The rotational barriers of several ortho and para substituted ethyl benzoates (EB) and benzoic acids (BA), used as models for internal rotation in PET polymers, have been calculated at the SCF and MP2 levels. Since there were no experimental data available on internal rotations in EB and BA, barriers to rotation in benzaldehyde and three similar compounds were calculated as benchmarks and compared with experimental data. Calculated barriers in the benzaldehydes were within 1 kcal/mol of the experimental values. The tentative conclusion from the rotation barrier results on model compounds is that PET-type polymers made from high-rotation-barrier monomers like ortho-hydroxy-terephthalic acid should have larger internal barriers to rotation than unsubstituted PET, leading to more rigid polymers with higher glass transition temperatures, Tg, and an altered degree of crystallinity. We found an unexpectedly low barrier to rotation of 2.3 kcal/mol for ortho XCH3, compared to about 5 kcal/ mol for ortho XH. Based on these model studies, we would expect that a PET-type polymer made from ortho-methyl substituted terephthalate would be more flexible, would have a lower glass transition temperature and different degree of crystallinity with accompanying change in void space.

Ab initio molecular orbital calculations were performed at the SCF and MP2 levels, using a 6-31G basis set, for complexes of CO2 with carbonyl compounds. The specific interaction between CO2 and the carbonyl oxygen can be described as a Lewis acid-base reaction. Two different geometries, one having C-2v symmetry and the other having C-s symmetry, were studied. The C-s symmetry was found to yield stronger binding of the CO2 complexes. The degeneracy of the nu(2) bending mode in free CO2 was lifted when the CO2 was bound. The calculated nu(2) splittings at the SCF level, using a 6-31G basis set, were comparable to literature values determined by IR spectroscopy of CO2-impregnated polymers. When steric hindrance was present, the binding energy of CO2 to carbonyls was reduced, resulting in lower nu(2) splittings. The interaction energy between benzene and CO2 was determined to be much lower than that associated with a carbonyl oxygen and CO2. The preference of CO2 for the carbonyl group over the benzene ring, along with the role that steric hindrance plays, allows an understanding of the specific interactions of CO2 with polymers.

An unprecedented small thioaurite cluster compound (with metallic Au-0 core) has been isolated in high yield by decomposition of polymeric Au(I)SG compounds, where GSH is the ubiquitous tripeptide glutathione, N-gamma-glutamyl-cysteinyl-glycine. The Au:SG clusters appear to share the high stability and robustness of their hydrophobic n-alkyl analogues but are highly water soluble. The most abundant cluster produced by these methods can be easily separated from its homologues by gel electrophoresis. Its total molecular weight is ca. 10.4 kDa, and the mass of its strongly bound inorganic core is 5.6 kDa, suggesting the composition AU(28)(SG)(16) This composition is also consistent with the X-ray diffraction pattern of the crystalline molecular solid. Distinct features in the optical absorption spectroscopy are inherently different from either larger clusters or smaller gold duster compounds. The compound is optically active, as evidenced by circular dichroism in the near-iii, visible, and near-UV regions. The C-13 NMR spectra suggest that the bonding environment of the GS-adsorbate is similar to that of the n-alkyl-adsorbate dusters, and the nonsulfhydryl properties are retained. The cluster is thus envisioned as a large metallic-cluster compound with distinctive optical properties encapsulated by a bioactive peptide monolayer.

Energies of the 15 lowest Rydberg states of the metastable H-4 cluster have been determined using Koopman's theorem at the equilibrium geometry of the par ent ion. H-4(+) To represent the core orbitals of H-4, a 6-31 G** basis set has been used at the SCF, MP2, MP4(SDTQ), CI-SD, and CI-SDT levels. The Rydberg orbitals have been modeled using a basis set analogous to that used to model the Rydberg orbitals of H-3 in previous theoretical calculations. To test the validity of the calculations for the H-4 molecule, ab initio calculations were repeated for the Rydberg orbitals of the H-3 molecule at the stable geometry of the H-3(+) core. Predicted transitions were within 2% of the rotational band spectra of H-3 observed by Herzberg. The metastable H-4 cluster formed from charge neutralization of H-4(+) decomposes into two H-2 molecules. Previous calculations have predicted that one of the two H-2 products will be vibrationally hot while the other will be relatively cold and that a large recoil energy of approximately 9 eV is expected for the relative kinetic energy of the two H-2 products. The present work suggests(2) that if Rydberg states are involved in the charge neutralization process, the recoil energy could be reduced due to radiative transitions among the Rydberg states.

Solutions of gamma-crystallin, and various enzymes, at neutral pH and 24-26 degrees C, became turbid upon exposure to UV radiation at 295 or 308 nm. SDS-PAGE analysis revealed interchain cross-linking and aggregate formation compared to dark control solutions as reported previously. When alpha-crystallin was added to the protein solutions in stoichiometric amounts, UV irradiation resulted in significantly less turbidity than in the absence of alpha-crystallin. For example, addition of 0.5 mg of alpha-crystallin to 0.5 mg of gamma-crystallin in 1.0 ml solution yielded only 25% of the turbidity seen in the absence of alpha-crystallin. Addition of 2.0 mg of alpha-crystallin resulted in 20% of the turbidity. Given the molecular weights of alpha- and gamma-crystallin (about 800 kDa and 20 kDa, respectively), a gamma/alpha 1:1 weight ratio corresponds to a 40:1 molar ratio, and a gamma/alpha 1:4 weight ratio corresponds to a 10:1 molar ratio. Hence, the molar ratio of alpha-crystallin needed to effectively protect gamma-crystallin from photochemical opacification was gamma/alpha = n:1, where n was in the range 10-40. In terms of subunits, this ratio is gamma/alpha = 1:m, where m = 1-4. Thus, each gamma-crystallin molecule needs 1-4 alpha subunits for protection. Similar stoichiometries were observed for protection of the other proteins studied. The protection stems in part from screening of UV radiation by alpha-crystallin but more importantly from a chaperone effect analogous to that seen in thermal aggregation experiments.

Buffer solutions of the lens protein gamma-crystallin and the enzymes aldolase and liver alcohol dehydrogenase became turbid and formed solid precipitate upon exposure to an elevated temperature of 63 degrees C or to UV radiation at 308 nm. When alpha-crystallin was added to the protein solutions in stoichiometric amounts, heat or UV irradiation did not cause turbidity, or turbidity developed much less rapidly than in the absence of alpha-crystallin. Hence, normal alpha-crystallin functioned as a "molecular chaperone," providing protection against both UV and heat-induced protein aggregation. When alpha-crystallin was preirradiated with UV at 308 nm, its ability to function as a chaperone vis-a-vis both UV and heat-induced aggregation was significantly impaired, but only at relatively high UV doses. A major effect of preirradiation of alpha-crystallin was to cause interpeptide crosslinking among the alpha A2 and alpha B2 subunits of the alpha-crystallin macromolecule. In our experiments alpha-crystallin was exposed to UV doses, which resulted in 0.50 and 90% crosslinking as judged by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. alpha-Crystallin samples that were 50% and 90% crosslinked gave chaperone protection, which was increasingly impaired relative to unirradiated alpha-crystallin. The results are consistent with the notion that UV irradiation of alpha-crystallin results in loss of chaperone binding sites.

The decomposition of H-4 cluster formed from charge neutralization of H-4(+) ion has been studied by ab initio energy surface calculations. H-4 decomposition is considered to proceed from the Franck-Condon geometry dictated by charge neutralization of H-4(+) according to H-4(+) + e(-) --> H-4* --> H-2 + H-2, where H-4* refers to the Franck-Condon geometry. The minimum energy decomposition path (MEP) of H-4* is calculated in two-, four-, and six-dimensional space using the Gonzalez and Schlegel method. In addition, the two-dimensional potential surface of H-4 is calculated as a grid of ab initio energy points in the vicinity of the H-4* geometry, and a decomposition path is found graphically which agrees well with the MEP calculations. The results predict that one of the two H-2 product molecules will be vibrationally hot (3-4 eV) while the other will be relatively cold. Rotational excitation of products is expected to be of minor importance. A large recoil energy, on the order of 9 eV, is expected for the relative kinetic energy of the two Hz products. These predictions assume an adiabatic process with no involvement of H-4 Rydberg states.

The six normal mode vibration frequencies and infrared intensities for H+4 and three tetra-atomic lithium–hydrogen ion clusters have been calculated using ab initio SCF, MP2, and CISD methods. Vibration frequencies are also reported for all possible deuterated analogs of the four ions. The normal mode vibrational characteristics and structures of the ions are discussed and compared to each other. The H+4 ion can reasonably be regarded as a strongly bonded, triangular H+3 core with an additional H atom less strongly bound to one of the ring hydrogens in a planar C2v arrangement. The calculated normal mode vibration frequencies for H+4 and its deuterated analogs confirm this picture. The hydrogen–lithium clusters, LiH+3, Li2H+2, and Li+4, had structures and normal mode frequencies which did not yield the same ring/ligand picture as H+4. Instead, these ions behaved like two interacting diatomic fragments (e.g., H2 and LiH+ subunits in LiH+3) oriented perpendicular to each other in a planar arrangement. The potential energy surfaces for the hydrogen–lithium cluster ions displayed two extremely low frequency vibrational modes corresponding to angular motion of these diatomic subunits relative to each other.

Ab initio effective core potential computations are used to describe neutral and ionic sodium halide clusters. Energies, structural parameters, and vibration frequencies have been determined for neutral and singly charged Na(n)X(m) (n = 1-3; m = 1-2; X = Cl, Br, I) species. Dissociation energies have been estimated from energy differences between isolated products and reactants. Examination of Mulliken atomic charges, together with structural data, provides some insight into the nature of binding in these molecules and ions.

The six normal mode vibration frequencies and infrared intensities for H-4(+) and three tetra-atomic lithium-hydrogen ion clusters have been calculated using ab initio SCE MP2, and CISD methods. Vibration frequencies are also reported for all possible deuterated analogs of the four ions. The normal mode vibrational characteristics and structures of the ions are discussed and compared to each other. The H-4(+) ion can reasonably be regarded as a strongly bonded, triangular H-3(+) core with an additional H atom less strongly bound to one of the ring hydrogens in a planar C-2v arrangement. The calculated normal mode vibration frequencies for H-4(+) and its deuterated analogs confirm this picture. The hydrogen-lithium clusters, LiH3+, Li2H2+, and Li-4(+), had structures and normal mode frequencies which did not yield the same ring/ligand picture as H-4(+), Instead, these ions behaved like two interacting diatomic fragments (e.g., H-2 and LiH+ subunits in LiH3+) oriented perpendicular to each other in a planar arrangement. The potential energy surfaces for the hydrogen-lithium cluster ions displayed two extremely low frequency vibrational modes corresponding to angular motion of these diatomic subunits relative to each other.

Ab initio Hartree-Fock, Moller-Plesset perturbation theory (MP2), and quadratic configuration interaction, using single and double substitutions (QCISD), calculations were carried out for the NF3+ ion. Optimized structures were examined at the various levels of theory. Calculation of the inversion barrier height shows the importance of optimizing the geometry at the post-Hartree-Fock level and the inclusion of polarization functions. The best calculated inversion barrier was 13.3 kcal/mol, compared to an experimental value of 17.3 kcal/mol. The dissociation transition state was computed to determine the well depth of the NF3+ ion and its stability toward dissociation. The computed well depth was 28 and 48 kcal/mol at the SCF and MP2 levels, respectively. (C) 1994 John Wiley & Sons, Inc.

Experiments with an electron cyclotron resonance ion source have been employed in a hunt for the elusive CH2+ ion. The authors' experiments do not find evidence for the existence of stable CH2+ ions. Different levels of ab initio molecular orbital theory have been employed to obtain potential energy curves for CH2+. Although SCF calculations show a small minimum in the potential, post-Hartree-Fock computations indicate CH2+ states to be repulsive.

Experiments with calf lens protein fractions in aqueous buffer solutions at room temperature showed that beta H-, beta L- and gamma-crystallin fractions became opaque following ultraviolet exposure at 308 nm, while the alpha-crystallin fraction remained transparent. Transmission loss, due to UV-irradiation, for all of the crystallin samples was studied in the concentration range of 0.1 mg/mL to 1.0 mg/mL, and for alpha- and gamma-crystallin, in the range up to 5 mg/mL. With increased concentrations of beta H-, beta L- and gamma-crystallin, the rate of opacification increased. However, with alpha-crystallin, the loss of transmission was negligible for all of the concentrations and irradiation times studied. Opacification of the crystallins was accompanied by formation of higher molecular weight insoluble proteins as detected by SDS-PAGE.

Photolysis of calf lens protein alpha-crystallin in aqueous solutions has been monitored by observing changes in fluorescence decay following UV irradiation at 308 nm. The fluorescence decay was biexponential in dark controls and in photolyzed solutions. The recovered lifetime components in pH 7.4 phosphate buffer at 23 degrees C were 3.5 and 0.5 ns before irradiation and 2.7 and 0.5 ns after irradiation. As the UV dose increased, the relative weighting coefficient of the 2.7-ns decay component decreased, and that of the 0.5-ns component increased, resulting in an overall lifetime shortening. Similar results were obtained in 5 M guanidine hydrochloride solutions where lifetime components of 2.7 and 0.5 ns were observed. These observations were in contrast to the behavior of tryptophan monomer solutions which did not show any change in fluorescence decay kinetics upon UV photolysis but only a reduced fluorescence intensity. Steady-state fluorescence spectra and fluorescence quantum yields were also measured at 23 degrees C for unirradiated bovine alpha-crystallin and gave phi F = 0.11 +/- 0.01 in pH 7.4 buffer and phi F = 0.10 +/- 0.01 in 5 M guanidine hydrochloride solutions. The combined steady-state and fluorescence decay data were consistent with assignment of the long-lived fluorescence decay component in alpha-crystallin to emission from Trp-9, which is known to photolyze relatively rapidly. The short decay component was assigned to Trp-60, which photolyzed much more slowly. We thus provide an example of using steady-state photochemical data to assign fluorescence decay components in a multi-tryptophan protein.

The primary photophysical and photochemical processes of aqueous tryptophan (Trp) are studied by picosecond transient absorption spectroscopy. Upon excitation of Trp at 292 nm, transient absorption appears within 1.5 ps in the whole visible region and remains for at least 400 ps. Under the present excitation conditions, the transient absorptions due to solvated electrons and the excited triplet state of Trp are relatively weaker, and the observed transient spectrum is predominantly due to excited singlet-state absorption S(n) <-- S1.

The fluorescence decay of tryptophan residues in the bovine lens protein gamma-II crystallin has been measured in aqueous buffer solutions. Results were obtained as a function of emission wavelength, temperature, dissolved oxygen, and denaturing solvent. The protein displayed complex fluorescence decay which fit a biexponential model with a long component (ns) and a short component (few hundred ps). Measured fluorescence quantum yields data for gamma-II crystallin allowed calculation of radiative and non-radiative rate constants. The radiative rate constant was consistent with that observed in other indole derivatives, while the nonradiative rate constant was quite large and accounted for the short lifetime in gamma-II. The temperature dependence of the non-radiative decay in gamma-II crystallin yielded a small activation energy of only 1-2 kcal/mol, compared to 4 kcal/mol for the reference compound NATA whose barrier is known to derive from the rotamer model.