Michelle Francl

Charges fit to electrostatic potentials. II. Can atomic charges be unambiguously fit to electrostatic potentials

The present work examines the conditioning of the least‐squares matrix for obtaining potential derived charges and presents a modification of the CHELP method for fitting atomic charges to electrostatic potentials. Results from singular value decompositions (SVDs) of the least‐squares matrices show that, in general, the least‐squares matrix for this fitting problem will be rank deficient. Thus, statistically valid charges cannot be assigned to all the atoms in a given molecule. We find also that, contrary to popular notions, increasing the point density of the fit has little or no influence on the rank of the problem. Improvement in the rank can best be achieved by selecting points closer to the molecular surface. Basis set has, as expected, no effect on the number of charges that can be assigned. Finally, a well‐defined, computationally efficient algorithm (CHELP‐SVD) is presented for determining the rank of the least‐squares matrix in potential‐derived charge fitting schemes, selecting the appropriate subset of atoms to which charges can be assigned based on that rank estimate, and then refitting the selected set of charges.

NMR and molecular modeling study of the conformations of taxol and of its side chain methylester in aqueous and non-aqueous solution.

Abstract The conformations of the antimitotic agent taxol and its side chain methyl ester have been studied by NMR-spectroscopy and molecular modeling in hydrophobic (CDCl3) and hydrophilic (water, d6-DMSO) solvents. For the side chain methyl ester (4), the coupling constant JH2′-H3′ changes from ≈2 Hz in chloroform to ≈5 Hz in d6-DMSO or water : d6-DMSO, 1 : 1 (v/v). The conformational equilibrium for 4 thus shifts from one favoring conformers with a gauche torsion angle (chloroform), to one predominantly of conformers having this torsion angle anti. In the case of taxol, JH2′-H3′ changes from 2.7 Hz in CDCl3 to ≈8 Hz in water, water - sodium dodecyl sulfate (SDS) and/or d6-DMSO. Again, gauche conformations are implicated in chloroform, but molecular modeling suggests the anti conformer 27 to be dominant in aqueous media and in d6-DMSO. No nuclear Overhauser effects (nOes) between the side chain and the taxane ring-system are observed in chloroform solution. In water and/or d6-DMSO, however, nOes between the side chain (Ph3′ and H2′) and the OAc4 methyl group are detected.

Distance Dependence of Nonadiabaticity in the Branching Between C–Br and C–Cl Bond Fission Following 1[n(O),π∗(C=O)] Excitation in Bromopropionyl Chloride

These experiments on bromopropionyl chloride investigate a system in which the barrier to C–Br fission on the lowest 1A‘ potential energy surface is formed from a weakly avoided electronic configuration crossing, so that nonadiabatic recrossing of the barrier to C–Br fission dramatically reduces the branching to C–Br fission. The results, when compared with earlier branching ratio measurements on bromoacetyl chloride, show that the additional intervening CH2 spacer in bromopropionyl chloride reduces the splitting between the adiabatic potential energy surfaces at the barrier to C–Br fission, further suppressing C–Br fission by over an order of magnitude. The experiment measures the photofragment velocity and angular distributions from the 248 nm photodissociation of Br(CH2)2COCl, determining the branching ratio between the competing primary C–Br and C–Cl fission pathways and detecting a minor C–C bond fission pathway. While the primary C–Cl:C–Br fission branching ratio is 1:2, the distribution of relative...

Competing C–Br and C–C bond fission following 1[n(O),π*(C=O)] excitation in bromoacetone: Conformation dependence of nonadiabaticity at a conical intersection

These experiments investigate the competition between C–C and C–Br bond fission in bromoacetone excited in the 1[n(O),π*(C=O)] absorption, elucidating the role of molecular conformation in influencing the probability of adiabatically traversing the conical intersection along the C–C fission reaction coordinate. In the first part of the paper, measurement of the photofragment velocity and angular distributions with a crossed laser‐molecular beam time‐of‐flight technique identifies the primary photofragmentation channels at 308 nm. The time‐of‐flight spectra evidence two dissociation channels, C–Br fission and fission of one of the two C–C bonds, BrH2C–COCH3. The distribution of relative kinetic energies imparted to the C–Br fission and C–C fission fragments show dissociation is not occurring via internal conversion to the ground electronic state and allow us to identify these channels in the closely related systems of bromoacetyl‐ and bromopropionyl chloride.In the second part of the work we focus on the m...

Through the eyes of a chemist.

Michelle Francl explores the concepts that could help non-chemists see the world more like those trained in the subject.

Scents and sensibility

“The bacon is about to burn.” I nudge my husband, a neophyte cook who is deeply engaged with measuring ingredients for quiche. I am nowhere near the stove, but a subtle shift in scent, a hint of charcoal drifting over the smell of rendering fat, tells me as surely as peering in the pan that the rashers are at the transition state between crispy and charred. Cooking is a sensual experience. Every sense is engaged, not just in appreciating what is being prepared, but in the preparation itself. I taste a spoonful of pasta sauce and add more basil. I knock on the bottom of a loaf of bread, listening for the hollow sound that signals it is ready. In contrast, the art of cooking up molecules is a sterile, almost asensory experience. Step from the kitchen into the lab and we isolate ourselves from sensory feedback, insulating ourselves at every turn from our materials. We pull on lab coats and gloves, carefully nudge samples with the tip of a spatula into vials. Even our sense of hearing is dulled by the roaring of the hoods and chugging of pumps. Above all we shut out the particularly chemical senses: taste and smell. It is not without reason that we insulate ourselves so thoroughly. We often deal in the unknown, where even the briefest exposures have been known to be fatal. In the late nineteenth century, while developing the synthesis of dimethyl mercury, at least two members of Edward Frankland’s research group breathed too deeply of their reaction mixtures, eventually losing their lives to organomercury poisoning1. But I suspect our unease goes deeper than the eminently reasonable safety concerns. Even when we are dealing with substances that are well characterized and known to be safe, we are hesitant to get in close contact with our molecules. Though most of us could synthesize and purify aspirin with materials found around the lab, I’m willing to wager that few chemists would hold it in their hands, let alone swallow what they have made. For better or for worse, we were once far more intimate with our materials. Before the development of techniques that enabled us to directly probe molecular structure, colour, taste and smell provided a rich set of clues to the identity of molecules. Synthetic papers reported the odours and tastes of products and intermediates — along with melting points and colour — as part of the routine characterization of compounds. I found hundreds of references to the taste and smell of compounds scattered among the roughly 180 papers published in Justus Liebigs Annalen Der Chemie during 1870, employing a rich array of descriptors: camphorous, stinging, unbearable, strangely lovely. In 1919 two Stanford chemists, Ernest Oertly and Rollin Myers, catalogued the tastes of nearly 100 (mostly) sweet-tasting molecules culled from reports in the literature covering the period from 1848 to Emil Fischer’s 1906 report of the “insipid and weakly bitter” taste of l-valine2. And yes, Frankland reported the tastes of a number of organomercury compounds, many of which were found to be nauseating and to linger on the palate3. Chemists sensibly stopped tasting their reaction mixtures, substituting thin-layer chromatography and NMR spectroscopy for their tastebuds. But even with modern air handling, we cannot always avoid smelling what we are making — witness the number of papers claiming less noisome methods for using thiols. Yet since the middle of the last century, the scent of a compound is only rarely reported in the literature. And although I can’t imagine thiols smell sweeter now than they did in 1900, contemporary descriptions are very bland compared to those from fifty years ago; odours of compounds are now reported as being ‘sweet’ or simply ‘bad’. There were, however, pockets of rebellion. In 1975 Kay Robert Brower and Rollie Schafer assessed the ability of organic chemists to determine functional groups by smell4, eerily reminiscent of Nobel-Prize-winning physiologist Edgar Adrian’s 1942 experiments with ‘lightly anaesthetized’ hedgehogs as detectors of smells5. Over the past twenty years, chemists have exercised what sociologist Erving Goffman called ‘tactful blindness’ when it comes to smells. Even when the targets are intended to smell, such as the violet-scented ionone derivatives synthesized by Giovanni Vidari and co-workers, they scrupulously avoid any mention of a person smelling the compounds6. The scant twelve lines detailing the ‘olfactory evaluation’ of the ionones — in which five perfume experts smell the compounds as they come off a gas chromatograph — are tucked away on thyl cetate (nail olish rover) Li co ric e/ an is e

Role of Active Site Residues and Solvation in RNase A

The molecular mechanism of catalysis by RNase A has been investigated using high level ab initio molecular orbital calculations in conjunction with molecular dynamics simulations. Despite recent high-resolution X-ray and neutron diffraction studies on RNase,’** the detailed reaction pathway, the roles of active site residues, protonation states, and hydrogen bonding arrangements have remained unclear. His 12 had been thought to be responsible for deprotonating 02’ in order to form the 2’,3’-cyclic phosphate intermediate, but the recent crystal structures show it to be hydrogen bonding to an equatorial phosphate oxygen while Lys 41 H-bonds to 02’. However, since the crystals were formed at low pH, both His 12 and His 119 are protonated. In the active form of the enzyme (at neutral pH) His 12 is expected to be initially unprotonated. We have found that when slightly rotated from its crystallographic position, a deprotonated His 12 can hydrogen bond to 02’. Molecular dynamics simulations of an RNase Alsubstrate complex with His 12 deprotonated are being undertaken to check this possibility. Ab initio calculations have been performed to examine further the roles of His 12 and Lys 41 in the deprotonation of 02’. Using imidazole as a model for His, ammonium for Lys, and either water or methanol for the 02’ hydroxyl, we have calculated the optimized geometry, total energy, and proton affinity (i.e., the difference in energy between a deprotonated and protonated molecule). The preferred protonation states can then be ascertained by comparing the proton affinities of individual components that hydrogen bond to each other (TABLE 1) in order to find the most probable position of the shared proton. For example, since the proton affinity of imidazole (a) is only 240 kcal/mole whereas that of methanol ( c ) is 409 kcal/mole, if imidazole hydrogen bonds to methanol the shared proton would prefer to stay on imidazole. The use of water (b , d ) rather than methanol as a model for 02’ proved too simplistic. This implies that, without other contributing factors, His 12 could not deprotonate 02’. However, if an ammonium ion representing Lys 41 H-bonds to 0 2 ’ ( e ) , the proton affinity of 02’ drops considerably. If the geometry of the H,N-H . . . 02’-CH3 complex is fully optimized, the proton will transfer from N to 0 2 ’ (f) and the proton affinity drops to 233 kcal/mole, 7 kcal below that of imidazole. Thus it appears that 02’ can be deprotonated by His 12, but it is simultaneously reprotonated by Lys 41. However, we hypothesize that the 02’ reprotonation by Lys 41 does not occur for two reasons. First, the 02’-P bond will start to form as soon as 0 2 ’ starts to be deprotonated by His 12. Second, as soon as His 12 is protonated it will prefer to H-bond to the negatively charged equatorial 0 6 rather than to 02’, thereby weakening P 0 6 and strengthening

Beyond CHELP: improved potential derived charges for sugars.

The partitioning of the overall molecular charge distribution into atom centered monopole charges, while quantum mechanically ill-defined, is nevertheless a technique which finds applications in several broad classes of chemical problems. Charges derived from fits to electrostatic potentials have an intuitive appeal since, in principle, these could be derived from either theoretical or experimental data. It has been noted, however, that such potential derived charges can be conformationally dependent in ways that do not appear to reflect the changes in the molecular wavefunction. Both the algorithm used for selecting points at which the molecular electrostatic potential will be fit and the density of points used in the fit have been suggested to influence the resultant charges. Recently [Stouch TR, Williams DE (1992) J Comp Chem 13: 622–32; Stouch TR, Williams DE (1993) J Comp Chem 14: 858–66] it has been noted that numerical difficulties may make it impossible to fit all the atomic charges in a molecule. Singular value decomposition (SVD) of the linear least squares matrices used in fitting atom based monopoles to molecular electrostatic potentials provides a tool for evaluating the integrity of the calculated charges. Based on the SVD analysis for a selected group of molecules we have noted particularly that increasing the molecular size reduces the fraction of charges which can be validly assigned. Users of PD derived charges, especially those who are using those charges for tasks other than reproduction of the MEP, should be aware that there is a high probability that a significant portion of those charges are statistically unreliable. Therefore, charges in many biological molecules, such as sugars, prove to be difficult to obtain by potential derived (PD) methods such as CHELP or CHELPG. Results from the SVD can be used to both assess PD charges and to generate an improved, albeit incomplete, set. Improved PD fits are presented for a series of simple saccharides. Abbreviations: HF, Hartree-Fock; LLS, linear least squares; MEP, molecular electrostatic potential; PD, potential derived; SVD, singular value decomposition

Zen and the art of molecules

The oddest scientific talk I ever gave was on the quantum mechanics of intramolecular rotation in 3-trifluoromethyl phenanthrene. The molecule’s behaviour is unarguably eccentric. The trifluoromethyl rotor seems to be disordered in the X-ray crystal structure, suggesting the presence of two conformers that are close in energy. The observed disorder does not, however, stem principally from two competing minima of the rotor, instead it reflects the libration of this high-barrier rotor: the two positions observed in the X-ray solution roughly represent the limits of the libration1. But the science was perhaps the least strange aspect of the experience. For one thing, I gave the talk barefoot.

How to counteract chemophobia

Michelle Francl ponders ways in which we can talk about chemistry without triggering chemophobia.