EP1935035A2 - Production of self-organized pin-type nanostructures, and the rather extensive applications thereof - Google Patents

Abstract

The invention relates to methods and devices comprising a nanostructure (2; 4, 4a) for improving the optical behavior of components and apparatuses and/or improving the behavior of sensors by increasing the active surface area. The nanostructure (2) is produced by means of a special RIE etching process, can be modified regarding the composition of the materials thereof, and can be provided with adequate coatings. The amount of material used for the base layer (3) can be reduced by supplying a buffer layer (406). Many applications are disclosed.

Description

 Self-organized needle-like nano-structures in their applications

Field of the invention (s).

The invention relates generally to the production of structured surfaces, and more particularly to the production of needle-like structures having nano-dimensions in the range, for example, below the wavelengths of visible light

Structures may be referred to as nanostructures. Your options and technical applications are in the foreground here.

State of the art.

Many electronic, opto-electronic, sensory and micromechanical components have silicon as part of doped and / or undoped, crystalline and / or polycrystalline and / or amorphous form. In order to meet the specific component-specific requirements, therefore, silicon must usually be processed accordingly, which often includes a structuring of the silicon.

For patterning of silicon, a mask of photoresist is generally produced, with the help of which the removal is controlled by an etching process. In order to produce small structures by means of a resist mask, the photoresist must be exposed with an exposure mask, which has correspondingly small structures. In the range below the usual wavelengths of light that are available for the exposure of the photoresist, this is possible only with increased effort. Often, however, structures having high aspect ratio features are needed, ie, the depth or height of the features is large relative to their lateral dimension. If, for example, depressions and thus elevations with nano dimensions are required on a silicon surface with an aspect ratio of 2, then a local removal of material must be carried out, which leads to a depression of, for example, 200 nm with a lateral dimension of 100 nm. In the case of an etching process based on a photoresist mask, therefore, it must likewise be produced with a comparable lateral dimension and must furthermore have the required etching selectivity in order to achieve the desired aspect ratio during the subsequent silicon etching process. Alternatively, high-resolution masks are also written using an electron beam (e-beam). These Although solutions are very versatile, they are also time-consuming and expensive. Therefore, there are always efforts to find alternative methods that also allow structuring in the nanometer range.

In many conventional processes, however, no large aspect ratio of the nanostructures is achieved, especially when a low defect density is desired. In the prior art, the nanostructure typically has an increased contamination density, that is to say undesired surface impurities, and / or an increased number of crystal defects, if initially crystalline silicon having a low crystal defect density was present. Therefore, these known methods can be used only to a limited extent or with worse results with regard to the overall performance of the component. In some of these conventional processes, microstructuring utilizing self-assembly to produce structured silicon surfaces has also been using plasma-enhanced reactive ion techniques, also known as RIE processes, based on SF ₆ (sulfur hexafluoride) and oxygen, with metal particles providing micromasking and thus ensured structure formation. (WO 02/13279 A2, US 6091021, US 6329296).

A disadvantage of this method is the use of metals in the plasma, which can lead to undesirable contamination of the silicon. The harmful effects of least metal traces in the semiconductor manufacturing process, especially in integrated circuits are known. In addition to the fouling effect of RIE plants due to metal addition, the overhead of these processes is also detrimental to use in manufacturing processes that require high yield and low process costs.

Due to the procedural difficulties in the conventional production of structured surfaces and their further processing, an application of nanostructures in components, such as sensors, optoelectronic components, and the like in a cost-effective and reliable manner is thus also far from widespread. Overview of the invention

The present invention is based on the object of specifying methods and components, wherein nanostructures can be reliably and inexpensively integrated while at the same time having good machinability and at the same time lead to an improvement in the performance of the components.

In one aspect of the invention, a method of anti-reflection of photoelectronic devices by self-assembled nanostructures and corresponding devices is disclosed, wherein the generation of suitable

Nanostrukturen based on a method that is described in detail in a parallel PCT patent application with today's filing and the file number PCTVE P2006 / 067248. Insofar as there are doubts in this application or additions are required, the said PCT is to be used.

According to the invention, an antireflection coating of photoelectronic components, such as e.g. Photodiodes, photocells, and the like as part of an integrated circuit or as a discrete component, not or not exclusively in a conventional manner with λ / 4 layers, but alternatively or additionally made by RIE-etched nanostructures having a much better broadband characteristics ,

Photosensitive devices in integrated circuits and as discrete components, such as photocells, are used to convert light into electrical energy or vice versa. Photodiodes e.g. to convert light signals into electrical signals. High sensitivity is desirable for these photosensitive devices. For example, photodiodes in microelectronics represent an integratable sensor whose area should be as small as possible or which should detect even the smallest amounts of light. Other components such as photocells should have a high efficiency. In all cases it is advantageous to minimize the reflection losses. In principle, the conversion of the photons into electrons in the semiconductor material, such as silicon, takes place by itself. The charge carriers are drawn off by an electric field, which is formed by a pn junction. For this to happen, however, the photons must first penetrate into the semiconductor, such as the silicon. Due to the large differences in the impedances or the

However, refractive indices of air and the semiconductor material, such as silicon, reflect a multiplicity of photons at the interface and do not reach the location of the transformation. This reduces the sensitivity of the photodiode or the efficiency of the photocell. Now there is no air / semiconductor junction in an integrated photodiode since semiconductor devices are usually protected by a passivation layer. It is often an arrangement of air / SiO 2 / Si or air / SisINU / Si when considering silicon-based devices. One has a three-layer system with two interfaces. The proportion of light entering the semiconductor is influenced by reflections at these interfaces and their interferences and depends on the layer thickness, the material and the wavelength. In the best case (constructive interference for the transmission) almost all light can be exploited, in the worst case (destructive interference) you lose 30% ... 50%, depending on the wavelength, namely just as much as at the bare semiconductor / air Boundary surface is reflected. Conventionally, the thickness and the material of the intermediate layer is selected so that constructive interference and thus maximum antireflection occurs, as described, for example, in DE-A 103 93 435.

However, the following limitations or disadvantages remain:

1. The layer thicknesses must be tightly tolerated.

2. The anti-reflection succeeds only for a certain wavelength satisfactorily; one also finds other wavelengths that represent other interference orders for which there is good antireflection, but these are not arbitrary.

3. In order to reduce the reflection in a broad wavelength range to near zero, an antireflection by simple λ / 4 layers fails. It would be necessary materials with finely graded refractive indices between 1 (air) and the maximum corresponding refractive index of the semiconductor, such as silicon, for the desired wavelength range. For example, the refractive index of

Silicon is strongly wavelength dependent, but is usually above 3.5. However, such materials for refractive index adjustment are not established in conventional semiconductor technology and are therefore currently not available. 4. The applied layers must have a very low absorption in the wavelength range used

In typical application fields for integrated photodiodes, generally no broadband light sources are used, but preferably light emitting or laser diodes. These only emit at a certain wavelength and a conventional λ / 4 antireflection layer could be tuned to them. Often, however, such integrated photosensors in the same design for different applications are used and they imply different wavelengths of light. An example is the pick-up or pick-up systems modern CD / DVD combi drives. These have to work optimally without conversion at three different wavelengths (blue, red, infrared), whereby these wavelengths can no longer be represented by different orders of constructive interference. Even with optimum adaptation of λ / 4 layers, the well-established materials silicon dioxide or silicon nitride do not achieve a desired high transmission. With silicon as the semiconductor material and the use of silicon dioxide as the intermediate layer, at least a reflection content of 8% remains. With nitride, this proportion is significantly smaller in the visible, but for wavelengths below 400 nm, the absorption is significant and in turn leads to light losses.

The first solution according to the invention (claim 1) indicates a photoelectronic component which has an optically active window for the entry and / or exit of radiation. Furthermore, a nanostructure with statistically distributed structural elements provided on a surface of the optically active window is provided with a

The end region and a foot region are formed in the device, wherein the tip of the end region has a lateral extent of less than 10 nanometers and the foot region has a lateral extent of 50 nanometers or more. Furthermore, an aspect ratio of the structural elements (height of the structural elements with respect to the lateral extent at the foot) is on average greater than four.

The invention (claim 1) thus provides an associated with little effort and therefore low cost antireflective surface for integrated optoelectronic circuits and discrete photoelectric devices ready, which also eliminates the disadvantages mentioned in point 1 to 4 or at least significantly reduced.

The circumscribed nanostructure with the "needle-like" structures as the essentially "pyramidal" structural elements, the "atomically pointed" end and a much wider towards the foot area lateral

Have expansion, succeeds a gradual adjustment of the refractive index between the window material and the surrounding medium. In particular, since the lateral dimensions at the foot region are smaller than the wavelength of the visible region, the "material mixture" for the light may be considered as a "continuous" mixture, with the proportion of the surrounding medium of the pure window material at the foot of the structural elements increasing with height. For example, air increases, so that the refractive index for the wavelength range of interest decreases quasi-continuously, so that thus for the light substantially no interface with a discontinuous jump in the refractive index more occurs. The nanostructure may be fabricated in a method compatible with bipolar, CMOS, or BiCMOS integrated or discrete device technology, as described in more detail below. The nanostructure can be used alternatively or in addition to an antireflection coating. It is not more elaborate than this, but has one, over a wide

Wavelength range non-wavelength good quality reflective coating. In one embodiment, this covers the entire wavelength range of interest for silicon photodiodes. Another advantage of the antireflection coating by the nanostructure described above is their low angle of incidence dependence compared to λ / 4 layers or regular structures.

As explained in more detail below, when using the nanostructure in photodiodes or photocells, the property is very advantageous that the nanostructure, if it is made of a semiconductor material, such as silicon, a high degree of "freedom from defects" of the used areas can be achieved, ie , in the generation of the structural elements by plasma etching substantially no additional crystal defects are caused. Thus, the generated electron-hole pairs find no additional recombination centers and can be further efficiently sucked by the resulting pn junction electric field, so that no sensitive reduction of the sensitivity is caused.

In a further advantageous embodiment, the photoelectronic component further has a passivation layer which leaves the optically active window free and forms a boundary with the latter, the structure elements extending substantially to the border. By means of this measure, the effective area of the nanostructure can thus be locally precisely defined, whereby, for example, well-established masking methods can be used. The formation of the structural elements takes place essentially also at the boundary region between the passivation layer and the nanostructure, which is to be understood such that the structural elements extend at least up to a distance from the passivation layer which corresponds to half the thickness of the passivation layer. Even with a pronounced step, which can be caused by the passivation layer, a high area coverage of the window is still achieved by the structural elements and thus maintained the high level of anti-reflection at the entrance or exit of radiation.

In one embodiment, the structural elements are constructed from single-crystalline semiconductor material. In this way, the electrical function of the device may remain substantially unaffected by the nanostructure. As before has been mentioned, for example, the effect of the charge carrier collection can be maintained, wherein the coupling of the radiation is significantly improved. Furthermore, the nanostructure can also be produced directly in the considered semiconductor material, wherein the composition and the doping can already be determined previously. If required, the nanostructure can also be formed at an early stage of the production process, with subsequent adjustment of certain properties of the semiconductor material, such as doping, composition, etc. For this purpose, the nanostructure can be "preserved" in a suitable material, such as silicon dioxide, so that bake-out processes, implantation, introduction of other types of atoms, for example germanium into one

Silicon base semiconductors, etc can be performed with a high degree of compatibility with conventional processes.

In an advantageous embodiment, the nanostructure has a monocrystalline base layer on which the structural elements are arranged. A crystal defect density of the features is substantially equal to the crystal defect density of the base coat. Thus, the quality of the semiconductor base material can also be provided in the structural elements.

In one embodiment, the semiconductor material is silicon. In this case, a more efficient self-organized etching process, as described in detail below, can be applied directly in a silicon-based device. In other cases, a silicon layer, crystalline or polycrystalline or amorphous can be applied and then the nanostructure can be efficiently produced by etching. The same can also be achieved in an i5 polysilicon wafer.

In an advantageous embodiment, the structural elements are at least partially constructed of an insulating material. As a result, an electrical passivation can be achieved in cases in which influencing the electrical behavior of a component is not desired, or a high degree

Resistance to a variety of ambient media is desired. For example, the insulating material may be silicon dioxide, silicon oxynitride or silicon nitride.

In an advantageous embodiment, a height of the structural elements is in the range from 400 nanometers to 1000 nanometers, in particular also beyond 1500 nanometers. With these dimensions in the height direction in combination with the previously specified lateral dimensions, excellent optical antireflection properties result in the visible spectrum and also in the infrared range. In a further advantageous embodiment, the photoelectronic component further has a planarization layer in the optically active window, wherein the structural elements of the nanostructure are embedded in the planarization layer.

The leveling layer, which may also be referred to as a protective layer, fills in the voids between the acicular structural elements to be protected, for example silicon tips, so that the structural elements are stabilized. For further processing, a closed layer is thus formed. Due to the smooth surface thus produced, mechanical stresses can be intercepted with little risk of destruction of the nanostructure. It is much easier to apply another layer to this smooth surface and to remove it again.

Depending on the material used this protective layer attacks differently in the

Functioning of the nanostructure. The surface-enhancing function of a nanostructure is completely prevented by a dense layer. A porous layer, on the other hand, can be used to pass only certain substances to the surface of the nanostructure, e.g. plays a role in chemical sensors. For all optical applications, it is advantageous that the properties of the reflection and transmission or of the scattering only slightly deteriorate or even improve. For this purpose, the layer preferably has a low absorption. Reflection losses remain minimal when the refractive index is as close as possible to the refractive index of the surrounding medium, such as air. For example, in one embodiment, the refractive index of the material of the flattening layer is 1, 5 or smaller.

In a further embodiment, a second nanostructure is further provided on a second surface of the optically active window. This is advantageous in applications where the optical window is not directly in the base material of the

Component is produced, or if recesses are to be covered in the base material through the window. For example, a suitable window can be made on a separate substrate and then transferred to the device. Advantageously, the nanostructure and the second nanostructure are embedded in a protective layer.

In a further aspect of the invention, a sensor device is provided. The sensor component comprises a sensor surface, which is formed by a nanostructure with statistically distributed structural elements, the structural elements having a Have end region and a footer. The tip of the end region has a lateral extent of less than 10 nanometers and the foot region has a lateral extent of 50 nanometers or more. The aspect ratio of the structural elements (the height of the structural elements for lateral expansion at the 5-foot area) is on average greater than 4.

By providing the nanostructure, the surface is significantly increased as a sensor surface, so that a higher sensitivity is achieved. Furthermore, with volatile media, such as gases, the residence time near the sensor surface can be extended. Further, if optical methods of detection are used, the sensor surface can be used at least partially as an optical window, with the advantages described above being achieved in addition to the improved sensor behavior.

In one embodiment, the nanostructure of the sensor component is formed from an insulating material, such as silicon dioxides or the like. In other embodiments, the sensor device further comprises a porous planarization layer for embedding the nanostructure. In this way, a high mechanical stability and also a protection from environmental influences is achieved, 0 wherein the porosity of the filling material still allows contact of the surface with certain substances.

In another aspect of the invention, an optical coating structure for use in optical devices or optical devices is provided. The optical coating structure comprises a base layer and one on the

Base layer applied nanostructure with statistically distributed structural elements. The structural elements have an end region and a foot region, wherein the tip of the end region has a lateral extension of less than 10 nanometers and the foot region has a lateral extension of 50 nanometers or more, and o wherein an aspect ratio of the structural elements, ie, the height of the Structural elements and the lateral extent at the foot, the average is greater than 4.

The coating structure of the invention can be used in a variety of devices and devices, wherein the absorption behavior and / or the emission behavior can be significantly improved due to the aforementioned

Properties of the nanostructure formed in the coating structure. The optical coating structure can be manufactured separately, for example on silicon surfaces, and can then be installed with suitable carrier materials in the actual application. Further advantageous embodiments of the optical coating structure are also described in claims 19 to 26.

In a further embodiment, the optical coating structure has a conformal metal layer on the structural elements, so that the absorption behavior and / or emission behavior of the nanostructure can be adjusted independently of the base material of the structural elements of the nanostructure. In one embodiment, the metal layer of the targeted heat dissipation. Thus, the improved emission of heat radiation achieved by the nanostructure and the metal layer, for example, can be significantly improved, the cooling effect of components, which overall low construction volumes are made possible.

In another aspect, an optical device having a liner as a broadband optical absorber is provided, the liner comprising an optical coating structure of the type previously described.

In a further aspect, there is provided a digital projector with digital light processing having an absorption surface comprising an optical coating structure in the manner previously described.

In a further aspect, there is provided apparatus comprising means for wavelength-independent conversion of optical radiation into heat, wherein the means for converting optical radiation to heat comprises an optical coating structure in the manner previously described,

In another aspect, there is provided an optical element for emitting optical radiation having an exit window comprising an optical coating structure in the manner previously described. In one embodiment, the exit window is coupled to a light emitting diode or a LASER.

In a further aspect, a reflection reference device is provided for determining low reflection values, wherein the device comprises an optical coating structure in the manner described above.

In the aspects described above, the versatility of the nanostructure according to the invention in different applications is advantageously exploited in order to improve the performance of many optical devices. In a further aspect of the present invention, a photoelectronic component having a reflection-reducing layer provided in the optically active window region of statistically regularly distributed, crystal defect-poor silicon needles with a height of 400 to 1500 nm and an aspect ratio greater than 4: 1 is provided.

Also in this case, the silicon needles lead to an improved performance with respect to the coupling and / or decoupling of radiation in a wide wavelength range, wherein the small amount of crystal defects does not affect the electronic behavior significantly negatively. With a height of 400 nm, for example, an excellent visible light reflection coating can be obtained, and the efficiency can be increased even more with an increase in the height of the silicon needles.

Advantageously, the silicon needles have a tip whose lateral extent is less than 10 nm. The tip of the silicon needles can thus be called "atomic pointed". In conjunction with these small dimensions of the end regions of the silicon needles, the foot region of the silicon needles can have a lateral extent of 50 nm or more, which on the one hand results in an advantageous embodiment

"pyramid-like shape" of the needles is generated and on the other hand, the lateral dimensions at the foot of the needles still remain below the wavelength of visible light. The pyramid-like needles are not too tight. The gap is at least 50 nm wide. In this way, a quasi-continuous material distribution is generated for the wavelength range of interest, so that a continuous change in the refractive index along the height direction of the silicon needles is achieved. Thus, for radiation: irrrorvisible area and for infrared radiation passing through the nanostructure in the height direction, there is a steady change in refractive index without encountering substantial discontinuous interfaces that would result in high reflectivity.

In one embodiment, the photoelectronic device has a passivation layer that exposes and forms a boundary with the active window region, the silicon needles being formed to a distance from the boundary that is less than half the thickness of the passivation layer.

In another aspect of the invention, an optical window is provided that includes silicon and has broadband IR-to-IR transparency, both Surfaces of the window needle-like structures in nanodimensions with an aspect ratio greater than 4: 1 possess (claim 39).

As previously described in the context of the optical coating structure, the optical properties of devices and devices requiring radiation exchange with infrared radiation can be improved. In this case, both the entrance surface and the exit surface of the window are provided with a nanostructure, which thus lead to a quasi-continuous change in the refractive index in the propagation direction of the radiation on at least one, preferably both sides and thus significantly reduce reflection losses. The front side is preferred.

In an advantageous embodiment, the needle-like structures with nanodimensions with a SOG layer (spin-on-glass) are protected from mechanical action. Thus, the optical window in the final phase of the production and during further processing, for example when mounted in a device or a component, effectively protected against mechanical or other environmental influences. For example, the protective material may have hydrogen silsesquioxane (HSQ), resulting in good processing with favorable optical properties, such as low refractive index, low absorption.

In an advantageous embodiment, the needle-like structures with nanodimensions are limited to certain areas of the window by means of conventional masking technology, so that untreated, mechanically stable and easily sealable regions against air, liquids and vacuum remain. In this way, the optical window can be used in a very flexible manner in many application situations.

In a further aspect, the invention relates to a method for adapting the

Refractive index of an optically active window of a photoelectric device. The method includes forming a nanostructure in the surface area of the window by means of a self-assembled plasma etching process for etching a silicon base layer, and setting an aspect ratio of structural elements of the nanostructure generated in the silicon base layer to a value of four or higher depending on an operating wavelength range of the photoelectric component. The use of a self-assembled plasma etching process for the production of nanostructures, as already described above, achieves a high degree of compatibility with many manufacturing processes in the semiconductor industry. In this case, by adjusting the aspect ratio by suitable means, as will be described below, an adaptation of the optical properties to the requirements of the photoelectronic component can be achieved in an efficient manner.

In an advantageous embodiment, during the plasma etching process, the needle-like structural elements are produced by using the working gases oxygen (O 2) and sulfur hexafluoride (SF ₆ ) without the use of additional means for targeted mask formation in a single process step, the silicon base layer being deposited during the process maintained at a constant temperature in the range of substantially 27 ° C, in particular in the peripheral region ± 5 degrees Celsius and is operated with a plasma power in the range of about 100 to 300 watts, with higher plasma powers are set at higher process pressures and the ratio of working gas flows in Depending on the geometric system parameters is set so that the oxygen in the reaction point on the silicon base layer shows a self-masking effect, which in the gas flow for SF ₆ at 50 to 150 sccm and for O2 at 20 to 200 sccm to achieve is and the process time is only a few minutes.

With this embodiment, needle-like structural elements in the nanostructure (in the above sense) with high aspect ratio and nano-dimensions on silicon surfaces can be produced, avoiding or reducing the masking effort and improved behavior with regard to crystal defects and surface chemical contaminants conventional methods and a high degree of compatibility with other manufacturing processes is achieved. For this purpose, a reactive plasma atmosphere with at most two different gas components, ie, with oxygen and a reactive SFε gas for the etching of silicon is generated by setting process parameters which develop a self-masking effect for producing a nanostructure. The etching takes place here without further working gases and is carried out as a one-step process, ie, after generation of the plasma atmosphere, the silicon surface is exposed to the action of Ätzplasmas without further process steps take place. In particular, no further measures are taken to achieve a targeted micro-masking of the silicon surface. Furthermore, the aspect ratio of the resulting in the plasma atmosphere Needle-like structures set to a value of 4 or greater by controlling the process time.

Thus, according to the invention, masking of the Si surface, whether by photoresist or other substances such as aluminum, gold, titanium, polymers, water, or any surface contaminants, etc., is dispensed with. The needle-like structures produced by the method according to the invention have a form suitable for optical applications in the visible light and also in the infrared range, wherein the form of the needle-like structures formed by the self-organized masking of the etching, which in addition to the aspect ratio of greater than 4 also has a "pyramid-like" section, wherein a very tapered needle end is formed, however, at the foot of the "needle-like structure" a relatively flat leaking area is generated.

Overall, a very favorable antireflection behavior in the visible range and also up to 3000 nm or more could already be detected for the resulting nanostructure with an average length or height of the structures of about 400 nm. Without intending to limit the invention by the following discussion, investigations by the inventors indicate that efficient self-organized masking (self-masking) is achieved by the etching process itself rather than by pre-existing or specially added materials. Corresponding investigations on the basis of Auger electron spectroscopy (AES) and energy dispersion X-ray spectroscopy (EDX) indicate that the masking effect is caused by SiO _x , so that a high shielding effect is achieved by the locally formed silicon oxide. Overall, this leads to a moderately low silicon consumption during the production of the needle-like structures with a simultaneously high aspect ratio, so that the method according to the invention can advantageously and efficiently be used in semiconductor production with a high degree of process compatibility.

No defects for targeted mask formation are exploited. In place of the targeted masking before an etching process thus occurs the previously described self-organized, caused by the special process conditions masking during the etching process. The combination of the self-masking with the etching during the RIE process thus enables the generation of self-organized pyramid structures in the nanometer range by the plasma. It is thus possible, a smooth silicon surface in a statistically regular, quasi-ordered needle structure in the nanometer range, ie with lateral dimensions in the range below the usual wavelengths of light, for example, to convert the wavelength range of visible light.

Furthermore, it is possible by the inventive method in a single etching step, both the number of contamination defects, for example, by

Etch by-products are usually caused, as well as crystal damage, which are found in conventional plasma-assisted processes, significantly reduce or substantially avoid within the measurement accuracy. Thus neither RHEED, CV measurements, TEM or PDS could detect such defects. Even a simple photodiode whose surface was processed by this process showed no peculiarities indicating increased defect densities. Thus, the nanostructure produced can be provided by a single plasma etching step in a quality that does not require further material removal.

The structures produced by the method show no edge shading at high edges. It is thus possible, for example, to structure surfaces of a few μm (microns), even if the surface is surrounded by a 5 μm high structure.

The structuring of the silicon is done by the plasma in the RIE process. These structural elements are greatly deepened by the etching process, resulting in the structures in the nanometer range with enormous aspect ratios.

When using the working gas consisting of SFε and O2, the needle-like structures with low defect rate, so low crystal defect density and low surface contamination, regardless of the crystallographic

Orientation of the silicon base surface, whereby a high degree of flexibility for the integration of the method according to the invention is provided in corresponding manufacturing processes for silicon-containing components.

In other embodiments, another working gas combination with O _{2 is used} as a component. For example, carbon fluorides (C _n F _m with n for example 1, 2 or 4, and with m for example 2, 4 or 8) can be used in combination with oxygen as the second gas component. SFε or the other aforementioned reactive gases are in each case next to oxygen, the second of the two gas components and in this case the actual etching gas, whereas O2 increases the etching rate and causes the self-masking (passivation). It also produces a high selectivity to SiO 2 in the etching behavior, so that an efficient limitation of the silicon surface to be structured by means of a corresponding mask layer is possible. The temperature of the silicon base layer and the ratio of the working gases at the reaction point on the Si surface are appropriately set. Thus, an efficient adjustment of the further process parameters, such as the specified flow rates can take place, since the temperature, which typically represents a "sensitive" parameter, is specified in a very precise manner.

The process pressure and the plasma power are also suitably matched to one another in order to obtain the desired aspect ratio with simultaneously reduced contamination rate and low crystal defect density.

In particular, while maintaining the oxygen component in the manner indicated, the ratio of the working gases is adjusted so that etching removal and self-masking balance each other. As a result, both the structuring in the intended sense, as well as "defect-free" achieved.

In the method according to the invention, the absolute parameter values can be adapted efficiently to the proportion of the open silicon surface. If the Si surface is covered to a high surface area by a mask layer, for example oxide or silicon nitride, this can be compensated for by increasing the reactive gas content, for example the SF ₆ content, in particular also with an increase in the SF ₆ content, at the same time Reduction of the oxygen content and simultaneous increase of the process pressure.

It is possible by the process described above to produce nanometer structures with high, variable aspect ratios in a short time using a simple RIE system with parallel plate reactor. This is possible over a large area as well as with targeted adaptation of the process parameters even in the smallest areas, so that individual components, such as sensor areas, optically active areas of optoelectronic components, and the like can be specifically provided with a corresponding nanostructure, without adversely affecting other component areas. Unstructured areas may be simply, e.g. protected by an oxide mask. Furthermore, due to the low contamination rate and the low crystal defect density, an immediate further processing after the production of the nanostructure can take place without requiring complicated preparation and / or post-processing processes.

In a further embodiment, a protective layer is formed for the nanostructure with a substantially planar surface. This can be appropriate Separation methods are used, which in itself allow a highly non-compliant material deposition, such as the spin-coating of low viscosity materials. Other methods include applying a suitable material with a suitable deposition technique followed by leveling with removal of excess material by CMP (chemical mechanical polishing).

In a further embodiment, a spin-on-glass (SOG) adapted in its properties to the requirements of the nanostructure to be passivated is applied in several steps and annealed (baked) after each application until the planar surface is formed (claim 53). Thus, efficient non-compliant deposition techniques can be used to apply low viscosity material, with subsequent curing occurring in layers, so that the desired final thickness of the cured material can be precisely adjusted. For example, For example, SOG layers of hydrogen silsesquioxane (HSQ) can be applied.

In a further embodiment (claim 48 or claim 65), an additional layer is applied before the plasma etching process, which as a buffer layer has a relation to the silicon base layer with the same process parameters modified etching behavior that forms a self-organized structure with only a relatively low aspect ratio, which in course of the process is reduced.

This contributes to a reduction of silicon consumption in the generation of self-assembled nanostructures in the field of window openings of integrated circuits with optoelectronic components and of discrete optoelectronic components. Especially at the beginning of the etching process, the structure formation is still quite slow, but even during this time silicon is removed. Only then accelerates the formation of the needles with a correspondingly large aspect ratio. The loss of material in this first etching phase can have detrimental effects, for example, when it comes to doped regions in which the doping drops with a gradient from the surface. For a negligible influence on the device data, this material removal should be as low as possible. By providing the buffer layer, material removal of the silicon can thus be initially prevented, although a certain masking nanostructure can nevertheless be formed in the buffer layer, which then, while maintaining a desired nonuniformity, then also leads to a locally different etching behavior in the buffer layer Silicon leads, with the unwanted initial loss of material in the silicon remains low.

For example, the etching removal of a doped surface layer of the region in the window of an optically active component is reduced, and moreover, the aspect ratio can be easily varied. By applying an additional layer of defined thickness with a different etching behavior than that of silicon, the loss of etched etching can be reduced. Since the structure generation in the etching step is based on the plasma and thus a physical component of the etching step is responsible for the production, this also acts in other materials such as SiO ₂ . Since the plasma-generated structures in the SiO 2 have only a very small height, no nanostructures with a high aspect ratio are formed in it. There is no self-masking effect by another chemical component. Thus, the oxide layer is approximately evenly removed, but still receives a plasma-based nanostructure of very low height on the surface. The etching rate for SiO 2 is much lower than that for silicon in the aforementioned RIE etching process. Shortly before the SiO ₂ layer has been completely removed, this holey nanostructure and the different etching rate result in rapid or immediate formation of the large aspect ratios in the silicon. The oxide layer is first removed at the locations of the smallest thicknesses, where the etching process begins at a much greater etching rate and forms a hole in the silicon.

If necessary, the entire oxide layer is removed after some time, but it has then already formed a nanostructure with a considerable aspect ratio in the silicon, the tips of the nanostructures are still close to the level of the former SiO2 / Si interface.

The process time of the etching step and the thickness and type of the buffer layer can be optimally adapted to each other. The etching step can last just as long until the buffer layer has been completely removed. But no longer, because otherwise more than necessary from the material is removed.

Due to the nature of the buffer layer material and the thickness of the buffer layer, the self-assembled nanostructure can be influenced in the aspect ratio and in its position at a distance below the starting surface, since the structure formation by the plasma is dependent on the material used and, depending on the etching rate, the buffer layer more or less long serves as an additional etching mask. In an advantageous embodiment, the aspect ratio of the structural elements of the nanostructure is adjusted via the thickness of the buffer layer. Thus, the optical properties of the nanostructure can be adjusted in a very efficient manner by a very precisely adjustable process parameter, ie, the layer thickness of the buffer layer, so that a greater degree of flexibility can be achieved in the selection of the etching parameters.

In a further advantageous embodiment, the buffer layer is a SiO ₂ layer, which in one embodiment has a thickness of 20 nm to 100 nm. In other embodiments, other materials may be used which include a

To cause delay of material removal in the actual silicon base layer. If e.g. a high level of compatibility of the buffer layer and the silicon base layer is desired, an additional silicon layer may be deposited on the base layer and serve as a buffer layer. Other materials, such as SiN, can also be used.

In an advantageous embodiment, the buffer layer is not completely removed, so that the resulting structural elements can have material of the buffer layer at their tip, which can lead to an improved resistance of the structural elements. In other cases, the buffer layer is in the

Substantially completely removed, so that silicon structural elements remain, the height of which substantially corresponds to the initial height of the silicon base layer.

In another aspect of the invention, there is provided a method of protecting a layer having high aspect ratio acicular structural elements, and

Part of a chemical sensor is provided, wherein in its properties adapted to the requirements of the layer to be passivated gas-permeable porous spin-on glass is applied in several steps and annealed after each application until a smooth surface is formed.

Thus, the nanostructure can also be used efficiently in sensor applications, since on the one hand a high protective effect is achieved and on the other hand the contact with gases is still possible.

In an advantageous embodiment, in a last step, an overlap with a non-porous layer takes place, which is removed again in the last mask process. In this way, a pronounced protection is achieved during the finishing of the nanostructure, which thus ensures a high degree of flexibility in the manufacturing process, whereby the non-porous covering layer can be removed without much effort before the last masking and thus structuring.

Further advantageous embodiments of the nanostructure with the needle-shaped structural elements are described in subclaims 56 and 57.

In another embodiment, the spin-on glass is liquid pervious, resulting in a broad array of sensor element applications.

In another aspect of the invention, a method of passivating the surface of a semiconductor device comprising silicon is provided. The method comprises exposing a portion of the surface locally, forming primary acicular structures having high aspect ratios in nanodots.

Dimensions with lateral dimensions in the range below the wavelengths of visible light by means of a reactive ion etching process and modifying the area provided with the primary needle-shaped structures to produce secondary, also needle-like structures.

By modifying the acicular structures after their fabrication based on silicon, efficient self-assembled plasma etching processes, such as those described above, can be used, in which case desired material properties of the acicular structures are then adjusted by the modification process. Nevertheless, a variety of different nanostructures can be produced by means of an etching recipe. In one variant, the needle-like structures can thus be provided with an insulating surface.

In one embodiment, a silicon layer is deposited to provide the surface. In this way, any carrier material can be used without substantially affecting the actual process of structuring, wherein the desired material properties can then be further adapted by the modification.

In one embodiment, modifying the region provided with the primary acicular structures includes thermal oxidation. In another embodiment, modifying the region comprises nitriding the region. In other cases, dopants and / or other types of semiconductors may be applied, such as germanium and the like.

In one embodiment, the silicon in the primary needle-like structures is substantially completely converted to silicon dioxide.

In a further embodiment, the primary needle-like structures are made by reactive ion etching (RIE) using the working gases oxygen and SF ₆ in a single process step without using additional means for targeted mask formation in the structuring process only by adjusting the process parameters so that the oxygen produces a self-masking effect on the silicon-containing surface at the reaction point and self-assembly of the needle-like structures takes place. Thus, by a highly efficient etching process, the primary structures can be formed, which can then be subjected to the desired modification due to the low defect and the low surface contamination without much effort.

Brief description of the drawings

Embodiments will now be described with reference to the accompanying drawings, in which:

1a is an electron micrograph of a RIE etched silicon surface in section in a region which is partially covered by an oxide layer,

1 b is an electron micrograph with obliquely incident electron beam from which the homogeneity of the distribution of the silicon needles and the depth of the spaces between the needles are visible,

Fig. 1c is an electron micrograph in

High-resolution radiation from the tip of a silicon needle, Fig. 1d is oriented vertically,

2a shows a conventional photodiode antireflection in a schematic sectional view,

2b is a photodiode according to the invention coated in a schematic sectional view,

2c shows a diagram of the reflections on silicon surfaces which are coated in different ways,

2d shows the transmission electron micrograph of a single Si tip of the RIE-treated surface,

2e the diagram of the spectral sensitivities of identically constructed diodes with different passivation layers,

3a shows a vertical section through a photodiode without protective layer on the nanostructure in a schematic representation,

3b is a vertical section through a photodiode with protective layer on the nanostructure in a schematic representation, FIG. 3 c shows a diagram with the values of the reflection before and after the application of the SOG protective layer to a silicon nanostructure, FIG.

4a shows a sequence of steps of the RIE etching of the Si surface without a buffer layer with increased silicon consumption,

4b shows a sequence of steps of the RIE etching of the Si surface with buffer layer with minimal Si consumption,

5 shows a flow of a modification of a silicon nanostructure to a SiO ₂ structure according to an illustrative embodiment,

6a shows measurement results of the optical reflection of modified

Represent silicon surfaces compared to untreated surfaces, wherein the measurement results were obtained by means of an integrating sphere and describe the reflections of the surfaces in all spatial directions,

6b shows the extremely small and wavelength-independent reflection of the modified silicon surface in detail, wherein very low and wavelength-independent reflection values occur in the visible range and wherein the noise above 800 nm is due to the detector change in the measuring device,

Fig. 6c shows the direct absorption measurement by the photothermal

Deflection Spectroscopy (PDS) of a modified silicon surface, wherein an amplitude of "1" corresponds to 100% absorption (0.9 eV = 1350 nm to 2.2 eV = 560 nm) and

Fig. 7 shows the transparency of a sample with unilaterally modified surface, wherein the theory curve neglects the absorption of the silicon. Detailed description

With reference to FIGS. 1a to 1c, an exemplary nanostructure and illustrative processes for the production thereof will first be indicated. Similar processes and nanostructures can then also be used in other applications, such as photoelectric components, sensor components, as optical coating structures in optical devices, as optical windows, and the like, as already described above and also described in subsequent embodiments. Further, the nanostructures, such as those prepared by the methods described below, may also be subjected to further protective layer application, surface modification, and the like.

1a shows a silicon-containing component 1 with a nanostructure 2, which has an i5 monocrystalline silicon base layer 3, on which needle-like silicon structures 4, which are alternatively (but synonymously) referred to as structural elements of the nanostructure 2 in this application, are formed. In this application needle-like silicon structures are to be understood as "pyramid-like needles" or structural elements that have a tip with lateral dimensions of a few 20 nanometers, wherein the tip increases significantly downwards in its lateral dimension, so that in the lower region of the structures a lateral Dimension of a few tens of nanometers or more, with relatively shallow leakage (against the slope of the many sidewalls of the center region).

In this embodiment, the silicon base layer 3 is delimited by a mask layer 5, which may be composed of silicon dioxide, silicon nitride or the like, wherein the needle-like silicon structures 4 are formed up to an edge region 5a of the mask layer 5 at a small distance from the edge region 5a. Here, a small distance is to be understood as a distance that is less than half

Thickness of the mask layer 5. In the embodiment shown, the

Silicon base layer 3, a part of a silicon wafer 6 inch diameter with a (IOO) surface orientation, which has a p-doping, which results in a resistivity of 10 ohm ^* cm.

However, as previously stated, the base layer 3 may have any crystal orientation with any predoping. In alternative examples, the base layer 3 may be formed substantially of amorphous or polycrystalline silicon. FIG. 1 b shows an enlarged section of the nanostructure 2, wherein the angle of incidence of the probing electron beam has an inclination angle of approximately 17 ° in order to more clearly show the size relationships in the lateral direction and in the height or thickness direction of the pyramid-like structures or structural elements 4 , As can be seen from FIGS. 1a and 1b, the silicon structures 4 have a height which is on average about 1000 nm, so that in some embodiments a height is reached that is greater than the wavelengths of the visible light. In other embodiments, the height of the structural elements is 400 nm to 1000 nm, in special cases also up to 1500 nm.

Due to the tilted electron beam of 17 °, the height entered as a measure in FIG. 2 is to be converted from 603 nm to the real height. It is also possible to convert the vertical extent by up to 60% for lower pyramid-like needles, which achieve their effects from about 400 nm. This is done by compression of Figure 2 in the height direction to 40% of the height shown.

But also pyramidal structures 4 with a mean height in the range of 400 nm show excellent optical properties in many applications. Thus, for example, for an average height of 400 nm, an excellent antireflection coating in the previously measured visible wavelength range up to the currently measured 3000 nm could be observed.

As can be seen from FIG. 1 a, an average maximum height of the silicon structures 4 may also be at substantially about 1000 nm or more.

On the other hand, FIGS. 1a and 1b show that the lateral dimension of the silicon structures 4 in a foot region 4b is typically less than 200 nm or less than 100 nm, so that on average an aspect ratio of height to lateral dimension of 4, or even higher is achieved becomes.

The results shown in FIGS. 1a and 1b, which are based on a p-doped 6-ZoII (100) Si wafer, a 10 ohm * cm resistance and an area fraction of the oxide mask, ie, the mask layer 5, of greater than 90%. (up to 93%) were prepared in a single-step plasma etching process in a STS320 plant with the following parameters: SFe gas flow: IOO sccm

O2 gas flow: 20 sccm

Gas pressure: 70 mTorr

Si slice temperature: 27 degrees Celsius Plasma power: 10O W

Etching time: 2 min

Self-adjusting BIAS (i.e., DC potential between the plasma atmosphere and the surface to be etched): varies by 350 V (self-adjusts).

The 6 "(inch, inch) disc lay in the RIE STS 320 system on an 8" disc and the plasma can also act next to the 8 "disc.

Specifying a power density is only possible in a first approximation in an estimation. The plasma power can be set in the range of 100 W to 300 W, which corresponds to a power density of about 4 W / cm ² to 12 W / cm ² for a 6 inch disk.

In other embodiments, gas flow rates of 50 to 150 sccm have been provided for the reactive gas, ie SF ₆ , C _n F _m or HCI / BCU. For oxygen gas flow rates of 20 to 200 sccm are provided. Further, in some embodiments, the temperature of the substrate, and thus the base layer 3, is set at 27 C ± 5 ° C.

In alternative examples, comparable results were obtained for the nanostructured surface. Output parameters and process parameters are given below.

For an area fraction of 0.1% silicon and 99.9% oxide mask, with the following parameters

150 sccm SF ₆ 20 sccm O ₂ 91 mTorr 27 ⁰ C

100 watts

Bias 28 V

4 minutes etching time (process time) For 100% silicon area, ie a bare silicon wafer 65 sccm SF ₆ 23 sccm O2 50 mTorr 27 ⁰ C 100 watts 10 min to 20 min etching time (process time).

For bare silicon wafers, process time of up to 20 min is also useful. Then the process results in an extremely high-quality antireflective coating of the surface nanostructured with the needles.

From the above indications, corresponding parameter values for other etching systems and other degrees of coverage of the silicon base layer 3 to be structured with the pyramidal structures can be determined. For example, a lower degree of coverage of the silicon base layer may be taken into account by a lower gas flow rate of the reactive gas.

In further examples, which are not detailed here, the absolute parameter values can be efficiently adjusted to the proportion of the open (or free) silicon surface. If the Si surface is covered to a high surface area by a mask layer, for example oxide or silicon nitride, this can be compensated for at least by an increase in the reactive gas fraction, for example the SF ₆ fraction, in particular also when the SF ₆ fraction is increased. simultaneous reduction of oxygen content and simultaneous increase of process pressure.

If no mask is given, the reactive gas content is low, and vice versa.

With the above settings, the Si needles or structural elements 4 having a height of up to about 1000 nm were generally randomly distributed at the regions not masked by the mask layer 5.

As mask layer 5, silicon oxides or silicon nitrides are suitable.

Machined disks with similar structures (without oxide mask) become completely black and showed a reflection of less than 0.4% for the wavelength range of 400 nm to 1000 nm, at the same time excellent homogeneity of this property over the entire wafer (disk). In particular, in a wavelength range extending in both directions beyond 180 nm to 3000 nm wavelength, the investigations showed a still excellent anti-reflection behavior with reflections below 2%. The reflections recorded here (practically only) the reflections in all solid angles.

Furthermore, the crystal damage caused by the plasma-assisted single-stage structuring process and the contamination are very low and are below the detection limit in the exemplary embodiments shown. No residual substances could be detected after the plasma structuring process and the crystal quality of the silicon structures is almost identical to the crystal quality of the silicon base layer before the etching process.

1c shows an illustration of a single tip 4a or of an end region of a structural element 4. As can be clearly seen, the needles are almost atomically pointed at their end region 4a, ie the lateral dimensions of the end region 4a are only a few nanometers and thus less than 10 nanometers. In the illustration of FIG. 1 c, the crystal direction is also entered perpendicular to the surface of the silicon base layer 3. This direction corresponds to a [100] direction, since for the embodiment shown, the surface orientation is a (100) orientation. As can be seen, the end region extends substantially along the [100] direction with only a slight deviation of less than 10 °, so that the structural elements are aligned almost perpendicularly with only a few degrees deviation from the surface of the base layer 3. Furthermore, individual lattice planes of the monocrystalline needle can be clearly recognized without crystal defects caused by the etching being recognizable. In the base layer configuration shown, the appearing lattice planes correspond to the (111) planes.

Due to the highly fissured after the process surface of the base layer 3 whose surface increases significantly, causing the properties change significantly. The enlarged surface offers a much larger attack surface for attaching molecules and can thus significantly increase the sensitivity of sensors, for example. For example, it has been found that gases remain localized in the nanostructure 2 for quite a long time.

In the optical region, the pyramidal structures 4 are interesting in that they are smaller than the light wavelength (VIS / NIR) in their lateral size and by their needle shape, ie by the small lateral dimension of the end region 4a and the relatively large dimension at the foot 4b the pyramidal structure, and the high aspect ratios give off an almost perfect gradient layer. The refractive index changes gradually from the refractive index of the silicon to the refractive index of the medium surrounding the nanostructure 2, for example air.

s The nanostructure 2 thus enables an impedance matching or refractive index adjustment, which leads to an excellent broadband reflection suppression. Furthermore, it is known that strong bends, as they have the needle tips 4a, are particularly suitable for field emission.

This results in a broad field for the use of the nanostructure 4 in many microcomponents and also in other fields, such as solar cells, sensors and the like, as already described above and further explained below.

s The examples thus provide methods and structures in which silicon structures with large and adjustable aspect ratio occur, wherein due to the (special) parameter setting in the self-masking plasma etching process in a single etching step contamination and formation plasma-related crystal defects is kept low, so that at low Cost for the single-stage o structuring process, the resulting structure can be used immediately without further post-processing steps are required when needle-like silicon structures in high monocrystalline form are required.

Furthermore, no elaborate surface preparations or additional measures to produce a micro-masking are required. A pre-conditioning, i.e. preparation of the surface for obtaining the nanostructures, may be omitted:

With the aid of a standard RIE etching process for silicon, without any additional structuring measure (e-beam, interference lithography, or the like)

Self-assembly generates a large number of virtually crystal-defect-free, needle-shaped structures with a high aspect ratio and with nanodimensions on the surface of a silicon wafer or another silicon base layer, as a result of which, among other things, a broadband antireflection coating can be achieved.

With reference to FIGS. 2a to 2f, applications for anti-reflection of photoelectronic components by self-organized nanostructures and corresponding components are described, wherein the structures described above, for example also the methods described above, can be used. The anti-reflection of photoelectronic components, such as photodiodes as part of an integrated circuit or as a discrete component is not made in a conventional manner with λ / 4 layers, but by means of RIE-etched nanostructures, which have a much better broadband characteristics.

Examples are thus photosensitive components in integrated circuits and discrete components, such as photocells. The results achieved by conventional means and the difficulties involved have already been presented.

Realizable methods for producing an antireflecting surface for integrated optoelectronic circuits with little effort and therefore low costs are possible, compatible with discrete and integrated component technologies, without the mentioned disadvantages.

Among other things, the corresponding methods are compatible with a bipolar, CMOS or BiCMOS technology for integrated or discrete components. It can be used alternatively or in addition to an antireflection coating. It is no more expensive than this, but has a wavelength-independent anti-reflection good quality over the entire wavelength range of interest for silicon photodiodes. Another advantage of the antireflection coating is its low Einfallswinkelabhängikeit compared to λ / 4 layers or regular structures.

Particularly important for photodiodes is the property of the defect-free ranges achieved in the stated method, since the electron-hole pairs generated otherwise recombine and can not be sucked by the electric field, which represents a sensitive reduction in sensitivity.

Fig. 2a shows schematically a conventional photoelectric device 200 with a silicon substrate 201, an n-well 202 and corresponding contacts 205. Further, an optical window 203 is provided, which is non-reflective with an anti-reflection layer 204.

Fig. 2b shows the photoelectric device 200 according to an embodiment of the present invention. Here, in the region of the optical window 203, a nanostructure, such as the structure 2, is provided in order to achieve a broadband antireflection coating. As described above, the self-assembled nanostructures, such as structure 2, have a geometric shape suitable for use as an antireflection coating. The lateral distances are smaller than the wavelength in the medium, so that no scattering losses occur. The nanostructures (FIGS. 1a to 1c) are sufficiently high with> 500 nm (at about 1000 nm). From 400 nm, a very good antireflection coating is already observed, which can be slightly improved with increasing height.

The reduction of the reflection on such a surface can be explained by an impedance matching, i.e. an adjustment of the refractive index, between the two materials. The structures create a gradual impedance transition between the materials. This transition must be sufficiently wide (here the height of the structures) to act accordingly. The gradual transition works on the principle of the effective medium, whereby two substances are mixed together so that it appears for use as a substance with mixed properties of the two starting materials. Since the nanostructures have a needle shape in the sense defined above, there is virtually a continuous transition from one medium to another (here from air to silicon). From the electrical engineering is known, cf. Pozar, Microwave Engineering (Second Edition), John Wiley and Sons, New York 1998, that at least at one end of the transition, a particular nonlinear shape is particularly effective.

The structures used herein have such a shape. They are very pointed, but run very flat, resulting in a first small, but in the end very strong impedance change.

FIG. 2 c shows corresponding reflection curves for different components 200 with and without the nanostructure 2.

The measured reflection spectra confirm the drastic reduction of the reflection losses.

FIG. 2 d shows the result of crystallographic investigations, which show that the nanostructures 2 remained monocrystalline during their production. In Fig. 2d corresponding network levels can be seen in an upper portion of a

Structural element, as shown in Fig. 1c. There are no additional crystal defects with respect to the base layer 3, and the needle is oriented substantially along the [100] direction. So there are no additional generation and recombination centers. Thus, the risk of unwanted photocurrent losses or increased dark currents is sufficiently reduced.

FIG. 2 e shows spectral sensitivity measurements on diodes which are so non-reflective according to the invention and confirm the increased sensitivity in a large wavelength range. In particular, eliminates the strong, due to interferences oscillations, which are common in normal passivation of integrated photodiodes.

One illustrative embodiment relates to a process for the antireflection of silicon photodiodes, which is characterized in that in the surface region of the diode window crystal defect-free, needle-like structures in nanometer dimensions with an aspect ratio of 4 to 1 and greater by means of a reactive ion etching (RIE) using the working gases oxygen and SF ₆ are generated without the use of additional means for targeted mask formation in a single process step, the

During the process, silicon wafer is maintained at a constant temperature in the range of 27 ° C ± 5 ° C and is operated with a plasma power in the range of about 100 to 300 watts, requiring higher plasma powers at higher process pressures and the ratio of working gas fluxes in Dependent on the geometric system parameters was previously determined empirically, is set so that the oxygen in the reaction point on the silicon wafer shows a self-masking effect, which can be achieved in the range of gas flows for SF ₆ : 50 to 150 sccm and for O2 20 to 200 sccm and the process time is only a few minutes.

In a further embodiment, the component, for example, the component 200, in the optically active window region, a reflection-reducing layer of statistically regularly distributed, low-defect silicon needles with a height of 400 to 1500 nm and an aspect ratio greater than 4: 1, as by the application of the RIE method according to the previous method arise.

With reference to Figures 3a to 3c, further embodiments of the invention will now be described which specify methods for protecting sensitive nanostructures.

Extremely fine structures (nanostructures) are not very robust to mechanical forces. Therefore, it is necessary for a variety of applications to protect them from mechanical damage. This protection is provided by a filling substance reached to a smooth surface. For this purpose, a spin-on-glass (SOG) with adapted properties can be used.

Protective coatings for easily scratchable surfaces have been around for some time. Be it hard coatings for plastic glass or CDs. There are also efforts to protect nanostructures, cf. EP-A 1 215 513.

Basically, a protective layer to prevent the destruction of a functional element, without affecting its function too strong. It is usually a number of boundary conditions to be taken into consideration, causing the

Realization of such a layer complicates. This is especially true for the protection of a layer consisting of acicular silicon tips in nanometer dimensions with a high aspect ratio of 4: 1 and larger - referred to as nanostructure for short - as e.g. can be prepared by the RIE process crystal-free self-organizing, as already described above.

The examples provide a method for protecting such nanostructures which provides mechanical protection in further processing of silicon wafers with such layers without substantially altering the particular properties of these layers, such as reflection, adhesion of chemicals, etc.

This achieves the advantages that the protective layer fills up the cavities between the needle-like silicon tips to be protected, thus stabilizing the structures. For further processing, a closed layer is formed.

Due to the smooth surface produced in this way, mechanical stresses can be absorbed without destroying the nanostructure. It is much easier to apply another layer to this smooth surface and to remove it again.

Depending on the material used, this protective layer intervenes differently in how the nanostructure works. The surface-enlarging function of a nanostructure is completely prevented by a dense layer. A porous one

Layer, on the other hand, can be used to pass only certain substances to the surface of the nanostructure, which plays a role in chemical sensors, for example. For all optical applications, it is important that the properties of the reflection and transmission or the scattering worsen only slightly or even improve. For this purpose, the layer preferably has a low absorption. Reflection losses remain minimal when the refractive index is as low as possible.

FIG. 3 a schematically shows a photoelectric component 300 with a silicon substrate 301, an n-well 302 and corresponding contacts 305. Furthermore, an optical window 303 is provided, which is antireflected with a described nanostructure 2.

For example, for the anti-reflection of the device, for. B. a photodiode, prepared by a CMOS process, etched into the surface of the silicon by the RIE method in the manner already described, a nanostructure 2. This process step is usually followed by others. Among other things, the bonding pads for the contacting 305 of the components 300 are still freed from the circuit passivating layer. This is usually made of SiO ₂ or Si _ß lNU and is usually applied by the CVD method. This procedure is more or less compliant. Top structures are preserved. It does not form a smooth surface. To remove the passivation layer, resist mask and etching step are used. However, the applied lacquer can not be easily removed from the nanostructure 2; Lackreste limit their functionality.

Fig. 3b shows the device according to an illustrative embodiment.

To protect the nanostructure 2, therefore, a layer 305 of spin-on-glass (SOG) is previously deposited by spin-coating, e.g. Hydrogen silsesquioxane (HSQ). Since this substance is liquid when applied, the spaces between the

Nanostructures filled void-free. An annealing step hardens this glass, but also leads to a certain shrinkage, so that this procedure is advantageous to repeat. After a few such steps, the nanostructure is completely encased and the surface is even and resistant to mechanical damage.

The protected nanostructure can now be further processed without any problems using the standard processes of CMOS technology. The application of a resist layer and its removal is not a problem. Due to the low refractive index of 1.38 and the low absorption in the wavelength range of 150 nm or 180 nm to 1100 nm, the optical function of the nanostructure 2 is only slightly deteriorated.

FIG. 3c shows corresponding measurement results for the reflection of the optical window 303 for situations with an ARC (antireflecting) layer, with bare silicon, FIG. each without the structure 2, and for the components 300 of FIG. 3a and 3b. It remains at a broadband anti-reflection, which is significantly better with 3.5% reflection than the smooth bare silicon interface with> 30%.

One embodiment relates to a method for protecting a layer consisting of 4: 1 and larger with nanometer dimensions of acicular silicon tips with a high aspect ratio, wherein a spin-on-glass adapted in its properties to the requirements of the layer to be passivated is applied in several steps is tempered after each application until a smooth surface is formed.

In another embodiment, SOG layers of hydrogen silsesquioxane (HSQ) are applied to layers of acicular silicon tips present in windows of photoelectric devices.

A further embodiment relates to a method for protecting a layer, which consists of acicular silicon tips with a high aspect ratio and is part of a chemical sensor, wherein applied in several steps in its properties adapted to the requirements of the layer to be passivated gas-permeable porous spin-on glass and is tempered after each application until a smooth surface is formed, and in a last step, an overlap with a non-porous layer takes place, which is removed in the last mask process again.

Another embodiment relates to a method for protecting a layer which consists of acicular silicon tips with a high aspect ratio and is part of a chemical sensor, wherein applied in several steps in its properties adapted to the requirements of the layer to be passivated liquid-permeable porous spin-on glass and is tempered after each application until a smooth surface is formed, and in a last step, an overlap with a non-porous layer takes place, which is removed in the last mask process again.

With reference to Figures 4a and 4b, further embodiments are described in which material removal during formation of a nanostructure with acicular tips and high aspect ratio is reduced, thus improving manufacturing tolerances and yield. The "pyramid-like nanometer structures" (short nanostructures) self-organized by the previously explained RIE (Reactive Ion Etching) process, as described above, take some time to complete their formation. Since the etching process is already effective from the beginning, there is a considerable amount of material removal. This effect can be reduced or completely suppressed by an adapted buffer layer.

This achieves the advantages that the etching removal of the doped surface layers of the region in the window of the optically active components or of sensor elements is reduced and, moreover, the aspect ratio can easily be varied.

4a shows a typical process sequence in individual intermediate stages for the production of a previously described nanostructure 2 (see FIGS. 1a to 1c), starting from a planar silicon surface 411a on a silicon body 401 with increasing etching time, the nanostructure 2 in an unmasked region 403 of the silicon surface 401a, finally, a material layer 405 of the initial silicon volume 401 is "consumed".

FIG. 4 b schematically shows the production of the nanostructure 2 by means of a buffer layer 406, which has a lower etch rate compared to silicon 401. Thus, initially a much less pronounced structuring 406a is created, which is then driven into the silicon 401, wherein the consumption of the silicon 401 can be significantly reduced or even prevented. As a result of the etching properties and the thickness of the buffer layer 406, the aspect ratio of the nanostructure 2 (as 403a) can thus also be set, as described above. The remnants of the buffer layer, shown as 406a, may be removed using the etch selectivity between the buffer layer 406 and the silicon 401, or may be retained, as shown.

With reference to FIG. 5, further embodiments are described in which a base nanostructure or a primary structure is modified to obtain desired surface properties.

By exposing existing or applying an additional silicon layer, the subsequent structuring of this layer by means of a self-organization of needle-shaped structures with dimensions in the nanometer range below the usual wavelengths of light and with a high aspect ratio (nanostructures) generating RIE process without the use of additional means for masking the structuring process in intended areas and modifying, for example thermal oxidation, this structured layer is a suitable surface, for example a SiO ₂ layer, produced with approximately the same structure. This layer has a broadband effect of the antireflection coating and can also contribute in sensor components to increase the sensitivity by increasing the attachment surface of atoms and molecules.

In a further embodiment, the invention relates to the generation of a passivation layer made of a desired material, such as SiO 2, on photosensitive or light-emitting components as well as on sensor components. These can be monolithically integrated both discretely and with semiconductor circuits. The passivation layer consists on its upper side of structures with needle-shaped tips of a large aspect ratio, and thus has a broadband effect of the anti-reflection in the usual wavelength range.

The method of the invention makes it possible with the

Semiconductor device technology adequate means such surface relief, characterized by needle-shaped structures with large aspect ratios in nanodimensions, d. H. in the range below the usual wavelengths of light, with suitable surface material, such as thermal SiO 2 to produce.

Thus, existing methods of semiconductor technology can be applied, no additional interference is generated and a layer with broadband antireflection or large deposition surface is achieved.

When oxidation modification is used, at the high temperatures of thermal oxidation, oxygen diffuses in all directions into the silicon lattice because of the needle shape of the individual tips. The process takes place everywhere on the large surface. Therefore, needle structures are oxidized particularly quickly. The method is simple to use and offers the possibility of forming a secondary nanostructure consisting of SiO 2 with relatively little effort, which can be produced in a significantly more complicated or restricted manner in other ways.

The SiO ₂ layer grows in two directions. On the one hand it expands into the silicon and on the other hand the whole structure grows because of the volume increase of the SiO2. The silicon is completely converted into SiO2, at least in the tip region. The surface relief of the silicon becomes less Transfer change to the new SiO 2 layer, while the interface Si / SiO 2 is greatly leveled compared to the original Si surface.

In other embodiments, other modification processes are performed, such as nitriding, where nitrogen is incorporated into the silicon to alter the surface properties. It is also possible to introduce dopants or substances for surface modification or also for material changes that sometimes extend deep into the needles.

FIG. 5 illustrates a process sequence. In the upper part of Fig. 5 is a

Nanostructure 2, which can be produced by previously beschiebe process, formed in a region 503 a of a silicon-based layer 503.

In the lower part of the figure, the nanostructure 2b is shown after undergoing a modification process, which in this embodiment may include thermal oxidation, plasma oxidation, wet chemical oxidation, and the like.

Due to the strongly fissured surface, their surface increases considerably, which significantly changes their properties. Gases stay localized in the structure for a long time. The increased surface area provides a much larger attack surface for attaching molecules and can thus significantly increase the sensitivity of sensors.

In the optical domain, the structures are interesting in that their lateral size is smaller than the wavelength of the light (VIS / NIR) and, due to their shape and the high aspect ratios, give off an almost perfect gradient layer. They thus allow an impedance matching which leads to an excellent broadband reflection suppression, without scattering the light.

Since a passivation layer is necessary for most semiconductor components and this can be realized by SiO ₂ , the invention is also suitable for optical components. It allows the application of a passivation layer without causing the usual reflection losses of 3.5% (SiO2 / air transition).

A further embodiment relates to a method for passivating the surface of semiconductor components made of silicon by means of a SiO ₂ layer, which has acicular structures with large aspect ratios in nano-dimensions, ie in the range below the usual wavelengths of light, characterized in that the Surface of the silicon is exposed locally and then by means of a reactive ion etching primary needle-like silicon structures are generated with nano-dimensions and this structured silicon surface is then converted by thermal oxidation completely into secondary, also needle-like SiO 2 structures.

A further embodiment relates to a method for passivating the surface of silicon semiconductor devices by means of a SiO ₂ layer, which has acicular structures with large aspect ratios in nano-dimensions, ie in the range below the usual wavelengths of light, characterized in that a silicon layer on the surface is then deposited and then by means of a reactive ion etching primary needle-like structures in this silicon layer with nanodimensions are generated and this structured silicon layer is then converted by thermal oxidation completely or partially into secondary, also needle-like SiO 2 structures.

In a further embodiment, the necessary primary nanostructures in the silicon by reactive ion etching (reactive ion etching - RIE) using the working gases oxygen and SFε in a single process step without using additional means for targeted mask formation in structuring process only by adjusting the process parameters so that Oxygen at the reaction point on the silicon wafer shows a Selbstmaskiernde effect and takes place a self-organization of the needle-like structures.

With reference to FIGS. 6a to 6c, further embodiments of the invention will now be described in which silicon-based nanostructures are used as broadband optical absorbers.

Silicon surfaces with a self-assembled nanostructure created by an RIE process can serve as an excellent absorber, absorbing nearly all light in the range of 180-1100 nm. Likewise, they are well suited for the radiation delivery. By applying a thin additional layer, the wavelength range of the absorption and emission can be significantly extended.

This aspect relates to the use of structured surfaces of silicon crystal bodies which guarantee the highest possible absorption of light for a large wavelength range. For this purpose, the interface properties between two media must be changed so that between They no impedance jump, so no discontinuity of the refractive index occurs, but the different impedances continuously merge into each other.

Thus, the advantages are achieved that the needle-shaped silicon tips with a high aspect ratio in a statistically homogeneous distribution on the surface form an effective medium, which ensures the continuous transition of the two material properties. As a result, an absorption of more than 99% can be achieved in the entire visible range, for the modification of a silicon surface. Even beyond the visible range, such good absorption is achieved.

FIG. 6 a shows measurement results of the optical reflection of modified silicon surfaces in comparison to untreated surfaces.

FIG. 6b shows the extremely small and wavelength-independent reflection of the modified silicon surface in detail.

FIG. 6c shows the direct absorption measurement by photothermal deflection spectroscopy (PDS).

From a wavelength of 1100 nm, silicon becomes transparent and no longer absorbs any light. In order to still act as an absorber in the wavelength range above 1100 nm, the structured silicon surface can be coated, for example, with a thin metal layer. The metal takes over the function of the absorbent material, the surface modification is given by the structure in the silicon.

The invention not only acts in one direction, ie from material A to material B, but equally well in the opposite direction, from material B to material A. Thus, it also serves to improve the emission in the affected wavelength range.

The particular advantage of broadband and efficiently absorbing self-assembled nanostructures on the silicon surface can be advantageously exploited in many applications. Such layers may preferably be used in optical devices or components. An example of this is the lining of precision-optical devices or the absorption surface in digital projectors with mirror technology (digital light processing), in which the most complete absorption of the incident light is required in order to achieve the highest possible contrast value. Besides, it is for the color correct Representation necessary that the absorption properties over a large wavelength range are constant. Other applications arise everywhere, where it must be ensured that incident light wavelength-independent as completely as possible is converted into heat. Due to the good broadband properties, the invention can also be used as a reflection standard for very low reflection values. Another application is the improved radiation output, as occurs in optical components such as LEDs, or LASER. Due to the metal coating, an emission of heat radiation is possible. This can be used for targeted heat dissipation or for more efficient cooling. An interesting application in this regard is the reduction of a cooling surface of a component by the improved heat dissipation.

One embodiment relates to self-organized needle-like structures in nano-dimensions with dimensions smaller than the wavelengths of light and with an aspect ratio greater than 4: 1, which are produced using the working gases oxygen and SFe (without the use of additional means for targeted mask formation) on silicon surfaces. during the etching process in a single process step, as already explained, wherein these nanostructures are used in the form of layers as broadband optical absorbers for the lining of precision optical devices.

In a further embodiment, the needle-like structures are used as absorption surface in digital projectors with mirror technology (digital light processing).

In a further embodiment, the needle-like structures are used for devices in which the optical radiation is converted as completely as possible into heat independent of wavelength.

In a further embodiment, the needle-like structures are for purposes of improved radiation delivery as used in optical devices, e.g. LEDs, or LASER occurs, used.

In another embodiment, the needle-like structures for reflection standards are used for very low reflectance values.

In another embodiment, the needle-like structures are coated with a thin metal layer. In a further embodiment, the metal layer of the targeted heat dissipation.

With reference to FIG. 7, further embodiments are described in which IR windows with high transmission are provided.

The broadband anti-reflection of silicon by a self-assembled nanostructure created by the RIE method, as described above, can be excellently used as an IR (infrared) window. Nearly all light is transmitted in the range above 1100 nm.

Silicon can be used as an IR window. At a wavelength of greater than 1000 nm, silicon begins to become transparent and absorbs less and less light. Since the air / silicon interface has a reflection of more than 30% and a window always has two interfaces, an untreated piece of silicon, despite its transparency in the infrared, transmits only about 50% of the incident light quantity, the other half being lost by reflection.

By providing a nanostructure-based IR window as described above, the advantages are achieved that the self-assembled nanostructures created by the RIE process form an effective medium that provides for the smooth transition of the two material properties. As a result, a transmission of more than 90% can be achieved in the infrared range with the modification of the silicon surfaces. The modified surface fulfills its task by changing the interface properties between the silicon and air or vacuum such that no impedance jump occurs between them, but the different impedances continuously merge into one another. The material is not absorbent for the desired wavelength range. The interfacial modification of silicon serves to suppress reflection and thus improve transmission.

Important here is the shape of the needle-shaped structures of the surface. The structures form an effective medium, which ensures the continuous transition of the two material properties. A one-sided surface modification already achieves a transmission of about 70%.

FIG. 7 shows the transparency of a sample with a surface modified on one side. For infrared light from 1200 nm, the theoretical values of 70% are well hit. A problem with a two-sided surface modification is the low mechanical strength of the structures produced, so that the handling of the window is difficult.

The surface modification can be limited to specific areas with conventional photoresist masking techniques, so that mechanically stressed areas can be easily separated from optically transparent areas. Thus, the disadvantage of difficult handling is eliminated, a stable, possibly also air, liquid or vacuum-tight installation of such a window is readily possible.

One embodiment relates to an optical window made of silicon with improved broadband transparency in the IR range, wherein at least one of the two surfaces having the RIE method needle-like structures in nanodimensions with a high aspect ratio greater than 4: 1, using the working gases oxygen and SFε were generated in a self-organizing manner in a single process step, as explained above.

Both surfaces of the window can be provided with the pyramid-like needles.

In a further embodiment, the nanostructures are protected against mechanical destruction by means of an SOG (spin-on-glass) layer.

In another embodiment, the protection is hydrogen silsesquioxane (HSQ).

In a further embodiment, the reflection-reducing nanostructuring is limited to certain areas of the window by means of conventional masking technology in order to provide untreated, mechanically stable and easily sealable areas against air, liquids and vacuum.

_{* * * **** **** * *}

Claims

claims:

1. Photoelectronic component with

an optically active window for entry and / or exit of radiation,

a nanostructure, provided on a surface of the optically active window, with randomly distributed structural elements having an end area and a foot area, the tip of the end area having a lateral extent of less than 10 nanometers, and

Foot region has a lateral extent of 50 nanometers or more, and wherein an aspect ratio of the structural elements, that is, the height of the structural elements and the lateral extent at the foot, is on average greater than 4.

2. A photoelectronic device according to claim 1, further comprising a passivation layer which leaves the optically active window and forms a boundary therewith, wherein the structural elements extend substantially to the limit.

3. Photoelectronic component according to claim 1 or 2, wherein the structural elements are constructed of single-crystalline semiconductor material.

4. The photoelectronic component according to claim 3, wherein the nanostructure comprises a monocrystalline base layer on which the structural elements are arranged, and wherein a crystal defect density of the structural elements is substantially equal to the crystal defect density of the base layer.

5. The photoelectronic component according to claim 3, wherein the semiconductor material is silicon.

6. Photoelectronic component according to claim 1 or 2, wherein the structural elements are at least partially constructed of an insulating material.

7. The photoelectronic component according to claim 6, wherein the insulating material is silicon dioxide or silicide or silicon oxynitride.

8. The photoelectronic component according to one of claims 1 to 7, wherein a height of the structural elements in the range of 400 nanometers to 1500 nanometers.

9. The photoelectronic component according to claim 1, further comprising a planarization layer in the optically active window, wherein the structural elements of the nanostructure are embedded in the planarization layer.

10. The photoelectronic component according to claim 9, wherein the material of the planarization layer has a refractive index of 1, 5 or smaller.

11. A photoelectric device according to any one of claims 1 to 10, further comprising a second nanostructure at a second interface of the optically active window.

12. The photoelectric component according to claim 11, wherein the at least one nanostructure is embedded in a protective layer, in particular also the second nanostructure.

13. A sensor device comprising: a sensor surface formed by a nanostructure having randomly distributed structural elements, wherein the structural elements have an end region and a foot region, wherein the tip of the end region has a lateral extent of less than 10 nanometers and the foot region has a lateral extent of 50 nanometers or more, and wherein an aspect ratio of the structural elements, that is, the height of the structural elements and the lateral extent at the foot, is on average greater than 4.

14. Sensor device according to claim 13, wherein the sensor surface is formed of an insulating material.

15. Sensor device according to claim 13 or 14, further comprising a porous planarization layer for embedding the nanostructure.

16. Sensor component according to one of claims 13 to 15, wherein the structural elements are at least partially constructed of silicon.

17. Sensor device according to one of claims 13 to 16, wherein a height of the structural elements in the range of 400 nanometers to 1500 nanometers.

18. Optical coating structure for use in optical devices or optical devices, comprising a base layer and a randomly distributed nanostructure applied to the base layer

Structural elements, wherein the structural elements have an end region and a foot region, wherein the tip of the end region has a lateral extent of less than 10 nanometers and the foot region has a lateral extent of 50 nanometers or more, and wherein an aspect ratio of the structural elements, ie, the height the structural elements and the lateral extent at the foot area, on average greater than 4.

19. An optical coating structure according to claim 18, wherein the structural elements are constructed of single crystal semiconductor material.

The optical coating structure according to claim 19, wherein the base layer is single crystalline, and wherein a crystal defect density of the structural elements is substantially equal to the crystal defect density of the base layer.

21. The optical coating structure of claim 19, wherein the semiconductor material is silicon.

22. An optical coating structure according to claim 18, wherein the structural elements 5 are at least partially constructed of an insulating material.

The optical coating structure of claim 22, wherein the insulating material is silicon dioxide.

lo 24. An optical coating structure according to any one of claims 18 to 23, wherein the heights of the structural elements in the range of 400 nanometers to 1000 nm, in particular up to 1500 nanometers.

The optical coating structure according to any one of claims 18 to 24, further comprising i5 a planarization layer, wherein the structural elements of the nanostructure are embedded in the planarization layer.

26. The optical coating structure according to claim 25, wherein the material of the leveling layer has a refractive index of 1, 5 or smaller.

20

27. An optical coating structure according to any one of claims 18 to 26, wherein a conformal metal layer is deposited on the structural elements.

28. An optical coating structure according to claim 27, wherein the metal layer of the 25 serves targeted heat release.

29. An optical device having a liner as a broadband optical absorber, the liner comprising an optical coating structure according to any one of claims 18 to 28.

30

30. Digital projector with mirroring technology (digital light processing) having an absorption surface comprising an optical coating structure according to any one of claims 18 to 28.

31. Apparatus comprising means for wavelength-independent conversion of optical radiation into heat, wherein the means for converting optical radiation to heat comprises an optical coating structure according to any one of claims 18 to 28,

32. An optical element for emitting optical radiation with an exit window, which comprises an optical coating structure according to one of claims 18 to 28.

33. The optical element of claim 32, wherein the exit window is coupled to a light emitting diode or a laser.

34. Reflection reference device for determining low reflectance values, wherein the device comprises an optical coating structure according to one of lo claims 18 to 28.

35. Photoelectronic component having a reflection-reducing layer provided in the optically active window region of statistically regularly distributed, crystal defect-poor silicon needles with a height of 400 to 1500 i5 nm and an aspect ratio greater than 4: 1.

36. The photoelectronic component according to claim 35, wherein the silicon needles have a tip whose lateral extent is less than 10 nm.

37. The photoelectronic component according to claim 36, wherein a foot region of the silicon needles has a lateral extent of 50 nm or more.

38. A photoelectronic device according to any one of claims 35 to 37, which has a passivation layer which leaves free the active window area and forms a boundary therewith, wherein the silicon needles are formed to a distance to the boundary, which is smaller than a thickness , in particular a half thickness of the passivation layer.

39. An optical window, which has silicon and has a broadband transparency in the IR range, wherein at least one, preferably both surfaces of the window have needle-like structures in nanodimensions with an aspect ratio greater than 4: 1.

5

40. An optical window according to claim 39, wherein the needle-like structures are protected with nanodimensions with an SOG (spin-on-glass) layer from mechanical action.

lo 41. An optical window according to claim 40, wherein the protective material comprises hydrogen silsesquioxane (HSQ).

42. The optical window according to claim 39, wherein the needle-like structures with nanodimensions are delimited by means of conventional masking technology to specific areas of the window and thus remain untreated, mechanically stable and easily sealable areas against air, liquids and vacuum.

43. A method of adjusting the refractive index of an optically active window of a photoelectric device, the method comprising:

(a) generating a nanostructure in the surface region of the window by means of a self-organized plasma etching process for etching a silicon base layer;

(b) setting an aspect ratio of structural elements of the nanostructure generated in the silicon base layer to a value of four or higher depending on an operating wavelength range of the photoelectric device.

10

44. The method according to claim 43, wherein the structural elements are produced as needle-like structural elements during the plasma etching process by using the working gases oxygen (O ₂ ) and sulfur hexafluoride (SF ₆ ) without the use of additional means for targeted mask formation in a single process step, wherein the silicon base layer during the

Held at a constant temperature in the range of 27 ° C ± 5 ° C and working with a plasma power in the range of about 100W to 300W, with higher plasma power at higher process pressures are necessary and the ratio of working gas flows depending on the o geometric Anlagenparametem is set so that the oxygen in the

Reaction point on the silicon base layer shows a self-masking effect, which can be achieved in the range of gas flows for SF ₆ : between 50 to 150 sccm and for O ₂ : between 20 to 200 sccm and the process time is only a few minutes. 5

45. The method of claim 43 or 44, further comprising forming a protective layer for the nanostructure having a substantially planar surface.

46. Method according to claim 45, wherein in several steps a spin-on-glass (SOG) adapted in its properties to the requirements of the nanostructure to be passivated is applied and annealed after each application until the planar surface is formed.

47. The method of claim 46, wherein SOG layers of hydrogen silsesquioxanes (HSQ) are applied.

48. A method for a method according to any one of claims 43 to 47, wherein an additional layer is applied prior to the plasma etching, which as a buffer layer has a relation to the silicon base layer (3) with the same process parameters in such a changed etching behavior that a

5 forms a self-organized structure with only a relatively low aspect ratio, which is reduced in the course of the process.

49. The method of claim 48, wherein the aspect ratio of the structural elements of the nanostructure is adjusted across the thickness of the buffer layer.

10

50. The method of claim 48 or 49, wherein the buffer layer is a SiO ₂ layer.

51. A method according to any one of claims 48 to 50, wherein the buffer layer has a thickness of from 20nm to 100nm.

52. The method according to any one of claims 48 to 51, wherein the buffer layer is not completely removed.

53. Method of protecting a layer having high aspect ratio acicular structural elements and forming part of a chemical sensor, wherein in several steps a gas permeable porous spin adapted in its properties to the requirements of the layer to be passivated

Applied on-glass and annealed after each application until a smooth

Surface is formed.

54. The method of claim 53, wherein in a last step, an overlap with a non-porous layer takes place, which is removed in the last mask process lo again.

55. The method of claim 53 or 54, wherein the needle-shaped structural elements have a tip with a lateral extent of less than 10 nm.

i5 56. The method according to any one of claims 53 to 55, wherein the needle-shaped structural elements are constructed of monocrystalline silicon.

57. The method according to any one of claims 53 to 55, wherein the needle-shaped structural elements comprise silicon dioxide.

20

58. The method of any one of claims 53 to 57, wherein the spin-on glass is liquid pervious.

59. A method of passivating the surface of a semiconductor device comprising silicon, the method comprising:

local exposure of a portion of the surface,

5

Forming primary needle-shaped structures with large aspect ratios in nano dimensions with lateral dimensions in the range below the wavelengths of visible light by means of a reactive ion etching process and

10

Modifying the region provided with the primary acicular structures to produce secondary, also needle-like, insulative surface structures.

i5

60. The method of claim 59, wherein a silicon layer is deposited to provide the surface.

61. The method of claim 59 or 60, wherein modifying the region provided with the primary acicular structures comprises thermal oxidation.

20

62. The method of claim 59, wherein modifying the region comprises nitriding the region.

63. The method of claim 61, wherein the silicon in the primary needle-like structures is substantially completely converted to silica.

64. The method according to any one of claims 59 to 63, wherein the primary needle-like structures by reactive ion etching (reactive ion etching - RIE) using the working gases oxygen and SFε in a single process step

30 without the use of additional means for targeted mask formation

Structuring process only by adjusting the process parameters so that the oxygen in the reaction point on the silicon-containing surface shows a self-masking effect and a self-organization of the needle-like structures takes place, can be generated.

35

65. A method for producing self-assembled nanostructures in the region of a window opening of an integrated circuit with optoelectronic component or a discrete optoelectronic component, wherein for the production of needle-like silicon structures with nanometer dimensions and

5 with aspect ratios greater than 4: 1 in a reactive ion etching (RIE) process using only the working gases oxygen and SF ₆ in a single process step and without the use of additional micro-masking at the beginning of the etch process, the process parameters reducing silicon consumption in the process Generation of the nanostructure so lo be set so that the oxygen in the reaction point on the silicon wafer has a self-masking effect and wherein prior to the etching process, an additional layer (406) is applied, which has as a buffer layer compared to the silicon at the same process parameters so far changed etching behavior in that self-organized, not yet needle-like structures i5 form in the buffer layer with only a relatively low aspect ratio, which initially become holey in the course of the process in the manner of a mask layer (406a), and continue to be almost completely removed Rd, in favor of the needle-like structures (403) in the silicon, which with the nanometer dimensions and with the aspect ratio greater than four to one through the holey

20 mask layer (406a) are formed, wherein this mask layer is at least almost completely removed.

66. The method of claim 65, wherein the aspect ratio of the silicon nanostructure is adjusted across the thickness of the buffer layer.

25

67. The method of claim 65 and 66, wherein the buffer layer is a SiO ₂ layer.

68. The method of claim 65, wherein the buffer layer has a thickness of 20 to 30 100nm.

69. The method of claim 65, wherein the buffer layer is not completely removed and remains on the tips of the needle structures.

35

* * * **