Solvent-pumped evaporation-driven concentration of an initial distribution of solutes on a porous substrate is considered in one and two dimensions. Approximate analytic solutions to the isotropic advection-dispersion equations are first found for a Gaussian kernel and an infinite domain, following the smoothed particle approximation. Analytic solutions for more general initial distributions are then found as sums of Gaussians, and comparison is made with numerical solutions. In each case, initial distributions are advected toward the stagnation point and concentrated. Two-dimensional analysis is then extended to describe anisotropy in permeability and diffusion, and hydrodynamic dispersion. Radial-flow experiments are performed using filter papers and water-soluble dyes. Diffusion coefficients, temperature and humidity profiles, and the evolution of spot distributions are measured. The results confirm minor anisotropy in permeability and diffusion, limited hydrodynamic dispersion, and largely uniform evaporation. Péclet numbers over 2500 are demonstrated. Evaporation-driven concentration provides a mechanism for solute transport over long timescales. Potential applications lie in the design of paper spray microanalytical devices operating by solvent pumping rather than capillary flow.

Paper-based evaporation concentrators with linear and radial geometries are compared. A new method of finding approximate analytic solutions of the advection-dispersion equation is proposed, based on the behavior of concentrators with infinite sources. Analytic approximations are compared with numerical solutions, and the advantage of radial concentration is highlighted: linear concentration rates scale with the square root of the Péclet number Pe while radial rates scale with Pe itself, leading to faster radial concentration beyond a critical value. Experiments are performed with Brilliant Blue FCF dye, using optical transmission and the Beer-Lambert law for quantitation. Dye concentrations are chosen for operation in the linear absorbance regime. Radial concentration is demonstrated under ambient conditions on filter paper disks with 60 mm diameter evaporation areas fed from a perimeter source, in a reverse of the well-known "coffee stain" experiment. Airflow enhanced concentration in strips and wedges is compared directly, using laser-patterned chromatography paper. The advantage of radial concentration is confirmed (and enhanced by diversion of concentrate to the corners of strips) and concentration factors greater than ∼ 500 (the dynamic range of measurement) are obtained in ∼2 h using 30 mm long columns.

This paper reviews mass spectrometers based on micro-electro-mechanical systems (MEMS) technology. The MEMS approach to integration is first briefly described, and the difficulties of miniaturizing mass spectrometers are outlined. MEMS components for ionization and mass filtering are then reviewed, together with additional components for ion detection, vacuum pressure measurement and pumping. Mass spectrometer systems containing MEMS sub-components are then described, applications for miniaturized and portable systems are discussed, and challenges and opportunities are presented.

Miniaturized mass spectrometers are becoming increasingly capable, enabling the development of many novel field and laboratory applications. However, to date, triple quadrupole tandem mass spectrometers, the workhorses of quantitative analysis, have not been significantly reduced in size. Here, the basis of a field-deployable triple quadrupole is described. The key development is a highly miniaturized ion optical assembly in which a sequence of six microengineered components is employed to generate ions at atmospheric pressure, provide a vacuum interface, effect ion guiding, and perform fragmentation and mass analysis. Despite its small dimensions, the collision cell efficiently fragments precursor ions and yields product ion spectra that are very similar to those recorded using conventional instruments. The miniature triple quadrupole has been used to detect thiabendazole, a common pesticide, in apples at a level of 10 ng/g.

RATIONALE:A recently developed miniature electrospray ionisation mass spectrometer has been coupled to a preparative flow chemistry system in order to monitor reactive intermediates and competing reaction paths, screen starting materials, and optimise reaction conditions. Although ideally suited to the application, mass spectrometers have rarely been used in this way, as traditional instruments are too bulky to be conveniently coupled to flow chemistry platforms.METHODS:A six-port switching valve fitted with a 5 μL loop was used to periodically sample the flow stream leaving the reactor coil. Mass spectra corresponding to the sample loop contents were observed approximately 10 s after activating the valve. High fluidic pressure was maintained throughout to ensure that gaseous products remained in solution. As an illustrative example of how this apparatus can be employed, the generation of benzyne and its subsequent reaction with furan were investigated. Benzyne was prepared via diazotisation of anthranilic acid using tert-butyl nitrite.RESULTS:Unexpectedly, the explosive diazotised intermediate was detected by the mass spectrometer at low coil temperatures or short residence times. The optimum reactor temperature and residence time for production of the desired Diels-Alder product are 50 °C and 3-5 min, respectively. There are competing reaction pathways leading to the formation of acridone and several other by-products.CONCLUSIONS:On-line mass spectrometry allowed the flow conditions to be quickly tuned for safe operation and optimal generation of the desired product. The validity of this approach was corroborated by off-line liquid chromatography/mass spectrometry (LC/MS) analysis of flow samples.

A miniature mass spectrometer capable of detecting analytes eluting from a high-performance liquid chromatography (HPLC) system is described and demonstrated for the first time. The entire instrument, including all pumps and the computer, is contained within a single enclosure that may be conveniently accommodated at the base of the HPLC stack. The microspray ion source, vacuum interface, ion guide, and quadrupole ion filter are all microengineered. These components are fabricated in batches using microelectromechanical systems (MEMS) techniques and considered to be consumables. When coupled to a standard HPLC system using an integrated passive split, the limit of detection for reserpine while scanning the full mass range is 5 ng on-column (1 pg of which is passed to the microspray). The mass range is m/z 100-800, and each spectrum is typically acquired at a rate of 1 scan per second.

Microfabricated quadrupole mass spectrometers with Brubaker prefilters are demonstrated for the first time. Complete filters are assembled from two dies, each carrying two pairs of rods providing the prefilter and main filter sections. The rods are held in precision silicon mounts that are fabricated using wafer-scale deep reactive-ion etching and anodic bonding to glass substrates. Improvements to ion transmission are obtained by tuning the bias potential applied to the prefilter. The effect is explained in terms of a simple analytic theory for ion motion in the prefilter. Mass filtering with a range of m/z = 0-1200 and a resolution of m/¿m ¿ 150 at 10% of peak height is demonstrated using 2-4-mm-long prefilter electrodes, 30-mm-long main electrodes (both of 650 ¿m diameter), and a radio-frequency drive at ¿6.5 MHz.

An electrospray-ionization mass spectrometer (ESI-MS) whose main components are all fabricated using silicon microelectromechanical systems (MEMS) techniques is demonstrated for the first time. The ion source consists of a microengineered alignment bench containing a V-groove mounting for a nanospray capillary, an ion-extraction electrode, and a pneumatic nebulizer. The vacuum interface consists of two plates, each carrying a 50-mu m-diameter capillary, that are selectively etched and bonded together to provide a differentially pumped internal cavity. The quadrupole filter consists of a microfabricated frame that provides mountings for stainless-steel rods measuring 650 mu m in diameter and 30 mm in length. Two different quadrupoles are compared: a first-generation bonded silicon device and a second-generation silicon-on-glass device with a Brubaker prefilter. Differential pumping of a MEMS component is demonstrated for the first time, atmospheric pressure ionization and ion transfer into vacuum are characterized, ESI-MS operation is demonstrated, and spectra are presented for a variety of compounds.

The limitations of conventional machining and assembly techniques require that designs for quadrupole mass analyzers with rod diameters less than a millimeter are not merely scale versions of larger instruments. We show how silicon planar processing techniques and microelectromechanical systems (MEMS) design concepts can be used to incorporate complex features into the construction of a miniature quadrupole mass filter chip that could not easily be achieved using other microengineering approaches. Three designs for the entrance and exit to the filter consistent with the chosen materials and techniques have been evaluated. The differences between these seemingly similar structures have a significant effect on the performance. Although one of the designs results in severe attenuation of transmission with increasing mass, the other two can be scanned to m/z=400 without any corruption of the mass spectrum. At m/z=219, the variation in the transmission of the three designs was found to be approximately four orders of magnitude. A maximum resolution of M/ΔM=87 at 10% peak height has been achieved at m/z=219 with a filter operated at 6 MHz and constructed using rods measuring (508±5) µm in diameter.

Recombination of D atoms resurfacing from the octahedral subsurface site on Ni(111) has been investigated as a model for the reaction of sub-surface D. The desorbing flux is made up of two components: translationally hot D2 formed in the reaction of energetic D atoms from the subsurface with chemisorbed D and a low energy component attributed to recombination of thermalised D. The D2 product shows no preferred scattering direction, consistent with a ‘hot precusor’ mechanism for reaction. Resurfacing D atoms scatter across the D/Ni(111) surface, losing any stereochemical memory of the resurfacing site before they react with chemisorbed D.

The adsorption of disilane (Si2H6) on D terminated Si(100) has been investigated using photodesorption to probe the coverage over a range of surface temperatures. Incident Si2H6 molecules can become molecularly (nondissociatively) chemisorbed after being trapped in a physisorbed precursor state. At temperatures higher than approximately 116 K, nearly all the physisorbed molecules desorb rather than chemisorb. Photofragmentation products are readily observed when chemisorbed Si2H6 is irradiated with 6.4 eV photons but surprisingly, none are observed in the case of physisorbed Si2H6. Consequently, Si can be deposited by reactions of the photofragmentation products with the surface, but only at temperatures lower than 116 K. The photodesorption cross section is also extremely dependent on the chemical nature of the Si(100) surface. It is high when the surface is terminated with a monodeuteride phase but much smaller when dideuteride species are present, suggesting that the presence of surface gap states is of importance. We consider how desorption might be initiated by electron attachment or by electron-hole recombination at the surface.

The recombination of surface and subsurface D atoms on Ni(111) has been studied using resonance-enhanced multiphoton ionisation (REMPI) to measure the internal state and translational energy distributions of the desorbing product. By detecting D2 formed during temperature-programmed desorption we were able to examine the reaction between subsurface and surface D atoms, and the recombination of two D atoms chemisorbed on the surface. Translational energy distributions for D2 formed by recombination of surface D are very sensitive to coverage. Desorption from a low coverage surface produced a translational energy release of 2.6 kT, but a thermal rotational distribution, reflecting an entrance channel barrier to dissociative chemisorption on the clean Ni(111) surface. Sticking probabilities predicted from detailed balance are consistent with molecular beam adsorption measurements. Desorption from D coverages above 0.5 ML resulted in a sub-thermal energy release, desorption being mediated by a molecular precursor state with D2 dissociation occurring via a non-activated, trapping-dissociation channel. In contrast, the reaction of subsurface D produces translationally hot D2, with a mean energy approaching 8 kTs at 180 K. This is consistent with the energetics for direct recombination of a chemisorbed D atom with a metastable subsurface D atom, which overcomes an activation barrier to resurface of between 0.35 and 0.47 eV depending on D concentration. The energy release decreases at higher temperature, probably as a result of a reduction in the energy of resurfacing D as the subsurface D concentration drops. This low energy component is attributed to accommodation of resurfacing D which is unable to react directly, followed by slow thermal desorption via the high coverage, surface D recombination channel. No internal rotational or vibration excitation was found in D2 formed by reaction of subsurface D.

The photochemistry of Si2H6 adsorbed on H terminated Si(100) has been investigated using the time-of-flight (TOF) technique and high resolution electron energy loss spectroscopy (HREELS). Intact Si2H6 desorbs via a photochemical mechanism during illumination with either 5.0 or 6.4 eV photons. Two cross sections differing by two orders of magnitude are required to describe photodesorption of the first weakly chemisorbed monolayer. It is likely that molecules adsorbed in this state adopt one of two orientations. We suggest that, as in the Antoniewcz model, the initial excitation involves temporary electron trapping and that the molecular orientation dependence of this process is reflected by the two cross sections. Photofragmentation is initiated by 6.4 eV photons but not by 5.0 eV photons and results in the desorption of mostly SiH4 and H2. TOF distributions corresponding to these two products are bimodal. The fast component corresponds to those fragments that are ejected with sufficient kinetic energy to overcome the desorption barrier directly whereas the slow component represents those fragments that become trapped and then thermally desorb. While the absence of fragmentation at a photon energy of 5.0 eV is consistent with the fact that the threshold for direct excitation is found to be at 5.2 eV, there is strong evidence that an indirect mechanism, again involving the formation of a temporary anion, is responsible for fragmentation. Specifically, HREELS shows that there are two resonances centered at 1.5 and 2.7 eV above the vacuum level. It appears that desorption can be initiated by electron trapping in either of these two resonances whereas fragmentation only occurs when a hot electron is trapped in the higher energy resonance. In common with photodesorption of intact Si2H6, photofragmentation is best described by two cross sections, although in this case they differ by a remarkable three orders of magnitude. There are at least two major Si2H6 fragmentation channels. It is proposed that one of these yields SiH4, Si and H2 when a short-lived electronic excitation causes the nuclei to accelerate along a reaction coordinate leading to these products whereas the other yields SiH4 and SiH2 as a result of randomized vibrational excitation.

A quasi-equilibrium model has been developed to describe the uptake kinetics of atomic hydrogen on Si(100) at 373 and 635 K. A new model is required because a simple consideration of only adsorption at dangling bond sites and Eley-Rideal abstraction cannot be reconciled with a saturation coverage of one monolayer (ML) at 635 K. We argue that although diffusion at low temperatures can be explained by hot precursor dynamics, slow vibrational relaxation in the chemisorption potential does not lead to an almost coverage independent sticking probability. Instead, we propose that a high sticking probability is maintained by reaction with doubly occupied dimers to give dihydride species which then migrate across the surface by an isomerization reaction. Subsequently, dihydride units either react with unoccupied dimer sites or molecular hydrogen is lost by desorption from two dihydride units (the β2 desorption channel). At 635 K the dihydride species may be regarded as a mobile chemisorbed precursor, although only the bonding arrangement is propagated by isomerization. When the atomic hydrogen source is turned off, the small steady-state dihydride coverage is rapidly lost to leave a saturated monohydride phase. A saturation coverage of 1.5 ML is obtained when abstraction and adsorption reactions are in dynamic equilibrium at 373 K. It is shown that the rate constant for abstraction from monohydride is essentially the same as for abstraction from dihydride. The measured initial rate constant for loss of adsorbed hydrogen during exposure to atomic deuterium is (1.8 ± 0.1) times larger than that for loss of adsorbed deuterium during atomic hydrogen exposure at 635 K. At this temperature, the initial rate constant for loss of deuterium during exposure to atomic hydrogen is found to be the same regardless of initial coverage, but the rate decreases more slowly than expected for first-order kinetics as the reaction proceeds. The uptake model also implies that a small amount of deuterium is lost as D2 and HD via the β2 thermal desorption channel at 635 K.

The mechanisms leading to desorption of molecular hydrogen from Si(100)-2 x 1 and Si(111)-7 x 7 surfaces have been elucidated and refined by detailed examination of the thermal desorption kinetics with particular emphasis on low and very low coverages. In the case of hydrogen desorption from Si(100)-2 x 1, a lattice-gas model incorporating the interactions that are responsible for pairing and clustering of adsorbed hydrogen atoms has been employed to fit temperature programmed desorption (TPD) peaks resulting from initial coverages between 0.01 and 1.0 monolayer (ML). From analysis of our low coverage data, we find that the pairing and clustering energies are (3.2 +/- 0.3) kcal mol(-1) and (3.4 +/- 0.5) kcal mol(-1), respectively. A subtle shift of the TPD peak maximum position as the initial coverage increases from 0.2 to 1.0 ML indicates that the pre-exponential factor and activation energy are weakly coverage dependent. me discuss how this is consistent with coupling of a dihydridelike transition state to its neighbors. The rate of molecular hydrogen desorption from Si(111)-7 x 7 is found to be very nearly second order in total hydrogen coverage when the initial coverage is low. This result is consistent with a two site model involving preferential adsorption of hydrogen atoms at rest atom sites rather than adatom sites. (C) 1998 American Institute of Physics.