On the convergence of the affine-scaling algorithm

The affine-scaling algorithm, first proposed by Dikin, is presently enjoying great popularity as a potentially effective means of solving linear programs. An outstanding question about this algorithm concerns its convergence in the presence of degeneracy. In this paper, we give new convergence results for this algorithm that do not require any non-degeneracy assumption on the problem. In particular, we show that if the stepsize choice of either Dikin or Barnes or Vanderbei, et al. is used, then the algorithm generates iterates that converge at least linearly with a convergence ratio of , wheren is the number of variables and (0,1] is a certain stepsize ratio. For one particular stepsize choice which is an extension of that of Barnes, we show that the cost of the limit point is within O(/(1–)) of the optimal cost and, for sufficiently small (roughly, proportional to how close the cost of the nonoptimal vertices are to the optimal cost), is exactly optimal. We prove the latter result by using an unusual proof technique, that of analyzing the ergodic convergence of the corresponding dual vectors. For the special case of network flow problems with integer data, we show that it suffices to take = 1/(6mC), wherem is the number of constraints andC is the sum of the cost coefficients, to attain exact optimality.This research is partially supported by the U.S. Army Research Office, contract DAAL03-86-K-0171 (Center for Intelligent Control Systems), by the National Science Foundation, grant NSF-ECS-8519058, and by the Science and Engineering Research Board of McMaster University.     

Comparison of particle size and flow rate optimization for chromatography using one-monomer molecularly imprinted polymers versus traditional non-covalent molecularly imprinted polymers

The dependence of enantio-selective chromatographic performance on particle size, as measured by separation factor, was investigated for one-monomer molecularly imprinted polymers (OMNiMIPs) compared to traditionally formed EGDMA/MAA molecularly imprinted polymers (MIPs). Five particle size ranges were compared (<20 μm, 20-25 μm, 25-38 μm, 38-45 μm, and 45-63 μm), revealing that the particle sizes above 25 μm provided the highest separation factor, and thus the best enantiomer separation, for both imprinted polymers. Other chromatographic parameters such as the number of theoretical plates and resolution exhibited only minor changes for the OMNiMIPs as the particle size changed, except for particles 20 μm and below. However, the number of theoretical plates and resolution for EGDMA/MAA are higher for particles in the 20-25 μm range. Thus, chromatographic factors for the EGDMA/MAA polymers are better in this range, despite better enantioselectivity for particle sizes above 25 μm. In contrast, OMNiMIPs generally show the most favorable performance for particle sizes in the 38-45 μm range. It was also found that decreasing flow rate resulted in improved enantioselectivity for both MIPs for all particle sizes.     

Uridine binding by Zn(II) macrocyclic complexes: diversion of RNA cleavage catalysts

Zn(II) complexes of 1-oxa-4,7,10-triazacyclododecane (12[ane]N3O), 1,5,9-triazacyclododecane (12[ane]N3), and 1-hydroxyethyl-1,4,7-triazacyclononane (9[ane]N3OH) promote cleavage of the RNA analogue, 2-hydroxypropyl-4-nitrophenyl phosphate (HpPNP) at pH 8.0, I=0.10 M (NaCl), 25 degrees C with second-order rate constants of 8.9x10(-3), 9.0x10(-3), and 3.3x10(-3) M-1 s-1, respectively. Cleavage of HpPNP by these catalysts is inhibited by uridine with inhibition constants (Ki) of 1.2, 0.46, and 45 mM, respectively, under these conditions. Binding constants derived from these inhibition constants are 2-200-fold larger than those for binding of related Zn(II) complexes to phosphate diesters under similar conditions, suggesting that uridine sequences in RNA will inhibit Zn(II)-catalyzed cleavage by competing with phosphate diester binding sites. Further studies are carried out that utilize pH-potentiometric titrations to monitor uridine binding to five Zn(II) macrocyclic complexes in aqueous solution at 25 degrees C, I=0.10 M (NaCl). The data are consistent with binding of the Zn(II) complexes to the N3-deprotonated form of uridine to give log KU.-values of 5.29, 4.57, 4.56, 3.47, and 2.65 for the Zn(II) complexes of 12[ane]N3, 12[ane]N4, 12[ane]N3O, 15[ane]N3O2, and 9[ane]N3OH, respectively (12[ane]N4=1,4,7,10-tetraazacyclododecane, 15[ane]N3O2=1,4-dioxa-7,10,13-triazacyclopentadecane). For the five Zn(II) complexes studied, there is a linear relationship between uridine anion binding constants and hydroxide binding constants.     

Electrically actuated, pressure-driven microfluidic pumps

In order to make the lab-on-a-chip concept a reality, it is desirable to have an integrated component capable of pumping fluids through microchannels. We have developed novel, electrically actuated micropumps and have integrated them with microfluidic systems. These devices utilize the build-up of electrolysis gases to achieve pressure-driven pumping, only require small voltages (approximately 10 V), and have approximate dimensions of 5 cm x 3 cm x 2 cm. Furthermore, these micropumps are composed of relatively inexpensive materials, and the reversible sealability of their poly(dimethylsiloxane) body to different microfluidic arrays enables repeated uses of the same pump. Under an applied potential of 10 V, three different micropumps had average flow rates of 8-13 microL min(-1) for water being pumped through five different 2 cm-long, 5500 microm(2) cross-sectional-area channels in poly(methyl methacrylate), in approximate agreement with predicted pump rates. We have also evaluated pump operation at the lower applied potential of 8 V and observed an average flow rate of 6.1 microL min(-1) for a pump-channel system. The current micropump design is capable of sustaining pumping pressures in the range of 300 kPa. The various advantages of these micropumps make them well suited for use in lab-on-a-chip analysis techniques.     

Six‐coordinate Co2+ with imidazole,NH3, and H2O ligands: Approaching spin crossover

Octahedral, six‐coordinate Co²⁺ can exist in two spin states: S = 3/2 and S = 1/2. The difference in energy between high spin (S = 3/2) and low spin (S = 1/2) is dependent on both the ligand mix and coordination stereochemistry. B3LYP calculations on combinations of neutral imidazole, NH₃, and H₂O ligands show that low‐spin isomers are stabilized by axial H₂O ligands and in structures that also include trans pairs of equatorial NH₃ and protonated imidazole ligands, spin crossover structures are predicted from spin state energy differences. Occupied Co d orbitals from the DFT calculations provide a means of estimating effective ligand strength for homoleptic and mixed ligand combinations. These calculations suggest that in a labile biological system, a spin crossover environment can be created. © 2006 Wiley Periodicals, Inc. Int J Quantum Chem, 2007     

Determination of fluoride in wine by fluoride selective ion electrode, standard addition method: collaborative study

The accuracy, precision, and reproducibility of a rapid method for determination of fluoride in wine, using a fluoride selective ion electrode, were established by a collaborative study involving 12 laboratories, 5 in Europe and 7 in the United States. The laboratories assayed 6 Youden pairs of fluoride-fortified, red and white wine samples with fluoride concentrations ranging from 0.2 to 3.0 mg/L. The relative standard deviations of repeatability ranged from 1.94 to 4.88%; relative standard deviations of reproducibility ranged from 4.15 to 18.40%. HORRAT values ranged from 0.30 to 0.97. The average recovery was 99.97%. Based on the statistical results of this collaborative study, the Study Director recommends that this method be adopted First Action.     

Hafnium oxide and zirconium oxide atomic layer deposition: initial precursor and potential side-reaction product pathways with H/Si(100)-2 x 1

Hybrid density functional calculations have been carried out using cluster models of the H/Si(100)-2 x 1 surface to investigate the mechanistic details of the initial surface reactions occurring in the atomic layer deposition of hafnium and zirconium oxides (HfO2 and ZrO2). Reaction pathways involving the metal precursors ZrCl4, Zr(CH3)4, HfCl4, and Hf(CH3)4 have been examined. Pathways leading to the formation of a Zr-Si or Hf-Si linkage show a significant sensitivity to the identity of the leaving group, with chloride loss reactions being both kinetically and thermodynamically less favorable than reactions leading to the loss of a methyl group. The energetics of the Zr(CH3)4 and Hf(CH3)4 reactions are similar with an overall exothermicity of 0.3-0.4 eV and a classical barrier height of 1.1-1.2 eV. For the reaction between H2O and the H/Si(100)-2 x 1 surface, the activation energy and overall reaction enthalpy are 1.6 and -0.8 eV, respectively. Due to contamination, trace amounts of H2O may be encountered by metal precursors, leading to the formation of minor species that can lead to unanticipated side-reaction pathways. Such gas-phase reactions between the halogenated and alkylated metal precursors and H2O are exothermic with small or no reaction barriers, allowing for the possibility of metal precursor hydroxylation before the H/Si surface is encountered. Of the contaminant surface reaction pathways, the most kinetically favorable corresponds to the surface -OH deposition. Interestingly, for the hydroxylated metal precursors, a unique reaction pathway resulting in the direct formation of Si-O-Zr and Si-O-Hf linkages has been identified and found to be the most thermodynamically stable pathway available, being exothermic by approximately 1.0 eV.     

Manganese detection in marine sediments: anodic vs. cathodic stripping voltammetry

Three different electroanalytical techniques for the detection of manganese in marine sediments are evaluated. The anodic stripping voltammetry of manganese at an in situ bismuth-film-modified boron-doped diamond electrode and cathodic stripping voltammetry at a carbon paste electrode are shown to lack the required sensitivity and reproducibility whereas cathodic stripping voltammetry at a bare boron-doped diamond electrode is shown to be reliable and selective with a limit of detection, from applying a 60 s accumulation period of 7.4 × 10⁻⁷ M and a sensitivity of 0.24 A M⁻¹. The method was used to evaluate the manganese content of marine sediments taken from Šibenik, Croatia.     

Evaluation of minimum detectable amounts of iodine in biological samples by neutron activation analysis

Evaluation of minimal detectable amounts (MDA) of iodine in various or specified irradiation and measurement conditions is described. Interfering contributions due to Na and Cl to the spectral background as well as the dead time of detection system were calculated in details based on an experimental approach. The MDA of iodine for some important biological samples with various Na and Cl concentrations can be evaluated and predicted according to the specified irradiation and measurement arrangements to meet the requirement of analytical purposes.