DE102011013262A1 - Electron-ionization source has photoelectrons that are generated from atmospheric-stable layer by ultra-violet source and are actuated with distance of acceleration in dimension of average free path length of gas - Google Patents

Abstract

The electron-ionization source has photoelectrons that are generated from an atmospheric-stable layer by an ultra-violet source (4) and are actuated with a distance of acceleration in the dimension of the average free path length of the gas. The layer emitting for the photoeletrons is made of cerium hexaboride, lanthanum hexaboride, strontium titanate, silicon dioxide, barium titanate, zinc oxide, tungsten trioxide, tin oxide or mixture of yttrium, samarium and thorium. The layer made from cobalt silicide or nickel silicide is inserted under the borides.

Description

Summary

The present invention describes an ionizer for the detection of trace gases for variable pressures up to atmospheric pressure and its preparation. He uses photoelectrons whose energy can be variably adjusted to ionization energies to obtain adjustable sensitivity. An electrical acceleration path is adapted to the middle of the molecules as well as to the free path lengths of the ionization, in order to avoid energy losses due to shocks during the Bescheunigung.

introduction

For chemical and biochemical analysis systems, the ionization of molecules with sum detection (photoionization detection, PID), separation of ions by mass / charge (mass spectrometry, MS) and drift behavior in the electric field under gas-fluid countercurrent (ion mobility spectrometry, IMS) is used.

MS is the most accurate and quantitatively reliable method, but requires high vacuum (HV), whereas IMS and PID are only subtractive, but semiquantitative, respectively, but provide only qualitative differences, but can be used at normal pressure , The aim of all processes is always high separation capacity (partly by coupling with chromatography-like or thermodesorptive precursors) and low equipment costs such as the elimination of high-vacuum pumping systems. One of the main considerations is therefore the ionizer, for the adjustable sensitivity and selectivity - if possible via model-dependent parameters - and for the overall system, the safe operation and portability are desired.

With regard to the model proximity, MS is advantageous in itself, since some of the molecules are led to the analyzer immediately after ionization. However, higher sensitivities also result if, as in the IMS, ions interact with one another, eg. B. in specially provided reaction chambers, from which analysis areas z. B. be fed by switching grid. However, many possible pathways force, if possible, the establishment of more complex and experiential models.

State of the art

The ionization of analyte molecules is effected either by bombardment with electrons or by interaction with ionizing electromagnetic radiation. The former are described in the Lafferty, No. 856,666, materials for thermionic emission, but also as part of any electron-emitting source of vacuum. These materials consist of a boride of calcium, barium, strontium, thorium, uranium or mixtures thereof or rare earths such. Lanthanum and cerium. A special combination of hexaborides of the form (CeB ₆ ) _1-xy (SrB ₆ ) _x (Ba _{6 6} ) _y are in the patent DT 2442 508 described.

For the ionization electron sources are known on different physical principles of action. The cathode emission is mostly based on field emission, photoemission, thermionic emission or their combination.

Incandescent emitters are among others in the patent U.S. Patent 2,639,399 . EP 0316 196 described and widely used. This z. B. tungsten or Lab6 directly heated. Indirectly heated LaB6 bars are in the U.S. Patent 2,639,399 described. These electron sources require an enormous temperature (T> 1000 ° C) and have the disadvantage of high heat output coupled with a limited lifetime. To improve the life in the patent DT 2553047 smallest LaB6 beads are proposed directly on a heater structure, but for a working temperature between 1500-1600 ° C.

With electron impact sources also variable energies can be adjusted. However, their annealing emitters require protection by high vacuum: In an electrostatic acceleration unit manufactured by precision engineering, the electrons are brought to about 70 eV and pass into a downstream reaction chamber. The acceleration arrangement works because the mean free path of the electrons with respect to the collisions with gas molecules (or atoms) under high vacuum conditions in the centimeter range and the electrons can be brought to their final energy in the electrostatic field without a significant proportion of them on the way loses its energy through gas shocks.

Electron or ion sources operating at lower temperatures or higher pressures can be realized by field emitters. The field emission in vacuum is described by the Fowler-Nordheim equation in which the work function of the emitter materials is taken as the determining factor. For Si, W, Mo or Au with a work function of about 4.5 eV, field strengths of at least 10 ⁶ V / cm result, which require peak rounding in the nanometer range. HH Busta, "Vacuum Microelectronics", J. Microchem. Microeng., Vol. 2, p. 43-74, 1992 describes a variety of different field emitters. Other forms and Manufacturing variants are in the patents DE 2816832 . DE 60013521 and EP 0790633 . DE 10 2007 012 97 disclosed. Although field emitters do not require heating power, their stability in high pressure operation is critical. On the one hand, the peaks become very hot due to the high surface charge currents, which leads to oxidation or changes in shape. On the other hand, at higher electrical voltages at the tip surrounding gas atoms, especially the main constituents such as oxygen or nitrogen, ionized and move back depending on the charge to the top, which then damage them by sputtering.

For atmospheric use z. B. in IM5 radioactive films of nickel isotope ⁵³ Ni are often used, the β-particle energies reach up to 67 keV and thus achieve ranges of a few cm in normal air. The energies are very sharply distributed but far exceed those of the single primary ionizations of known substances. If radioactive sources are not desired, alternative methods such as the atmospheric pressure chemical ionization source (APCI) are available. DE 033993 . DE 049350 as well as other corona and corona spray discharges.

Characteristic of the above mentioned methods is their blurring with respect to accurate energies in the range of less eV, whereby the possibility is denied to selectively analyte molecules and suppress carrier gases or normal air constituents. Individual molecular groups could otherwise be selectively detected with respect to the minimum necessary ionization energies and possibly aromatics and amines of alcohols and alkanes are distinguished and these in turn of oxygen, water, CO and CO2.

This succeeds better with photoionization via gas discharge lamps or non-resonant laser processes in the energy range of the first ionization energies (cf. US 5338931 . DE 19609582 ), which by exchange or switchable operation would allow a variation of the maximum ionization energies in the range between 9 eV to 13 eV. A recent method of resonant laser two-photon absorption or ionization with photoelectrons of adjustable energy from gold layers increases this variability.

The disadvantage here, however, the high equipment complexity, which can rule out at least portability in the foreseeable future.

Photoionization detectors have no drift path and measure only the generated total current. Decisive for their design are low power consumption and miniaturized size, since they are mostly used in portable devices. The "Pen-Ray" or "PID" lamps used there are UV sources with 8.4 to 11.8 eV photon energy and address the important analyte groups of organic molecules. Their high cost price and short life, especially for 11.8 eV in miniaturized designs are currently opposed to a further spread of PID. The currently developed UV LEDs with cut-off wavelengths of up to 255 nm (4.5 eV) are still unsuitable for these applications.

Another basis for the invention below is the photoemission. The patent EP 0642 147 describes a structure for the photoemission. Photons of energy (hv in electron volts eV) are absorbed in the photoabsorption layer to excite photoelectrons. These are accelerated by an internal internal field and can go into vacuum through an external voltage. However, the materials used limit the use to vacuum or high vacuum, which will corrode immediately upon exposure to air or moisture.

The combination of photoemission and field emission over metal and semiconductor field emitter tips was developed by H. Neumann, "The theory of photo-field emission from metals" Physica 44, p. 587, 1971 and H. Neumann, Ch. Kleint "Field Enhanced Photoemission from Tungsten", Ann. Physics 27, 5, 237, 1971 , shown. Field emitter tips with small tip radii were used. This combination has also been used for other materials such as LaB6 S. Mogren, R. Reifenberger, "Field emission and photofield emission energy distribution from LaB6", Surf. Science, 254, p. 169, 1991 , examined. The field emitter or the photofield emitter are technologically difficult to produce due to the necessary small tip radius. The materials used require conditioning at higher temperatures under vacuum, which must be maintained for the entire operating time.

For the requirements discussed above, therefore, a device for energy-variable ionization of gases is sought, can be operated at variable pressures as directly as possible in air.

In this way, the most diverse groups of organic molecules can be successively "switched on" with one source up to the air components in a charge flow measurement. The necessary sharpness of the energy distribution should then be in the range of distinction of these groups of a few electron volts.

The new solution consists of three building blocks: First, an array of electrons in length dimensions in the range of the free path length to the desired final energies can accelerate with sharp energy distribution. Secondly, a teaching on materials that can be used better than gold and without conditioning or vacuum to generate photoelectrons, and third, a series of technology steps and structural designs to produce a new electron emitter as an ionization source.

First, it is learned from the example of the electron impact source that the energy uncertainty of the electrons decreases when the dimensions of the mechanical structure make a gas impact of the electrons during their acceleration likely. The most probable path length which the electrons traverse between the collisions is referred to below as the mean free path of the electrons. It decreases linearly with pressure and is also energy dependent, with the largest cross sections for lossy shocks resulting in energy losses at energies close to the first ionization energies of the struck gases. The newly developed solution now results from the fact that electrons must be brought to their final energy on the length dimensions below or at least in the same order of magnitude as the mean free path of the adjacent gases. This minimum length, representative of nitrogen at ambient conditions, is about 110 nm. The solution therefore contains an acceleration path for electrons of this dimension, which can advantageously be realized with a surface micromechanical microstructure.

On the other hand, free electrons must be generated. If thermal emission ceases to be an option due to the high pressures sought, the field emission or photoemission would be absent. High-performance field emitters have long been known in electron microscopy, but may only be operated under very good high-vacuum conditions (generally 10-7 mbar). In addition, nano-field emitters are known in the scientific literature, for. Based on GaN, diamond and silicon or based on carbon nanotubes (CNT). So far, however, such devices have found no input in practice, because they are accelerated for various reasons such as the backsputter effects by positively ionized gas molecules that are against the direction of electron movement towards the emitting cathode or the formation of natural oxides upon exposure to ambient air to etched structures (shutdown and Pressure rise in vacuum equipment) not reliable and without elaborate conditioning work.

Also in the photoemission, the choice of materials is limited because of the formation of natural passivation layers. The use of gold, which does not form a stable oxide layer, but with its work function of 4.5 eV, requires corresponding UV sources such as the gas discharge lamps mentioned. If, on the other hand, UV LEDs with lower photon energies are used, corresponding materials must be found again.

The persecuted basic idea shows : A thin photoelectron-emitting layer ( 2 ) is applied to a substrate ( 1 ) applied. Above the emissive layer, a perforated accelerating electrode is arranged on the gas side and electrically isolated from the emissive layer, which is at a suitable potential U _A. The distance between the emissive layer and the acceleration electrode should be of the same order of magnitude or less than the mean free path d _{mf of} the gas molecules in the analyte space (FIG. 5 ) be. In this arrangement, the electrons receive a sharp energy of e. U _A , which varies only by the scattering of the work function of the emitter between Φ _min and Φ _max , as far as it remains smaller than the cut-off wavelength energy. The photoemission is in this arrangement by a UV source ( 4 ), preferably with a radiation wavelength energy just above the required work function of the material. In the preferred arrangement, the light source ( 4 ) outside the analyte space and irradiates the substrate ( 1 ) which is transmissively selected for the wavelength. Since the preferred radiant energies are in the range 2.8-4.3 eV, quartz substrates meet this requirement. Alternatively, the UV source can also be used in the analyte room ( 4a ) or through a further window through the analyte space. In this arrangement, the substrate ( 1 ) and the emitting material ( 2 ) as a massive body z. B. be performed in ceramic, sintered compact or a common microstructure substrate such as silicon.

Suitable photoemissive materials are those with sufficient electrical conductivity, which - in addition to the group of noble metals - do not oxidize in the air or react with atmospheric moisture. These are, above all, the group of metal oxides, since these are "already oxidized". In addition, there are largely oxidation-stable compounds similar to the lanthanum boride. The following materials are suitable for their work function or cut-off wavelength (Φ [eV] or hν [nm]: The lowest values show strontium oxide (2.0 eV) when used in moderate humidities, but since it tends to form hydroxide, it is recommended in its place the use of strontium titanate, but which shows facet dependent values between 2.8 to 5.2 eV For barium titanate, the same values and resistances apply.Higher but comparatively more stable values are obtained with the choice of borides of CeB6 and Lanthan LaB6 reached (2.5 and 2.66 eV, respectively) .The diffusion of boron into the underlying substrates can be successfully achieved by introduction Thinnest cobalt and Nickelsilizidschichten be prevented as a diffusion barrier layer, which can achieve stable emission behavior. On <100> silicon, thin CoSi ₂ also gives preference to the low work function LaB6 facet. From the group of pure metal oxides, zinc oxide ZnO (3.2-3.7 eV), tungsten oxide WO3 (4.2-5.3 eV) and tin oxide SnO2 (4.3 eV) are suitable with increasing work function values, the two former Materials during the deposition of the coating vary more synthetically (facet or texture formation).

Special consideration is given to those materials whose oxide work function is close to that of the pure surface. The elements samarium, yttrium and thorium are suitable in terms of energy and moisture stability from the lanthanides and actinides.

je nach Leitfähigkeit des verwendeten Materials. 1

1

Sea 79 in Göpel

Bei Lanthanborid liegt die optische Absorptionslänge d bei Wellenlängen von 300–350 nm in derselben Größenordnung (ca. 43

), so dass die optimalen Quantenwirkungsgrade bei Dicken von 30–40

erreicht sind.Is the backside radiation of these materials after ( 4 ), the layer thicknesses are to be optimized with regard to high absorption and the range of the electrons in solids. These are around 10 eV for energies at 25-35

depending on the conductivity of the material used. 1

1

Sea 79 in Göpel

In the case of lanthanum boride, the optical absorption length d at wavelengths of 300-350 nm is of the same order of magnitude (approx

), so that the optimal quantum efficiencies at thicknesses of 30-40

are reached.

As a third solution part, the acceleration structure is to be designed under the framework of mean free path lengths of the molecules of about 100 nm under normal conditions. A closer look leads to the ionization lengths, which must also be taken into account. These are, as described above, energy dependent and gas specific. For ethene, for example, the ionization length (100%) is 10 μm for 1 bar and 14 eV electron energy. For these energies, the ionization probability also decreases with reduced energy. In comparison of average free and ionization length it is clear that the choice of the mean free path or a few multiples thereof as the reference length for the distance of the accelerating electrode from the emitter layer d _Acc, this is sufficiently small dimensioned to a reliable acceleration of the electrons to the desired final energies to reach from some to about 20 eV.

The entire acceleration structure can be generated in surface micromechanical technique, such as as an embodiment shows. For this purpose, a suitable quartz substrate, z. B. Lithosil, a thin Lanthanborid- or emitter layer ( 2 ) and then with an insulator ( 6 ) of thickness d _precipitate , preferably of SiO 2, in one, coated by low-temperature deposition method for thin films such as PECVD. The advantage of low deposition temperatures is the protection of the emissive layers from recrystallization and usually increased oxide formation. After the insulator layer, a conductive layer ( 3a ), which as a later acceleration electrode ( 3 ) serves. The conductive layer ( 3a ) is now structured, thereby partially exposing the underlying insulator material. This is then removed by etching until the surface of the emitter material under the openings in ( 3 ) is exposed. In an advantageous embodiment, SiO 2 is removed as an insulator by means of buffered hydrofluoric acid by wet-chemical means. The etching stops on Lanthanhexaborid, which has a sufficient etch resistance to rf acid (HF) and thus protects the underlying quartz glass. The electrode layer is then made of a precious metal z. B. made of platinum or gold.

With regard to the surface currents between emitter and acceleration layer, it is also advantageous to leave the insulator regions, which serve as spacers after the etching, with the smallest possible circumference. For this purpose, the perforation of the conductive layer ( 3a ) partially omitted, as in was drawn. When using the strongly corrosive HF forms such a column structure. The width of the distance between these columns is then limited by their mechanical deflection under the action of the electric field of U _A.

Another embodiment, not shown here, provides for the repeated deposition of the emissive layer after the pillar structure has been produced. The holes in the conductive layer then act as a shadow mask for deposition. This method is suitable for emissive layers, which are otherwise attacked by the etching process.

In addition to the parasitic surface currents, possible field emission currents from sharp structural edges are to be avoided. In the embodiment according to is done on the insulating layer ( 6 ) a photoresist is structured in the negative image of the later electrode layer ( ). The exposed isolation areas are first etched isotropically, resulting in round pits. These are as undirected as possible with the precious metal ( 3a ), which also covers the photoresist strips ( ). The photoresist and the precious metal thereon are removed by Lift Off. The resulting structure shows , It is rounded down and shows due to the protective oxide, a higher electrical breakdown strength than air. Upwards, however, may be the result of demolition tips. These are planarized by Chemo Mechanical Polishing (CMP). The entire structure is subsequently using Reactive Ion Etching (RIE) etched down to the emitter layer. shows the finished structure, which has no edges on the side which faces the emitter layer. The sharp transitions above do not exert excessive influence on nearby cylinders. Particularly advantageous is the last deep etching step using the anisotropic DRIE process when it is terminated after a directional soil-exposing oxygen step. In this case, sidewall passivation of fluorinated hydrocarbons is insulating and hinders moisture-dependent surface conductivity.

The surface conductivity is further reduced during operation of the device when it is equipped with a heater to reduce adsorbed water films or passively heated.

With regard to the dielectric strength of the free electron sources, the following can be achieved: With dBeschl. = 100 nm and a field emission limit of about 0.6 V / nm theoretically result in 60 volts maximum acceleration voltage. For aspired UA up to 14 volts (CO) this is sufficient.

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• DT 2442508 [0005]
• US 2639399 [0007, 0007]
• EP 0316196 [0007]
• DT 2553047 [0007]
• DE 2816832 [0009]
• DE 60013521 [0009]
• EP 0790633 [0009]
• DE 10200701297 [0009]
• DE 033993 [0010]
• DE 049350 [0010]
• US 5338931 [0012]
• DE 19609582 [0012]
• EP 0642147 [0015]

Cited non-patent literature

• HH Busta, "Vacuum Microelectronics", J. Microchem. Microeng., Vol. 2, p. 43-74, 1992 [0009]
• H. Neumann, "Theory of Photofield Emission from Metals" Physica 44, p. 587, 1971 [0016]
• H. Neumann, Ch. Kleint "Field Enhanced Photoemission from Tungsten", Ann. Physics 27, 5, 237, 1971 [0016]
• S. Mogren, R. Reifenberger, "Field emission and photofield emission energy distribution from LaB6", Surf. Science, 254, p. 169, 1991 [0016]

Claims (13)

Electron ionization source, characterized in that photoelectrons a) are generated from an atmosphere-stable layer by a UV source and b) are brought to the ionization energies of the gases with an acceleration distance in the order of the mean free path of the gases or less. Electron ionization source, characterized in that for the photoelectron emitting layer layer of CeB6 and LaB6, SrTiO3, SrO2, BaTiO3, ZnO, WO3, SnO2 or Y, Sm, Y Th or Th or mixtures thereof containing the same. Electron ionization source for gases according to 2), characterized in that a layer of CoSi2 or NiSi is introduced under said borides. Electron ionization source according to 1), characterized in that the acceleration path consists of an emissive layer on which an insulator and a conductive layer are applied and in these two layers holes which extend to the emissive layer are structured 4), a portion of the insulator material is removed below the patterned conductive layer. 4) characterized by edge rounding at the conductive layer. 4), characterized in that the edges of the conductive layer in the area of the holes are rounded by introducing them into recesses of the insulator. 4), characterized in that a fluorine-containing layer is carried by sidewalls of the insulator in the region of the holes. 1), characterized in that it has a heating of the emissive layer. 4) characterized in that, for the production thereof, the emissive layer acts as an etch stop with respect to an underlying substrate. 4) characterized in that the emissive layer is applied or reinforced after the etching step by the shadow mask of the conductive layer. Electron ionization source according to 1), characterized in that the UV source is an LED. 1) characterized in that the emissive layer is formed on an optically transparent material having a thickness in the range of the absorption length of the light and irradiates the UV source back into the ionizer.

Images