Semiconductor device manufacturing: process – Forming bipolar transistor by formation or alteration of... – Self-aligned
Reexamination Certificate
2003-03-17
2004-09-14
Gurley, Lynne A. (Department: 2812)
Semiconductor device manufacturing: process
Forming bipolar transistor by formation or alteration of...
Self-aligned
C438S391000, C438S674000, C438S681000, C438S685000, C438S686000
Reexamination Certificate
active
06790737
ABSTRACT:
BACKGROUND OF THE INVENTION
Field of the Invention
The invention relates to a method for producing metal layers on semiconductor substrates.
Economic success in the semiconductor industry is substantially influenced by further reduction of the minimum feature size that can be fashioned on a microchip. Reducing the minimum feature size makes it possible to increase the integration density of the electronic components, such as transistors or capacitors, on the microchip and hence to increase the computing speed of processors and also to increase the storage capacity of memory modules. In order to keep down the area requirement of the components on the chip surface, a depth of the substrate is also utilized for capacitors. To that end, first a trench is introduced into a wafer. Afterward a bottom electrode is produced, for example by the regions of the wafer that adjoin the wall of the trench being doped in order to increase the electrical conductivity. A thin layer of a dielectric is then applied to the bottom electrode. Finally, the trench is filled with an electrically conductive material in order to obtain a counter electrode.
The counter electrode is also referred to as a top electrode. The configuration of electrodes and dielectric results in that the capacitor is virtually folded. With consistently large electrode areas, that is to the same capacitance, it is thereby possible to minimize the lateral extent of the capacitor on the chip surface. Such capacitors are also referred to as “deep trench” capacitors.
In memory chips, the charged and discharged states of the capacitor correspond to the two binary states 0 and 1, respectively. In order to be able to reliably determine the charge state of the capacitor and thus the information stored in the capacitor, the latter must have a specific minimum capacitance. If the capacitance or, in the case of a partly discharged capacitor, the charge falls below the limit value, the signal disappears in the noise, that is to say the information about the charge state of the capacitor is lost. After writing, the capacitor is discharged by leakage currents that bring about a charge balancing between the two electrodes of the capacitor. With decreasing dimensions, the leakage currents increase since tunneling effects gain in importance. In order to counteract a loss of information through the discharge of the capacitor, the charged state of the capacitor is checked at regular intervals and if appropriate refreshed, that is to say a partly discharged capacitor is charged again up to its original state. However, technical limits are imposed on these so-called “refreshing” times, that is to say they cannot be shortened arbitrarily. During the period of the refreshing time, therefore, the charge of the capacitor is permitted to decrease only to an extent such that reliable determination of the charge state is possible. For a given leakage current, the capacitor must therefore have a specific minimum charge at the beginning of the refreshing time, so that, at the end of the refreshing time, the charge state is still high enough above the noise to be able to reliably read out the information stored in the capacitor. In order to be able to achieve a sufficient capacitance of the capacitors with low leakage currents even in the case of advancing miniaturization, a multiplicity of solution approaches are being pursued. Thus, by way of example, the surface of the electrodes is provided with a structure in order that, as the length and width of the electrodes decrease, the surface thereof is made as large as possible. Furthermore, new materials are being used. Thus, attempts are being made to replace the silicon dioxide, which has been used hitherto as a dielectric, by materials with a higher dielectric constant.
As electrode material, polysilicon is currently used to fill the trench. With further miniaturization, that is to say a smaller diameter of the trench, the layer thickness of the conductive material decreases, so that the electrical conductivity of the polysilicon no longer suffices to provide the required charge.
In order to combat a loss of capacitance of the capacitors in the context of advancing miniaturization, electrodes made of metals having higher electrical conductivity, for example platinum or tungsten, are used instead of the currently used electrodes made of doped polysilicon. As a result, it is possible to suppress depletion zones in the electrodes and thus to fabricate thinner electrodes which nevertheless provide the required charge density on the electrodes.
U.S. Pat. No. 5,905,279 describes a trench capacitor in which, in addition to polysilicon, further electrically conductive materials, such as WSi, TiSi, W or Ti are also used to fill the trenches.
Trench capacitors have a very high aspect ratio of usually more than 60. The term aspect ratio denotes the ratio of the extent of the capacitor in its longitudinal direction, that is to say into the depth of the substrate, to the diameter of the opening of the capacitor at the surface of the substrate. The high aspect ratio leads to difficulties in the construction of the trench capacitor. A trench which has been introduced into the wafer for the construction of a trench capacitor has, on the one hand, a very small opening at the substrate surface, through which substances can be transported into the trench in order to be deposited there, but, on the other hand, a very large extent into the depth of the substrate, in which case the material to be deposited has to be able to penetrate down to the bottom of the trench. During the deposition of layers in the trench, for example in order to produce a dielectric disposed between the bottom electrode and the top electrode, the layer thickness is intended to be as uniform as possible in the entire trench. Only a few methods are suitable for fabricating such layers. The deposition is usually effected with the aid of a chemical vapor deposition (CVD) or atomic layer deposition (ALD) method. In this case, use is made of gaseous precursors that are converted into the desired compounds at the substrate surface. In the CVD method, the reactants are simultaneously situated in the gas space above the substrate. The material to be deposited is deposited as a result of the conversion of the reactants on the substrate surface. With this method, relatively thick layers can be produced in comparatively short times, but fluctuations in the layer thickness, caused for example by flow effects, have to be accepted. In the ALD method, the layers are constructed by depositing individual layers of the various reactants. Thus, only ever one reactant is situated in the gas space above the substrate, and is deposited in a monomolecular layer on the substrate. Afterward, excess reactant is removed from the gas space, for example by pumping away or flushing with an inert gas, after which a further reactant is introduced into the gas space above the substrate. The further reactant reacts with the reactant previously bonded as a monomolecular layer on the substrate, and likewise forms a monomolecular layer. This makes it possible to fabricate very uniform layers with a defined layer thickness. Both CVD and ALD methods require gaseous reactants. Furthermore, on the one hand the reactants must be sufficiently reactive to be able to produce a layer in tenable process times; on the other hand the reactants must also be stable enough not to decompose before the actual deposition. In the case of the ALD method, the reactant must be able to form a monomolecular layer that remains stable until the deposition of the further reactant. This greatly restricts the selection of the reactants. Such precursor compounds are not available for a relatively large number of metals. Furthermore, the reaction products liberated during the reaction of the reactants must not attack the substrate. Thus, by way of example, WF
6
as a gaseous precursor compound is used for the fabrication of thin tungsten layers on a silicon substrate, in which case, fluorine is liberated during the conve
Jaeger Wolfgang
Rogalli Michael
Schneegans Manfred
Greenberg Laurence A.
Gurley Lynne A.
Infineon - Technologies AG
Mayback Gregory L.
Stemer Werner H.
LandOfFree
Method for fabricating thin metal layers from the liquid phase does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for fabricating thin metal layers from the liquid phase, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for fabricating thin metal layers from the liquid phase will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3270018