Previtamin D: Z-E photoisomerization via a Hula-twist conical intersection.

On photoisomerization of previtamin D - a steroid Z-triene - produced in situ by ring opening of 7-dehydrocholesterol in a cold matrix, it was found in A. M. Müller et al. [Angew. Chem., Int. Ed., 1998, 37, 505-507] that the product (tachysterol) had rotated not only its central double bond but also an adjacent single bond. This is called a Hula twist (HT) due to the alternative description, in which it is just one central CH group that rotates. It was pointed out that the results directly support the calculated molecular structure at a conical intersection, which mediates the Z-E isomerization of polyenes. With a more sophisticated technique, Saltiel et al. (J. Phys. Chem. Lett., 2013, 4, 716-721) confirmed this tachysterol rotamer as the main product but found two additional conformers. They believed to have seen also three previtamin D conformers, suggested to be a result of hot-ground-state reactions from the primary rotamer, and interpreted all tachysterol products to be a result of a double-bond twist (DBT), not a HT. On the basis of published circular dichroism data and consideration of other reactions, it is here shown that under these conditions hot-ground-state reactions are unimportant or even negligible and that there is practically only a single conformer of previtamin D after ring opening. All products can be easily understood on the basis of an HT-type conical intersection, which is thus further supported. Invoking a published pretwist model even rationalizes product ratios. The two twists in HT are concerted. Furthermore HT is fully consistent with the NEER principle (nonequilibration of excited rotamers) and even offers additional possibilities for conformer control.

Vibrational anharmonicities and reactivity of tetrafluoroethylene.

Compared to ethylene and its nonfluorinated derivatives, C(2)F(4) is peculiar in many reactions. It very easily adds to radicals and prefers formation of four-membered rings over Diels-Alder reactions. This has been rationalized by the preference of fluorine for carbon sp(3) hybridization, which is possible on opening of the double bond. Another property, the thermal dissociation of the C ═ C bond, has been explained by the stabilization of the product (CF(2)) by back-bonding. Here, it is attempted to correlate such properties with vibrational constants, in particular for C ═ C stretching and twisting and for carbon pyramidalization. The only force constant found to be lowered compared to ethylene is that for trans pyramidalization (ν(8)), and CC bond softening on ν(8) distortion is indicated by the conspicuously large magnitude of anharmonic constant, x(18). Both observations can be rationalized by a valence-bond model that predicts a trans bent structure on weakening the CC bond. Conclusions are drawn about the dissociation path and peculiarities of the potential. Other anharmonicities, both experimental and calculated and some in (12)C(13)CF(4) and (13)C(2)F(4), are also discussed. In particular some strong Fermi resonances are identified and their effects accounted for.

Photodissociation of group-6 hexacarbonyls: observation of coherent oscillations in an antisymmetric (pseudorotation) vibration in Mo(CO)(5) and W(CO)(5).

On dissociation of M(CO)(6), M = Cr, Mo and W, by a femtosecond UV laser (<270 to 360 nm), pronounced coherent oscillations are observed in the pentacarbonyl products on probing by long-wavelength (810 nm) ionization in the gas phase. They are vibrations in the ground state, driven by the slope from a conical intersection on relaxation from the initially formed excited state (S(1)). Surprisingly, with M = Mo and W we also find a fundamental of an antisymmetric (b(2) in C(4v)) vibration. From positive and negative displacements along such a coordinate one would expect the same signal, so that there should be only overtones. Vibrational selection rules are therefore considered for time-resolved spectroscopy. The reason for the symmetry breaking is suggested to result from the fact that the phase in superposition of wave functions is established by the pump process and this phase is conserved in probing, independently of the probe delay. An antisymmetric fundamental can be observed, if there is a small tunneling splitting in a state involved in the probe process. The observations also imply some conclusions on the dissociation and relaxation processes and the potentials: with longer wavelengths, the wave packet enters on the same surface but from a different direction to S(1). Only a very minor fraction of the available energy appears as coherent oscillation. There is no equipartition at the end, and a second CO is cleaved off in few picoseconds, even if there is only very little excess energy. Triplets do not contribute, even in the tungsten system and at longest wavelengths. The dissociation mechanism involves passage of the wave packet from all initial states over an avoided crossing to a repulsive ligand-field surface. It predicts that in some other molecules, the barrier caused thereby is larger and for long photolysis wavelength the lifetime is long enough for intersystem crossing to take place; it also predicts wavelength dependences in these cases. It is again emphasized that there is no vertical internal conversion; instead, the molecule is controlled by slopes and intersections of potentials. Also lifetimes can be considered as a control parameter in photochemistry.

Biological homochirality as result from a single event.

The parity-violating effect of the weak force is much too small to be statistically significant for the origin of biological homochirality. Other physical interactions such as with circularly polarized light (CPL) are larger and are discussed in the literature as an effect on small molecules such as amino acids or sugars. To enable their polymerization, they must have existed in large quantities, so that statistical fluctuations may have been smaller than the CPL effect. However, the subsequent formation of the first polymer with good self-reproducing ability was very improbable and may indeed have happened only once. (Such a polymer is isotactic, containing only homochiral constituents.) In this case, the enantiomer excess was 100%, the sign emerging by pure chance without any deterministic influence. The single-event hypothesis can also easily explain why many natural sugars and related compounds (e.g. ascorbic acid) belong exclusively to the l-series, while with others both enantiomers are represented, although Nature uses only d-glucose and d-ribose. It is also pointed out that in the sugar series the sign of physical effects not only varies from species to species but even in the easily equilibrating anomers. It is, however, not excluded that a chemical deterministic effect may have supported the formation of a self-reproducing polymer: enantiomer enrichment by adsorption on a mineral surface.

Time-resolved electron diffraction from selectively aligned molecules.

We experimentally demonstrate ultrafast electron diffraction from transiently aligned molecules in the absence of external (aligning) fields. A sample of aligned molecules is generated through photodissociation with femtosecond laser pulses, and the diffraction pattern is captured by probing the sample with picosecond electron pulses shortly after dissociation-before molecular rotation causes the alignment to vanish. In our experiments the alignment decays with a time constant of 2.6+/-1.2 ps.

Does life originate from a single molecule?

Life did not emerge in a single step. In chemical evolution, the first formation of a self-replicating molecule was probably one of the most critical bottlenecks, which was overcome only with a very low probability. If only one such event was successful, present-day life originates from a single molecule. In this case, homochirality in DNA and RNA is explained almost without further assumptions. By contrast, the enantiomer excess, produced by the deterministic mechanisms suggested so far, is smaller than the statistical standard deviation, unless the postulated initial number of molecules is very--in some mechanisms unreasonably--large. A certain chiral nonuniformity of natural monosaccharides other than (deoxy)ribose supports the idea that homochirality originates not from such small molecules but from an early RNA-like oligomer. This nonuniformity seems also hard to explain by any deterministic mechanism.

Vacuum ultraviolet pulses of 11 fs from fifth-harmonic generation of a Ti:sapphire laser.

We demonstrate that in a short Ar cell, generation of the fifth harmonic from 12 fs pulses at 810 nm directly results in ultrashort vacuum UV pulses at 162 nm. They have a spectral width of approximately 5 nm and a duration of 11+/-1 fs (1.4 times the transform limit), as measured by cross correlation with the fundamental pulses. Their energy (estimated to 4 nJ) turned out to be sufficient for use as a pump in time-resolved experiments.

Sub-10-fs supercontinuum radiation generated by filamentation of few-cycle 800 nm pulses in argon.

Focusing 12 fs pulses of 800 nm with moderate energy (0.35 mJ) into atmospheric-pressure argon (Ar) gives rise to filamentation (self-focusing) and a supercontinuum with a very broad pedestal, extending to 250 nm. According to the present understanding, the short wavelengths are produced by self-phase modulation in the self-steepened trailing edge of the pulse. Pulses in this spectral range might thus be intrinsically short. Indeed we demonstrate this by extracting the light near the end of the filament, terminating self-focusing by a pressure gradient at a pinhole, beyond which the Ar is pumped away. We obtain pulses of 9.7 fs in the region of 290 nm without the necessity of compression.

Forward and backward pericyclic photochemical reactions have intermediates in common, yet cyclobutenes break the rules.

Photochemical pericyclic reactions are believed to proceed via a so-called pericyclic minimum on the lowest excited potential surface (S(1)), which is common to both the forward and backward reactions. Such a common intermediate has never been directly detected. The photointerconversion of 1,3-butadiene and cyclobutene is the prevailing prototype for such reactions, yet only diene ring closure proceeds with the stereospecificity that the Woodward-Hoffmann rules predict. This contrast seems to exclude a common intermediate. Using ultrafast spectroscopy, we show that the excited states of two cyclobutene/diene isomeric pairs are linked by not one, but by two common minima, p* and ct*. Starting from the diene side (cyclohepta-1,3-diene and cycloocta-1,3-diene), electrocyclic ring closure passes via the pericyclic minimum p*, whereas ct* is mainly responsible for cis-trans isomerization. Starting from the corresponding cyclobutenes (bicyclo[3.2.0]heptene-6 and bicyclo[4.2.0]octene-7), the forbidden isomer is formed from ct*. The path branches at the first (S(2)/S(1)) conical intersection towards p* and ct*. The fact that the energetically unfavorable ct* path can compete is ascribed to a dynamic effect: the momentum in C=C twist direction, acquired--such as in other olefins--in the Franck-Condon region of the cyclobutenes.

Ultrafast relaxation and coherent oscillations in aminobenzonitriles in the gas phase probed by intense-field ionization.

4-Aminobenzonitrile derivatives have two excited states of similar energy: besides the benzene-like L(b) state (also termed "locally excited" or LE state) one with charge-transfer (CT) character that is slightly higher in the isolated molecules. The CT state can be lowered by solvents of suitable polarity, so that dual fluorescence can be observed in them. It is controversial along which coordinate this state is displaced, although the amino-group twist is a wide-spread assumption. We investigated a number of such compounds by transient ionization in the gas phase, initially exciting the higher-lying L(a) state (S(2)). Here we briefly review the previous results on 4-(dimethylamino)benzonitrile (the prototype of this class of molecules), 4-piperidino-, pyrrolidino- and pyrrolyl-benzonitrile and compare them with new results on 4-aminobenzonitrile and on the bridged derivative N-methyl-6-cyano-1,2,3,4-tetrahydroquinoline (NMC6). Although in the latter two molecules the CT state has never been detected before, we find the same relaxation path for all compounds: From S(2), the wave packet passes through a conical intersection (CI); from there part of it reaches the S(1) (L(b)) state directly, whereas another part temporarily populates the CT state (also in NMC6), from where it goes around the CI also to the L(b) well. The wave packet directly reaching the L(b) well oscillates there along coordinates involving amino-group twist and wagging or molecular arching and a quinoidal distortion. These coordinates must be components of the CI displacement vector. A vibration involving bond-length alternation of the benzene ring is ascribed to a momentum caused by the electronic symmetry change in the CI, i.e., to the nonadiabatic coupling vector. Also the CT state involves amino-group twist, as to conclude from the anisotropy of the corresponding signal. The six-membered aliphatic ring in NMC6 hinders the twist and raises the CT state to an energy that is, however, still below the L(a) state, so that it can be temporarily populated in a barrierless process. Also in aminobenzonitrile the CT state is between L(a) and L(b) and is reached from L(a) without a barrier. The twist is rationalized by vibronic interaction with a higher state that is pi-antibonding between the amino group and the aromatic ring.

Coherent oscillation and ultrafast internal conversion of tetrafluoroethene after excitation at 197 nm.

C2F4 was excited by using a 150 fs pulse in its longest-wavelength band to the Rydberg (3 s) state and then probed by photoionization techniques at 810 nm. The molecule relaxes in two consecutive steps (time constants 29 and 118 fs), probably via the pipi* state, which is lowered in energy by stretching and twisting the C=C bond. A coherent oscillation (350 fs) was found, which we assign to an overtone of the twist vibration (47.6 cm(-1)) in this state. we also conclude that dissociation to singlet and some triplet CF2 only takes place in the hot ground state of C2F4, from where also the C2F4 triplet state is populated. The potentials and their conical intersections are discussed with respect to relaxation and dissociation, including also some new considerations of thermal processes.

Ultrafast temporary charge transfer in pyrrolidinyl-benzonitrile and pyrrolyl-benzonitrile in the gas phase.

4-(N-pyrrolidinyl)benzonitrile and 4-(N-pyrrolyl)benzonitrile were excited by an ultrashort pulse at 270 nm to their La (S2) states and then probed by ionization at long wavelengths. Parent and fragment ion signals show components with time constants < 100 fs which we attribute to ultrafast relaxation to the Lb (S1) state. From this short time we infer a conical intersection between the Lb and La surfaces. The wave packet can branch there, one part going temporarily to a strongly displaced state. Its shift was concluded from an anisotropy observed only there. The only excited state known to have a large displacement is the charge-transfer (CT) state. The positive anisotropy indicates that the CT state belongs to the 2A species for both molecules. For pyrrolylbenzonitrile, this is in contrast to previous assignments. The anisotropy, and a coherent oscillation observed in pyrrolidinylbenzonitrile, support the idea that the amino-group twist is an important component of the CT reaction coordinate.

Dissociation of multiply ionized isocyanic acid through electron impact.

The dissociation of singly to triply ionized isocyanic acid (HNCO) has been investigated by two- and three-dimensional covariance mapping techniques through electron impact ionization at an electron energy of 200 eV. The absolute cross sections for the various dissociation channels of up to triply ionized HNCO have been measured. The HNCO dications dissociate mostly into ion pairs, while the HNCO trications dissociate mostly into ion triples, both through all the possible bond cleavages and charge allocations. Some major ion-pair dissociation channels of HNCO2+ are supposed to be sequential dissociation through initial charge separation. The metastable decay traces caused by HNCO(2+)-->H(+)+NCO+ and HNCO(+)-->HCO(+)+N have been observed on the covariance map.

Ultrafast charge transfer via a conical intersection in dimethylaminobenzonitrile.

The L(a)-like S2 state (2A) of 4-(dimethylamino)benzonitrile was pumped at 267 nm in the gas phase at 130 degrees C. Nonresonant multiphoton ionization at 800 nm with mass-selective detection then probed the subsequent processes. Whereas ionization at the Franck-Condon geometry only gave rise to the parent ion, fragmentation increased on motion towards the charge-transfer (CT) state. This useful difference is ascribed to a geometry-dependent resonance in the ion. The time constants found are interpreted by ultrafast (approximately 68 fs) relaxation through a conical intersection to both the CT and the L(b)-type S1 state (1B). Then the population equilibrates between these two states within 1 ps. From there the molecule relaxes within 90 ps to a lower excited state which can only be a triplet state (T(n)) and then decomposes within 300 ps. Previous experiments either investigated only 1B --> CT relaxation-which does not take place in the gas phase or nonpolar solvents for energetic reasons--or, starting from S2 excitation, typically had insufficient time resolution (>1 ps) to detect the temporary charge transfer. Only recently temporary population of the CT state was found in a nonpolar solvent (Kwok et al., J. Phys. Chem. A. 2000, 104, 4188), a result fully consistent with our mechanism. We also show that S2 --> S1 relaxation does not occur vertically but involves an intermediate strong geometrical distortion, passing through a conical intersection.

Ultrafast [2 + 2]-cycloaddition in norbornadiene.

Excitation of norbornadiene (bicyclo[2.2.1]hepta-2,5-diene) at 200 nm populates two states in parallel, the second pi(pi*) state and a Rydberg state. We monitored both populations by transient nonresonant ionization. From the pi(pi*) state the molecule relaxes in consecutive steps with time constants 5, 31 and 55 fs down to the ground-state surface, whereas the Rydberg population merges to the other path on the pi(pi*) surface within 420 fs. The relaxation steps are discussed in terms of conical intersections (CoIns) between different surfaces Information on them is inferred from known spectroscopy and, for the last CoIn, from published calculations on Dewar benzene-->prismane conversion and on ethylene photodimerization for which norbornadiene with its two nonconjugated double bonds is a model. The calculation predicts symmetry breaking for this CoIn, the two ethylenes forming a rhombus Although this distortion is hindered in norbornadiene by ring strain, this CoIn seems easily accessible as indicated by the short time (<55 fs) found for passing through it.

Structure of the conical intersections driving the cis-trans photoisomerization of conjugated molecules.

High-level ab initio calculations show that the singlet photochemical cis-trans isomerization of organic molecules under isolated conditions can occur according to two distinct mechanisms. These mechanisms are characterized by the different structures of the conical intersection funnels controlling photoproduct formation. In nonpolar (e.g. hydrocarbon) polyenes the lowest-lying funnel corresponds to a (CH)3 kink with both double and adjacent single bonds twisted, which may initiate hula-twist (HT) isomerization. On the other hand, in polar conjugated systems such as protonated Schiff bases (PSB) the funnel shows a structure with just one twisted double bond. The ground-state relaxation paths departing from the funnels indicate that the HT motion may take place in nonpolar conjugated systems but also that the single-bond twist may be turned back, whereas in free conjugated polar molecules such as PSB a one-bond flip mechanism dominates from the beginning. The available experimental evidence either supports these predictions or is at least consistent with them.

Low-Temperature Photochemistry of Previtamin D: A Hula-Twist Isomerization of a Triene.

What is the standard course of the cis-trans isomerization of nonpolar polyenes? Many results support a pathway in which only a central CH group rotates out of the plane, while the remaining parts of the molecule reorient within the plane ("hula-twist" mechanism). The intermediate structure through which the molecule passes corresponds to the geometry predicted for the conical intersection of the S1 and S0 potential energy surfaces (see picture).