A microfluidic device is described in which an electrospray interface to a mass spectrometer is integrated with a capillary electrophoresis channel, an injector and a protein digestion bed on a monolithic substrate. A large channel, 800 µm wide, 150 µm deep and 15 mm long, was created to act as a reactor bed for trypsin immobilized on 40–60 µm diameter beads. Separation was performed in channels etched 10 µm deep, 30 µm wide and about 45 mm long, feeding into a capillary, attached to the chip with a low dead volume coupling, that was 30 mm in length, with a 50 µm i.d. and 180 µm o.d. Sample was pumped through the reactor bed at flow rates between 0.5 and 60 µL/min. The application of this device for rapid digestion, separation and identification of proteins is demonstrated for melittin, cytochrome c and bovine serum albumin (BSA). The rate and efficiency of digestion was related to the flow rate of the substrate solution through the reactor bed. A flow rate of 1 or 0.5 µL/min was found adequate for complete consumption of cytochrome c or BSA, corresponding to a digestion time of 3–6 min at room temperature. Coverage of the amino acid sequence ranged from 92% for cytochrome c to 71% for BSA, with some missed cleavages observed. Melittin was consumed within 5 s. In contrast, a similar extent of digestion of melittin in a cuvet took 10–15 min. The kinetic limitations associated with the rapid digestion of low picomole levels of substrate were minimized using an integrated digestion bed with hydrodynamic flow to provide an increased ratio of trypsin to sample. This chip design thus provides a convenient platform for automated sample processing in proteomics applications. Copyright © 2000 John Wiley & Sons, Ltd.

A microchip structure etched on a glass substrate for packed column solid-phase extraction (SPE) and capillary electrochromatography (CEC) is described. A 200 microm long, octadecylsilane (ODS) packed column was secured using two different approaches: solvent lock or polymer entrapment. The former method was utilized for SPE while the latter approach was applied for CEC. In SPE, the ODS packed chamber gave a detection limit of 70 fM for a nonpolar BODIPY (493/503) dye when concentrated for 3 min at an electroosmotic flow rate of 4.14 nL/min, compared to 30 pM for this detector without the SPE step. SPE beds showed reproducible, linear calibration curves (R(2) = 0.9989) between 1 and 100 pM BODIPY at fixed preconcentration times. Breakthrough curves for the 330 pL (ODS-packed) bed indicated a capacity for BODIPY dye of 8.1 x 10(-14) mmol, or 0.25 mmol dye per liter of bed. The ODS-chamber could also be used to analyze dilute amino acid and peptide solutions. In the CEC format, two neutral dyes (BODIPY and acridine orange) were baseline-separated in an isocratic run with a theoretical plate count of 84 (420 000 plates/m) and a reduced plate height of about 1. A labeled peptide was also analyzed by CEC, using the acidic eluent (84% acetonitrile, and 26% aqueous trifluoroacetic acid (0.05%)) preferred for peptide separations on ODS-coated silica particles.

Introduction of a DNA interlayer adjacent to an Al cathode in a polymer light-emitting diode leads to lower turn-on voltages, higher luminance efficiencies, and characteristics comparable to those observed using a Ba electrode. The DNA serves to improve electron injection and also functions as a hole-blocking layer. The temporal characteristics of the devices are consistent with an interfacial dipole layer adjacent to the electrode being responsible for the reduction of the electron injection barrier.

be achieved at relatively lower fi elds. [ 3 ] However, this would require the design and the synthesis of new dielectric materials with higher relative permittivity and very low dielectric losses. In addition, because the energy density is proportional to E 2 , it would benefi t by charging the capacitor with as high of an electric fi eld as possible.Another approach, then, would be to use materials with higher barrier heights or electrodes with carefully selected work functions to minimize or retard the tunneling current, so that high energy densities, by charging the capacitors at high fi elds, can be achieved, without increasing the leakage current or reducing the effi ciency. But, again, this would require synthesis of new materials.We propose a novel approach to achieve higher energy densities by re-engineering the architecture of capacitors. Our new capacitor device is a layered structure that incorporates thin electron and hole blocking layers deposited between the conducting electrodes and the dielectric material as schematically illustrated in Figure 1 . [ 4 ] The purpose of these blocking layers is to achieve an "effective" increase in barrier height that will minimize or delay tunneling current until extremely high electric fi elds are reached, without the need to develop new dielectric materials. To explain our concept, let us fi rst restate Zhang and co-workers' observations in terms of the lowest unoccupied molecular orbital (LUMO) and the highest occupied molecular orbital (HOMO) levels. [ 3 ] Applying a positive bias voltage to the electron side of the capacitor raises the work function of the conducting electrode on the electron side, W fe . This reduces the energy gap ( E g ) between W fe and the LUMO level of the capacitor dielectric, thus reducing the barrier height. The higher the applied voltage, the smaller the E g , which in turn increases the fl ow of electrons over the barrier height or LUMO level of the capacitor dielectric. The insertion of an electron blocking layer (EBL), between the electrode and the dielectric fi lm, with a higher LUMO level than the capacitor dielectric ( Figures 1 and 2 ), Electrostatic capacitors offer high power density, lower loss, and higher operating voltage than their electrolytic and supercapacitor counterparts. However, these capacitors suffer from the low energy density (<2 J cm −3 ), limiting their applications in high power integrated systems such as pulsed power and high frequency inverters.In an electrostatic capacitor, the energy stored is in the form of an energy density in an electric fi eld. [ 1,2 ] Consequently, for linear dielectrics under an electric fi eld (dielectric constant independent of the applied fi eld), the stored energy density is linearly proportional to the dielectric constant ( ε ) and quadratically proportional to the applied electric fi eld ( E ). The level of the electric fi eld that can be applied, short of the catastrophic electrical failure (dielectric fi eld strength) of the material, may depend on either the inherent pro...