DE19715501C1 - Method for structuring thin metal layers. - Google Patents

Abstract

The method consists of the following steps: (a) provision of a structurable layer (2); (b) production of depressions (3) in this layer; (c) application of a layer (5) of copper, a noble metal or a noble metal alloy; (d) removal of the layer (5), except from the depressions (3). The method is characterised by the fact that removal of the layer (5) takes place by means of a laser beam. Alternatively method includes the following steps: 1) production of a thin layer (7) of copper, a noble metal or noble metal alloy on a substrate; 2) application of a structurable layer (8); 3) structurisation of the layer (8) so that the layer (7) is exposed within certain regions; (4) removal of the layer (7) from these regions. This alternative method is characterised by the fact that the step (4) is carried out with use of a laser beam.

Description

The present invention relates to a method for Structuring of thin metal layers, especially in the Area of semiconductor technology in the metallization of integrated circuits is used.

For the metallization of integrated circuits, in particular with CMOS technology, aluminum is currently predominantly used or aluminum with small additions of silicon and copper used. With increasingly closer integration with However, structure sizes below 1 µm occur with the metal aluminum due to its insufficient conductivity and the occurring Electromigration problems.

Currently, copper is used worldwide as Metallization material is being investigated as it has a lower specific resistance and a higher strength with respect to Has electromigration (see, for example, Jian Li, "Copper-Based Metallization for ULSI Application ", MRS Bulletin June 1993 P. 18). This allows higher integration densities and faster Cycle times of the manufactured circuits. The deposit thinner Copper layers are made by means of CVD or sputtering. However, the structuring of these layers prepares Conductor tracks, bond pads and vias from the following Reasons problems.

The usual structuring of aluminum, i. H. Photolithography and subsequent etching in plasma, can only be partially resolved Transfer copper. Copper as a transition element forms im In contrast to aluminum, it does not contain any volatile compounds Corrosive gases that contain fluorine or chlorine (see e.g. Y. Igarashi, "High-Reliability Copper Interconnects through Dry Etching Process ", Japan. J. Appl. Phys. Vol. 34 (1995) pp. 1012-1015). Copper must therefore be etched at high temperatures.

There are other possibilities for etching copper in Use of cyanides and the use of the reverse reaction of the CVD Deposition, as for example in M. Schober, "Low Temperature Dry Etching of Copper Using a New Chemical Approach, "Proceedings of MAM '97, Materials of Advanced Metallization, p. 30 is. However, these chemicals are toxic and therefore for reasons work safety and environmental protection as much as possible avoid.

A third method, called the Damascene technique, is described in p. Lakshminarayanan, "Contact and Via Structures with Copper Interconnects Fabricated using Dual Damascene Technology ", IEEE Electron Device Letters, Vol. 15, No 8, Aug 1994, p. 307 explained. In this technique, the copper is in beforehand structured trenches deposited. With CMP (Chemical Mechanical Polishing) the excess copper is subsequently outside the trenches removed by grinding. Disadvantages of this technique are the dishing effect and the high buildup of liquid Hazardous waste.

In contrast, from US-A-5075200 a method is known in which a superconductor precursor on a pre-structured layer is deposited over the entire surface and then by means of ionic Electron or gamma radiation the areas outside the Structures including mask are removed. Furthermore, from JP A-08-236240 a method known in which a metal in a Notch is made and the protruding parts through Polishing to be removed. In addition, US-A- 5266446 a process in which excess metallic Material is removed by chemical-mechanical polishing. In Similarly, in US-A-5130229 excess metallic Material removed by chemical-mechanical polishing.

From the post-published document US-A-5635423 is a Damascene process known in which a conductive wiring is planarized with the aid of chemical-mechanical polishing. Furthermore, from DE 44 20 347 C2 is a Method for the adjustment of electronic components, which are made of a substrate and at least one on the substrate arranged, the electrical parameters of the respective Component-determining layer exist and where the Adjustment of parameters without removing substrate material by a the thickness of the layer reducing the area of material removal takes place, known. In this process, the flat Material is removed with at least one laser beam, whose focus is at a defined distance above the surface of the layer to be removed is located.

In addition, the general use of laser technology is in the production of printed circuit boards from the article metal surface 8/1991, pages 349 to 358.

The object of the present invention is to provide a Process for structuring thin metal layers without Provide use of toxic or expensive chemicals. the end For reasons of cost, the procedure should be carried out with as few as possible Get by process steps. For a high yield, the process should easily controllable and independent of those that are difficult to control Be side effects.

This task is with the characteristics of the applicable Claims 1 and 2 solved. Advantageous further training of the Invention are the subject of the subclaims.

In the method proposed according to the invention, a structurable layer is first provided. This can be done, for example, by depositing an SiO ₂ layer on a substrate. In this layer, depressions are created that define the structure of the later metal layer. The depressions usually consist of a series of trenches and holes, which correspond to the course of later conductor tracks and the position of later plated-through holes (vias, contact holes). The depressions can be produced, for example, by means of photolithography technology and subsequent etching. The thin metal layer is then applied either directly to this structured layer or to a thin intermediate layer (or layer sequence) applied to it. The structuring of the thin metal layer takes place in that it is removed again outside the depressions by laser ablation.

With laser ablation is the evaporation of a material through called high-energy laser pulses. Pulsed lasers will though used for a long time for material processing (cf.e.g. R. F. Haglund, in J. C. Miller "Laser Ablation" Springer Verlag, 1994, Vol 28; P. 11 ff.). Laser ablation is used in semiconductor manufacturing however, usually not applicable because the to be generated Structures are often smaller than the wavelength of the laser light, and a successive writing exposure takes too long. the However, the present invention enables in an advantageous manner the use of this technology to create structures below the resolution limit of the laser radiation used.

Another way of structuring thinner According to the present invention, metal layers consist in first of all to apply the thin metal layer to a substrate. A structurable layer is then placed on top of this layer applied and structured in such a way that the underlying Metal layer is exposed in certain areas. The thin one Metal layer is eventually made from the exposed areas removed by laser ablation. In this case, too, remains a structured thin metal layer.

In both of the above cases, any type of carrier is to be considered under substrate understand, for example a wafer with an already applied layer sequence.

The method has the advantage that it is not toxic or expensive Chemicals for structuring thin metal layers are necessary. The procedure is with a few process steps and therefore inexpensive to carry out and has no difficult controllable side effects.

The method can be used in an advantageous manner, conductor tracks and Vias in copper without toxic or expensive Chemicals are produced.

The invention is explained below on the basis of exemplary embodiments explained in more detail in connection with the drawings. Show it

Fig. 1 shows an example for the process according to claim 1 based on individual process steps; here

Figure 1a shows a silicon wafer (1) with a patternable layer (2) after the generation of recesses (3) and the applying an adhesive layer (4).

FIG. 1b shows the silicon wafer from FIG. 1a after a thin copper layer ( 5 ) has been deposited over the entire surface;

1c shows the silicon wafer of Figure 1b with the finished conductor track (6) after the removal of the copper layer by means of laser ablation outside the wells..;

FIG. 2 shows an example of the method according to claim 2 based on individual process steps; here

2a shows a silicon wafer (1) with an oxide (2) and an adhesive layer (4) on which a thin copper layer (7) and a structurable reflection layer (8) are deposited.

FIG. 2b shows the substrate from FIG. 2a after structuring the reflective layer ( 8 ); FIG. and

Fig. 2c, the substrate of Fig. 2b with the finished conductor (10) after removal of the copper layer (7) by laser ablation of the exposed portions (9).

In an exemplary embodiment of the method according to the invention, the structure for the conductor tracks and vias in an intermetallic dielectric ( 2 ), usually SiO ₂ , is initially produced with photo lithography and subsequent etching in accordance with the Damascene technique. The created depressions ( 3 ) are then coated with an adhesive layer ( 4 ), which also acts as a diffusion barrier. The adhesive layer usually consists of TiN.

Fig. 1a shows a silicon wafer ( 1 ) on which an oxide layer ( 2 ) is applied. A trench (3 ) defined by means of photolithography was etched out of the oxide layer ( 2 ) and an adhesive layer ( 4 ) was then applied.

Subsequently, copper ( 5 ) is deposited over the entire surface with a typical thickness of 0.2 to 2 μm, as shown in FIG. 1b. Possible deposition methods are CVD (Chemical Vapor Deposition) from organometallic precursors (e.g. CupraSelect®), PVD (Physical Vapor Deposition) such as sputtering and vapor deposition, electroless deposition or galvanic deposition. After the copper plating, the ablation takes place, the essential step of the method according to the invention. For this purpose, the wafer is heated to a basic temperature in the range of approx. 200 ° C to 500 ° C in a closed process chamber. The lower temperature limit is defined here by an increasingly higher threshold energy of the ablation, the upper limit by possible damage to already finished layers in the wafer ( 1 ). The process chamber must be evacuated or filled with an inert atmosphere in order to prevent oxidation of the copper by the oxygen in the air. Alternatively, it can also contain a reactive process gas.

The copper-plated wafer is now all over or in sections exposed with successively overlapping fields from the laser. A laser with short-term, high-energy pulses. The short pulses with a typical duration from 1-100 ns are necessary to a small volume of the metal to evaporate suddenly without the environment by heat dissipation to warm up. An excimer laser is preferably suitable for this. Also an Nd: YAG laser or a diode laser, if possible with Frequency doubling or quadrupling can be used will. The absorption of the laser radiation depends on the wavelength addicted. For example, copper reflects in the visible Spectral range with λ <500 nm is very good, while in the ultraviolet spectral range, e.g. B. at the line of the KrF with λ = 248 nm, the absorption of copper is good. It goes without saying even that of those skilled in the art for successful ablation in each case will choose a laser wavelength at which the to be removed Metal has significant absorption.

Exceeding an energy density threshold, which is in the order of magnitude of 0.5 J / cm ^2, is decisive for successful ablation. With a lower energy density, only a modification of the metal surface takes place. If the energy density is too high, the underlying layer will be damaged. When selecting the adhesive layer, it must therefore be ensured that the threshold of the energy density of the destruction of the adhesive layer is well above the energy density used.

For example, good results get thinner with ablation Copper layers obtained under the following conditions:

Layer thickness of copper; 800 nm Layer thickness of the TiN adhesive layer: 25 nm. Base temperature of the wafer: 400 ° C Vacuum in the process chamber: 0.1 Pa (1. 10 ^-3 mbar) Used laser: KrF excimer laser (wavelength 248 nm) Energy density of the laser beam: 0.5 J / cm 2

In order to achieve the necessary energy density, the laser beam has to be reduced in its cross-section with an optical element (one or more lenses or mirrors), ie it has to be focused. The energy density is adjusted by controlling the intensity of the laser, by varying the focus, or by using an adjustable beam attenuator. With each laser flash, the copper is evaporated on the vertically illuminated surface, while the copper remains in the trenches and holes. Fig. 1c shows a trench produced in this manner, filled with metal, ie a complete conductor track (6). The desired structuring has thus already taken place.

The process chamber leaves an aerosol of copper dust, which can be easily filtered out, or a vaporous one Connection when using a reactive process gas. It there is no toxic waste or by-products. As another If necessary, process steps are followed by cleaning and separation an upper diffusion barrier (TiN) and another Intermetallic dielectric.

The selectivity of the ablation between the metal to be removed on the flat surface and the metal to be preserved in the trench is based in the simplest case on the dependence of Energy density threshold from the layer thickness of the metal (here: Copper). To be able to use this effect, the Deposition process in the depressions for the conductor tracks and Contact holes create a thicker layer than outside. at non-selective deposition and surface-controlled process this is for depressions with an aspect ratio (depth / width) greater than 1 is the case. A selective deposition process or one inserted between deposition and ablation Tempering step with a flowing of the copper layer can Increase the layer thickness in the trenches as well or additionally.

In the following, a further embodiment of the invention is explained, with which surfaces of copper of uniform thickness and laterally large can be structured. Here the structuring takes place by depositing an additional (structurable) layer ( 8 ) (e.g. SiO ₂ or TiN) on the copper ( 7 ). This is shown in Fig. 2a, in which a silicon wafer ( 1 ) with an oxide layer ( 2 ), an adhesive layer (TiN) ( 4 ), the copper layer ( 7 ) and the additional layer ( 8 ) is shown. The thickness and the refractive index of this layer ( 8 ) are chosen so that the light cannot couple into the underlying copper (reflective layer). This layer is structured with normal lithography steps in such a way that the copper layer ( 7 ) is exposed in certain areas ( 9 ) (cf. FIG. 2b). The covered areas define the later structure of the metal layer. The copper is then ablated from the opened areas ( 9 ) ( FIG. 2c). The desired conductor tracks or areas ( 10 ) remain. The residues of the auxiliary layer ( 8 ) remaining on the conductor tracks ( 10 ) can then be removed.

In both exemplary embodiments, it is advantageous to also design the adhesive layer ( 4 ) used under the copper layer as a reflection layer. In this way, the light transmitted through the thin copper layer is reflected back into the copper, increasing the effectiveness of the laser pulse and reducing the necessary energy density.

The process chamber can either be operated with a vacuum during the ablation (a vacuum of 0.1 Pa (1. 10 ^-3 mbar) is sufficient) or contain a process gas. The additional gas can be an inert gas (nitrogen or noble gas), which removes the evaporated copper from the reaction space through its flow, or a reactive gas, which combines with the hot and excited atoms or clusters of copper and a compound with a higher vapor pressure which does not condense on adjacent surfaces.

For improved coupling of the laser radiation into the partially reflective metal, an antireflection _{layer (e.g. SiO 2} ) of suitable thickness can be applied to the areas to be removed, if necessary. This auxiliary layer is then ablated together with the copper.

To increase the effect of the ablation, the copper can be used during a small amount of foreign matter is added to the deposit or as a separating layer between the metal and the adhesive layer be deposited beforehand. These foreign substances are involved in Irradiation with the laser suddenly evaporates and supports the residue-free removal of the metal. For good adhesion and permanent stability of the copper in the finished component this foreign substance, however, with a tempering step (longer time at elevated temperature, possibly in forming gas) completely remove his. As foreign substances come here z. B. Hydrocarbons into consideration.

In line with current technical developments, the Embodiment copper as metal for metallization described. Others are also used in special components Precious metals like gold, silver, platinum, palladium and theirs Alloys used, which in the structuring by chemical etching cause problems. Also for these metals is the method according to the invention suitable.

Claims (25)

1. A method for structuring thin metal layers, especially in the field of semiconductor technology, with the following steps:

• 1. - Provision of a structurable layer ( 2 );
• 2. Creation of depressions ( 3 ) in the structurable layer ( 2 );
• 3. Application of a thin metal layer ( 5 ) made of copper, a noble metal or a noble metal alloy over the entire surface; and
• 4. - subsequent removal of the thin metal layer ( 5 ) outside the depressions ( 3 ); characterized in that the thin metal layer ( 5 ) is removed by laser ablation.

2. Process for structuring thin metal layers, especially in the field of semiconductor technology, with the following steps:

• 1. - Application of a thin metal layer ( 7 ) made of copper, a noble metal or a noble metal alloy to a substrate;
• 2. Application of a structurable layer ( 8 );
• 3. Structuring of the structurable layer ( 8 ) in such a way that the thin metal layer ( 7 ) underneath is exposed in certain areas ( 9 ); and
• 4. - subsequent removal of the thin metal layer ( 7 ) from exposed areas ( 9 );
  characterized in that the thin metal layer ( 7 ) is removed by laser ablation.

3. The method according to claim 1, characterized in that the thin metal layer ( 5 ) is deposited so that its layer thickness in the depressions ( 3 ) is greater than outside the depressions. 4. The method according to any one of claims 1 or 3, characterized in that the depressions ( 3 ) are produced with an aspect ratio greater than 1. 5. The method according to any one of claims 1, 3 or 4, characterized in that the thin metal layer ( 5 ) is applied by a selective deposition process. 6. The method according to any one of claims 1, 3 to 5, characterized in that a tempering step is carried out between the application and the removal of the thin metal layer (5). 7. The method according to any one of claims 1, 3 to 6, characterized in that an intermetallic dielectric is used as the structurable layer (2). 8. The method according to any one of claims 1 to 7, characterized in that the structurable layer ( 2 , 8 ) is a reflective layer for the wavelength of the laser radiation used. 9. The method according to any one of claims 1 to 8, characterized in that the thin metal layer ( 5 , 7 ) consists of gold, silver, platinum or palladium or an alloy of these. 10. The method according to any one of claims 1 to 9, characterized in that the thickness of the thin metal layer ( 5 , 7 ) is 0.2 to 2 microns. 11. The method according to any one of claims 1 to 10, characterized in that the thin metal layer ( 5 , 7 ) is heated to a base temperature in the range of 200-500 ° C to carry out the laser ablation. 12. The method according to any one of claims 1 to 11, characterized in that the thin metal layer ( 5 , 7 ) is exposed over the whole area to the laser radiation for performing the laser ablation. 13. The method according to any one of claims 1 to 11, characterized in that the thin metal layer ( 5 , 7 ) for carrying out the laser ablation is exposed in sections with successively overlapping fields of the laser radiation. 14. The method according to any one of claims 1 to 13, characterized marked, that the laser ablation takes place in a process chamber that is evacuated. 15. The method according to any one of claims 1 to 13, characterized marked, that the laser ablation takes place in a process chamber that has a contains inert gas. 16. The method according to any one of claims 1 to 1, 3, characterized marked, that the laser ablation takes place in a process chamber that has a contains reactive process gas. 17. The method according to any one of claims 1 to 16, characterized marked, that for laser ablation an excimer laser, a Nd: YAG laser or a diode laser with a pulse duration in the range of 1-100 ns each is used. 18. The method according to any one of claims 1 to 17, characterized in that an adhesive layer ( 4 ) is deposited as a diffusion barrier before the thin metal layer (5 , 7) is applied. 19. The method according to claim 18, characterized in that the adhesive layer ( 4 ) represents a reflection layer for the wavelength of the laser radiation used, or a separate reflection layer is deposited. 20. The method according to any one of claims 1 to 19, characterized in that an anti-reflective layer for the wavelength of the laser radiation used is applied to the areas of the thin metal layer (5 , 7) to be removed before the laser ablation. 21. The method according to any one of claims 1 to 20, characterized marked, that the metal during the application a small amount of Foreign substances are mixed in, which later through a Annealing step can be removed again. 22. The method according to any one of claims 1 to 20, characterized in that before the application of the metal layer ( 5 , 7 ) a separating layer, which can later be removed again by a cleaning step, is deposited. 23. The method according to any one of claims 1 to 22, characterized in that the structuring of the structurable layer ( 2 , 8 ) takes place by photolithography and subsequent etching. 24. The method according to any one of claims 1 to 23, characterized in that an upper diffusion barrier is further deposited on the metal layer (5 , 7 ). 25. The method according to any one of claims 1 to 24, characterized marked, that a further structurable layer is deposited.