7a–d, indicate that bond strengthening follows the order:[18R . 23RR]> [18R . 23RS]> [18R . 23SS].

These results point to a significant effect of the configurationof the diol moiety on the intracomplex forces involved in theisomeric [18R . 23RR], [18R . 23RS], and [18R . 23SS] adducts.The OH . . .O hydrogen bonding in these complexes isresponsible for the bathochromic shifts observed in thecorresponding spectra (Scuderi et al., 2002a). Different spectralshifts for diastereomeric complexes are often due to thesuperimposing effects of attractive dispersive (polarization)and repulsive (steric) interactions (Latini et al., 2000).

As pointed out before, the IR/UV double resonance spectraof the complexes between 1S and the alaninols 17R and 17Sexhibit spectral features due to structures involving not only theexpected intermolecular hydrogen bonding (either OH . . .O orOH . . .N), but also extensive intramolecular OH . . .N andOH . . . p hydrogen bonding (Le Barbu et al., 1998). Similarintramolecular interactions are present in the isomeric[18R . 23RR], [18R . 23RS], and [18R . 23SS] adducts as well.

Figure 8 reports the optimized [18R . 23RR], [18R . 23RS],and [18R . 23SS] structures, calculated at the MM3 force-fieldlevel of theory. Two different stable structures were found for

[18R . 23RR] and [18R . 23SS], characterized by intramolecularOH . . .O hydrogen bonding in the diol moiety in a gauche([Rrr]g and [Rss]g in Fig. 8) or the anti conformation ([Rrr]a and[Rss]a in Fig. 8). Two different structures have been identified forthe [18R . 23RS] isomer as well, one with the OH . . .O(H)-C(R)

intermolecular arrangement ([Rrs]R in Fig. 8) and the other withthe OH . . .O(H)-C(S) one ([Rsr]S in Fig. 8).

According to the OH . . .OH . . .OH interaction dis-tances of the structures of Figure 8, the a band of [18R . 23RR],which displays the largest red-shift, is assigned to the tightestgauche structure [Rrr]g. No significant structural and energeticdifferences can be appreciated in the gauche and anti conformersof [18R . 23SS] and [18R . 23RS]. In particular, irrespective of thespecific conformation and orientation of the diol moiety, theOH . . .O distance in these structures ranges around 2.01–2.03A and is much closer to that of [Rrr]a (2.04 A) than of [Rrr]g

(1.96 A). This analogy may explain the minor red-shifts observedin [18R . 23RS] and [18R . 23SS], relative to [18R . 23RR]. Similarconclusions are reached by investigating the 1cR2PI-TOFspectra of supersonically expanded isomeric complexes of(R)-(�)-(24R), (S)-(þ)-1,2-propanediols (24S), (2R,4R)-(�)-

FIGURE 7. Mass resolved 1cR2PI excitation spectrum of: (a) the bare

(R)-(þ)-1-phenyl-1-propanol (18R); (b) the [18R . 23SS] cluster; (c) the

[18R . 23RS] cluster; (d) the [18R . 23RR] cluster (total stagnation

pressure¼ 2� 105 Pa).

& SPERANZA ET AL.

600

(25RR), (2S,4S)-(þ)-2,4-pentanediols (25SS), with 18R (Scuderiet al., 2003). It is concluded that the spectral features of thesecomplexes depend not only on the basicity of the diol moiety, butalso on the specific configuration of their chiral carbons and therelative position of their hydroxyl groups.

The effect of changing the nature of the chromophore hasbeen investigated by comparing the R2PI spectra of diastero-meric [18R . 8] and [18R . 9] complexes with the correspondingspectra with (R)-(þ)-1-phenylethanol (26R) and (R)-(�)-indanol(27R), as chromophores. As for [18R . 8], the diastereomeric[26R . 8] and [27R . 10] complexes exhibit spectral signaturescharacterized by a significant red-shift Dn of the 00

0 S1 S0

transition relative to that of the bare chromophore. These spectralshifts reflect the combined effect of the electrostatic anddispersive interactions between the solvent molecule and theHOMO and LUMO of the chromophore. As expected, thehomochiral clusters exhibits a red-shift (Dnhomo) less negativethan that displayed by the heterochiral analog. Their differences,D(Dn)¼Dnhomo�Dnhetero¼þ12 cm�1 (for [26R . 8]) (Le Barbuet al., 2001) and þ30 cm�1 (for [27R . 10]) (Scuderi et al.,2002b), are consistent with a S1 S0 energy gap of theheterochiral complex, which is smaller than that of thecorresponding homochiral adduct.

A strict correspondence is also observed between the1cR2PI spectral shifts Dn of the selected diastereomeric clusters

and the relevant 1cR2PI mass fragmentation patterns. In the1cR2PI ionization experiments, some excess energy is impart-ed to the adduct to an extent which somewhat reflectsits HOMO-LUMO energy gap (and, therefore, Dn). SincejDnhomoj< jDnheteroj, one should expect that homochiral clustersdisplay 1cR2PI/TOF fragmentation patterns which are qualita-tively similar, though less abundant than those exhibited by thecorresponding heterochiral adducts. The experimental 1cR2PI/TOF mass spectra for 8 and 10 as solvent molecules are fullyconsistent with this expectation (Table 3) (Piccirillo et al., 1997).

B. Energetics of Molecular Complexes

Further insights into the forces operating in the molecularcomplexes of 18R and 26R with chiral secondary alcohols isobtained from the measurement of their binding energies(Brutschy, 1990; Latini et al., 1999b, 2000; Giardini Guidoniet al., 2000, 2001a; Mons et al., 2000). The used procedure issummarized in Figure 2. Accordingly, the binding energy D0

00 ofa molecular complex is derived from the difference between itsdissociative ionization threshold (hn1þ hn3) and the ionizationthreshold of the bare 18R or 26R chromophore (hn1

0 þ hn20)

(eq. 1). The binding energy D000 of the molecular complex in the

S1 excited state is taken as equal to D000�Dn, using the

FIGURE 8. H-bond distances (in A) in the isomeric [18R . 23SS], [18R . 23RS], and [18R . 23RR] structures.


601

appropriate Dn terms of Table 2 (eq. 2). The dissociation ionizedenergy D0

þ of cluster is estimated from the difference between itsdissociative ionization threshold (hn1þ hn3) and its ionizationthreshold (hn1þ hn2) (eq. 3). The relevant results are listed inTable 4.

The experimental D000 and D0

0 values are lower limits of theactual binding energies, because of some underestimate of theadiabatic ionization threshold of 18R or 26R. In particular,significant geometry change of the 26R ionic state with respect tothe excited state (unfavorable Frank Condon factors) leads tovery low undesirable values of D0

00 and D00 for [26R . 8R] and

[26R . 8S]. Nevertheless the DD0¼D0(hetero)�D0(homo) dif-ference between the hetero and homo complexes bindingenergies are correctly determined from the dissociation thresholddifference (Brutschy, 1990; Mons et al., 2000).

The D000 value of the complex between 18R and 6 (2.6�

0.2 kcal mol�1) well conforms to the approximate value of 2–3 kcal mol�1 indirectly estimated for the dissociation energy ofthe complex between 1R and 2 (Latini et al., 1999a). Concerningthe diastereomeric complexes, the homochiral adducts are invari-ably more stable than the heterochiral ones. This trend extends tothe corresponding S1 excited complexes as well. This observa-tion, coupled with the appreciable deviation from linearity of thecorresponding Dn values (Fig. 6), corroborates the view that theinteraction forces in these complexes are affected by stericcongestion to a different extent. Their sensitivity to steric factorsis demonstrated by the diverging observations that (Table 4): i) inthe diastereomeric pairs with 8R and 8S, the larger red-shift is

associated with the less stable heterochiral [18R . 8S] complex.This trend is confirmed by replacing 18R with 26R (Table 4)(Giardini Guidoni et al., 2001b) or with 1 (Dnhomo¼�125 cm�1;Dnhetero¼�136 cm�1) (Le Barbu et al., 2001); ii) in thediastereomeric pair with 9R and 9S, the less stable heterochiral[18R . 9S] complex is blue-shifted with respect to the homochiral[18R . 9R] complex. This trend is confirmed by replacing 18Rwith 26R (Dnhomo¼�101 cm�1; Dnhetero¼þ5 cm�1) (GiardiniGuidoni et al., 2001b). This implies that the relatively high stericcongestion, making the heterochiral complexes with 8S lessstable than the homochiral analogs, is insufficient to altersignificantly the nature of the attractive intracomplex forces. Thisis no longer true in going from the homochiral complexes with 8Rto the homochiral complexes with 9R. In fact, steric congestion inthe latter complexes is so high to modify to some extent the natureof the attractive intracomplex forces and, therefore, themagnitude of its binding energy in both the ground and theexcited states, relative to those operating in the homologouscomplexes with 8R. This trend goes to extremes in the least stableheterochiral adducts with 9S. The low binding energy of thesecomplexes and the considerably less red-shift accompanyingtheir S1 S0 transition point to an exceedingly high stericcongestion, which dramatically perturbs the spatial arrangementof the two moieties so as to favor attractive O-H . . . p-ringelectrostatic forces.

Inspection of the data in Table 4 reveals that the D0þ values

always exceed the corresponding D000 interaction energies. The

slow rise of the ion current observed in the 2cR2PI spectra

TABLE 4. Spectral shifts (Dn) and binding energies (D0

0 0, D0

0, and D0

þ (see text)) of the molecular complexes between

chromophores C and solvent molecules M (1 kcal mol�1 ¼ 349.77 cm�1)

TABLE 3. Relative ion abundance from the R2PI-TOF mass spectra of supersonically expanded 1:1 [C .M] complexes

& SPERANZA ET AL.

602

suggests significant geometry change of the complex in itsexcited and the ionic state and, therefore, the phenomenologicalD0þ terms of Table 4 are probably not representative of the actual

binding energies of the ionized molecular complexes, but rathermust be considered as lower limits (Le Barbu et al., 2001).Nevertheless, they provide an additional phenomenological toolfor chiral recognition in the isolated state.

C. Reaction Thresholds

R2PI studies have dealt with side chain fragmentation of alkylaromatic radical cations and their cluster. Spectral studies on thesolvation and related chemistry of 18R (Piccirillo et al., 1997;Satta et al., 2000; Scuderi et al., 2002a, 2002b, 2003) and 26R(Filippi et al., 2001) showed that after photoionisation theclusters fragment by homolytic Ca-Cb bond cleavage. Theefficiency of the process is enhanced when a hydrogen-bondinteraction is present between the chromophore and the solventmolecule, whose intensity is related to the proton affinity of thesolvent. A similar behavior is found in the heterolytic Ca-Cb

fragmentation occurring in phenylethylammine (PEA) (Yaoet al., 2000) clustered with polar solvents, where extensivefragmentation was found even for threshold ionisation. This lowenergy Ca-Cbs-bond dissociation was ascribed to an intersectionbetween the potential energy surfaces of the lowest energyelectronic states of the radical cations, which are perturbed by thepresence of a solvent molecule.

As shown in Figure 9, the intersection between the PotentialEnergy Surface (PES), adiabatically correlated with the ionicbiradical . [C6H5CHOH]þ . and the PES adiabatically correlatedwith the closed shell ion [C6H5CHOH]þ generates a small barrierto the ethyl radical loss. The effect of monosolvation on theenergetics and dynamics of photo-dissociation of Ca-Cb bond inthe (R)-(þ)-1-phenyl-1-propanol ion [18R]þ are thoroughlyexamined by measuring their fragmentation thresholds (Catoneet al., 2004; Piccirillo et al., 2004).

The ionization and fragmentation thresholds of 18R and itsclusters with water [18R . solv] (solv¼H2O) or chiral diols

(solv¼ 23RR, 23SS) have been inferred from the onsets of therelevant photoionization and photofragmentation efficiencycurves, respectively, using a two-color R2PI (2cR2PI) sequence.Equations 5–8 resume the low energy dissociation pathsinvolved in the ionization of the S1 excited [18R]* and[18R . solv]*.

Ionization: ½18R��þhn2 ! ½18R�þþe ð5Þ

½18R � solv��þhn2 ! ½18R � solv�þþe ð6Þ

Dissociation: ½18R��þhn4 ! ½18R�C2H5�þ

þ � C2H5þeð7Þ

½18R � solv��þhn4 ! ½18R�C2H5 � solv�þ

þ � C2H5þeð8Þ

The photoionization and photodissociation thresholds wererecorded fixing the energy of the excitation laser (hn1) at theelectronic origin of the most stable isomers, at 37618 cm�1 for18R, 37516 cm�1 for [18R . 23RR], 37581 cm�1 for [18R . 23SS](as indicated in Fig. 7) and 37665 cm�1 for [18R .H2O].

The relevant results are reported in Table 5. Accordingly,microsolvation appears to have a crucial role in controlling theethyl-loss barrier involved in these photochemical reactions. Themarked difference between the ethyl loss barrier in the bare andsolvated ions is because of the ability of solv to stabilize thepositive charge on the Ca atom by interacting with the OH groupof the ionized chromophore. In the clusters, the hydrogen bondbetween the aromatic alcohol and solv seems to activate theinductive and electrostatic forces able to strongly modify theoverall shape of the potential energy surfaces in their intersectionregions. The effect of hydration on the ethyl-loss barrier can beestimated by considering the difference of 4080 cm�1 betweenthe fragmentation onsets of [18R]þ, and of [18R .H2O]þ ions(Table 5).

We report in Figure 10, a section of semiempirical PM3 PESof [18R .H2O]þ to show an overview of the main features, whichcontrol the fragmentation pattern of the mono-hydrated cluster(Piccirillo et al., 2004). It appears that the intermolecularhydrogen bond length is reduced during the fragmentationprocess (see the minimum energy path shown). A distinct issue ofFigure 10 is the appearance of a saddle point connecting the

FIGURE 9. Pictorial crossing between the two lowest-energy electronic

states of [18R]þ (full lines) and [18R . solv]þ (broken lines).

TABLE 5. Ethyl-loss fragmentation energies DE# for the unsol-

vated [18R]þ and monosolvated [18R . solv]þ ions, and the IPs of

the corresponding neutral systems


603

regions of bound [18R .H2O]þ cluster with that of the Ca-Cb

fragmentation products. The PM3 barrier height of the minimumenergy path to cleave the Ca-Cb bond in the isolated [18R]þ islocated at approximately 5606 cm�1 above equilibrium mole-cular energy. A marked decrease of the ethyl-loss barrier isobserved for the cluster: in this case, the minimum pathwayconnecting [18R .H2O]þ to [C6H5CHOH .H2O]þ is located atapproximately 3215 cm�1 above the equilibrium cluster energy.

The differences in the experimental thresholds give a lowerlimit of the ethyl-loss barriers. This is because of poor Frank–Condon factors between neutral and ionic electronic states.Nevertheless, the experimental observed decrease of the Ca-Cb

fragmentation energy in [18R .H2O]þ with respect to theunsolvated chromophore is qualitatively reproduced by semi-empirical calculations. This fact supports the hypothesis that onewater molecule is able to alter substantially the topology of theintersection region of the two ionic potential energy surfacesinvolved in the dissociation.

The role of multiple hydrogen bonds on the fragmentationcan be inspected by looking at the difference between thedissociation thresholds of [18R]þ and of those of the diol-clusters,which are measured to be 6330 cm�1 for the homochiral clusterand of 6730 cm�1 for the heterochiral complex (Table 5). Hencethe two hydroxyl groups of the diols have a much stronger effectin lowering the fragmentation threshold with respect to the watersolvent. This can be rationalized assuming that the stabilizationof the positive charge on the Ca is due to cooperative interactions,as already discussed for the neutral state systems. These

interactions cause both an increase of the intermolecular bindingenergies and a lowering of the ethyl loss barriers. At variance with[18R .H2O]þ, (Do

þ¼ 5.4� 0.2 kcal mol�1, Table 4, DE#¼9.7� 0.2 kcal mol�1, Table 5), the fragmentation of the covalentCa-Cb bond in [18R . 23RR]þ or [18R . 23SS]þ is energeti-cally preferred relative to cluster dissociation (loss of solvmolecule).

A closer inspection of Table 5 reveals that the activationbarrier of the homolytic Ca-Cb bond cleavage in the heterochiral[18R . 23SS]þ complex (DE#¼ 740� 40 cm�1) is significantlylower than that in the homochiral [18R . 23RR]þ one(DE#¼ 1140� 40 cm�1) in spite of the identical basicity of theisolated diol enantiomers. This difference may be attributed tostructural factors making solv in the latter adduct a better H-bondacceptor than in the former adduct.

D. Chiral Effects in RydbergElectron-Bound Complexes

A large number of molecules of biological interest are stronglypolar, and thus they can efficiently capture low energy electrons,giving rise to Rydberg electron-bound complexes. As seenbefore, a Rydberg electron is localized in a very diffuse orbitalmainly outside the neutral molecular frame and its bindingenergy is generally very low (0.5–100 meV). For this reason, theformation of the Rydberg electron-bound anion does not perturbthe neutral parent structure. On the contrary, if valence anions are

FIGURE 10. Projection of the multidimensional PM3 potential energy surface of the [18R .H2O]þ, in the

Ca-Cb and O ...H hydrogen bond coordinates. [Color figure can be viewed in the online issue, which is

available at www.interscience.wiley.com.]

& SPERANZA ET AL.

604

formed, the excess electron enters a valence orbital, whichparticipates into the chemical bonding and the geometricalstructure of the neutral parent system is modified. Analogousconsiderations can be made if we consider the electron transferprocess on a molecular complex.

For each molecule and molecular complex, it is possible tomeasure the dependence of the anion creation rate as a function ofthe Rydberg atom principal quantum number n. This n-dependence allows the experimental measure of the excesselectron binding energy EADB

exp, with the help of a curve-crossing model, according to the equations reported above. Fromthe computed electrostatic parameters (m, Q, a), a semi-empiricalelectrostatic model provides the corresponding dipole-boundelectron affinity EADB

calc. RET spectroscopy has proven to be avery useful tool to investigate the weak noncovalent interactionsbetween polar biological molecules (Van Berkel, 1997; Abdoul-Carime & Desfrancois, 1998; Carles et al., 2001b; Desfrancois,Carles, & Schermann, 2000), but only recently it has been appliedto investigate chiral clusters (Lecomte et al., private commu-nication). In particular, the diastereomeric complexes of 26R with(R)-(�)-(28R) and (S)-(þ)-2-pyrrolidinmethanol (28S) (prolinol)have been investigated.

Figure 11 shows the B3LYP/6-31þþG**-calculated opti-mized structures of neutral diastereomeric [26R . 28R] and

[26R . 28S] comlex. Two structures have been predicted for boththe homochiral (IHOMO and IIHOMO) and heterochiral (IHETERO

and IIHETERO) complexes. Their electrostatic parameters, bind-ing energies and computed dipole-bound electron affinitiesEADB

calc are reported in Table 6.The two more stable structures IHOMO and IHETERO are

characterized by a double hydrogen bond between 26R and 28R or28S. The 26R molecule acts as proton donor towards the nitrogenof prolinol, and as acceptor towards the alcoholic proton of 28R/S.In the two less stable structure IIHOMO and IIHETERO, the prolinolmaintains an intramolecular H-bond between the alcoholicoxygen and nitrogen and, thus, only one hydrogen bond withthe 26R molecule is possible, in which the oxygen of 28R/Saccepts a proton.

These calculations can be compared with the experimentalRET spectroscopy data obtained on separate mixtures of 26Rwith 28R, and 26R with 28S. Figure 12 reports the n-dependencyof the relative formation rate for the two diastereomericcomplexes. As can be seen, a small discrimination between thetwo sets of data is possible: the plot of [26R . 28R] data is peakedat Rydberg quantum number n¼ 12, while the one of [26R . 28S]at n¼ 13. From these values, the experimental dipole-boundelectron affinities result to be: EADB

exp ([26R . 28R]¼ 19 meVand EADB

exp ([26R . 28S]¼ 16 meV. The EADBexp ([26R . 28R])

FIGURE 11. Calculated structures of the neutral homochiral (a) and heterochiral (b) complexes between

(R)-1-phenylethanol (26R) and 2-pyrrolidinmethanols (28R and 28S). Red, oxygen; black, carbon; blue,

nitrogen; light-blue, hydrogen. [Color figure can be viewed in the online issue, which is available at

www.interscience.wiley.com.]


605

value appears in better agreement with the calculated EADBcalc

one corresponding to the predicted less stable structure IIHOMO

(Table 6). This may indicate that conformational effectsfavors formation of structures II in the supersonic beamexpansion.

V. CONCLUDING REMARKS

The comprehension of enantioselective phenomena in chemicaland biochemical processes requires the knowledge of theintrinsic interactions operating in related simplified models inthe gas phase, in the absence of any solvation and ion pairingeffects. The energy of noncovalent interactions under ambientconditions is comparable to the average thermal energy of thekinetic motion of molecules. Thus, noncovalent aggregates areoften dynamic systems of limited stability and high structuralflexibility. Sometimes, however, the combination of severalnoncovalent bonds confers a pronounced stability even to veryflexible aggregates. This combination represents the keystone ofliving matter: bio-macromolecules ‘‘communicate’’ throughmultiple noncovalent interactions, which must be sufficientlystrong to ensure preferential bonding among them, but not sostrong to allow thermodynamic sinks which would be incompa-

tible with the existence of living matter itself. Thus, under-standing the way bio-macromolecules ‘‘communicate’’ in livingsystems or the forces holding molecules together in supramole-cular assemblies require a deep knowledge of noncovalentinteractions in simplified models. Of particular interest forunderstanding chiral recognition and rate acceleration of bio-macromolecules are those involving chiral species. A key stepalong this direction requires the study the single interactionsacting between the individual chiral components of tailor-mademolecular clusters in the gas phase, without any interferencefrom the environment.

In the present excursion through the world of noncovalentclusters in the gas phase, the reader can readily realize theexceptional experimental and conceptual know-how needed forgenerating weakly-bonded aggregates in the gas phase and forexamining their structural and stability features at the micro-scopic level. The huge effort unfolded along this direction bymost advanced spectroscopic and mass spectrometric methodol-ogies is fully rewarded by precious information about thestructure, stability, reactivity, and dynamics of chiral molecularclusters in the gas phase. Mass-resolved laser spectroscopy,coupled with supersonic beam expansion, has allowed the precisedetermination of the structure and the energetics of diaster-eomeric molecular clusters at very low temperatures. These gas-phase studies were driven by the continuing interest in chiralspecies, whether involved in enantioselective reactions, or in theorigin of life, or in biochemical cycles.

REFERENCES

Abdoul-Carime H, Desfrancois C. 1998. Electrons weakly bound to

molecules by dipolar, quadrupolar, or polarization forces. Eur Phys J

D 2:149.

Agnew-Heard KA, Pena MS, Shamsi SA, Warner IM. 1997. Studies of

polymerized sodium N-undecylenyl-L-valinate in chiral micellar

electrokinetic capillary chromatography of neutral, acidic, and basic

compounds. Anal Chem 69:958.

Al-Rabaa AR, Breheret E, Lahmani F, Zehnacker A. 1995. Enantiodiffer-

entiation in jet-cooled van der Waals complexes of chiral molecules.

Chem Phys Lett 237:480.

Al-Rabaa AA, Barbu KL, Lahmani F, Zehnacker-Rentien A. 1997a. van der

Waals complexes between chiral molecules in a supersonic jet: A new

spectroscopic method for enantiomeric discrimination. J Phys Chem A

101:3273.

TABLE 6. Calculated energy value and electrostatic parameters for the optimized [26R . 28R]

isomers, IHOMO and IIHOMO, and [26R . 28S] isomers, IHETERO and IIHETERO

FIGURE 12. Relative dipole bound anion formation rates in RET

collisions between Rydberg Xe(nf) atoms and a supersonic beam of (R)-

1-phenylethanol (26R) with 2-pyrrolidinmethanols (28R and 28S).

& SPERANZA ET AL.

606

Al-Rabaa AR, Le Barbu K, Lahmani F, Zehnacker A. 1997b. Laser induced

fluorescence of jet-cooled complexes between chiral molecules: A

photophysical method for chiral discrimination. J PhotoChem Photo-

biol A: Chemistry 105:277.

Amirav A, Even U, Jortner J. 1981. Microscopic solvation effects on excited-

state energetics and dynamics of aromatic molecules in large van der

Waals. J Chem Phys 75:2489.

Amirav A, Even U, Jortner J. 1982. Electronic-vibrational excitations of

aromatic molecules in large argon clusters. J Phys Chem 86:3345.

Amirav A, Even U, Leutwyler S, Ondrechen MJ, Berkovitch-Yellon Z,

Jortner J. 1983. Energetics and dynamics of large van der Waals

molecules. Faraday Discuss Chem Soc 73:153.

Anderson JB. 1974. Molecular beams and low density gas dynamics.

Wegener PP, editor. New York: Dekker.

Babbitt RJ, Ho CJ, Topp MR. 1988. Fluorescence excitation spectroscopy of

the S2 state of coronene: Solid state, isolated molecule and van der

Waals clusters. J Phys Chem 92:2422.

Bahatt D, Even U, Shalev E, Ben-Horin N, Jortner J. 1991. Isomer-specific

radiative lifetimes of molecular heteroclusters. Chem Phys 156:223.

Bauchert J, Hagena OF. 1965. Mass determination of ionized agglomerates in

condensed molecular beams by the electric counterfield method. Z

Naturforsch 20:1135.

Bombach R, Honnegger E, Leutwyler S. 1985. Solute-solvent interactions in

microhydrate clusters: Carbazole. (H2O)n. Chem Phys Lett 118:449.

Breen JJ, Kilgore K, Stephan K, Hofmann-Sievert R, Kay BD, Keesee R,

Mark TD, Van Doren JM, Castleman AW, Jr. 1984. The use of similarity

profiles in studying cluster formation in molecular beams: Evidence for

the role of pre-existing dimers. Chem Phys 91:305.

Brutschy B. 1990. Reactions in molecular clusters following photoionization.

J Phys Chem 94:8637.

Carles S, Lecomte F, Schermann JP, Desfrancois C. 2000a. Gas-phase

experimental and theoretical studies of adenine, imidazole, pyrrole, and

water non-covalent complexes. J Phys Chem A 104:10662.

Carles S, Desfrancois C, Schermann JP, Smith DMA, Adamowicz L. 2000b.

Structures and electron affinities of indole-(water)N clusters. J Chem

Phys 112:3726.

Carles S, Desfrancois C, Schermann JP, Berges J, Houee-Levin C. 2001a.

Rydberg electron-transfer spectroscopy and ab initio studies of

dimethylsulfoxide-water neutral and anion dimers. Int 205:227.

Carles S, Lecomte F, Schermann JP, Desfrancois C, Xu S, Nilles JM, Bowen

KH, Berges J, Houee-Levin C. 2001b. Nondissociative electron capture

by disulfide bonds. J Phys Chem A 105:5622.

Castleman AW, Jr, Bowen KH. 1996. Clusters: Structure, energetics, and

dynamics of intermediate states of matter. J Phys Chem 100:12911.

Catone D, Giardini Guidoni A, Paladini A, Piccirillo S, Rondino F, Satta M,

Scuderi D, Speranza M. 2004. Homolytic Ca-Cb bond cleavage in a

chiral alkylarene radical cation: Effects of asymmetric microsolvation.

Angew 43:1868.

Chappell JS. 1997. Infrared discrimination of enantiomerically enriched and

racemic samples of methamphetamine salts. Analyst (Cambridge,

U.K.) 122:755.

Claverie O. 1978. Intermolecular interactions from diatomics to biopolymers.

Pullman B, editor. New York: Wiley.

Cohen R, Brauer B, Nir E, Grace L, de Vries MS. 2000. Resonance-enhanced

multiphoton ionization spectroscopy of dipeptides. J Phys Chem A

104:6351.

Courty A, Mons M, Calve JL, Piuzzi F, Dimicoli I. 1997. Chemistry of

aromatic cations on water clusters: Magic numbers as a mass

spectroscopic diagnosis of reactivity. J Phys Chem 101:1445.

Dao PD, Morgan S, Castleman AW, Jr. 1984. Resonance-enhanced

multiphoton ionization of van der Waals molecules: Studies of

spectroscopic shifts of phenylacetylene clustered with molecules and

atoms. Chem Phys Lett 111:38.

Dao PD, Morgan S, Castleman AW, Jr. 1985. Two-color resonance enhanced

multiphoton ionization of van der Waals molecules: Studies of

spectroscopy shifts and ionization thresholds of para-xylene clustered

with argon. Chem Phys Lett 113:219.

Dao PD, Morgan S, Breen JJ, Kilgore K, Wei S, Tzeng WB, Keesee RG,

Castleman AW, Jr. 1989. Spectroscopy of phenylacetylene bound to

clusters of ammonia and the surface cluster analogy. J Chem Phys

90:11.

Desfrancois C. 1995. Determination of electron binding energies of ground-

state dipole-bound molecular anions. Phys Rev A 51:3667.

Desfrancois C, Abdoul-Carime H, Schermann JP. 1996. Ground-state dipole-

bound anions. Int J Modern Phys 10:1339.

Desfrancois C, Carles S, Schermann JP. 2000. Weakly bound clusters of

biological interest. Chem Rev 100:3943.

Desfrancois C, Abdoul-Carime H, Adjoure C, Khelifa N, Schermann JP.

1994. From 1/r to 1/r2 potentials: Electron exchange between Rydberg

atoms and polar molecules. Phys Rev Lett 73:2436.

Desiderio C, Fanali S. 1998. Chiral analysis by capillary electrophoresis using

antibiotics as chiral selector. J Chromatogr A 807:37.

Dzygiel P, Rudzinska E, Wieczorek P, Kafarski P. 2000. Determination of

optical purity of phosphonic acid analogues of aromatic amino acids by

capillary electrophoresis witha-cyclodextrin. J Chromatogr A 895:301.

Even U, Jortner J. 1983. Spectroscopy of large molecular complexes in

supersonic jets. J Chem Phys 78:3445.

Filippi A, Giardini A, Latini A, Piccirillo S, Scuderi D, Speranza M. 2001.

Chiral discrimination of monofunctional alcohols and amines in the gas

phase. Int J Mass Spectrom 210:483.

Fleischhauer J, Harmata M, Kahraman M, Koslowski A, Welch CJ. 1997. The

determination of the absolute configuration of a chiral molecular

tweezer using CD spectroscopy. Tetrahedron Lett 38:8655.

Fuke K, Yoshiuchi H, Kaya K, Achiba Y, Sato K, Kimura K. 1984.

Multiphoton ionization photoelectron spectroscopy and two-color

multiphoton ionization threshold spectroscopy on the hydrogen bonded

phenol and 7-azaindole in a supersonic jet. Chem Phys Lett 108:179.

Fung KH, Henke WE, Hays TR, Selzle HL, Schlag EW. 1981. Ionization

potential of the benzene-argon complex in a jet. J Phys Chem 85:3560.

Giardini Guidoni A, Piccirillo S. 1997. Lasers and molecular beams applied to

the study of clusters of chiral molecules. Isr J Chem 37:439.

Giardini Guidoni A, Piccirillo S, Scuderi D, Satta M, Di Palma TM, Speranza

M. 2000. Chirality and intermolecular forces: Studies using R2PI

experiments in supersonic beams. Phys Chem Chem Phys 2:4139.


M, Filippi A, Paladini A. 2001a. R2PI study of intermolecular hydrogen

bond in solvent-free chiral complexes. Chirality 13:727.


M, Filippi A, Paladini A. 2001b. Laser production of gas phase

complexes of metal a-aminophosphonic acid mixtures and their role in

chiral recognition. Int J Photoenergy 3:217.

Gonohe N, Abe H, Mikami N, Ito M. 1985. Two-color photoionization of van

der Waals complexes of fluorobenzene and hydrogen-bonded com-

plexes of phenol in supersonic jets. J Phys Chem 89:3642.

Grady T, Harris SJ, Smyth MR, Diamond D. 1996. Determination of the

enantiomeric composition of chiral amines based on the quenching of

the fluorescence of a chiral calixarene. Anal Chem 68:3775.

Guebitz G. 1990. Separation of drug enantiomers by HPLC using chiral

stationary phases—A selective review. Chromatographia 30:555.

Hagena OF. 1974. Molecular beams and low density gas dynamics. Wegener

PP, editor. New York: Dekker.

Hagena OF, Henkes W. 1965. Determination of the effective ionization cross

section in condensed molecular beams. Z Naturforsch 20:1344.

Hagena OF, Obert W. 1972. Cluster formation in expanding supersonic jets.

Effect of pressure, temperature, nozzle size, and test gas. J Chem Phys

56:1793.


607

Hager JW, Smith MA, Wallace SC. 1986a. Autoionizing Rydberg structure

observed in the vibrationally selective, two-color, threshold photo-

ionization spectrum of jet-cooled aniline. J Chem Phys 83:4820.

Hager JW, Smith MA, Wallace SC. 1986b. Two-color threshold photoioniza-

tion of jet-cooled aniline: Vibrationally selective autoionization. J

Chem Phys 84:6771.

Helm RM, Clara M, Grebner TL, Neusser HJ. 1998. Hydrogen bonding in the

indole-water complex: A high resolution UV study of the hydrogen

donor conformer. J Phys Chem 102:3268.

Hollfelder F, Kirby AJ, Tawfik DS. 1996. Off-the-shelf proteins that rival

tailor-made antibodies as catalysts. Nature 383:60.

James TD, Sandanayake K, Shinkai S. 1995. Chiral discrimination of

monosaccharides using a fluorescent molecular sensor. Nature

(London) 374:345.

Jonkman HT, Even U, Kommandeur J. 1985. Clusters of organic molecules in

a supersonic jet expansion. J Phys Chem 89:4240.

Jortner J. 1984. Level structure and dynamics of clusters. Ber Bunsenges Phys

Chem 88:188.

Keese RG, Castleman AW, Jr. 1990. Atomic and molecular clusters. Bernstein

ER, editor. Amsterdam: Elsevier.

Kenny JE, Johnson KE, Sharfin W, Levy DH. 1980. The photodissociation of

van der Waals molecules: Complexes of iodine, neon, and helium. J

Chem Phys 72:1109.

Koppenhoefer B, Epperlein U, Schwierskott M. 1997. Enantioseparation of

four-membered rings by gas chromatography. An overview.’’ Frese-

nius’ J Anal Chem 359:107.

Kram TC, Lurie IS. 1992. The determination of enantiomeric composition of

methamphetamine by proton-NMR spectroscopy. Forensic Sci Int

55:131.

Kubo Y, Maeda S, Tokita S, Kubo M. 1996. Colorimetric chiral recognition by

a molecular sensor. Nature (London) 382:522.

Lahmani F, Le Barbu K, Zehnacker-Rentien A. 1999. Photophysical studies

on molecular chirality: Ground and excited state enantioselective

interactions between 2-naphthyl-1-ethanol and natural bicyclic com-

pounds. J Phys Chem A 103:1991.

Lakin NM, Pietraperzia G, Becucci M, Castellucci E, Coreno M, Giardini-

Guidoni A, van der Avoird A. 1998. High-resolution spectroscopy of 4-

fluorostyrene-rare gas van der Waals complexes: Results and compar-

ison with theoretical calculations. J Chem Phys 108:1836.

Latini A, Toja D, Giardini Guidoni A, Palleschi A, Piccirillo S, Speranza M.

1999a. Spectroscopic enantiodifferentiation of chiral molecules in the

gas phase. Chirality 11:376.

Latini A, Toja D, Giardini Guidoni A, Palleschi A, Piccirillo S, Speranza M.

1999b. Energetics of molecular complexes in a supersonic beam: A

novel spectroscopic tool for enantiomeric discrimination. Angew Chem

Int Ed Engl 38:815.

Latini A, Satta M, Giardini Guidoni A, Piccirillo S, Speranza M. 2000. Short-

range interactions within molecular complexes formed in supersonic

beams: Structural effects and chiral discrimination. Chem Eur J 6:1042.

Le Barbu K, Lahmani F, Zehnacker-Rentien A. 2002. Formation of hydrogen-

bonded structures in jet-cooled complexes of a chiral chromophore

studied by IR/UV double resonance spectroscopy: Diastereoisomeric

complexes of (�)-2-naphthyl-1-ethanol with (�)-2-amino-1-propanol.

J Phys Chem A 106:6271.

Le Barbu K, Brenner V, Millie Ph, Lahmani F, Zehnacker-Rentien A. 1998.

An experimental and theoretical study of jet-cooled complexes of chiral

molecules: The role of dispersive forces in chiral discrimination. J Phys

Chem A 102:128.

Le Barbu K, Zehnacker A, Lahmani F, Mons M, Piuzzi F, Dimicoli I. 2001.

Spectroscopic studies of enantiomeric discrimination in jet-cooled

chiral complexes. Chirality 13:715.

Lecomte F, Lucas B, Gregoire G, Schermann JP, Desfrancois C. 2005. (private

communication).

Leutwyler S, Bosiger J. 1990. Rare-gas solvent clusters: Spectra, structures,

and order-disorder transitions. Chem Rev 90:489.

Leutwyler S, Jortner J. 1987. The adsorption of rare-gas atoms on

microsurfaces of large aromatic molecules. J Phys Chem 91:5558.

Lubman DM, Li L. 1990. Laser and mass spectrometry, chapter XVI. Lubman

DM, editor. Oxford: Oxford University Press.

Maerk TD, Castleman AW, Jr. 1984. Experimental studies on cluster ions.

Adv At Mol Phys 20:65.

Milne TA, Greene FT. 1968. Direct mass-spectrometric sampling of high-

pressure systems. Adv Chem Ser 72:68.

Mons M, Piuzzi F, Dimicoli I, Zehnacker A, Lahmani F. 2000. Binding energy

of hydrogen-bonded complexes of the chiral molecule 1-phenylethanol,

as studied by 2C-R2PI: Comparison between diastereoisomeric

complexes with butan-2-ol and the singly hydrated complex. Phys

Chem Chem Phys 2:5065.

Muller-Dethlefs K. 1995. High-resolution laser photoionisation and photo-

electron studies. Chichester, England: J. Wiley and Sons.

Muller-Dethlefs K, Hobza P. 2000. Noncovalent interactions: A challenge for

experiment and theory. Chem Rev 100:143.

Nir E, Grace L, Brauer B, de Vries MS. 1999. REMPI spectroscopy of jet-

cooled guanine. J Am Chem Soc 121:4896.

Nir E, Imhof P, Kleinermanns K, de Vries MS. 2000. REMPI spectroscopy of

laser desorbed guanosines. J Am Chem Soc 122:8091.

Nir E, Janzen Ch, Imhof P, Kleinermanns K, de Vries MS. 2001. Guanine

tautomerism revealed by UV-UVand IR-UV hole burning spectroscopy.

J Chem Phys 115:4604.

Nir E, Janzen C, Imhof P, Kleinermanns K, de Vries MS. 2002. Pairing of the

nucleobases guanine and cytosine in the gas phase studied by IR-UV

double-resonance spectroscopy and ab initio calculations. Phys Chem

Chem Phys 4:732.

Parker KS, Townshend A, Bale SJ. 1995. Determination of the enantiomeric

composition of 1-phenylethylamine based on its quenching of the

fluorescence of 2,20-dihydroxy-1,10-binaphthalene. Anal Proc 32:329.

Parker KS, Townshend A, Bale SJ. 1996. Simultaneous determination of the

concentrations of each enantiomer of 1-phenylethylamine using their

quenching of the fluorescence of two chiral fluorophores. Anal

Commun 33:265.

Philips LA, Levy DH. 1986. Determination of the transition moment and the

geometry of tryptamine by rotationally resolved electronic spectro-

scopy. J Chem Phys 90:4921.

Piccirillo S, Coreno M, Bosman C, Giardini A, Mele A, Palleschi A. 1995.

Resonant two-photon ionization of van der Waals adducts of 4-

fluorostyrene with monomethylamine and monoethylamine: Intraclus-

ter chemical reactions. Chem Phys Lett 247:577.

Piccirillo S, Bosman C, Toja D, Giardini Guidoni A, Pierini M, Troiani A,

Speranza M. 1997. Gas-phase enantiodifferentiation of chiral mole-

cules: Chiral recognition of 1-phenyl-1-propanol/2-butanol clusters by

resonance enhanced multiphoton ionization spectroscopy. Angew

Chem Int Ed Engl 36:1729.

Piccirillo S, Satta M, Catone D, Scuderi D, Paladini A, Rondino F, Speranza

M, Giardini Guidoni A. 2004. Mass resolved laser spectroscopy of

micro-solvated R-(þ)-1-phenyl-1-propanol: A chiral molecule of

biological interest. Phys Chem Chem Phys 6:2858.

Pirkle WH, House DW. 1979. Chiral high-performance liquid chromato-

graphic stationary phases. 1. Separation of the enantiomers of

sulfoxides, amines, amino acids, alcohols, hydroxy acids, lactones,

and mercaptans. J Org Chem 44:1957.

Plutzer C, Nir E, de Vries MS, Kleinermanns K. 2001. IR-UV double-

resonance spectroscopy of the nucleobase adenine. Phys Chem Chem

Phys 3:5466.

Rizzo TR, Park YD, Levy DH. 1986. Dispersed fluorescence of jet-cooled

tryptophan: Excited state conformers and intramolecular exciplex

formation. J Chem Phys 86:6945.

& SPERANZA ET AL.

608

Rizzo TR, Park YD, Peteanu LA, Levy DH. 1985. Electronic spectrum of the

amino acid tryptophan cooled in a supersonic molecular beam. J Chem

Phys 85:4819.

Rizzo TR, Park YD, Peteanu LA, Levy DH. 1986. The electronic spectrum of

the amino acid tryptophan in the gas phase. J Chem Phys 86:2534.

Robertson EG, Simons JP. 2001. Getting into shape: Conformational and

supramolecular landscapes in small biomolecules and their hydrated

clusters. Phys Chem Chem Phys 3:1.

Robinson PJ, Holbrook KA. 1972. Unimolecular reactions. New York: Wiley.

Ryali SB, Fenn JB. 1984. Clustering in free jets—Aggregation by dispersion.

Ber Bunsen-Ges Phys Chem 88:245.

Satta M, Latini A, Piccirillo S, Di Palma TM, Scuderi D, Speranza M, Giardini

Guidoni A. 2000. Energetics of monohydrated chiral R(þ)-1-phenyl-1-

propanol: Supersonic beam experiments and density functional

calculations. Chem Phys Lett 316:94.

Schauer M, Bernstein ER. 1985. Molecular jet study of the solvation of

benzene by methane, ethane, and propane. J Chem Phys 82:726.

Scherzer W, Selzle HL, Schlag EW. 1992. Identification of spectra of mixed

structural isomers via mass selective hole-burning in the gas phase.

Chem Phys Lett 195:11.

Schreier P, Bernreuther A, Huffler M. 1995. Chapter 3. Analysis of chiral

organic molecules. Methodology and applications. Berlin: Walter de

Gruyter.

Schuette JM, Will AY, Agbaria RA, Warner IM. 1994. Fluorescence

characterization of the cyclodextrin/pyrene complex interaction with

chiral alcohols and diols. Appl Spectrosc 48:581.

Scuderi D, Paladini A, Piccirillo S, Satta M, Catone D, Giardini A, Filippi A,

Speranza M. 2002a. Chiral discrimination of 2,3-butanediols by laser

spectroscopy. Chem Commun 2438.

Scuderi D, Paladini A, Satta M, Catone D, Piccirillo S, Speranza M, Giardini

Guidoni A. 2002b. Chiral aggregates of indan-1-ol with secondary

alcohols and water: Laser spectroscopy in supersonic beams. Phys


Scuderi D, Paladini A, Satta M, Catone D, Rondino F, Tacconi M, Filippi A,

Piccirillo S, Giardini Guidoni A, Speranza M. 2003. Chiral recognition

of diols by complexation with (R)-(þ)-1-phenyl-1-propanol: A R2PI

approach in supersonic beam. Phys Chem Chem Phys 5:4570.

Smalley RE, Levy DH, Wharton L. 1976. The fluorescence excitation

spectrum of the helium-iodine (HeI2) van der Waals complex. J Chem

Phys 64:3266.

Smalley RE, Wharton L, Levy DH. 1977. Molecular optical spectroscopy

with supersonic beams and jets. Acc Chem Res 10:139.

Snoek LC, Kroemer RT, Simons JP. 2002. A spectroscopic and computational

exploration of tryptophan-water cluster structures in the gas phase. Phys


Snoek LC, Robertson EG, Kroemer RT, Simons JP. 2000. Conformational

landscapes in amino acids: Infrared and ultraviolet ion-dip spectroscopy

of phenylalanine in the gas phase. Chem Phys Lett 321:49.

Takeuchi M, Yoda S, Imada T, Shinkai S. 1997. Chiral sugar recognition by a

diboronic-acid-appended binaphthyl derivative through rigidification

effect. Tetrahedron 53:8335.

Talbot FO, Simons JP. 2002. Sugars in the gas phase: The spectroscopy and

structure of jet-cooled phenyl b-D-glucopyranoside. Phys Chem Chem

Phys 4:3562.

Taylor DR, Maher K. 1992. Chiral separations by high-performance liquid

chromatography. J Chromatogr Sci 30:67.

Van Berkel GJ. 1997. Electrospray ionization mass spectrometry. Cole RB,

editor. New York: Wiley.

Veenstra BR, Jonkman HT, Kommandeur J. 1994. Formation of mixed binary

clusters in a supersonic molecular beam: A nice case of RRK kinetics. J

Phys Chem 98:3538.

Weber Th, Riedel E, Neusser HJ, Schlag EW. 1991. Van der Waals bond

lengths and electronic spectral shifts of the benzene-krypton and

benzene-xenon complexes. Chem Phys Lett 183:77.

Weinkauf R, Schanen P, Yang D, Soukara S, Schlag EW. 1995. Elementary

processes in peptides: Electron mobility and dissociation in peptide

cations in the gas phase. J Phys Chem 99:11255.

Wittmeyer SA, Topp MR. 1989. Spectral hole burning in free perylene

and in small clusters with methane and alkyl halides. Phys Lett Chem

163:261.

Yang H, Bohne C. 1995. Chiral discrimination in the fluorescence quenching

of pyrene complexed to b-cyclodextrin. J PhotoChem Photobiol A

86:209.

Yao J, Im HS, Foltin M, Bernstein ER. 2000. Spectroscopy of neurotrans-

mitters and their clusters: Phenethylamine and amphetamine solvation

by nonpolar, polar, and hydrogen-bonding solvents. J Phys Chem A

104:6197.

Zheng YJ, Ornstein RL. 1996. A molecular dynamics and quantum

mechanics analysis of the effect of dmso on enzyme structure and

dynamics: Subtilisin. J Am Chem Soc 118:4175.

Maurizio Speranza is a Full Professor of General and Inorganic Chemistry at the

University of Rome ‘‘La Sapienza’’ from 1986. From 1969 to 1986, he has been a staff

member of the Institute of Nuclear Chemistry of the Research National Council of Italy. His

current research interests include structural, kinetic, and mechanistic studies in the area of

gas-phase ion chemistry, with particular emphasis on chiral systems, by mass spectrometric,

spectroscopic, and radiolytic methods. He is co-author of approximately 200 publications.

Mauro Satta was born in 1972 in Rome, Italy. He received his undergraduate degree in

chemistry from ‘‘La Sapienza’’ University in 1998 in Rome, where he also received his

Ph.D. degree in chemistry in 2002 under the supervision of Prof. F.A. Gianturco. He is

currently a permanent researcher of the IMIP of the CNR in Italy. His fields of research

include theoretical photo-dissociation dynamics of molecules and clusters in the gas phase

and theoretical investigations of intermolecular interactions in molecular adducts

containing p systems. He is also interested in the modeling of laser ablation process.

Susanna Piccirillo received her Ph.D. from the University of Rome ‘‘La Sapienza’’ in

1992. She spent postdoctoral years at the University of Paris Sud (with Prof. P. Brechignac)

and at CNR Potenza (with Prof. A. Giardini). Since 1996, she is a researcher at the


609

University of Rome Tor Vergata. Her research interests have been mainly devoted to the

spectroscopic characterization of molecular clusters in supersonic beam, and since 1995

mainly focused on chiral systems. She is also active in the field of materials science,

currently on carbon-based nanostructures.

Flaminia Rondino received a graduate degree in Chemistry in 2003 from the University of

Rome ‘‘La Sapienza.’’ She did graduate work on the laser photofragmentation of

noncovalently bound chiral clusters. At present, her current research interest as Ph.D.

student is the development of novel sources to bring molecules of biological interest in

supersonic beam.

Alessandra Paladini received her Ph.D. from the University of Rome ‘‘La Sapienza’’ in

2002. She spent a postdoctoral year at the University of Bielefeld (Germany), with Prof.

H.J.Loesch. At present, she has CNR scholarship at the University of Rome. Her research

interests have been mainly devoted to laser spectroscopy and mass spectrometry of clusters

of biological interest.

AnnaGiardini is a Full Professor of Molecular Spectroscopy and Director of the Dynamics

and Laser Photochemistry Group at University of Rome ‘‘La Sapienza.’’ Since 1958, she

performed research activity on ion-molecule reactions, electron scattering and ionization

experiments (e, 2e), selective absorption of one or more coherent laser photons on atoms,

molecules, and clusters in gaseous phase. Other research activities concern the study of

laser-matter interactions and production of thin films of ceramic materials by pulsed laser

ablation and deposition. Her activity is well known in Italy and worldwide. She is co-author

of more than 250 publications.

Antonello Filippi was born in Rome. After the High School of Chemistry, he worked, from

1986 to 1996, as a research assistant in the Institute of Nuclear Chemistry of the Research

National Council of Italy. In 1992 received his graduate degree in Chemistry from the

University of Rome ‘‘La Sapienza’’. Since 1996, he is a researcher at the University of

Rome ‘‘La Sapienza.’’ His research activity concerns the investigation of ion-molecule

reactions and molecule-molecule interactions in the gas phase (structure, dynamics,

and mechanism) by theoretical, mass spectrometric, spectroscopic, and radiolytic

methodologies.

Daniele Catone received a graduate degree in Chemistry in 2001 from the University of

Rome ‘‘La Sapienza.’’ His graduate work was focused on ultrafast femtosecond laser

spectroscopy of polymers and green fluorescent protein. His present research, as Ph.D.

student, is dedicated to the study of chiral molecules and their diastereomeric aggregates

through multi-photon laser spectroscopy and circularly polarized sinchrotron radiation.

& SPERANZA ET AL.

610

Chiral recognition by mass-resolved laser spectroscopy

Documents

Transcript of Chiral recognition by mass-resolved laser spectroscopy