The present invention generally relates to polishing pads, in particular for chemical-mechanical polishing (CMP) with the use of a slurry. CMP is a process step in the semiconductor fabrication sequence that has generally become an integral part of the manufacture of semiconductor wafers. The process is used in a variety of applications in the semiconductor fabrication sequence. A summary of the different applications would include that which is referred to as “oxide” or “ILD/PMD”, “STI”, “copper”, “barrier”, “poly” and “tungsten”, the terms generally indicating the material that is being removed. The common theme relating all of these applications is that CMP is required to expediently remove material and planarize the surface, while leaving it defect and contamination free. These applications generally require the use of different slurries, and their mechanism of removal is therefore also generally different. Because of that, the optimal condition of each of the applications tends to be different as well.
The manufacture of integrated circuits consists of a large number of steps performed in sequence and can be generally described by one of two process flows, where one flow is often referred to as “Aluminum back end” and the other is often referred to as “copper back end”. Of these two, the aluminum process is technologically older, while the copper process is newer. A general description of the aluminum back end is as follows:
Starting with bare silicon, the transistors are outlined on the wafer and are electrically insulated from each other by filling trenches etched in the silicon with an oxide, usually SiO2. The oxide overburden is removed and planarized using an STI process. The fabrication of the transistors is completed and they are covered with another SiO2 layer, often a doped oxide. This layer is planarized using a PMD process. Vias are etched and filled with tungsten to make contact to the transistors. The overburden is removed and the tungsten planarized using a tungsten process. Aluminum is deposited, patterned, and etched to create conductive interconnect lines. Subsequent alternating oxide and aluminum layers are created, where in each case the oxide layer is planarized using an ILD process. This is continued until the completion of all the layers.
A general description of the copper back end is as follows: Starting with bare silicon, the transistors are outlined on the wafer and are electrically insulated from each other by filling trenches etched in the silicon with an oxide, usually SiO2. The oxide overburden is removed and planarized using an STI process. The fabrication of the transistors is completed, often using a process which is the inverse of the method used to make the gates typically used in the aluminum process. The oxide is etched and filled with polysilicon. The overburden is removed and planarized using a poly process. An oxide layer is deposited over the gates and often etched for a tungsten deposition known as Local Interconnect. The CMP process here would also be a tungsten process. Another oxide layer is deposited and channels and vias etched in the oxide, which are filled with copper. The copper is then polished using a copper process. Subsequent layers of oxide and copper are deposited, but in this case the CMP is applied to the copper layer rather than the oxide layer. The barrier is a material which is deposited below the copper so as to prevent the copper from diffusing into the oxide and into the devices. This barrier material is typically Ti or TiN, and it is removed by a barrier CMP step which follows the copper step.
In any of these CMP processes, the silicon substrate is forcibly placed in direct contact with a moving polishing pad. A wafer carrier applies pressure against the backside of the substrate, usually while simultaneously forcibly applying rotation. During this process a slurry is made available, and is generally carried between the wafer and the pad by the motion of the pad. The elements contained in the slurry are chosen by the CMP application. In general, slurries that are designed to remove insulating materials consist of water, an abrasive and an alkali formulation designed to “hydrolyze” the insulating material. Copper slurries on the other hand, tend consist of water, an abrasive, an oxidizing agent, a complexing agent, and a chemical to passify the surface. A typical slurry often has very low removal rate on a material it was not designed to remove.
The presence of grooves is instrumental in delivering the slurry to the wafer-pad interface, where it is required for the process to be carried out. The slurry enables the polishing process to occur by chemically reacting with the material which is being polished. The pattern, pitch, width and depth of these grooves are generally known to be an important part of the process. Grooves are discussed in various patents. See, for example, U.S. Pat. Nos.: 6,645,061; 6,439,989; 6,241,596; 5,984,769; 5,921,855 and 5,489,233. The patterns recognized include substantially circular, spiral, multiple spiral, wavy concentric, off-center concentric, disjoint concentric, oscillating radial, arcuate, x (straight and parallel), x-y, grooves of different pitch and combinations thereof, deep and shallow, wide and narrow grooves and combinations thereof. Additional patterns include fractal, perforated, hexagons, triangles and tire-tread. The groove profile may be rectangular with straight side-walls or the groove cross-section may be “V”-shaped, “U”-shaped, triangular, or tetragonal. Also the groove design may change across the pad surface.
The purpose of grooves on CMP pads can be summarized as follows:
1. Grooves help prevent the wafer from hydroplaning. If the pad is smooth and without channels or perforations, a continuous boundary layer of slurry can form at the pad wafer interface, preventing intimate pad-wafer contact and significantly reducing removal rate.
2. Grooves ensure the transport of slurry to the center of the wafer. Because of the motion of the pad, slurry tends to reach the edges of the wafer without the need for grooves. But a plethora of data shows that the absence of grooves causes the rate to drop toward the center of the wafer, implying the center has been starved of slurry.
3. Grooves reduce the area of contact between the pad and the wafer, increasing the local pressure. This is not so important for the wafer where the mechanisms provided by most commercial CMP tools is adequate to deliver the desired downforce and thus pressure, but is very useful for conditioning, a process of forcibly applying a diamond or abrasive studded disk against the moving pad to create roughness. The mechanism for this action provided by most commercial CMP tools is often inadequate compared to what is desired. By reducing the pad/conditioner area of contact, a given provided force results in a higher local pressure.
4. Grooves provide air under the pad so as to avoid the phenomenon known as “stiction”. During the course of polishing, the surface of the pad in the wafer track (the donut shaped area created by the rotation of the pad) tends to become smoother, even in the face of conditioning. Together with the help of the slurry, the wafer and pad tend to “mate”, i.e. form a very close contact at all places. This results in a well known sticking force, which causes the requirement of higher force to lift the wafer off the pad surface after the end of the polishing process. This higher force can easily exceed the attraction force keeping the wafer attached to the carrier and result in the wafer coming loose and being left on the pad. This undesirable effect is strongly mitigated by the effect of grooves allowing air to enter under the wafer, equalizing the pressure and alleviating the vacuum effect.
5. Finally, and most important to the process of metal polishing, grooves act as channels for the removal of by-product and polishing debris from the pad surface. While for oxide polishing, a build-up of debris increases the likelihood of scratches and other defects, for metal polishing, the removal of by-products of the reaction is essential to proper continuation of the reaction. Without the removal of these by-products, the reaction will slow and the removal rate will slow. Also, the effect of “staining”, the build up of by product absorbed into the pad surface is worsened without grooves.
It is believed that the existing polishing pads can be further improved.
Accordingly, it is an object of the present invention to provide a polishing pad for mechanical polishing, which is a further improvement of the existing polishing pads of this type. In particular, the present invention seeks to provide improvement related to the reasons 2) and 5) listed in the background section for grooving. Briefly, 2) and 5) teach that grooves enhance the transport of slurry to and from the wafer center. This invention describes a methodology for grooving which seeks to improve that transport. This invention does not address reasons 1, 3 or 4.
Additionally, it is an object of the present invention to recognize two important restrictions on the design of the grooving pattern:
In keeping with these objects and with others which will become apparent hereinafter, one feature of the present invention resides, briefly stated, in a polishing pad which has a body with a working surface adapted to provide polishing of a workpiece, such as for example a wafer, wherein the working surface is provided with a singular or a plurality of grooves, and the grooves are formed so that the direction of the grooves with respect to the primary direction of the relative motion of the wafer and the pad are substantially random. In this usage, the term random is not meant to imply only disordered, but also that if we were to envision a point at the center of the wafer and consider the angles at which it impinges upon grooves, that during the course of a single revolution of the pad it would encounter grooves at almost all angles in roughly equal proportion.
When the grooves on the working surface on the pad are designed in accordance with the present invention, introduction of new slurry and evacuation of spent slurry and by-products is facilitated, particularly to and from the wafer center. Additionally, while there exist patterns that do not immediately appear to be radial and seem to provide randomness of direction such as an x-y grid pattern, such patterns still readily allow for the rapid dispersion of slurry due to centrifugal forces and are therefore not desirable. Additionally, in some cases pads can emit a noise while polishing. In the case where the grooves are primarily in one direction for a period of time (such as an x-y grid or straight parallel lines), the noise will tend to modulate in volume, creating an undesirable condition. Proper randomization of the groove direction will tend to cause the noise to be uniform in volume. Finally, closely spaced grooves which intersect (as they would in an x-y pattern) can weaken the structural integrity of the pad sufficiently so as to reduce its ability to planarize to below-acceptable levels. While this reduction of integrity has been sited as a method to engineer the pad properties (see U.S. Pat. No. 6,736,709), it generally is undesirable.
In accordance with one embodiment of the present invention, the grooves extend in substantially all directions of the plane of the working surface of the pad. The grooves can be of various widths, depths and pitches.
In accordance with still a further feature of the present invention, the working surface of the pads is subdivided into a plurality of individual portions that are space filling, and in each portion the grooves are formed by a plurality of substantially parallel lines. The lines can be of various pitch less than the STL and the grooves can be of various widths, depths and pitches.
In accordance with another embodiment of the present invention, the grooves are as a sine wave pattern on the working surface of the pad. The sine waves can be of various amplitude, wavelength and offset and the grooves can be of various widths, depths and pitches.
In accordance with still a further feature of the present invention, the working surface of the pads is subdivided into a plurality of individual portions, and in each portion the grooves are formed by a plurality of substantially concentric circles. The circles can be of various radii and the grooves can be of various widths, depths and pitches.
It is understood that the pads of this invention can be used for application of process on any of a number of substrates, such as a bare silicon wafer, a semiconductor device wafer, a magnetic memory disk or similar. The pad may be anywhere in the range of what by someone skilled in the art is considered soft (Modulus of Elasticity<1000 psi) to what is considered hard (Modulus of Elasticity>10,000 psi).
Pads of the present invention can be made by any one of a number of polymer processing methods, such as but not limited to, casting, compression, injection molding, extruding, web-coating, extruding, and sintering. The pads may be single phase or multiphase, where the second phase could include polymeric microballoons, gases or fluids. The second phase could also by an abrasive such as silica, alumina and calcium carbonate, alumina, ceria, oxides of titanium, germanium, diamond, silicon carbide or combinations thereof.
The novel features which are considered as characteristic for the present invention are set forth in particular in the appended claims. The invention itself, however, both as to its construction and its method of operation, together with additional objects and advantages thereof, will be best understood from the following description of specific embodiments when read in connection with the accompanying drawings.
A polishing pad for chemical and mechanical polishing is identified as a whole with reference numeral 1. It has a body 2 with a working surface 3. The working surface 3 is provided with a plurality of grooves as will be explained herein below.
The grooves can be produced by laser cutting, and other suitable methods. In the pad the grooves can maintain a pitch of less than 180 micrometers at least everywhere inside the wafer track. The grooves can be uniformly or non-uniformly spaced apart from one another. The grooves are open at a periphery of said working surface. The grooves can be composed of a material selected from the group consisting of a polyurethane. Each of the grooves can have a width of 5 mils to 50 mils. One or more lines of the pattern are composed of grooves, holes or a combination thereof. The grooves can be machined by a laser or by mechanical means.
It will be understood that each of the elements described above, or two or more together, may also find a useful application in other types of constructions differing from the types described above.
While the invention has been illustrated and described as embodied in polishing pad, it is not intended to be limited to the details shown, since various modifications and structural changes may be made without departing in any way from the spirit of the present invention.
Without further analysis, the foregoing will so fully reveal the gist of the present invention that others can, by applying current knowledge, readily adapt it for various applications without omitting features that, from the standpoint of prior art, fairly constitute essential characteristics of the generic or specific aspects of this invention.