Abstract
A method for optical transmission of a structure into a recording medium which can be transformed locally from a first undefined state into a second defined state by irradiating with photons from a photon source. The two states of the recording medium are manifested in different physical and/or chemical properties of the recording medium. At least one photon source having a photon flux of less than 10.sup.4 photons per second is selected for the irradiation with the photons. It was recognized that with such a low photon flux especially fine structures can advantageously be transmitted into the recording medium without the irradiation having to be partially blocked by a mask. In this manner, for a given wavelength (energy) of the photons, structures can be transmitted that are considerably smaller than the width, defined by the diffraction limit, of the probability distribution for the locations at which the emitted photons are incident.
Claims
1. A photolithography method for writing on a recording medium by optically imparting a structure to the recording medium, comprising irradiating only selected areas of the recording medium with photons from a plurality of photon sources configured to be actuated separately, the medium comprising a material having a physical or chemical property which changes upon irradiation with photons, the changes in the property constituting the structure imparted to the recording medium, and conducting the irradiation at a photon flux of less than 10.sup.4 photons per second, and wherein, for each photon source, at least one of frequency at which the photon source is operated and the photon source operation duration, x.sub.i, is determined as the solution to an equation system in which the photon dose D.sub.k with which the recording medium is irradiated at each location k is expressed as the sum of the contributions d.sub.ik(x.sub.i) which each photon source i makes to the photon dose D.sub.k.
2. The method according to claim 1, wherein the photon source is operated in working cycles in which it emits between 1 and 100 photons.
3. The method according to claim 1, wherein the photon source is set at a working distance of 1 m or less from the recording medium.
4. The method according to claim 1, wherein the photon source and the recording medium are moved relative to one another.
5. The method according to claim 1, wherein the recording medium is one in which the irradiated areas are transformed from an unwritten to the written state only above a predefined threshold dose of photons.
6. The method according to claim 1, wherein the recording medium is a photoresist for photolithography.
7. The method according to claim 1, wherein the photon sources are arranged in a matrix having a matrix spacing of 100 nm or less, preferably 50 nm or less.
8. The method according to claim 1, wherein in the equation system, the contributions d.sub.ik(x.sub.i) are expressed as the product of x.sub.i with the probability p.sub.ik that an individual photon emitted by the photon source i impinges onto or into the recording medium at the location k.
9. The method according claim 1, wherein the arrangement of photon sources is brought into n different positions relative to the recording medium, and D.sub.k is expressed as the sum of the contributions d.sub.ik(x.sub.i) which each photon source i makes to the photon dose D.sub.k at the position p=1, . . . , n.
10. The method according to claim 7, wherein the matrix spacing is 50 nm or less.
11. The method according to claim 1, wherein the selected areas are irradiated without use of a mask.
Description
BRIEF DESCRIPTION OF THE DRAWINGS
(1) FIG. 1 shows an exemplary embodiment of the method according to the invention.
(2) FIG. 2 shows a switching diagram for the separate actuation of individual photon sources in a matrix.
(3) FIGS. 3a and 3b show images, taken by electron microscopy, of a hexagonal arrangement of photon sources.
(4) FIGS. 4a and 4b show characterization of the light emission from the photon sources shown in FIG. 3.
(5) FIGS. 5a, 5b, 5c, and 5d show pyramid-shaped individual photon source.
(6) FIG. 6 shows integration of the individual photon source shown in FIG. 5 into a high-frequency structure.
(7) In the embodiments disclosed hereinbelow, the individual photon sources have a photon flux of less than 10.sup.4 photons per second.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
(8) FIG. 1 schematically shows an exemplary embodiment of the method according to the invention. The recording medium 2, to which a structure is to be imparted, is applied to a substrate 1. In order to impart the structure, a carrier substrate 3 having a plurality of individual photon sources 4 is brought into the vicinity of the recording medium 2. The recording medium 2 is then transformed from the unwritten to the written state only in the region of those sources 4 which are activated and emit light 5. Different structures can thus be written solely by way of a different actuation. In a simplified form of this exemplary embodiment, the individual photon sources 4 cannot be actuated separately but rather are either all activated or all not activated. The written structure is then defined by the two-dimensional arrangement of the sources 4 in the plane of the carrier substrate 3.
(9) FIG. 2 shows a switching diagram for the separate actuation of sources 4. A linear matrix of word lines W1, W2, . . . , Wn and a second linear matrix of bit lines B1, B2, . . . , Bn are provided. The two linear matrices together create a square grid. The sources 4 are arranged in the same periodicity as this square grid, wherein one source is connected to precisely one word line and precisely one bit line. When a voltage is applied between one word line and one bit line, precisely one source 4 is thus activated. The word lines and bit lines must not run in the same plane and must also not run in one plane with the sources 4. If the sources 4 are light-emitting diodes (LEDs), for example, in which the stack of n-contact, pn-transition and p-contact runs in the direction of the normal on the plane of the drawing, then for example the word lines may run below the n-contacts or in the plane of the n-contacts, so that each word line connects only n-contacts to one another. The bit lines may run above the p-contacts or in the plane of the p-contacts, so that each bit line connects only p-contacts to one another.
(10) In such a cross-bar array, between one word line and one bit line there is in principle just one current path through the source 4, which is switched directly between this word line and this bit line. Besides this direct path, there are also further parasitic paths through a plurality of further sources 4. However, each source on such a parasitic path is acted upon by a much lower voltage than the source that is switched directly between the actuated word line and the actuated bit line. The voltage on the sources on parasitic paths is lower than the bandgap-dependent minimum voltage that is required in order to cause an LED to illuminate. Therefore, only the source 4 that is switched directly between the actuated word line and the actuated bit line is caused to illuminate, even though there are parasitic paths through further sources.
(11) FIG. 3 shows images, taken by electron microscopy, of a hexagonal arrangement of sources 4. Sub-image b is a zoomed image of a portion of sub-image a. The sources 4 are LED structures, the common n-region of which is the n-doped GaN layer 4a on a sapphire substrate, this layer being shown in dark in FIG. 3. Located on the n-doped GaN layer 4a are etched columns of an undoped GaInN/GaN multilayer structure 4b, which each form a multi-quantum well (MQW). The MQW acts as an active medium which determines the wavelength of the photon source. On each MQW, p-doped GaN is grown as the p-region 4c.
(12) The arrangement was produced by first applying the multilayer structure 4b and the p-region 4c as flat layers to the GaN layer 4a. Photoresist was then structured by electron beam lithography so that the p-region 4c was exposed wherever a source 4 was to remain as a column. A nickel layer was applied and then the photoresist was removed by lifting off so that nickel 4d remained only on the p-regions 4c and was otherwise removed. The multilayer structure 4b and the p-region 4c were etched down to the GaN layer 4a wherever they were not protected by nickel.
(13) At the stage shown in FIG. 3, the arrangement still lacks the electrical contacting of the p-regions 4c to the outside world. This can be achieved by introducing an isolating material between the columns to such a height that only the regions 4c and 4d still protrude. One such suitable isolating material, for example, is hydrogen silsesquioxane (HSQ), which can be applied to the arrangement by spin coating and forms insulating SiO.sub.2 as the end product under the effect of heat. When the SiO.sub.2 layer is formed, the nickel is selectively removed and a transparent, electrically conductive contact layer composed of a nickel-gold alloy is applied, which connects the desired regions 4c to the outside world. N-contacts are produced by Ar ion beam etching of the SiO.sub.2 down to the n-GaN layer in the regions which were defined beforehand by lithographic processes. A metallization with Ti/Al/Ni/Au layers then takes place, these subsequently being alloyed.
(14) The arrangement can also be configured in such a way that the sources 4 can be addressed individually. To this end, the GaN layer 4a is structured in the form of word lines which each connect one row of columns to one another. Each column then stands in each case on precisely one word line, and between the word lines the GaN layer 4a is removed down to the sapphire substrate so that the word lines are isolated from one another. The transparent, electrically conductive contact layer is structured laterally in the form of bit lines, which in each case connect columns of regions 4c to one another and to the outside world. By applying a voltage between one word line and one bit line, precisely one source 4 can then be activated.
(15) FIG. 4a shows the spatial distribution of the photoluminescence which the layer system shown in FIG. 3 has emitted upon optical excitation with a light wavelength of 325 nm. The photoluminescence took place at a wavelength of 440 nm. The measurement of the photoluminescence is suitable as an intermediate test for quality control purposes during manufacture. If no photoluminescence shows, the arrangement produced is waste and the further process steps are superfluous.
(16) FIG. 4b shows the intensity distribution I of the electroluminescence over the energy E for the flat layer stack from which the arrangement shown in FIG. 3 was produced (prior to application of the photoresist (curve (i)), and for an individual source 4 (curve (ii)). The distribution has a full width at half maximum (FWHM) of 140 meV for curve (i) and a full width at half maximum of 100 meV for curve (ii). By reducing the spatial dimension of the source 4 from the flat layer to the column, the number of emitted photons per unit time is reduced so far that, given a sufficiently short pulse duration, the emission of individual photons can be brought about. The smaller the etched column, the higher the quantum confinement in the columns and the narrower the range of photon energies emitted by an individual photon source.
(17) FIG. 5 shows a further exemplary embodiment of an individual photon source 4 which is suitable for carrying out the method according to the invention. First, an n-doped GaN layer 52 is applied to a sapphire substrate 51 by a MOVPE process. An HSQ layer 53 is applied thereto by spin coating, said layer forming insulating SI).sub.2 as the end product under the effect of heat. A polymer polymethyl methacrylate PMMA is applied and serves as a positive resist for the electron beam lithography. Wherever the individual photon source 4 is to be located, the layer is exposed to an electron beam and then is selectively removed. The exposed SiO.sub.2 is removed by means of reactive ion etching. The PMMA is then removed. SiO.sub.2 is not only an insulating material but serves as a mask for the subsequent epitaxy. By a further MOVPE process, InN is selectively applied to the n-GaN through the opening created in the HSQ layer. The InN grows as a pyramid 54 out of the SiO.sub.2 mask. The pyramid is overgrown conformally by p-doped GaN 55. The p i n transition of the LED is formed between the p-doped GaN, the undoped InN and the n-doped GaN (FIG. 5a).
(18) A photoresist layer 56 is applied and is structured by lithography such that regions in which the SiO.sub.2 layer 53 is exposed are created to the left and to the right of the pyramid 54/55. In these regions, the SiO.sub.2 layer 53 and a region of the n-GaN layer 52 that is close to the surface are etched away by the argon ion etching illustrated by dashed arrows (FIG. 5b). The metals Ti/Al/Ni/Au are then applied, which in these regions, after alloying, form metal contacts 57 which connect the n-GaN layer to the outside world. The remaining metal is removed by lifting off the photoresist layer 56. A further photoresist layer 58 is applied and is structured laterally in such a way that the pyramid 54 is exposed (FIG. 5c).
(19) Metal is again applied, in this case Ni and Au, which surrounds the pyramid 54/55. The metal deposited on the photoresist layer 58 is removed by lifting off again. A metal contact 59 remains, via which the pyramid 54/55 can be actuated. The photon source is activated when a voltage is applied between the contacts 57 and 59.
(20) FIG. 6 shows images, taken by way of electron microscopy, of an individual photon source produced in this way, which is integrated in a high-frequency structure. Sub-images a-d show different zoom stages. The contact 59 is configured here as a metal tongue 59a, 59b which covers the (still visible) p-GaN-coated InN pyramid 54/55. In each case two of these tongues open into a macroscopic contact pad 59c for electrically connecting the pyramid 54/55 to the outside world. The contacts 57 likewise open into macroscopic contact pads 57a, 57b for connecting the n-GaN layer to the outside world. When a voltage is applied between the contact pads 59c and 57a, the source on the metal tongue 59a is activated.
