Mean field theories of ion distributions, such as the Gouy-Chapman theory that describes the distribution near a charged planar surface, ignore the molecular-scale structure in the liquid solution. The predictions of the Gouy-Chapman theory vary substantially from our x-ray reflectivity measurements of the interface between two electrolyte solutions. Molecular dynamics simulations, which include the liquid structure, were used to calculate the potential of mean force on a single ion. We used this potential of mean force in a generalized Poisson-Boltzmann equation to predict the full ion distributions. These distributions agree with our measurements without any adjustable parameters.

Extended abstract of a paper presented at Microscopy and Microanalysis 2006 in Chicago, Illinois, USA, July 30 – August 3, 2006

C.A. Crot, D.G. Schultz, M. Meron, A. Kilislioglu, P.D. Edirisinghe, K.A. Skinner-Nemec, L. Hanley 1 Department of Chemistry, University of Illinois at Chicago, Chicago, IL, USA 60607-7061 2 ChemMatCARS, Argonne National Laboratory, 9700 S Cass Ave, Argonne, IL, USA 60439 3 Istanbul University, Department of Chemistry, Avcilar 34320, Istanbul, Turkey 4 Biosciences Division, Argonne National Laboratory, 9700 S Cass Ave, Argonne, IL, USA 60439

Synchrotron X-ray reflectivity is used to study the electron density as a function of depth through the bulk water/2-heptanone interface. The measured interfacial width of 7.0±0.2Å is comparable to the value calculated from capillary wave theory (7.3Å) using the measured interfacial tension of 12.6mN/m. This result is consistent with capillary wave theory and molecular dynamics simulations that describe a molecularly sharp interface roughened by thermal fluctuations.

There are several advantages in using molecular ions for surface modification. The modification can be confined to the uppermost layer of the surface, the molecular character of the ion can be imparted to the surface, and sputter yields are often higher. These effects are demonstrated by the use of mass selected ion beams incident on well characterized surfaces. Energy transfer is examined by detecting the masses and energies of ions scattered off surfaces and performing molecular dynamics simulations. Surface modification is followed by chemical analysis with x-ray photoelectron spectroscopy and surface mass spectrometry. TRIDYN monte carlo simulations are used to support some of the modification experiments. Energy transfer is examined for Si(CD3)(3)(+) scattered off clean and hexanethiolate covered Au(111). Adsorbate desorption cross sections and substrate damage depths for NH3/CO/Ni(111) are compared for 10 - 1000 eV isobaric atomic and polyatomic ions, Xe+ and SF5+. The surface chemical modification of polystyrene thin films by 10 - 100 eV SF5+ and C3F5+ ions is also examined.

Relative dissociation energies (RDEs) are obtained for the major fragment ions produced by electrospray ionization/surface-induced dissociation of singly protonated triglycine, tetraglycine, leucine enkephalin, and leucine enkephalin arginine. A previously described data analysis method (Lim, H.; et al. J. Phys. Chem. B 1998, 102, 4753) is employed to analyze the energy-resolved mass spectra by subtracting out the distribution of energy transferred to the surface, integrating over the distribution of the incident ion energy, and taking into account the precursor ion initial internal energy and kinetic energy distributions. These variables are optimized by anchoring the RDE for the lowest energy fragment of a given precursor ion to its literature values and then using these optimized parameters to obtain the other RDEs. The RDEs of the four major fragments of triglycine vary from 2.4 eV for the b(2) fragment ion to 6.0 eV for the a(2) ion. The RDEs of the four major fragments of tetraglycine vary from 3.2 eV for the y(2) ion to 5.7 eV for the a(2) ion. The leucine enkephalin RDEs range from 1.1 eV for the b(4) ion to 2.1 eV for the b(2) ion. The leucine enkephalin arginine RDEs all lay between 2.5 and 3.5 eV. The overall trend of fragmentation order for all peptides is (y(n), b(n)) < a(n) and is consistent with the results from other experiments. The peptide RDEs presented here are only as accurate as the literature values to which they are anchored. Determination of absolute dissociation energies from SID data will require further refinement of the data analysis method.

Full understanding of the surface-induced dissociation (SID) of biological ions requires the determination of the energy channeling into the surface and the scattered ion kinetic and internal energies. Parent and fragment ion kinetic energy distributions were measured for five peptide ions scattered off a hexanethiolate monolayer on Au(lll). Singly protonated ions of triglycine, tetraglycine, cyclo(Pro-Gly), cyclo(His-Phe) and tentoxin were formed by electrospray ionization and scattered at 15-55 eV collision energies off the organic surface. The scattered parent ion kinetic energies were 24% of the incident ion energy for the linear peptides, 21% for the cyclic dipeptides and 17% for the four-peptide ring. These results suggest that ion size and/or structure influences the scattered kinetic energy. Using these values and assuming an average internal excitation efficiency of 17%, it is estimated that the final internal energy given to tbe surface is 59-66% of the initial collision energy. This energy transfer to the surface is very close to that previously estimated for a host of smaller polyatomic ions scattered from similar organic targets. However, comparison with small ion SID shows that the peptides leave the surface with a wider distribution of kinetic energies. Finally, the measured kinetic energy distributions show that the fragment ions for a given peptide leave the surface with a common velocity, suggesting that dissociation occurs away from the surface. All fragments were found to result from non-reactive, inelastic scattering off the organic surface. Copyright (C) 1999 John Wiley & Sons, Ltd.

Several basic topics are presented here that are related to the modification of organic surfaces by 10–100 eV molecular ions. First, experimental results are presented for energy transfer for Si(CD3)3+ and m/z 190–713 peptide ions scattered off hexanethiolate self-assembled monolayers adsorbed on Au(111). Next, experimental and computer simulation results are presented on the dissociation timescales of these same ion-surface pairs. Finally, the surface chemical modification of polystyrene polymer surfaces by CF3+ and C3F5+ ions is analyzed by X-ray photoelectron spectroscopy. These results combine to demonstrate that low energy molecular ion–organic surface modification is highly surface selective, transfers a large fraction of the incident ion energy to the uppermost layer of the surface, can present unique reactive species to the surface, and can transfer to the surface a portion of the ion's chemical functionality.

We have improved our previously described method for extracting activation energies of fragmentation for polyatomic ions from surface-induced dissociation (SID) data [Wainhaus, S. B.; et al. J. Am. Chem. Sec. 1997, 119, 4001]. Our method analyzes the energy-resolved mass spectra and the kinetic energy distribution spectra of the parent and fragment ions that scatter off the surface. It extracts the activation energies by integrating over the distribution of the initial ion energy and the energy transferred to the surface, taking into account both the average value and the width of these distributions. The new method gave improved activation energies for SiMe3+ --> SiMex+ (x = 0-2) fragmentation at a hexanethiolate-covered gold surface. We then used our data analysis method to analyze the activation energies for the fragmentation of thiophene ions at the hexanethiolate-covered gold surface. The activation energies for the formation of C2H2S+, CHS+, and C3H3+ from C4H4S+ were found to be 4.6 +/- 0.7, 6.9 +/- 0.7, and 6.5 +/- 0.7 eV, respectively. Our activation energy results followed the trend in the values from threshold photoelectron photoion coincidence data. However, the SLD values were similar to 50% higher than the threshold photoelectron photoion coincidence values; this discrepancy mostly resulted from delayed dissociation. This model may be used to extract quantitative activation energies from SID data once certain ongoing issues are resolved in future papers. Molecular dynamics simulations were also performed to assist in the data analysis and to test the assumptions of energy transfer in this system. Qualitative agreement in energy transfer was found between the experiments and simulations.

We provide experimental evidence that upon hyperthermal impact of Si(CD3)3+ ions with an organic surface, a portion of the ions undergo dissociation while still in contact with the surface. We use a tandem configuration of quadrupole mass spectrometers along with an energy analyzer to measure the kinetic energy distributions of the fragments that form as a result of the surface scattering of 25 eV Si(CD3)3+. These distributions are different for scattering from a clean Au(111) surface versus scattering from an organic surface composed of a self-assembled monolayer of hexanethiolate on Au(111). Parent and fragment ions recoil from the clean Au(111) surface with the same velocity, as is expected for fragmentation away from the surface. However, the same scattering products recoil from the organic surface with different velocities but similar energies, suggesting that the fragmentation dynamics are modified by surface interactions. We perform molecular dynamics simulations which predict residence times of ∼210 fs at the organic surface and ∼20 fs at the Au surface. The simulations also predict that 13% and 31% of the ions fragment within 1.1 ps of surface impact at the organic and Au surfaces, respectively. Thus, the experimental observation of dissociation at only the organic surface results from its longer ion-surface interaction time. The fragmentation time scale predicted by Rice–Ramsperger-Kassel–Marcus calculations is yet longer, suggesting that at least a portion of the surface-induced dissociation of Si(CD3)3+ may occur via a nonstatistical mechanism. Our interpretation draws heavily from an analogous “shattering” mechanism previously proposed for cluster-surface scattering [E. Hendell, U. Even, T. Raz, and R. D. Levine, Phys. Rev. Lett. 75, 2670 (1995)].

We scatter 10–70 eV SiMe3+ from clean Au(111), a hexanethiolate self-assembled monolayer on Au(111) (C6), and a NiO(111) layer grown on top of Ni(111). We examine both the scattered ion fragmentation patterns and the kinetic energy distribution spectra (KEDS) as a function of the incident ion energy E. Surface infrared and KEDS data indicate that we have prepared a saturated monolayer of hexanethiolate (C6) on Au(111) where the C6 carbon backbone is predominantly upright on the surface. C6 monolayers with a mixture of prone and upright C6 can also be prepared, but only the upright C6 monolayers are used for ion scattering experiments. The fragment ion distributions and the KEDS are then used to determine the channeling of the incident SiMe3+ion energy into the scattered ion internal energy Eint, and the scattered ion kinetic energy Escat. Overall, we find the order of Eint/E for SiMe3+ to be Au(111)≫NiO(111)>C6. From the Escat values, we find that MiSe3+ scattering off C6 is highly inelastic while scattering off Au is much more elastic. We use this information to calculate the energy (Esurf) which is channeled into the surface as a result of the collision for C6 and Au(111). By comparing the KEDS for the parent and fragment ions, we determine the importance of unimolecular dissociation off the various surfaces. In the accompanying paper, we present a series of classical dynamics simulations to explain and supplement these experimental results.

Experiments and classical dynamics simulations are performed to determine the energy transferred when C4H4S+, Si(CD3)3+, and Fe(C5H5)2+ collide with clean Au(111), hexanethiolate/Au(111) and various organic multilayer surfaces. For these various ion-surface pairs, the energy transferred to the surface Esurf can be described by a linear function of the incident energy E over the range E = 10–70 eV. For the hexanethiolate and organic multilayer surfaces, the slope of this line varies from 0.65 to 0.94 and indicates an efficient transfer of energy to the surface. For the clean Au(111) surface, Esurf is much smaller with slopes of 0.15 and 0.43. Classical dynamics simulations quantitatively duplicate the Esurf behavior for Si(CD3)3+ scattering off hexanethiolate and find that Esurf varies roughly with Ecos0.8(θi), where θi is the incident ion angle.

This paper presents classical dynamics simulations of Si(CD3)3+scattering from a hexanethiolate self-assembled monolayer on Au(111) and from a clean Au(111) surface. Simulations are performed with a united atom model using purely repulsive scattering potentials. These simulations predict the partitioning of the incident ion kinetic energy into the scattered ion kinetic energy and the internal modes of both the surface and the ion. For the organic surface, the simulations predict energy transfer to surface, ion internal, and ion kinetic energies of 0.78, 0.11, and 0.12 of the collision energy. The corresponding transfer efficiencies of 0.12, 0.21, and 0.65 were calculated for the Au(111) surface. These computational results compare well with the experimental results on the same systems which are reported in the preceding paper. The simulations predict near specular scattering for both surfaces. They also demonstrate that the ion penetrates only the topmost two to three layers of Me atoms of the organic surface and that it spends up to 250 fs in contact with the surface. Finally, these calculations determine the dependence of energy transfer on the incident ion angle.